Encyclopedia of Lingguistics: Volume 1

e n c y c l o p e d i a o f LINGUISTICS V o l u m e 1 Philipp Strazny Editor e n c y c l o p e d i a o f LINGUISTICS V o l u m e 1 A-L FITZROY DEARBORN An Imprint of the Taylor & Francis Group New York • Oxon This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Published in 2005 by Fitzroy Dearborn An Imprint of the Taylor & Francis Group 270 Madison Avenue New York, NY 10016 Published in Great Britain by Fitzroy Dearborn An Imprint of the Taylor & Francis Group 2 Park Square Milton Park, Abingdon Oxon, OX14 4RN UK Copyright © 2005 by Taylor & Francis Books, Inc. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopy- ing and recording, or in any information storage and retrieval system, without permission in writing from the publishers. Library of Congress Cataloging-in-Publication Data Encyclopedia of linguistics / Philipp Strazny, editor. p. cm. Includes bibliographical references and index. ISBN 1-57958-391-1 (set : alk. paper) -- ISBN 1-57958-450-0 (v. 1 : alk. paper) -- ISBN 1-57958-451-9 (v. 2 : alk. paper) 1. Linguistics--Encyclopedias. I. Strazny, Philipp. P29.E483 2005 410'.3--dc22 2004014173 ISBN 0-203-31920-6 Master e-book ISBN Board of Advisers Robert Beard Bucknell University Bill Foley University of Sydney John Goldsmith University of Chicago David Ingram Arizona State University Francis Katamba Lancaster University Michael Kenstowicz Massachusetts Institute of Technology Darlene LaCharité Université Laval Sydney Lamb Rice University Vincenzo Lombardo Universita di Torino Brian MacWhinney Carnegie Mellon University Horst Müller Universität Bielefeld Emanuel Schegloff University of California, Los Angeles Sarah Thomason University of Michigan Elly van Gelderen Arizona State University  Contents Alphabetical List of Entries ix Thematic List of Entries xv Contributors xxi Introduction xxix Entries A to Z 1 Index 1211  Alphabetical List of Entries Volume 1 Aramaic Arawak A Archaeology and Language Acoustic Phonetics Armenian Acquisition Artificial Intelligence Acquisition Theories Artificial Languages Aerodynamics of the Vocal Tract Aspect Affixation Assimilation and Coarticulation African American Vernacular English Assimilation and Dissimilation Afroasiatic Austin, John Langshaw Age and Language Australia Agreement Austria Ainu Austroasiatic Akan and Nyo Languages Austronesian Albanian Auxiliaries Algeria Altaic B Ambiguity Babylonian Traditional Grammar American Sign Language Balkans Amharic and Ethiopian Semitic Languages Baltic Languages Analogical Change Bambara, Mandenkan and the Mande languages Anaphora Bar-Hillel, Yehoshua Anatomy of the Articulatory System Basque Anatomy of the Auditory System Baudouin de Courtenay, Jan Ignacy NiecisBaw Ancient Egyptian Belgium Ancient Greek Benveniste, Emile Animals and Human Language 1: Overview Bever, Thomas Animals and Human Language 2: Dolphins Bilingual Acquisition Animals and Human Language 3: Parrots Bilingual Mixed Languages Animals and Human Language 4: Primates Bilingualism Aphasia Biosemiotics Applied Linguistics: Overview Bloomfield, Leonard Arabic Boas, Franz Arabic Traditional Grammar Body Language, see Paralanguage ix Alphabetical List of Entries Bopp, Franz D Brain Organization and Auditory Pathway Dakota and Siouan Languages Bresnan, Joan Deep Structure British Sign Language Definiteness Brugmann, Karl Deixis Bühler, Karl Determiner Burmese Developmental Stages Burushaski Dialectology Diglossia C Discourse Analysis Canada Discourse Strategies Cape Verdean Creole Dravidian Carib and Cariban Languages Dutch Caribbean Dyslexia Case Causation E Celtic Languages Emeneau, Murray Barnson Chafe, Wallace Emotion and Language Chao Yucn Ren Empty Categories China Empty Morphemes Chinese (Mandarin) Endangered Languages Chinese and Japanese Traditional Grammar English Chinese Pidgin English Epenthesis and Syncope Chinese Pidgin Russian Eskimo-Aleut Chinook Jargon Ethnicity and Language Chomsky, Noam Ethnography of Communication Clark, Eve V. Etymology Clause Euphemism Clause-Type Indicators European Traditional Grammar Clinical Linguistics Evolution of Language 1: Overview Code-Switching Evolution of Language 2: Cognitive Preadaptations Coherence in Discourse Evolution of Language 3: Physical Preadaptations Color Terms Evolution of Language 4: Social Preadaptations Communication Theory Ewe and Gbe Languages Comparative Method Compositional Semantics F Compounding Fanakalo Computational Linguistics Farsi Computer-Assisted Language Learning (CALL) Feature Theory Comrie, Bernard Ferguson, Charles Albert Configurationality Field Methods Connectionism Figurative Speech Constituency Test Fillmore, Charles John Context Finite-State Morphology Conversation Analysis Finnish and Finnic Languages Coordination Firth, John Rupert Coptic Egyptian Fishman, Joshua A. Corpus Linguistics Focality Courtroom Discourse Forensic Linguistics Crioulo, Gulf of Guinea France Crioulo, Upper Guinea French Language x  Alphabetical List of Entries Fromkin, Victoria Alexandra Hockett, Charles F. Function Words Humboldt, Wilhelm von Functional Approaches Hungarian and Ugric Languages Hymes, Dell Hathaway G Gender and Language I Gender: Class Marking Iconicity Generation Identity and Language Generative Grammar Ideology and Language Genericity Idiomaticity Genetic Relationship Idioms Genre Igbo and Igboid Languages Georgian and Caucasian Languages India German Indian Ocean Creoles Germany Indian Traditional Grammar Givón, Talmy Indo-European 1: Overview Gothic Indo-European 2: Germanic Languages Grammar, Traditional Indo-European 3: Indo-Iranian Languages Grammar, Theories Indo-European 4: Romance Grammatical Function Indo-European 5: Slavic Grammaticalization Indo-Pakistani Sign Language Great Britain Inflection and Derivation Great Vowel Shift Information Retrieval Greek, Modern Interpreting Greenberg, Joseph Harold Irony Grice, H. Paul Israel Grimm, Jacob Italian Guaymí and Chibchan Languages Italy Gullah Gumperz, John Joseph J Jakobson, Roman H Japanese Haas, Mary Rosamond Japanese Sign Language Haitian Creole Javanese Hale, Kenneth Jespersen, Otto Halle, Morris Jones, Sir William Halliday, Michael Alexander Kirkwood Juba and Nubi Arabic Handwriting Harris, Zellig Sabbetai K Haugen, Einar Kayardild and the Tangkic Languages Hausa and Chadic Languages Khmer and Mon-Khmer Languages Hawaiian Creole English Khoisan Hebrew: Biblical Kinship Terms Hebrew: Modern Kituba Hindi-Urdu Korean Hiri Motu Krio (and West African Pidgin English) Historical Linguistics Kriol (Roper River Creole) History of Linguistics: Overview Hittite L Hjelmslev, Louis Labov, William Hmong and Hmong-Mien (Miao-Yao) Languages Ladefoged, Peter xi Alphabetical List of Entries Lakoff, George Modification Langacker, Ronald Mohawk and Iroquoian Languages Language Change Mongolian Language: Contact—Overview Montague, Richard Language Death Mood Language Planning Moore and the Gur Languages Language Socialization Morpheme Lateralization and Handedness Morphological Typology Latin Morphology Lehmann, Winfred Philipp Murrinh-Patha and Daly Languages Leskien, August Lesser Antillean French Creole N Lévi-Strauss, Claude Nahuatl and Uto-Aztecan Languages Lexical Borrowing Naming Lexicalization Natural Classes Lexicon: Overview Navajo and Athabaskan-Eyak Languages Lingua Franca Neurolinguistics Literacy New Guinea Localization of Linguistic Information Nida, Eugene Albert Long-Distance Dependency Niger-Congo Long-Range Comparison Nigeria Louisiana Creole Nilo-Saharan Languages Lyons, John Nootka and Wakashan Languages Northwest Caucasian Languages Volume 2 Noun Incorporation Number Marking M Machine Translation O Malagasy Official Language Selection Malay-Indonesian and Malayic Languages Ojibwe and Algonquian Languages Malkiel, Yakov Okanagan and Salishan Languages Manner of Articulation Old Chinese Ma¯ ori and Polynesian Languages Old Church Slavonic Marshallese and Micronesian Languages Old English Martinet, André Old French Mass Media and Language Old High German Mathematical Linguistics Old Irish Maya and Mayan Languages Old Japanese McCawley, James David Old Norse Meaning Old Tibetan Medicine and Language Onomatopoeia Meillet, Antoine Oto-Manguean Languages Metaphor Metathesis P Metonymy Pacific Mexico Pama-Nyungan Middle (Classical) Japanese Pa¯ nini " Middle English Papiamento Migration and Language Paralanguage Miskito and Misumalpan Languages Parsing Modern Linguistics Paul, Hermann xii  Alphabetical List of Entries Peirce, Charles Sanders Saramaccan Personality and Language Saussure, Ferdinand de Philippine Spanish Creoles Scandinavia Philology Searle, John R. Philosophy of Language Second Language: Acquisition Phoneme Second Language: Learning Phonetic Transcription Second Language: Teaching Phonetics Semantic and Discourse Typology Phonology Semantics Phrase Structure Semiotics Pidgins and Creoles Semitic Languages Pike, Kenneth Lee Serbo-Croatian and South Slavic Languages Pintupi and Pama-Nyungan Languages Serial Verb Constructions Pitcairnese Sign Relationships Place of Articulation Signed Languages Plurality Sindhi Polish and West Slavic Languages Sino-Tibetan Languages Politeness Slobin, Dan Isaac Possessives Sociolect Pragmatics Sociolinguistics Predication Soundwave Analysis Professions for Linguists South Africa Proficiency Testing South America: Argentina Pro-forms South America: Brazil Proper Nouns Southeast Asia Prosody Soviet Union Psycholinguistics Soviet Union: Successor States Spain Q Spanish and Iberoromance Languages Quantification Spectral Analysis Questions Speech Acts Quine, Willard van Orman Speech Perception Speech Processing R Speech Processing (Neurobiology) Reading Speech Production Reading Impairment Speech Production: Neurobiology Reference Speech Synthesis Register Sprachbund Reichenbach, Hans Sranan Relevance in Discourse Sri Lanka Portuguese Rhetoric and Linguistics Standard Language Romanian Story Grammar Rules vs. Constraints Structural Interference Russell, Bertrand Arthur William Structuralism Russian and East Slavic Languages Stylistics Sumerian S Swadesh, Morris Sandhi Swedish and Scandinavian Languages Sanskrit Sweet, Henry Sapir, Edward Switch-Reference Sapir-Whorf Hypothesis Switzerland xiii Alphabetical List of Entries Syllable: Structure Universal Grammar Syntactic Categories Utterance-Centered Linguistics Syntactic Typology Syntax V T Variation Vietnamese Taboo Visual Word Recognition Tagalog and Philippine Languages Voice Tamil Vowel Harmony Tarascan Teaching: Curricula Teaching: Methods W Telugu Wackernagel, Jacob Tense and Aspect Marking Warao Tense: Syntax Wayampi and Tupi-Guarani Languages Tewa and Kiowa-Tanoan Languages Weinreich, Uriel Text Linguistics Western Caribbean Creole(s) Text Understanding Whorf, Benjamin Lee Thai and Tai Languages Wittgenstein, Ludwig Thematic Structure Wolof and Atlantic Languages Tibetan Word Time and Tense Word Order Tiv and Tivoid Languages Word Sense Disambiguation Tocharian Working Memory Tok Pisin Writing Systems Tone Languages Translation Trubetzkoy, Nikolai Sergeyevich Y Tuareg and Berber Languages Yémba and the Grassfields Bantu Languages Tungusic Yiddish Turing, Alan Yoruba and Yoruboid Languages Turkey Typology Z U Zipf, George Kingsley Udmurt Zulu and Southern Bantu Languages United States Zuni xiv  Thematic List of Entries ACQUISITION Ojibwe and Algonquian Languages Acquisition Okanagan and Salishan Languages Acquisition Theories Oto-Manguean Languages Bilingual Acquisition Tarascan Developmental Stages Tewa and Kiowa-Tanoan Languages Language Socialization Wayampi and Tupi-Guarani Languages Second Language: Learning APPLIED LINGUISTICS AFROASIATIC LANGUAGES Applied Linguistics: Overview Amharic and Ethiopian Semitic Languages Computer-Assisted Language Learning (CALL) Arabic Forensic Linguistics Aramaic Genre Biblical Hebrew Language Planning Coptic Egyptian Literacy Hausa and Chadic Languages Professions for Linguists Modern Hebrew Proficiency Testing Tuareg and Berber Languages Reading Second Language: Teaching Stylistics ALTAIC LANGUAGES Mongolian Teaching: Curricula Tungusic Teaching: Methods ARTIFICIAL LANGUAGES AMERICAN INDIAN LANGUAGES Artificial Languages Arawak Carib and Cariban Languages Chibchan Languages AUSTRALIAN LANGUAGES Dakota and Siouan Languages Kayardild and the Tangkic Languages Maya and Mayan Languages Murrinh-Patha and Daly Languages Miskito and Misumalpan Languages Pintupi and Pama-Nyungan Languages Mohawk and Iroquoian Languages Nahuatl and Uto-Aztecan Languages AUSTRO-ASIATIC LANGUAGES Navajo and Athabaskan-Eyak Languages Khmer and Mon-Khmer Languages Nootka and Wakashan Languages Vietnamese xv Thematic List of Entries AUSTRONESIAN LANGUAGES DRAVIDIAN LANGUAGES Javanese Tamil Malagasy Telugu Malay-Indonesian and Malayic Languages Ma¯ ori and Polynesian Languages HISTORICAL LINGUISTICS Marshallese and Micronesian Languages Analogical Change Tagalog and Philippine Languages Archeology and Language Comparative Method COMMUNICATION THEORY Dialectology Animals and Human Language 1: Overview Etymology Animals and Human Language 2: Dolphins Evolution of Language: Overview Animals and Human Language 3: Parrots Evolution of Language: Cognitive Preadaptations Animals and Human Language 4: Primates Evolution of Language: Physical Preadaptations Biosemiotics Evolution of Language: Social Preadaptations Interpreting Genetic Relationship Paralanguage Great Vowel Shift Translation Language Change Long-Range Comparison COMPUTATIONAL LINGUISTICS Migration and Language Artificial Intelligence Finite State Morphology HISTORY OF LINGUISTICS Generation Arabic Traditional Grammar Information Retrieval Babylonian Traditional Grammar Machine Translation Chinese and Japanese Traditional Grammar Parsing European Traditional Grammar Speech Processing History of Linguistics: Overview Text Understanding Indian Traditional Grammar Visual Word Recognition Modern Linguistics Word Sense Disambiguation INDO-EUROPEAN LANGUAGES DEAD LANGUAGES Albanian Ancient Egyptian Armenian Ancient Greek Baltic Languages Gothic Celtic Languages Hittite Dutch Latin English Middle English Farsi Middle (Classical) Japanese French Language Old Chinese German Old Church Slavonic Greek, Modern Old English Hindi-Urdu Old French Italian Old High German Polish and West Slavic Languages Old Irish Romanian Old Japanese Russian and East Slavic Languages Old Norse Serbo-Croatian and South Slavic Old Tibetan Sindhi Sanskrit Spanish and Iberoromance Languages Sumerian Swedish and Scandinavian Languages Tocharian Yiddish xvi  Thematic List of Entries ISOLATES (LANGUAGES) Semiotics Ainu Sociolinguistics Basque Structuralism Burushaski Syntax Japanese Typology Korean Writing Systems Warao Zuni MAJOR LANGUAGE FAMILIES Afroasiatic LANGUAGE CONTACT Altaic Bilingual Mixed Languages Austroasiatic Bilingualism Austronesian Endangered Languages Dravidian Language: Contact—Overview Eskimo-Aleut Language Death Hmong and Hmong-Mien (Miao-Yao) languages Lexical Borrowing Indo-European 5: Slavic Official Language Selection Indo-European 4: Germanic Languages Pidgins and Creoles Indo-European 3: Indo-Iranian Sprachbund Indo-European 2: Overview Structural Interference Indo-European 1: Romance Khoisan LEXICON Niger-Congo Clause Lexemes Nilo-Saharan Languages Grammaticalization Pama-Nyungan Idiomaticity Semitic Languages Lexicalization Sino-Tibetan Languages Lexicon: Overview MORPHOLOGY LINGUISTIC APPROACHES Affixation Communication Theory Agreement Computational Linguistics Compounding Connectionism Empty Morphemes Corpus Linguistics Gender: Class Marking Field Methods Inflection and Derivation Functional Approaches Morpheme Generative Grammar Number Marking Grammar, Theories Sandhi Grammar, Traditional Switch-Reference Historical Linguistics Tense and Aspect Marking Mathematical Linguistics Vowel Harmony Morphology Word Neurolinguistics Philology NEUROLINGUISTICS Philosophy of Language Aphasia Phonetics Brain Organization and Auditory Pathway Phonology Clinical Linguistics Pragmatics Dyslexia Psycholinguistics Lateralization and Handedness Rhetoric and Linguistics Localization of Linguistic Information Semantics Reading Impairment xvii Thematic List of Entries Second Language: Acquisition Halliday, Michael Alexander Kirkwood Speech Processing (Neurobiology) Harris, Zellig Sabbetai Speech Production: Neurobiology Haugen, Einar Working Memory Hjelmslev, Louis Hockett, Charles F. NIGER-CONGO LANGUAGES Humboldt, Wilhelm von Akan and Nyo Languages Hymes, Dell Hathaway Bambara, Mandenkan, and the Mande Languages Jakobson, Roman Ewe and Gbe Languages Jespersen, Otto Igbo and Igboid Languages Jones, Sir William Moore and the Gur Languages Labov, William Tiv and Tivoid Languages Ladefoged, Peter Wolof and Atlantic Languages Lakoff, George Yémba and the Grassfields Bantu Languages Langacker, Ronald Yoruba and Yoruboid Languages Lehmann, Winfred Philipp Zulu and Southern Bantu Languages Leskien, August Lévi-Strauss, Claude NORTH CAUCASIAN LANGUAGES Lyons, John Georgian and Caucasian Languages Malkiel, Yakov Northwest Caucasian Languages Martinet, André McCawley, James David PERSONS Meillet, Antoine Austin, John Langshaw Montague, Richard Bar-Hillel, Yehoshua Nida, Eugene Albert Baudouin de Courtenay, Jan Ignacy NiecisBaw Pa¯ nini " Benveniste, Emile Paul, Hermann Bever, Thomas Peirce, Charles Sanders Bloomfield, Leonard Pike, Kenneth Lee Boas, Franz Quine, Willard van Orman Bopp, Franz Reichenbach, Hans Bresnan, Joan Russell, Bertrand Arthur William Brugmann, Karl Sapir, Edward Bühler, Karl Saussure, Ferdinand de Chafe, Wallace Searle, John R. Chao, Yuen Ren Slobin, Dan Isaac Chomsky, Noam Swadesh, Morris Clark, Eve V. Sweet, Henry Emeneau, Murray Barnson Trubetzkoy, Nikolai Sergeyevich Ferguson, Charles Albert Turing, Alan Fillmore, Charles John Wackernagel, Jacob Firth, John Rupert Weinreich, Uriel Fishman, Joshua A. Whorf, Benjamin Lee Fromkin, Victoria Alexandra Wittgenstein, Ludwig Givón, Talmy Zipf, George Kingsley Greenberg, Joseph Harold Grice, H. Paul PHONETICS Grimm, Jacob Acoustic Phonetics Gumperz, John Joseph Aerodynamics of the Vocal Tract Haas, Mary Rosamond Anatomy of the Articulatory System Hale, Kenneth Anatomy of the Auditory System Halle, Morris Assimilation and Dissimilation xviii  Thematic List of Entries Manner of Articulation Indian Ocean Creoles Phonetic Transcription Juba and Nubi Arabic Place of Articulation Kituba Prosody Krio (and West African Pidgin English) Soundwave Analysis Kriol (Roper River Creole) Spectral Analysis Lesser Antillean French Creole Speech Synthesis Lingua Franca Louisiana Creole PHONOLOGY Papiamento Assimilation and Dissimilation Philippine Spanish Creole(s) Epenthesis and Syncope Pitcairnese Feature Theory Saramaccan Metathesis Western Caribbean Creole Natural Classes Sranan Phoneme Sri Lanka Portuguese Rules vs. Constraints Tok Pisin Syllable: Structure Tone Languages REGIONS Algeria PRAGMATICS Australia Code-Switching Austria Coherence in Discourse Balkans Conversation Analysis Belgium Deixis Canada Discourse Analysis Caribbean Discourse Strategies China Irony France Politeness Germany Relevance in Discourse Great Britain Speech Acts India Taboo Israel Text Linguistics Italy Utterance-Centered Linguistics Mexico PSYCHOLINGUISTICS New Guinea Age and Language Nigeria Handwriting Pacific Speech Perception Scandinavia Speech Production South Africa South America: Argentina PIDGINS AND CREOLES South America: Brazil Cape Verdean Creole Southeast Asia Chinese Pidgin English Soviet Union Chinese Pidgin Russian Soviet Union: Successor States Chinook Jargon Spain Crioulo, Gulf of Guinea Switzerland Crioulo, Upper Guinea Turkey Fanakalo United States Gullah Haitian Creole SEMANTICS Hawaiian Creole English Ambiguity Hiri Motu Causation xix Thematic List of Entries Color Terms Personality and Language Compositional Semantics Sociolect Context Standard Language Coordination Variation Definiteness Figurative Speech SYNTAX Genericity Anaphora Idioms Aspect Kinship Terms Auxiliaries Meaning Case Metaphor Clause Metonymy Configurationality Modification Constituency Test Naming Deep Structure Onomatopoeia Determiner Plurality Empty Categories Proper Nouns Focality Quantification Function Words Reference Grammatical Function Register Iconicity Sapir–Whorf Hypothesis Long-Distance Dependency Sign Relationship Mood Time and Tense Noun Incorporation Phrase Structure SIGN LANGUAGES Possessives American Sign Language Predication British Sign Language Pro-forms Indo-Pakistani Sign Language Questions Japanese Sign Language Serial Verb Constructions Signed Languages Story Grammar Syntactic Categories SINO-TIBETAN LANGUAGES Tense: Syntax Burmese Thematic Structure Chinese: Mandarin Universal Grammar Tibetan Voice Word Order SOCIOLINGUISTICS African American Vernacular English TAI LANGUAGES Courtroom Discourse Thai and Tai Languages Diglossia Emotion and Language TYPOLOGY Ethnicity and Language Morphological Typology Ethnography of Communication Semantic and Discourse Typology Euphemism Syntactic Typology Gender and Language Identity and Language URALIC LANGUAGES Ideology and Language Finnish and Finnic Languages Mass Media and Language Hungarian and Ugric Languages Medicine and Language Udmurt xx  Contributors Abazova, Alfia. Jersey City, New Jersey. Baptista, Marlyse. Linguistics Program, University of Georgia, USA. Ackema, Peter. University of Edinburgh, UK. Barcelona, Antonio. Departamento de Filología Inglesa, Aikhenvald, Alexandra. Research Center for Linguistic Universidad de Murcia, Spain. Typology, Institute for Advanced Study, La Trobe University, Melbourne, Australia. Barnard, Roger. Department of General and Applied Linguistics, University of Waikato, New Zealand. Altarriba, Jeanette. Department of Psychology, University at Albany, State University of New York, USA. Barron, Anne. Englisches Seminar, Universität Bonn, Germany. Altmann, Lori J.P. Department of Communication Barry, Betsy. University of Georgia, USA. Sciences and Disorders, University of Florida, USA. Barss, Andrew. Department of Linguistics, University of Anderson, Warren D. Department of Foreign Arizona, USA. Languages and Anthropology, Southeast Missouri State University, USA. Barton, Ellen. Linguistics Program, Department of English, Wayne State University, Michigan. Antieau, Lamont. Linguistics Program, University of Georgia, USA. Bashir, Elena. University of Chicago, Illinois. Archibald, John. Department of Linguistics, University Bastiaanse, Roelien. University of Groningen, The of Calgary, Alberta, Canada. Netherlands. Arends, Jacques. Department of Linguistics, University Basturkmen, Helen. Department of Applied Language of Amsterdam, The Netherlands. Studies and Linguistics, Auckland University, New Zealand. Aronoff, Mark. Department of Linguistics, Stony Brook Bat-El, Outi. Department of Linguistics, Tel-Aviv University, Stony Brook, New York, USA. University, Israel. Battaner-Moro, Elena. Departamento de Lengua Attardo, Salvatore. English Department, Youngstown Española y Lingüística General, Universidad Nacional de State University, Ohio. Educación a Distancia, Madrid, Spain. Axelrod, Melissa. Department of Linguistics, University Bauer, Laurie. Linguistics and Applied Language of New Mexico, USA. Studies, Victoria University of Wellington, New Zealand. Baldauf, Richard B., Jr. School of Education, University Baumgartner, Joanne Marie. Australian National of Queensland, Australia. University, Canberra, Australia. xxi Contributors Beckwith, Christopher I. Indiana University, USA. Carter, Michael G. Department of Oriental and East European Studies, Oslo University, Norway. Belnap, R. Kirk. Brigham Young University, Utah. Clankie, Shawn M. Center for Language Studies, Otaru Benner, Allison. Victoria, British Columbia, Canada. University of Commerce, Japan. Benremouga, Karima. Fine Arts and Language Arts, San Cloutier, Robert A. Amsterdam Center for Language and Jacinto College Central, Pasadena, TX. Communication/Department of Dutch Studies, University of Amsterdam, Amsterdam, The Netherlands. Berge, Anna. Alaska Native Language Center, University of Alaska, Fairbanks, USA. Cockerton, Camilla M. Happy Valley, Rakaia, New Zealand. Bernhardt, Karl. University of Buckingham, United Kingdom. Cogill-Koez D. Department of Linguistics, University of Bhatia, Tej K. Syracuse University, New York. New England, Armidale, Australia. Bird, Steven. Linguistic Data Consortium, University of Cohen, Ariel. Department of Foreign Literatures and Pennsylvania, Philadelphia. Linguistics, Ben-Gurion University of the Negev, Beer Sheva, Israel. Blume, Kerstin. Bergische Universität Wuppertal, Germany. Colarusso, John. McMaster University, Hamilton, Bodomo, Adams. Department of Linguistics, University Ontario, Canada. of Hong Kong, China. Cole, Jennifer. Department of Linguistics, University of Bolotin, Naomi. Cambridge, Massachusetts. Illinois at Urbana-Champaign, USA. Born, Renate. Department of Germanic and Slavic Colley, Michael. Linguistics Program, University of Languages, University of Georgia, USA. Georgia, USA. Bowden, John. Linguistics Department, Research School Comrie, Bernard. Max-Planck-Institut für evolutionäre of Pacific and Asian Studies, The Australian National Anthropologie, Leipzig, Germany. University, Australia. Corré, Alan D. Hebrew Studies, University of Wisconsin- Bretones, Carmen. International Computer Science Milwaukee, USA. Institute, Berkeley, California. Cots, Josep Maria. Departament d’Anglès i Lingüística, Brown, Jeffrey M. Texas A&M International University, Universitat de Lleida, Catalunya, Spain. USA. Croft, William. Department of Linguistics, University of Branigan, Holly. School of Philosophy, Psychology, and Manchester, United Kingdom. Language Sciences, University of Edinburgh. Burley, Lynn Ann. Department of Writing and Speech, Crowley, Terry. Department of General and Applied University of Central Arkansas, USA. Linguistics, University of Waikato, New Zealand. Camdzic, Amela. University College London, United Daniels, Peter T. Bronx, New York. Kingdom. Daniliuc, Radu. Department of Linguistics, Australian Campana, Mark. Kobe, Hyogo-ken, Japan. National University, Canberra, Australia. Carnie, Andrew. Department of Linguistics, University of Daniliuc, Laura. Department of Linguistics, Australian Arizona, USA. National University, Kambah, Australia. Carstairs-McCarthy, Andrew. Linguistics Department, Daza, Jaime Luis. Linguistics Department, Unversity of University of Canterbury, Christchurch, New Zealand. California, Los Angeles, USA. xxii  Contributors Dimitriu, Rodica. Department of English, University of González-Romero, Luisa. Departamento de Filología Iasi, Romania. Inglesa, Universidad de Huelva, Andalucía, Spain. Ding, Picus Sizhi. Macao, Hong Kong, China. Gordon, Lynn. Department of English, Washington State University, USA. Dooley Collberg, Sheila. Department of Linguistics, University of Arizona, USA. Graesser, Arthur. Department of Psychology, University of Memphis, USA. Douglas, Dan. Department of English, Iowa State University, USA. Green, Ian. Graduate School, University of Tasmania, Hobart, Australia. Dumas, Bethany K. Department of English, University of Tennessee, USA. Greenberg, Marc L. Slavic Department, University of Kansas, USA. Ebermann, Erwin. Department of African Studies, University of Vienna, Austria. Grenoble, Lenore A. Program in Linguistics and Cognitive Science, Dartmouth College, New Hampshire. Echeruo, Michael J.C. Department of English, Syracuse University, Syracuse, New York, USA. Gribble, Charles. Slavic Department, Ohio State University, USA. Eira, Christina. Adelaide Graduate Centre, University of Adelaide, Australia. Grohmann, Kleanthes K. Department of English Studies, University of Cyprus, Nicosia. Engelberg, Stefan. University of Wuppertal, Germany. Grondona, Veronica M. Department of English, Eastern Ennaji, Moha. Fez, Morocco. Michigan University, USA. Essegbey, James. Department of African Linguistics, Guillaume, Jacques. Paris, France. Leiden University, The Netherlands. Gutiérrez-Rexach, Javier. Department of Spanish & Evans, Nicholas. Department of Linguistics & Applied Portuguese, Ohio State University, USA. Linguistics, University of Melbourne, Australia. Falk, Julia S. La Jolla, California, USA. Haberlandt, Karl. Department of Psychology, Trinity College, Connecticut, USA. Feldman, Laurie Beth. University of Albany, State University of New York, USA. Hahn, Udo. Linguistische Informatik/Computerlinguistik, Freiburg University, Germany. Fettes, Mark. Simon Fraser University, Ottawa, Ontario, Canada. Hajicova, Eva. MFF UK, Prague, Czech Republic. Flier, Michael S. Department of Slavic Languages and Hall, David R. Linguistics Department, Macquarie Literatures, Harvard University, Massachusetts, USA. University Sydney, Australia. Foster-Cohen, Susan H. The Champion Centre and Hall, Kira. Department of Linguistics, Stanford University of Canterbury, Christchurch, New Zealand. University, California. Gagné, Christina L. Department of Psychology, Hammond, Michael. Department of Linguistics, University of Alberta, Edmonton, Canada. University of Arizona, USA. Gildea, Spike. Department of Linguistics, University of Harlow, Ray. Department of General and Applied Oregon, USA. Linguistics, University of Waikato, New Zealand. Goatly, Andrew. Department of English, Lingnan Harrington, Jonathan. Institute of Phonetics and Digital University, N.T., Hong Kong, China. Speech Processing, University of Kiel, Germany. xxiii Contributors Harrison, K. David. Linguistics Department, Swarthmore Jaggar, Philip J. School of Oriental & African Studies, College, Swarthmore, Pennsylvania, USA. London, United Kingdom. Hartig, Matthias. Sprach- und Literaturwissenschaft, Jensen, Cheryl. Summer Institute of Linguistics, Belém, Universität Paderborn, Germany. Pará, Brazil. Hartsuiker, Robert J. Department of Experimental Johnson, E. Marcia. Department of General & Applied Psychology, Ghent University, Belgium. Linguistics, University of Waikato, New Zealand. Heine, Bernd. Institut für Afrikanistik, Universität zu Johnson, Ellen. Department of English, Rhetoric, and Köln, Germany. Writing, Berry College, Georgia. Jones, Francis R. School of Modern Languages, Hendriks, Petra. Center for Language and Cognition University of Newcastle, United Kingdom. Groningen, University of Groningen, The Netherlands. Jongman, Allard. Department of Linguistics, University Heredia, Roberto R. Texas A&M International of Kansas, USA. University, USA. Justus, Carol F. Classics Department, University of Herrmann, Stefanie. University of Tübingen, Baden- Texas, Austin, USA. Württemberg, Germany. Karimi, Simin. Department of Linguistics, University of Holtgraves, Thomas M. Psychological Sciences, Ball Arizona, USA. State University, Indiana. Katz, Leonard. Haskins Laboratories, New Haven, Hongladarom, Krisadawan. Department of Linguistics, Connecticut. Faculty of Arts, Chulalongkorn University, Bangkok, Thailand. Kaye, Alan S. Linguistics Department, California State University-Fullerton, USA. Hopple, Paulette. Summer Institute of Linguistics, University of Texas, Arlington, USA. Kemper, Susan. Department of Psychology, University of Kansas, USA. Howe, Darin. Department of Linguistics, University of Calgary, Alberta, Canada. Kibrik, Andrej. Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany. Hua, Zhu. Department of Speech, University of Newcastle upon Tyne, United Kingdom. King, Robert D. Department of Linguistics, University of Texas, Austin, USA. Hudson, Grover. Linguistics Department, Michigan State Kirtchuk-Halevi, Pablo Isaac. Beer Sheva, Israel. University, USA. Knight, Chris. School of Social Sciences, University of Hurford, James R. Department of Linguistics, University East London, Dagenham, Essex, United Kingdom. of Edinburgh, Scotland, United Kingdom. Kroskrity, Paul V. Department of Anthropology, Hutchins, W. John. Norwich, United Kingdom. University of California, Los Angeles, USA. Ibarretxe-Antunano, Iraide. Facultad de Filosofía y Kull, Kalevi. Department of Semiotics, University of Letras, Universidad de Deusto - Deustuko Unibertsitatea, Tartu, Estonia. Bizkaia, Spain. Kullavanijaya, Pranee. Department of Linguistics, Inchaurralde, Carlos. Departamento Filologia Inglesa y Faculty of Arts, Bangkok, Thailand. Alemana, Universidad de Zaragoza, Spain. Labeau, Emmanuelle. School of Languages and Social Insler, Stanley. Department of Linguistics, Yale Sciences, Aston University, Birmingham, West Midlands, University, Connecticut. United Kingdom. xxiv  Contributors Lahousse, Karen. Katholieke Universiteit Leuven Mahmoodi-Bakhtiari, Behrooz. Center for the Great (Catholic University Leuven) & Fonds voor, Islamic Encyclopedia, Tehran, Iran. Wetenschappelijk Onderzoek–Vlaanderen (Fund of Scientific Research–Flanders). Marret, Carine. Nice, France. Lainio, Jarmo. Centre for Finnish Studies, Mälardalen Martín Arista, Javier. Departamento de Filologías University, Eskilstuna/Västerås, Swedan. Modernas, Universidad de La Rioja, Logroño, Spain. Lang, George. Dean of Arts, University of Ottawa, Canada. Martín Camacho, José Carlos. Facultad de Filosofía y Letras, Universidad de Extremadura, Cáceres, Spain. Lasnik, Howard. University of Maryland, USA. Mattina, Anthony. Department of Anthropology, Leiber, Justin. Philosophy Department, University of University of Montana, USA. Houston, Texas. Mattissen, Johanna. Institut für Linguistik (Department of Linguistics), University of Köln (Cologne), Germany. Leitner, Gerhard. Institut fuer englische Philologie, Freie Universität Berlin, Germany. Maxwell, Judith M. Department of Anthropology, Tulane University, Louisiana. Levis, John M. Department of English, Iowa State University, USA. Mc Laughlin, Fiona. African Languages and Linguistics, University of Florida, USA. Lieberman, Philip. Department of Cognitive and Linguistic Sciences, Brown University, Rhode Island. McDonald, David. Arlington, Massachusetts. Llurda, Enric. Departament d’Anglès i Linguistica, McGregor, William B. Department of Linguistics, Universitat de Lleida, Catalunya, Spain. University of Aarhus, Denmark. Loengarov, Alexander. Research Assistant of the Fund McMenamin, Gerald. Fresno, California. for Scientific Research–Flanders (Belgium). Department of Linguistics, K.U. Leuven, Belgium. Meisel, Jürgen M. Institut für Romanistik, Universität Hamburg, Germany. Lohr, Marisa. Trinity College, Cambridge, United Kingdom. Melançon, Megan E. Department of English, Speech, and Journalism, Georgia College and State University, USA. Louwerse, Max M. Department of Psychology/Institute for Intelligent Systems, University of Memphis, USA. Menchetti, Angiolo. Dipartimento di Scienze Storiche del Mondo Antico, Università di Pisa, Italy. Luo, Yongxian. Melbourne Institute of Asian Languages Miller, Catherine. Iremam-MMSH, University of Aix en and Societies, University of Melbourne, Australia. Provence, France. Luraghi, Silvia. Dipartimento di Linguistica, Università Minkova, Donka. Department of English, University of di Pavia, Italy. California, Los Angeles, USA. Mackert, Michael. German-English Language Services, Mithun, Marianne. Department of Linguistics, Morgantown, West Virginia, USA. University of California, Santa Barbara, USA. MacMahon, Michael K.C. Department of English Moore, John. Department of Linguistics, University of Language, University of Glasgow, Scotland, United California, San Diego, USA. Kingdom. Morgan, Michael. Kobe City University of Foreign Mahanta, Deepshikha. Central Institute of English and Studies, Kobe, Japan. Foreign Languages, Hyderabad, India. Mufwene, Salikoko. Department of Linguistics, Mahdi, Waruno. Berlin, Germany. University of Chicago, Illinois, USA. xxv Contributors Mukherjee, Joybrato. Department of English, Justus Paap, Kenneth R. Department of Psychology, New Liebig University, Giessen, Germany. Mexico State University, USA. Müller, Horst M. AG Experimentelle Neurolinguistik, Paltridge, Brian. OE Faculty of Education, University of Fakultät für Linguistik und Literaturwissenschaft, Sydney. Universität Bielefeld, Germany. Pandey, Anjali. Department of English, Salisbury Murray, Robert W. Department of Linguistics, University, Maryland, USA. University of Calgary, Alberta, Canada. Pandey, Anita. Department of English & Language Arts, Murray, Denise. National Centre for English Language Morgan State University, Maryland Teaching and Research, Macquarie University, Ryde, Australia. Papangeli, Dimitra. Faculteit der Letteren, Utrecht Institute of Linguistics, OTS, The Netherlands. Nash, David. Australian National University, Canberra, Papen, Robert A. Linguistique et de didactique des Australia. langues, Université du Québec, Montréal, Canada. Nau, Nicole. General Linguistics, Kiel University, Germany. Parasher, Shree Vallabh. Central Institute of English and Foreign Languages, Hyderabad, India. Newmark, Leonard. University of California, San Diego, USA. Parkvall, Mikael. Institutionen för Lingvistik, Stockholms Universitet, Sweden. Nichols, Lynn. Department of Linguistics, University of California, Berkeley. Pensalfini, Rob. School of English, Media Studies and Art History, University of Queensland, Brisbane, Nifadopoulos, Christos G. Bristol, United Kingdom. Australia. Noma, Hideki. Department of Korean Studies, Tokyo Pepperberg, Irene M. Department of Psychology, University of Foreign Studies, Japan. Brandeis University, Waltham, MA 02454. O’Rourke, Sean. Department of Linguistics, Yale Perkins, Michael. Human Communication Sciences, University, Connecticut. University of Sheffield, South Yorkshire, United Kingdom. Oakes, Michael P. University of Sunderland, United Pierce, Marc. Department of Germanic Languages and Kingdom. Literatures, University of Michigan, USA. Obeng, Samuel Gyasi. Linguistics Department, Indiana Piller, Ingrid. Linguistics Department, University of University, USA. Sydney, Australia. Ohala, John J. Department of Linguistics, University of Pineda, Baron. Department of Anthropology, Oberlin California at Berkeley, USA. College, Ohio. Pires, Acrisio. Department of Linguistics, University of Oliverio, Giulia. Alaska Native Language Center, Michigan, USA. University of Alaska-Fairbanks, USA. Ponchon, Thierry. Université de Reims, France. Ongstad, Sigmund. Oslo University College, Norway. Prakasam, Vennelakanti. Central Institute of English and O . la Orie, O . lanike. . Linguistic Program & Department of Foreign Languages, Lucknow, India. Anthropology, Tulane University, Louisiana. Quesada, J. Diego. Escuela de Literatura y Ciencias del Orr, Robert. Gloucester, Ontario, Canada. Lenguaje. Universidad Nacional. Heredia. Costa Rica. Otero, Carlos. University of California, Los Angeles, USA. Radev, Dragomir R. School of Information, Department of Electrical Engineering and Computer Science, and Oventile, Robert S. Pasadena City College, California. Department of Linguistics, University of Michigan, USA. xxvi  Contributors Rajagopalan, Kanavillil. State University at Campinas, Sayahi, Lotfi. Department of Languages, Literatures and Campinas, Brazil. Cultures, The University at Albany, State University of New York, USA. Rathert, Monika. Seminar für Sprachwissenschaft, Universität Tübingen, Baden-Württemberg, Germany. Schneider, Klaus P. Englisches Seminar, Universität Bonn, Germany. Reh, Mechthild. Department of African Languages and Ethiopian Studies, Hamburg University, Germany. Scott, Suzanne. Dunedin, New Zealand. Reid, Nicholas. School of Languages, Cultures and Sew, Jyh Wee. Singapore. Linguistics, University of New England, Armidale, Australia. Sezer, Engin. Bilkent University, Ankara, Turkey. Reid, Lawrence A. Tokyo, Japan. Sgall, Prague Petr. Faculty of Mathematics and Physics, Prague, Czech Republic. Reintges, Chris H. Center for Linguistics, Leiden University, The Netherlands. Sgarbas, Kyriakos N. Department of Electrical and Computer Engineering, University of Patras, Greece. Reiss, Charles. Linguistics Program, Concordia University, Montreal, Canada. Shekar, Chandra. California State University, Fresno. Riedinger, Edward A. Department of Spanish and Sicoli, Mark. Departments of Anthropology and Portugese, The Ohio State University, USA. Linguistics, University of Michigan, USA. Rodríguez-Arrizabalaga, Beatriz. Departamento de Siegel, Jeff. School of Languages, Cultures and Filología Inglesa. University of Huelva, Andalucía, Spain. Linguistics, University of New England, Australia, and Department of Second Language Studies, University of Rogers, Henry. Department of Linguistics, University of Hawai’i, USA. Toronto, Ontario, Canada. Signorini, Inês. Department of Applied Linguistics, Romaine, Suzanne. Merton College, University of Universidade Estadual de Campinas (UNICAMP), Oxford, United Kingdom. São Paulo, Brazil. Rubio, Gonzalo. Department of Classics and Ancient Mediterranean Studies, Pennsylvania State University, USA. Silzer, Peter J. Department of TESOL and Applied Linguistics, Biola University, USA. Rutkowski, PaweB. Department of Polish, Warsaw University, Poland. Siple, Patricia. Department of Psychology, Wayne State University, Michigan. Sackmann, Robin. Freie Universität Berlin, FB PhilGeist, Freie Universität Berlin, Germany. Sitzmann, Alexander. Institut für Germanistik, Universität Wien, Austria. Sanders, Alton. Department of Computer Science and Systems Analysis, Miami University, Ohio. Sivell, John N. Department of Applied Language Studies, Brock University, St. Catharines, Ontario, Canada. Sanders, Carol. Department of Linguistic, Cultural and International Studies, University of Surrey, Guildford, Smith, Ian R. Department of Languages, Literatures and United Kingdom. Linguistics, York University, Toronto, Ontario, Canada. Sanders, Ruth H. Department of German, Russian, East Smith, Allison. Department of English, Louisiana Tech. Asian and Hebrew, Miami University, Ohio. University, USA. Sansò, Andrea. Dipartimento di Linguistica, Universita’ Snoj, Marko. Institut za slovenski jezik ZRC SAZU, di Pisa, Italy. Slovenia. Savova, Lilia. Department of English, Indiana University Sotillo, Susana M. Department of Linguistics, Montclair of Pennsylvania, USA. State University, New Jersey. xxvii Contributors Steinkrüger, Patrick O. Zentrum für allgemeine Vovin, Alexander. Department of East Asian Languages Sprachwissenschaft, Berlin, Germany. and Literatures, University of Hawai’i, USA. Strazny, Philipp. Monona, USA. Vrooman, Michael. Department of Modern Languages and Literatures, Grand Valley State University, Michigan. Sujoldzic, Anita. Institute for Anthropological Research, Zagreb, Croatia. Wagner, Richard. Department of Psychology, Florida State University, USA. Sunderland, Jane. Department of Linguistics and English Language, Bowland College, Lancaster University, Walker, Douglas C. Department of French, Italian and Lancashire, United Kingdom. Spanish, University of Calgary, Alberta, Canada. Tadmor, Uri. Department of Linguistics, Max Planck Walker, James A. Department of Languages, Literatures, Institute for Evolutionary Anthropology, Leipzig, and Linguistics, York University, Toronto, Ontario, Canada. Germany/Jakarta, Indonesia. Wang, William S.-Y. Department of Electronic Tambovtsev, Yuri. Department of English and Engineering, City University of Hong Kong, China. Linguistics, KFNPU, Novosibirsk, Russia. Watt, William C. Social Science, University of Tatham, Mark. Colchester, United Kingdom. California, Irvine, USA. ten Hacken, Pius. School of European Languages, Waugh, Linda R. Department of French, Italian and University of Wales Swansea, UK. English, University of Arizona, USA. Thomason, Sarah G. University of Michigan, USA. Wehmeyer, Ann. Department of African and Asian Languages and Literatures, University of Florida, USA. Thomassen, Arnold J.W.M. University of Nijmegen, The Netherlands. Weiss, Sabine. Brain Research Institute, Cognitive Neuroscience Group, Medical University of Vienna, Austria. Toribio, Josefa. Department of Philosophy, The University of Edinburgh, Scotland. Wexler, Paul. Department of Linguistics, Tel-Aviv University, Israel. TSE, Grace Yuen Wah. School of Arts and Social Sciences, The Open University of Hong Kong, China. Whaley, Lindsay. Linguistics & Cognitive Science, Tsedryk, Egor. Department of French Studies, University Dartmouth College, New Hampshire. of Western Ontario, London, Canada and Department of Wilcox, Phyllis Perrin. Department of Linguistics, French Studies, Concordia University, Montreal, Canada. University of New Mexico, USA. Tserdanelis, Giorgos. The Ohio State University, USA. Wilson, Andrew. Linguistics Department, Lancaster Turovski, Aleksei. Tallinn Zoo, Tallinn, Estonia. University, United Kingdom. van Gelderen, Elly. English Department, Arizona State Wireback, Kenneth J. Department of Spanish and University, USA. Portuguese, Miami University, Ohio. van Wieringen, Astrid. Lab. Exp. ORL, Katholieke Woll, Bencie. Sign Language & Deaf Studies, Department Universiteit Leuven, Belgium. of Language and Communication Science, City University, London, United Kingdom. Vaux, Bert. Department of Foreign Languages and Linguistics, University of Wisconsin-Milwaukee, USA. Wong, Andrew. Department of Linguistics, University of Hawai’i at Manoa, USA. Veselinovic, Elvira M. University of Cologne, Germany. Woodman, Karen. School of Languages, Cultures and Viereck, Wolfgang. University of Bamberg, Germany. Linguistics, University of New England, Armidale, Australia. Voeltz, F.K. Erhard. Institut für Afrikanistik, Universität Zeshan, Ulrike. Research Centre for Linguistic Typology, zu Köln, Germany. La Trobe University, Bundoora, Australia. xxviii  Introduction The study of language goes far back in recorded history. and I made this encyclopedia very readable by eliminating Almost two-and-a-half millenia ago, the Indian grammari- technical terminology as far as possible and by making an Panini wrote his formal treatises on Sanskrit, Xun Zi each essay self-contained. appeared as China’s first major philosopher of language, and Plato and Aristotle initiated the Greek philosophy of How to Use This Book language. Since the renaissance, there has been an increas- ing focus on the description of individual languages, the The Encyclopedia of Linguistics is organized into a series exploration of familial relationships between languages, of 508 free-standing essays, between 1000 and 3000 words and the formulation of increasingly general theories of lan- in length. They range from factual narrative entries to the- guage structure. Now, almost every university or college matic and analytical discussions, and combinations of all has a language department or even a specialized linguistics these. Where debates and controversies occur, these are department, which means that an immense number of indicated and discussed. As far as possible, this book takes researchers are working in the field and have published an the field of linguistics up to the present, at least to the open- enormous body of primary literature. ing years of the twenty-first century. Given this wealth of published data, it is not surpris- Perhaps the most significant feature of the encyclopedia ing to meet beginning graduate students of linguistics is the easily accessible A to Z format. Cross-referencing in who already identify themselves as ‘syntacticians’ or the form of See Alsos at the end of each entry refers the ‘phoneticians’. This early specialization reflects genuine reader to other related essays. A thorough, analytical Index interest, but is also in part a mechanism by which stu- complements the accessibility of the entries, easing the dents block out a large number of possible inputs to be reader’s entry into the wealth of information provided. able to concentrate on a more manageable few. In their References at the end of each entry refer the reader to sem- subsequent research, most researchers narrow down inal writings as well as some of the most recent work on the their field of interest even more; they become immersed subject. Other special features include 12 language-distri- in the highly conceptual and terminological world of bution maps and a thematic Table of Contents in addition their specialty, and they often write articles comprehen- to an alphabetical Table of Contents. In addition, more sible only to their small group of peers. In other words, than 100 illustrations are dispersed throughout. A total of linguistics is a mature science and, as in other scientific 288 scholars from 34 countries have contributed their fields, there can be a communications gap both within expertise to this encyclopedia. the field and, even more so, between the active researchers and the general public. Contents The Encyclopedia of Linguistics provides an accessible overview of and introduction to the multiple facets of the These essays have been written by scholars who provide a study of language. To bridge this gap between professional general introduction to the material without presupposing linguists and the general public, my editorial colleagues knowledge about the subject and without going into a xxix Introduction theoretical depth that would raise questions that cannot be how this particular situation came about historically, and answered in the space given. The entries fall into the fol- discuss language-political issues relevant in this region. lowing thematic categories: linguistic topics (for example, We have made a concerted effort to cover languages code switching) (50%), languages (for example, Sumerian) around the globe. Should you note any imbalance in favor (30%), persons (for example, Noam Chomsky) (15%), and of ‘western’ regions, languages, persons, or topics, this sim- regions of the world (for example, Algeria) (5%). ply reflects that the European languages are the best-stud- ied languages in the world. Linguistic Topics Acknowledgments Among these essays are general introductions to major fields of inquiry, such as semantics, historical linguistics, First of all, I thank Steve Larue and Paul Schellinger of and neurolinguistics. Other articles concentrate on issues Fitzroy Dearborn Publishers for giving me the opportunity within those fields, introducing concepts that are important to participate. And to their staff Christy Prahl, Heather in linguistics regardless of theoretical perspective, such as Sabel, and Peter Daniels, who did most of the hard work to ‘affix’, or ‘reference’, or which are so commonly used that start this project. At Routledge Reference, a team led by the concept becomes unquestioned, such as ‘phoneme’ or Marie-Claire Antoine—including Kate Aker, Susan Gamer, ‘deep structure’. Other articles describe crucial phenomena and Josh Pasternak—successfully completed this work. that any theory of language has to explain, such as ‘acquisi- Deciding which topics to cover was one of the hardest tion’, ‘tone’, or ‘aphasia’. aspects in shaping this reference work, and I am deeply indebted to our advisers who helped shape the book in mul- tiple rounds of suggestions and constructive criticism. Languages Robert Beard and John Goldsmith unfortunately resigned These essays cover the major language families of the from the Advisory Board when the book took a direction world and discuss how individual languages are related to they could not agree with. Still, many thanks to them for each other. Articles concentrating on specific languages their invaluable input while they were part of the project. from Ainu to Zuni explain where these languages are spo- Among our authors, I would like to specifically acknowl- ken, by whom, and under what sociopolitical circum- edge Alexandra Aikhenvald, Jim Hurford, Mikael Parkvall, stances. They provide a glimpse of the language’s structure and Ulrike Zeshan for their helpful suggestions. And many and highlight particularly interesting characteristics with thanks to Elly van Gelderen, Marc Greenberg, and Ray examples. Harlow, for helping to get the project off the ground with their sample articles. When I first invited potential advisers, a well-known linguist declined, saying that ‘linguists can’t Persons write for lay people’. I firmly believe that our authors These essays highlight the major theoretical contributions proved this statement wrong. Thus, greatest thanks is due of noted linguists, discuss the influences that led up to their our authors, who did an impressive job of introducing their work, and put the contributions into historical perspective, specialty without recourse to the precise terminology they in addition to providing basic biographical sketches. are accustomed to using. I thank both authors and advisers for staying the course during the years from inception to publication of the Encyclopedia of Linguistics. Regions Last but not least, I am deeply indebted to Karin, Isabel, These essays map out the linguistic landscape of a particu- Max, and Mia, who supported me while I disappeared into lar region. They show which languages are spoken, explain the basement for months on end to work on this book. Philipp Strazny xxx  A Acoustic Phonetics Acoustic phonetics is the study of the acoustic charac- Fourier’s theorem enables us to describe speech teristics of speech. Speech consists of variations in air sounds in terms of the frequency and amplitude of pressure that result from physical disturbances of air each of its constituent simple waves. Such a descrip- molecules caused by the flow of air out of the lungs. tion is known as the spectrum of a sound. A spectrum This airflow makes the air molecules alternately crowd is visually displayed as a plot of frequency vs. ampli- together and move apart (oscillate), creating increases tude, with frequency represented from low to high and decreases, respectively, in air pressure. The result- along the horizontal axis and amplitude from low to ing sound wave transmits these changes in pressure high along the vertical axis. from speaker to hearer. Sound waves can be described The usual energy source for speech is the airstream in terms of physical properties such as cycle, period, generated by the lungs. This steady flow of air is con- frequency, and amplitude. These concepts are most verted into brief puffs of air by the vibrating vocal easily illustrated when considering a simple wave cor- folds, two muscular folds housed in the larynx. The responding to a pure tone. A cycle is a sequence of one dominant way of conceptualizing the process of increase and one decrease in air pressure. A period is speech production is in terms of the source-filter theo- the amount of time (expressed in seconds or millisec- ry, according to which the acoustic characteristics of onds) that one cycle takes. Frequency is the number of speech can be understood as a result of a source com- cycles in one second, expressed in hertz (Hz). An ponent and a filter component. The source component increase in frequency usually results in an increase in is determined by the rate of vocal fold vibration, which perceived pitch. Amplitude refers to the magnitude of in turn is affected by a number of factors, including the vibrations, with larger vibrations resulting in greater rate of airflow and the mass and stiffness of the vocal peaks of pressure (greater amplitude), which usually folds. The rate of vocal fold vibration directly deter- result in an increase in perceived loudness. mines the F0 of the waveform. The mean F0 for adult Unlike pure tones, which rarely occur in the environ- women is approximately 220 Hz, and approximately ment, speech sounds are complex waves with combina- 130 Hz for adult men. In addition to their role as prop- tions of different frequencies and amplitudes. However, erties of individual speech sounds, F0 and amplitude as first stated by the French mathematician Fourier also signal emphasis, stress, and intonation. (1768–1830), any complex wave can be described as a For speech, the source component itself has a com- combination of simple waves. A complex wave has plex waveform, and its spectrum will typically show a regular rate of repetition, known as the fundamental the highest energy at the lowest frequencies and a frequency (F0). Changes in F0 give rise to differences number of higher frequency components that in perceived pitch, whereas changes in the number of systematically decrease in amplitude. This source constituent simple waves and their amplitude relations component is subsequently modified by the vocal tract result in perceived differences in timbre or quality. above the larynx, which acts as the filter. This filter 1 ACOUSTIC PHONETICS 3000 heed hayed 2500 had hid head 2000 Frequency (Hz) heard hod 1500 hud hood hawed who'd hoed 1000 hod hawed Figure 1. Frequencies of the first two for- head hud had mants of 12 vowels of American English, 500 hoed hid hayed heard hood who'd averaged across 48 adult female speakers. heed F1 is in black, F2 is in gray. (Source: Hillenbrand et al., J. Acoust. 0 Soc. Am., 1995) enhances energy in certain frequency regions and sup- case of consonants with a constriction toward the very presses energy in others, resulting in a spectrum with front of the vocal tract, the extremely short section in peaks and valleys, respectively. The peaks in the spec- front of the constriction does not result in clearly trum (local energy maxima) are known as formant fre- defined spectra. As a result, bilabial [b, p] and labio- quencies. The lowest-frequency peak is known as the dental [f, v] consonants are described as having diffuse first formant, or F1, the next lowest is F2, and so on. spectra, without any clear concentration of energy. The vocal tract filter is determined by the size and From a linguistic point of view, a detailed description shape of the vocal tract and is therefore directly affect- of speech sounds in terms of their frequency, in addition ed by the position and movement of the articulators to amplitude and duration, can elucidate the factors that such as the tongue, jaw, and lips. shape sound categories and determine phonological Vowels are typically characterized in terms of the processes both within and across languages. In addition, location of the first two formants, as illustrated in acoustic phonetic analysis may serve to quantify atypi- Figure 1 for the vowels of American English. For a cal speech patterns produced by nonnative speakers or given speaker, each vowel typically has a unique for- speakers with specific speech disorders. mant pattern. However, variation in vocal tract size among speakers often leads to a degree of formant overlap for different vowels. References Consonants can also be described in terms of their Asher, R., and Eugenie Henderson (eds.) 1981. Towards a his- spectral properties. These sounds are produced with a tory of phonetics. Edinburgh: Edinburgh University Press. complete or narrow constriction in the vocal tract, Fant, Gunnar. 1960. Acoustic theory of speech production. The essentially creating a vocal tract with two sections: Hague: Mouton. one behind and the other in front of the constriction. Flanagan, James L. 1965. Speech analysis synthesis and per- ception. Berlin: Springer-Verlag. The length of the section in front of the constriction is Hardcastle, William, and John Laver (eds.) 1997. The handbook one of the primary determinants of the spectra of these of phonetic sciences. Oxford and Cambridge, MA: sounds. The longer this section (i.e. the farther back Blackwell. the constriction), the lower the frequency at which a Helmholtz, Hermann von. 1877. Die Lehre von den concentration of energy occurs. For example, conso- Tonempfindungen als physiologische Grundlage fur die Theorie der Musik. Braunschweig: F. Vieweg; as On the nants like k and g, which are produced at the back of sensations of tone as a physiological basis for the theory of the mouth, are typically characterized by a concentra- music, translated by Alexander J. Ellis, London: Longmans, tion of energy between approximately 1,500 and 2,500 Green, 1885. Hz, whereas more anterior consonants like t and d typ- Hillenbrand, James, Laura A. Getty, Michael J. Clark, and ically have a concentration of energy above 3,000 Hz. Kimberlee Wheeler. 1995. Acoustic characteristics of American English vowels. Journal of the Acoustical Society Similarly, the sibilants [ʃ,] produced in the middle of of America 97(5). 3099–111. the mouth have major energy around 2,500 to 3,500 Jakobson, Roman, Gunnar Fant, and Morris Halle (eds.) 1952. Hz, whereas the more anterior ones [s, z] have major Preliminaries to speech analysis. Cambridge, MA: MIT energy well above 4,000 to 5,000 Hz. However, in the Press. 2  ACQUISITION Johnson, Keith. 1997. Acoustic and auditory phonetics. Oxford Lehiste, Ilse. 1970. Suprasegmentals. Cambridge, MA: MIT and Cambridge, MA: Blackwell. Press. Kent, Raymond D., Bishnu S. Atal, and Joanne L. Miller (eds.) Lieberman, Philip, and Sheila E. Blumstein. 1988. Speech 1991. Papers in Speech Communication: speech production. physiology, speech perception, and acoustic phonetics. Woodbury, New York: Acoustical Society of America. Cambridge: Cambridge University Press. Ladefoged, Peter. 1962. Elements of acoustic phonetics. Stevens, Kenneth N. 1998. Acoustic phonetics. Cambridge, Chicago and London: The University of Chicago Press, 2nd MA: MIT Press. edition, 1996. ALLARD JONGMAN Acquisition Language acquisition is the study of the development that children are born with the ability to learn a lan- of a person’s language. It generally refers to the way guage, i.e. they are born with a ‘language acquisi-tion people learn their native, first, second, or other lan- device’. The latter is species-specific or only for guages. More specifically, it may refer to the time a humans, language-specific or only for their first/ language feature has been acquired. This may vary native language, and innate or only inborn. He also from the first emergence or onset of a language item to claims that this ability is unconscious and children the time of its accurate use. As a field of study, it is the learn their native language by exposure to it and by subject of linguistics, psychology, and applied linguis- using it, and not by being taught or corrected. He tics. Its object is to study (1) how languages are argues that as children acquire their native language, learned, (2) what are the developmental stages in this they are able to produce sentences that they have not process, and (3) what is the nature of language. To find heard before. answers to these questions, researchers apply longitu- While early work on children’s language acquisi- dinal and cross-sectional methods. In the first of these, tion focused on the development of children’s ability they study specific developments in the language of to produce novel sentences, more recently, researchers individuals or groups over a period of time. In the sec- have emphasized children’s acquisition of word mean- ond, they research a particular feature in the language ings and their linguistic and cognitive development, of a group at a given point in time. their acquisition of the phonology of their native lan- guage, and their language development in relation to First Language Acquisition their interaction with parents and peers. Some First language acquisition is the child’s learning of his researchers also see a parallel between the stages of or her first or native language. Traditionally, and espe- children’s language development regardless of the cially in monolingual societies, ‘first’ and ‘native’ lan- specific language they are learning. guage were used synonymously. With the expansion of cross-cultural communication, the two terms become First Language Acquisition and Cognitive more distinct. For example, children may acquire Development some knowledge of another language from a nurse or A child’s language development is closely related to a relative before they acquire their native language, his or her cognitive development. Here, the ability to e.g. the language of the country they live in. Thus, a identify and form categories and concepts is of crucial Chinese child born in the United States may first learn importance. ‘Categorization’ involves the treatment of Chinese from her parents, and learn English later from distinct linguistic phenomena, such as ‘worked’, English-speaking children and adults. To avoid the ‘studied’, ‘saw’, and ‘went’, as if they were part of the confusion arising from the use of ‘first’ and ‘native’, same phenomenon, or the same grammatical category, another term, ‘primary’, is sometimes used to indicate i.e. past tense. Young children do not have fully devel- a child’s first language chronologically. oped abilities in categorization. Many childhood first language errors, for example, ‘*I eated it’, point to the First Language Acquisition and the Language gaps in their ability to form categories. Furthermore, Acquisition Device even seemingly correct utterances do not imply that Noam Chomsky’s work aroused interest in the the child has achieved an adult stage in the mastery of way children learn their native language. He believes the corresponding language category. 3 ACQUISITION Closely related to the ability to categorize is the First Language Acquisition and Bilingualism ability to differentiate a category, for example, tense, Cognitive and social development, as well as the lan- from the mental structure, which it represents, in this guage acquisition device and the critical period affect case, time. These mental structures are known as the language development of the bilingual as well as ‘concepts’. To learn a language, a child must acquire the monolingual child. the concepts that underlie linguistic structures. It is not Bilingual first language acquisition is defined as the possible to master grammatical categories, such as parallel acquisition of two languages, which is, sup- tense, in any language without mastering concepts posedly, an evenly paced process. However, such a such as time, space, modality, causality, and number. perfect balance can rarely be achieved. Commonly, the Young children’s errors in tense indicate that they do child would use one language in one environment, and not grasp the concept of time. another in a different setting. Thus, inevitably, one lan- guage gains dominance over the other. This domi- nance may extend to some or all areas of First Language Acquisition and Social Development communication. As a result, the child’s other language Children’s social adjustment is as important as their may become secondary in both development and use. cognitive growth to their language development. As Furthermore, there may be some interference from the they acquire various language categories and the con- dominant language that causes errors in the child’s cepts they represent, children also learn about the cul- secondary language. However, there is no evidence tural, moral, religious, and other conventions of the that this results in massive confusion in one or both society they live in. They learn how to express their languages. Furthermore, there is no evidence that thoughts, feelings, and wishes in a socially acceptable bilingual children differ from other children in their manner. For example, children learn that it is not cognitive, social, or language development. always advisable to speak their minds. As they come to realize that words can serve to make friends as well as enemies, they learn that it is not always possible to Second Language Acquisition tell the truth. In this way, while acquiring their first Second language acquisition (SLA) is defined as the language, children also develop a social identity. process of becoming competent or proficient in a sec- How-ever, their progress is slow and not devoid of ond or foreign language, from the first use of a lan- some rather amusing or even embarrassing errors. guage item to its advanced applications at a later stage. Children learning their first language, therefore, have As a field of research, SLA is a fairly new interdisci- a long way to go, even after they have acquired the plinary subject with most of its empirical research basic concepts and their corresponding language cat- done since the 1960s. It is largely based on theories egories. By comparison, adult foreign language learn- and research methods developed in the fields of edu- ers, who are knowledgeable about the sociocultural cation, psychology, linguistics, anthropology, foreign aspects of their native language, are a step ahead of languages, English as a second language, and linguis- child learners, even though they may also be prone to tics. In the United States, researchers study the way similar social blunders because of sociocultural dif- nonnative speakers acquire English phonology, ferences. syntax, and pragmatics. The purpose of SLA studies is to describe and explain the way second languages are First Language Acquisition and the Critical Period learned in terms of both linguistic and communicative The ‘critical period hypothesis’ claims that there is a competence. To do this, researchers study learners’ period in child development during which language can performance and their intuitions about correct and be acquired with native-like proficiency. Some, like the incorrect use of language. The object of second lan- biologist Lenneberg, believe that this period lasts until guage acquisition is to find more effective ways of puberty, after which the brain loses some of its ability to teaching and learning foreign languages, and assumes adapt due to its laterization, i.e. the establishment of that such research can affect the way foreign lan- specific language functions in particular parts of the guages are learned. brain. After that, the decreased plasticity of the brain makes the acquisition of another language a psycholog- The Meaning of ‘Second Language’ in Second ically different and more difficult process. While there Language Acquisition is compelling evidence that supports those claims, there There are different interpretations and uses of both ‘sec- are also important facts that undermine their veracity. ond’ and ‘acquisition’ in SLA. ‘Second’ may be used to First, the strictly biological evidence is by no means distinguish it from ‘foreign’ or ‘third’ language acquisi- conclusive. Second, other factors, such as lack of tion. Traditionally, the terms ‘second’ and ‘foreign’ have motivation, may explain nonnative pronunciation. been used alternatively to refer to any language other 4  ACQUISITION than the first. More recently, with the emergence of Other researchers analyze what it means to know a English as a global language and the establishment of second language. From a linguistic perspective, Teachers of English to Speakers of other Languages Chomsky focuses on the results of SLA, which he (TESOL) as a worldwide professional organization, a defines as ‘competence’ and ‘performance’. According distinction is made between the two. ‘Second’ language to him, ‘competence’ in a second language is the mas- acquisition refers to the study of English by foreigners tery of the internalized grammar that the ideal speaker in countries where English is the native or the official or hearer, not a real one, has of the whole language. language, whereas ‘foreign’ language acquisition refers Such mastery enables him/her to produce grammati- to the study of English everywhere else. Furthermore, cally correct sentences as well as recognize existing this distinction extends to differences in what is learned and nonexisting sentences. For example, knowing the and how it is learned. Learners of English as a foreign rule, which makes ‘I speak English’ possible, a person language (EFL) prefer standard varieties of English, can produce ‘I speak French’ even though he or she whereas learners of English as a second language (ESL) may not have seen this before. Furthermore, he or she try to blend with their sociolinguistic environment. All would know that the form ‘*I speaking English’ is of these differences are reflected in the goals and meth- nonexistent. ‘Performance’ in Chomsky’s Generative ods of EFL and ESL. A further distinction is made Transformational Grammar, on the other hand, refers between ‘second’ and ‘third’ language acquisition, to a person’s actual use of a language in the under- which marks the learner’s relative proficiency rather standing and production of sentences. Unlike commu- than the order in which he or she acquired these lan- nicative competence, which is internal and invisible, guages. Sometimes, the term ‘alternative’ is used to performance is external and observable. Furthermore, refer to any nonnative language. performance does not mirror competence, since peo- ple may know how to produce a sentence but may err The Meaning of ‘Acquisition’ in Second Language when they try to do so. Thus, performance can also be Acquisition defined as the grammar that a person uses to under- Acquisition is often used to refer to different aspects of stand and produce language, which is both correct and the process and results of learning a second language. appropriate. Performance could be used to investigate While trying to find out about the process, i.e. how sec- competence through the analyses of samples of spoken ond languages are learned, researchers often compare or written discourse. Within performance, Widdowson different learning experiences that lead to SLA, such as distinguishes ‘usage’, which refers to the learner’s learning a language through organized instruction or in ability to apply grammar rules accurately in the pro- an immersion situation. From a sociolinguistic per- duction of grammatically correct sentences. ‘Use’, on spective, acquisition through organized instruction the other hand, signifies their ability to apply linguis- occurs in classrooms with the help of teachers and tic and sociolinguistic knowledge appropriately and instructional materials. Acquisition through immersion communicate effectively in diverse contexts. occurs in social situations using contextual clues. Yet another distinction is made from the psycholinguistic Second Language Acquisition Research perspective. Klein identifies ‘spontaneous’ and ‘guid- A large part of SLA research is learner based. It ed’ acquisition. The first focuses on everyday commu- describes and analyzes the nature of learner language nication, whereas the second targets the mastery of the and learner differences, learning processes, and language system. Similarly, from a psychological point pedagogical input and output. It does so to provide of view, Krashen distinguishes between ‘acquisition’ answers to important second language research ques- and ‘learning’. In his analysis of the process of master- tions, which may offer effective solutions to crucial ing a second language, he reserves ‘acquisition’ for the second language classroom problems. subconscious process of learning a language by being exposed to it. ‘Learning’, according to him, is the con- Learner Language in Second Language Acquisition scious process of mastering a language by studying it. Research Ellis finds this distinction problematic and considers its Researchers study learner language by examining demonstration difficult. Furthermore, he states that samples of oral and written texts. Their goal is to iden- researchers disagree about what kinds of performance tify errors, establish developmental patterns and constitute evidence of ‘acquisition’. For some, such sequences, trace variability, and explore use. Errors evidence can be found in the ways learners speak and were first believed to be the result of native language write. For others, it is the learner’s intuitions about the ‘transfer’ or ‘interference’. This view was promoted by second language that matter. Yet another group of numerous contrastive analyses conducted from the scholars seek evidence of acquisition by assessing the 1940s to the 1960s. Such studies compared two lan- learner’s introspections. guages to find out what similarities and differences 5 ACQUISITION existed between the two. Lado thought that similar ele- affect the way learners acquire a second language. ments would be easy to learn, while dissimilar ones Ellis contends that social factors shape learners’ atti- would be difficult to master. The belief that linguistic tudes, which, in turn, may affect motivation and learn- difference could be a predictor of difficulty gave rise ing outcomes. Social factors include natural and to the ‘contrastive analysis hypothesis’. That and the educational settings. For example, natural settings, behaviorist approach to learning, which claimed that where English is the native or the official language, learning is a process of habit formation, led to the offer opportunities different from educational settings, belief that SLA should be a process of overcoming such as the foreign language classroom where the habits from the native language and consolidating cor- native language is the medium of instruction. rect habits in the target language. While social factors influence second language In 1967, Pitt Corder proposed a new definition of acquisition indirectly, input, output, and interaction errors. He thought they were systematic deviations seem to have a direct impact. ‘Input’ is the learner’s from the norm, which reflect the learner’s current access and exposure to the second language, both writ- stage of second language development. Errors, he ten and oral. Exposure to the foreign language may claimed, are different from mistakes, which can easily engage learners in ‘interactions’ with native or nonna- be self-corrected. ‘Error analysis’ treated errors as a tive speakers, or it may involve them in listening to sign of the learner’s hypothesis testing, which would tapes, films, radio, and TV programs. Researchers ultimately lead to the formation of the correct form vary in their assessment of the importance of input and and its underlying rule. Thus, errors were seen as part interaction. Behaviorist theories emphasize the impor- of the learner’s language at every stage of its develop- tance of input. Chomsky, on the other hand, claims ment. To emphasize its unique features, Corder that there is no necessary correlation between lan- referred to learner language as ‘idiosyncratic dialect’. guage input and learner output. Krashen believes that Nemser called it an ‘approximative system’ and learners acquire language in a natural order as a result Selinker coined the term ‘interlanguage’. Thus, the of being exposed to ‘comprehensible input’ addressed notion of learner language evolved from a faulty, to them. In contrast to Krashen, Swain proposes the deviant product of the target language to a continuous, ‘comprehensible output hypothesis’, which claims that approximative progress towards its mastery. Both con- comprehending input alone will not prepare students trastive and error analysis were criticized for their to produce language. According to him, it is correct exclusive reliance on the analysis of a linguistic prod- production resulting from challenging practice in uct, i.e. errors, to yield insights into a psycholinguistic speaking and writing that facilitates acquisition. Both process, i.e. second language acquisition. the comprehensible input and comprehensible output Another feature of learner language is its passage hypotheses have been criticized on the grounds that through a sequence of developmental stages, which the processes of comprehensible input and output and are universal. Thus, many of the initial utterances that the process of SLA are not the same. learners produce may be simple formulae, for General factors, such as social setting, input, out- example, ‘What’s this’. These are followed by struc- put, and interaction, result in a variety of individual tures of greater morphological and syntactic com- differences in SLA. Furthermore, individual factors, plexity, for example, ‘I wonder what this might be’. such as age, language aptitude, motivation, cognitive The existence of developmental stages in SLA, which style, and learning strategies can have similar effects. are similar to those in first language acquisition, These factors affect second language learning in ways along with some variations in the specific order in that are mostly independent of the learner. For exam- which particular features occur, have renewed interest ple, a learner can do nothing about his or her age, or in grammar instruction. In its systematic develop- language aptitude. Few learners may have the oppor- ment, learner language also exhibits certain variabili- tunity to switch from one educational setting to anoth- ty. For example, learners may use the third person er. Given the appropriate guidance, however, some singular ‘s’ correctly sometimes and omit it at other learners may be able to improve their motivation and times. In addition to lexical and syntactic variability, learning strategies over time. For example, ‘extrinsic they often have problems on the pragmatic level, i.e. motivation’, which derives from external rewards, may they may use language or act in a socially inappro- evolve into ‘intrinsic motivation’, which derives from priate manner. personal interests. Learner strategies, which contribute to the learner’s conscious efforts to learn, may also Factors in Second Language Acquisition change. For example, learners may expand their SLA is also influenced by the environment in which it ‘cognitive strategies’ by learning new concepts. They occurs. Social factors, language input, and interaction may also perfect their ‘metacognitive strategies’ by 6  ACQUISITION THEORIES developing their study skills, or enhance their ‘social Hatch, E. (ed.) 1978. Second language acquisition. Rowley: strategies’ by practicing their knowledge in authentic Newbury House. Klein, Wolfgang. 1986. Second language acquisition. Avon: social settings. Cambridge University Press. Krashen, Steven. 1981. Second language acquisition and sec- References ond language learning. Oxford: Pergamon. Larsen-Freeman, Diane, and Michael Long. 1991. An introduc- Bennet-Castor, T. 1988. Analyzing children’s language. tion to second language acquisition research. London and Oxford: Blackwell. NewYork: Longman. Brown, Douglas H., and Susan Gonzo. 1995. Readings on sec- Lenneberg, E. 1967. Biological foundations of language. New ond language acquisition. Englewood Cliffs, NJ: Prentice- York: Wiley. Hall Regents. Long, Michael. 1995. The least a second language theory needs Chomsky, Noam. 1965. Aspects of the theory of syntax. to explain. Readings on second language acquisition, ed. by Cambridge, MA: MIT Press. Douglas H. Brown and Susan Gonzo. Englewood Cliffs, NJ: Cook, Vivian. 1993. Linguistics and second language acquisi- Prentice-Hall Regents. tion. New York: St. Martin’s Press. Richards, Jack, et al. 1992. Longman dictionary of applied lin- Ellis, Rod. 1994. The study of second language acquisition. guistics. Harlow, Essex: Longman. Oxford: Oxford University Press. Widdowson, Henry. 1978. Teaching language as communica- Fletcher, P. 1985. A child’s learning of English. Oxford: tion. Oxford: Oxford University Press. Blackwell. Fletcher, P., and M. Garman (eds.) 1979. Language acquisition. LILIA SAVOVA Cambridge: Cambridge University Press. See also Acquisition Theories Acquisition Theories The goal of acquisition theories is to explain how it is tion) from language-using older members of our that any normal child, born into any linguistic com- species (i.e. they are severely neglected, or are raised munity, learns the language (or languages) of that by other animals) that language does not emerge with- community. For many theorists, the challenge is also out at least some linguistic exposure—input—during to explain what appears to be the relatively short time the first few years of life. This strongly suggests that period in which acquisition is achieved, the fact that it there is a critical or sensitive period during which the appears to be done without either overt teaching or mechanism or mechanisms responsible for language sufficient information from the input (what the child development is/are primed to receive input. However, hears [or sees, in the case of sign language]), and to the resilience of language development to quite wide follow a path that seems remarkably similar in all chil- variations in input within any given language commu- dren, despite variation in early childhood experiences nity, as well as the similarities among children learn- and in the types of languages they are exposed to. ing quite different languages, suggest that these There is also a consensus that language acquisition is mechanisms, whatever they are, must either be quite largely independent of cognitive development, despite tolerant of such variation, or be primed in such a way the fact that some deficits in cognitive development that the crucial input for language acquisition is can have an effect on certain aspects of language always made available. development. Whether the language is a spoken lan- There are important roles for both ‘experience- guage or a sign language, whether the language is dependent’ (nurture) and ‘experience-expectant’ highly inflected like Finnish or uninflected like (nature) learning in language development, and theo- Mandarin, whether the child is raised in poverty or ries are distinguishable in terms of the relative contri- luxury, by highly educated or illiterate adults, or even butions they see for these two types. On the one hand, other children, it seems that normally developing there are those researchers who see a large didactic role children pass through roughly the same stages in the for input (experience-dependent learning), and on the same sequence, and achieve the steady state of other, there are those who see a much smaller role for acquired language by about the same age. input and a much larger role for genetic predispositions We know from unfortunate natural experiments in that are triggered by linguistic experience (experience- which children are raised in isolation (or near isola- expectant learning). Theories are also distinguishable 7 ACQUISITION THEORIES in terms of whether they are trying to explain how lan- In the course of acquisition, generalization (and guage emerges on a day-by-day basis in any given overgeneralization) allows new knowledge to permeate child or whether they are trying to explain how what is across-the-board, and ‘bootstrapping’ allows learning in perceived as a gulf between experience as a child and one area of language to trigger new learning in another. adult knowledge of language could be bridged in prin- Semantic bootstrapping involves understanding an ciple. The latter are engaged in trying to solve the ‘log- utterance in context and using it in conjunction with ical problem of language acquisition’. innate expectations of language principles to crack the Another dimension of difference between theories code of the syntax. (For example, a child who does not concerns the nature of the experience-expectant yet know the required complements of the verb ‘put’, (innate) aspects of language acquisition. There are can work these out from understanding utterances such those researchers, most notably the generative lin- as ‘Put the cup on the table’ in context.) Syntactic boot- guists in a broadly Chomskyan paradigm, who argue strapping involves working from an already understood for a dedicated language acquisition device (LAD), structure to fill in meanings, semantic information, by that has evolved to serve the precise purpose of lan- deduction. (For example, if you hear ‘John glopped his guage acquisition. This device is primed specifically friend on the head and he fell down’, you may not know to receive linguistic input, and requires a minimal exactly what ‘glopped’ means, but you can work out a amount of it to set to work building mental represen- lot of what it must mean.) tations for language in the mind of the child. The fact The Principles and Parameters (P&P) model has that the required triggering input is so minimal pro- been a highly influential linguistic approach in lan- vides an explanation for the consistency of language guage acquisition research, even while researchers in acquisition paths across otherwise fairly widely psychology and anthropology have been pursuing sig- varying life experiences. nificantly different lines of investigation. In linguis- The most elaborated version of the LAD account— tics, the P&P model lies at the intersection of the Principles and Parameters approach—suggests generative (specifically Chomksyan) linguistic that children are born with a Universal Grammar theorizing about the nature of adult mental representa- (UG), which means they are (unconsciously) antici- tions for language and accounts of how children pating those features that are common to all lan- acquire language. It has evolved as an account of how guages (the principles), as well as limited options for language could actually develop across time, even those things that differ among languages (the param- while its roots are in the ‘logical problem of language eters). Upon exposure to actual input from a given acquisition’ because it assumes that what cannot be language, children are able to ‘decide’ which sort of learned from the input must be genetically prespeci- language they have encountered. So, for example, fied. It advances various arguments in support of the some languages have basic subject–verb–object thesis that the input is in fact incapable of providing organization in which complements are attached to sufficient information for language to be learned the right of the heads of phrases (thus objects follow entirely through experience-dependent mechanisms, verbs, relative clauses follow noun heads, and noun and thus that there is an essential problem of the phrases follow prepositions), while other languages ‘poverty of the stimulus’. The account is also strongly are subject–object–verb where the reverse order of modular in the sense that it sees language as being complements is found. A child exposed to a language acquired by a specially dedicated mechanism (the of the first type need only process a simple structure LAD), independent of other types of mental represen- (say one with a verb followed by an object) and it will tations or mental functioning. trigger the expectation that all the other head-comple- Despite its power within linguistics, accounts pred- ment structures will be in the same order. When all icated on some version of the UG story actually attract the open parameters have been set, the child possess- only a minority of adherents within the broad field of es the ‘core’ grammar of the specific language he or child language research. Most researchers are con- she is exposed to. At the same time, however, the vinced that language in its entirety can be worked out child has also been acquiring those aspects of the lan- by the child on the basis of the input, coupled with guage that are not anticipated by UG, using experi- innate (nonspecific) predispositions for analyzing their ence-dependent learning. These aspects are often environment. As a result of their analysis, children together referred to as ‘peripheral’ grammar. Some possess the capacity to produce and comprehend lan- researchers in this paradigm have assumed that guage. They may also have stored mental representa- all principles and parameters are operational or sensi- tions for how language works, but, unlike the P&P tive to the input from the beginning of life. Others account, most of the accounts that rely on processing have suggested that some may at least emerge with of the input as the way in which acquisition occurs maturation. also regard the endpoint as processes for production 8  ACQUISITION THEORIES and comprehension, without independent mental rep- for example, argues that the behavior of a child’s care- resentations. Ann Peters and Dan Slobin, for example, taker provides all the cues that the inherently social have argued that children possess strategies, operating child needs to acquire language. The way in which principles, for carving up the speech stream into language is used in the here and now, in conjunction repeating bits, storing those bits, recognizing the with actions that match what is being said, ‘scaffolds’ commonalities across them, and thereby building up a the child’s understanding of the language. Others see performance grammar for comprehension and produc- difficulties with this approach, not only because not tion called Basic Child Grammar, which subsequently all successful language learners receive the kind of becomes elaborated into an adult grammar. careful scaffolding it seems to require but also Another of the processing theories of language because the parental complexity of language seems to acquisition is Elizabeth Bates’ and Brian Mac follow rather than lead increasing complexity in the Whinney’s Competition Model. Unlike the operating child’s language. In these and other ways, the account principles approach, this does not argue for a steady seems too simple to account for the complexity of the state in the form of a grammar at the end of acquisi- task. A similar approach was advocated by Jean tion, but rather for a permanently dynamic response to Piaget, who saw language development as the logical input throughout life. The impression that language extension not of social behavior, but of cognitive has been definitively learned comes simply from the development. In Piaget’s initial proposal, embedding fact that new input changes the child’s system very lit- of sentences was seen as analogous to nesting boxes, tle if a child remains in the same speech community and the former dependent on the latter. Although most (although it will usually change with exposure to a of his specific predictions have not been supported by new dialect or to a new language). Acquisition takes subsequent research, it is clear that at least certain place as children respond to the distribution of various aspects of language are intimately related to cognitive cues to meaning (word order, inflectional informa- development, although the direction of influence is tion, intonation and stress, etc.) and respond proba- not clear. One area of current research concerns the balistically to conflict among them. Since each child emergence of the capacity to make informed guesses processes the input independently, individual differ- about what other people know (‘theory of mind’), and ences between children are expected, and advocates its relationship to language development. There is of this approach argue that the differences found considerable discussion about whether the capacity to among children support the model. As they learn embed clauses under main verbs such as ‘think’ and more of their language, they pay attention to more and ‘know’ drives or is driven by the capacity to under- more of the cues and let the stronger ones win out stand that others may not share the same assumptions over the weaker ones for their language. So, at an as oneself. early stage, a child learning English may assume that Finally, it is worth noting that significant debate nouns at the beginning of sentences are agents, but surrounds the issue of whether, when an individual when they begin to pay attention to passive morphol- learns a second language, they go about it in the same ogy, they will have to adjust their assumptions accord- way as a first language learner, or whether it is a fun- ingly. As should be clear from this example, the damentally different process. Evidence suggests that ability to derive meanings for language from context, there is some kind of critical period for second lan- in advance of actually understanding how language is guage development, as children seem to be able to do structured is crucial (as it is in semantic bootstrap- it so much better than adolescents and adults. ping). Active application of this and other distribu- However, it is not yet known when bilingual first lan- tional models, aimed at demonstrating that the input guage development should be seen as having given is sufficient to account for language acquisition, is way to second language development. Nor is it clear seen in the computer-modeled connectionist net- that adults, when given the kind of input and motiva- works. Attracted by the architectural similarity tion of children, are always incapable of the same level between computer networks and neurons, these of success. It is also unclear whether second language researchers argue that experience with language learners are able to reaccess the learning capacities teaches the network so effectively and so quickly that they had as children learning their first language, or it gives the impression of prior knowledge. whether these are permanently overridden and made As already indicated, children frequently seem to unavailable by the presence of the first language. rely on pragmatic expectations of what language ought to mean in a given situation as a basis for learn- References ing how it is structured. Some theories of language Crain, Stephen, and Diane Lillo-Martin. 1999. An introduction acquisition place an even greater emphasis on the role to linguistic theory and language acquisition. Oxford: of the sociopragmatic environment. Jerome Bruner, Blackwell. 9 ACQUISITION THEORIES de Boysson-Bardies, Benedicte. 1999. How language comes to Owens, Robert. 1996. Language development: an introduction, children. Cambridge, MA: MIT Press. 4th edition. Boston: Allyn and Bacon. Fletcher, Paul, and Brian MacWhinney (eds.) 1997. The hand- Pinker, Steven. 1994. The language instinct: how the mind cre- book of child language. Oxford: Blackwell. ates language. New York: William Morrow and Company Inc. Foster-Cohen, Susan. 1999. An introduction to child language Ritchie, William and Tej Bhatia (eds.) 1999. Handbook of child development. Harlow, Essex: Longman/Pearson Education. language acquisition. New York: Academic Press. Hoff-Ginsberg, Erika. 2000. Language Development, 2nd edi- Smith, Neil. 1999. Chomksy: ideas and ideals. Cambridge: tion. Pacific Grove, CA: Brooks/Cole Publishing Company. Cambridge University Press. Jackendoff, Ray. 1995. Patterns in the mind: language and SUSAN H. FOSTER-COHEN human nature. New York: Basic Books. Karmiloff, Kyra and Annette Karmiloff-Smith. 2001. Pathways to language: from fetus to adolescent. Cambridge, MA: See also Acquisition; Developmental Stages; Slobin, Harvard University Press. Dan Isaac Aerodynamics of the Vocal Tract Speech production may be viewed as a means of con- and inversely with the size of the channel it is verting slow pressure variations in the vocal tract into flowing through. (This principle is important the rapid pressure variations that constitute sound. because turbulence, the source of acoustic noise, is When there is a difference in air pressure across a a function of air velocity.) valve and it opens, sound is created by the rapid move- 5. When air is flowing through a channel, the pres- ment of air. A stop burst is created by the sudden sure it exerts on the walls of a channel at 90° release of air under pressure, the continuous noise of a with respect to the direction of airflow varies fricative is created by a similar but slower release, and inversely with the velocity of the airflow (this is repetitive release of air through the glottis by the the Bernoulli effect). vibrating vocal cords creates voicing. Speech aerody- namics studies in detail as to how these sounds are cre- Four ‘air stream mechanisms’ for generating the ated, how the air pressures and airflows are generated, pressures needed in speech are used in languages of and their phonological consequences. the world. They are named for the principal piston-like From an aerodynamic point of view, the vocal tract structure creating the volume change, plus ‘egressive’ is a system of chambers connected to each other by or ‘ingressive’ to denote whether air moves out of the valves, with some of them having access to the air cavity or into it when the pressure is released to the outside. Most of the chambers are equipped with atmosphere: piston-like structures that can change the volume of Pulmonic egressives: The air in the lungs is com- their chamber. pressed by the expiratory muscles acting to decrease There are a few general physical principles that the volume of the rib cage; the pressure increases (by govern the generation of pressures and flows in the principle 1). Sound is created when this air is released vocal tract: at the glottis and/or at any supraglottal constriction, e.g. [p], [l], [m], [o]. All languages use pulmonic egressive 1. For a given mass of air, pressure varies inverse- sounds and the vast majority like English, Japanese, ly with the volume of the chamber containing it Finnish, and Hawai’ian use them exclusively. (Boyle–Mariotte’s Law). Glottalic egressive (‘ejectives’): With air trapped in 2. For a given volume of a chamber, pressure varies the oral cavity between a closed glottis and a supra- directly with the mass of air in it. glottal constriction (and, of course, velic closure), the 3. The quantity of airflow (the ‘volume velocity’) air is compressed by the elevation of the larynx (the through an aperture varies as a function of the glottis acting like a piston) and some pharyngeal con- size of the aperture and the pressure difference striction. Such sounds may include stops, affricates, across it. and fricatives, e.g. [p’], [ts’], [f’]. Ejectives are found 4. The velocity of airflow (the so-called ‘particle in approximately 18% of the world’s languages, e.g. velocity’) varies directly with the volume velocity Amharic, Quechua, and Navaho. 10  AFFIXATION Glottalic ingressive (‘implosives’): The oral cavity requires that there be air flowing between the vocal is enlarged (by lowering the larynx and possibly also cords. Obstruents partially or completely block the the jaw) so that the pressure in the oral cavity decreas- airflow exiting the oral cavity. But if glottal airflow es. Most implosives are voiced, the lowering larynx continues, then the pressure in the oral cavity will thus acting like a leaky piston, but voiceless implo- rise (by principle 2, above). Then, the difference in sives, although quite rare, do exist. Voiced implosives the pressure across the glottis will be reduced and occur in about 10% of the world’s languages, e.g. may fall below the value needed to maintain the air- Kalabari, Sindhi, and Xhosa. flow required for vocal cord vibration (by principles Velaric ingressive (‘clicks’): Air is trapped between 3 and 5, above). Passive yielding of the vocal tract the tongue and the palate (or the lips) and the small walls to the impinging pressure permits voicing to cavity is enlarged by the tongue moving down and/or persist for around 65 msec (more for labial stops, backwards. The air pressure is greatly lowered and which have more surface area behind the constric- when released by the tongue or lips, producing a ‘pop’ tion and less for velars). For voicing to be main- or fricative sound. The sound that comic strips try to tained beyond this, some active expansion of the oral convey with ‘tsk’! or ‘tut-tut’ (a sound expressing dis- cavity is required, e.g. by lowering the larynx and/or approval or regret) is the alveolar click [!]. The sound the jaw. This leads to many phonological conse- of a ‘kiss’ is the bilabial click [>]. Clicks are found quences, among them the greater incidence of ‘miss- widely in different cultures, but usually as interjec- ing [•]’ in the stop inventories of languages utilizing tions or calls to animals. As speech sounds, they are a voicing contrast on stops, e.g. as in Dutch, Czech, quite rare and are found only in certain Southern and Thai, the greater tendency for geminate stops to African languages such as Zulu, Xhosa, Ndebele, and be voiceless or to undergo devoicing morphophone- !Xóõ. Given that clicks’ sound-making structures are mically. so localized in one part of the vocal tract, clicks can be Many other common sound patterns in language, combined simultaneously with sounds made by the e.g. the development of affricates from stops, as well other air stream mechanisms, e.g. they can be accom- as the way specific speech sounds are modified in cer- panied by voicing or voicelessness, nasalization, etc. tain cases of speech disorders, e.g. cleft palate, laryn- Voicing and trills: Trills and voicing constitute aero- gectomy, can be explained by reference to speech dynamically driven, self-sustaining oscillations. The aerodynamics. cycle of vocal cord vibration could start with the vocal cords lightly adducted in the presence of airflow. This References causes the subglottal pressure to build up behind the glottal closure, the increased pressure forces the cords Ladefoged, P., and I. Maddieson. 1996. The sounds of the world’s languages. Oxford: Blackwell. apart, and the resulting high rate of airflow through the Ohala, J.J. 1983. The origin of sound patterns in vocal tract con- cords creates a negative pressure between them (by straints. The production of speech, ed. by P.F. MacNeilage, principle 5, above), which ‘sucks’ the cords together 189–216. New York: Springer-Verlag. again (along with an elastic recoil force of the vocal Ohala, J. J. 1997. Aerodynamics of phonology. Proceedings of cord tissue); the cords close again, and the cycle the 4th Seoul International Conference on Linguistics [SICOL], 92–7. 11–15 August 1997. repeats itself. Trills are created in a similar way. JOHN J. OHALA Obstruent and voicing: Obstruents inherently favor voicelessness for the following reason: voicing See also Speech Production Affixation Affixation is a morphological process that adds phono- (1) a. fond fondness logical material to a word in order to change its b. start restart meaning, syntactic properties, or both. Some examples of affixation in English are given in (1). c. car cars 11 AFFIXATION In this article, an overview of different types of ment is not sufficient evidence for the existence of a affixation processes is given, followed by a discussion circumfix in English, because the form can be analyzed of how to distinguish affixation from other processes. transparently as a result of suffixation of -ment to the Then, a number of theoretical problems related to verbal base enrich, which is in turn the result of prefix- affixation are presented. ation of en- to the adjective rich. A better candidate for a circumfix is Dutch ge…te, exemplified in (3). Types of Affixation (3) steen (‘stone, rock’) gesteente (‘type of The phonological material added in affixation is called stone/rock’) the affix. The affix is attached to the base. Different types of affixation can be distinguished on the basis of The change in meaning in (3) can be expressed as a the position of the affix with respect to the base or on generalization to a type. It can be argued that ge…te is the basis of how affixation affects the meaning of the a circumfix, because neither *gesteen nor *steente are base. In this section, the emphasis is on the former. The possible Dutch words. A theory that excludes circum- latter is discussed in the section on delimitation issues. fixes would have to postulate one of these nonexisting Suffixation is the most common type of affixation. forms as an intermediate form in the derivation. In suffixation, the affix is added to the end of the base. Examples are (1a) and (1c) above. In (1a), the Delimitation Issues suffix-ness is added to the adjective fond to produce the noun fondness. In (1c), the suffix -s is added to As affixation is the most salient morphological the noun car to produce the plural of the noun. In process, it is no surprise that many of the foundation- most languages, suffixation is the most widespread al discussions in morphology are directly related to form of affixation. In languages such as Turkish and affixation. Finnish, it is the only type of affixation. A first set of issues concerns the function of In many respects, prefixation is the mirror image of affixation. Traditionally, inflection and derivation are suffixation. In prefixation, the affix is attached to the distinguished. Inflection adapts a word to its syntactic beginning of the base. An example is (1b) above, where environment. A prototypical example is the agreement the prefix re- is added to the base start. Prefixation is of a verb with its subject in person and number, as less widespread than suffixation, but some languages, illustrated for sleep in (4). e.g. Khmer, only have prefixation. In languages that have both suffixation and prefixation, the former usual- (4) a. The child sleeps. ly has a larger range of functions. Thus, in English, all b. The children sleep. inflection is expressed by suffixation. In word forma- tion, we find both, but prefixation only rarely changes Derivation is a type of word formation that involves the syntactic category of the base, as when the verb affixation. It forms new lexemes, marked by a differ- enrich is formed from the adjective rich. ent syntactic category and/or a different lexical mean- In most languages, affixation only involves suffixes ing, as illustrated in (1a-b). Another contrast is the one and prefixes. Infixation, in which the affix is attached between inflection and stem formation. An example of inside the base, is much rarer. An example is the infix the latter can be observed in Dutch kinderstoel (‘high- -um- in Tagalog, illustrated in (2). chair’), a compound of kind (‘child’) and stoel (‘chair’). The form kinder is different from the singu- (2) a. bilih (‘buy’) bumilih (‘bought’) lar ( kind) and the plural ( kinderen). The suffix -er can b. gradwet (‘graduate’) grumadwet be said to produce a secondary stem used in com- (‘graduated’) pounds. Another set of delimitation issues concerns the The infix in (2) is used to express the past tense of nature of the affix as a phonologically dependent, mor- the verb. It attaches itself after the first consonant or phological unit. Affixes contrast on the one hand with cluster of consonants of the base. This is its anchor clitics, and on the other with stems used in com- point. In general, the anchor point of an infix is at most pounding. Clitics are phonologically dependent parti- one syllable removed from the left or right boundary cles, i.e. particles that must attach to a host word much of the base. like an affix, whose distribution is, however, deter- Whether other types of affixation exist depends on mined by syntactic rules rather than morphological the theoretical framework adopted. Thus, circumfixa- rules. An example is the Italian clitic pronoun -ti tion is the simultaneous addition of a prefix and a (‘you’) in incontrarti (‘to meet you’). While these cli- suffix. The mere observation of a form such as enrich- tics are generally not considered as affixes, Turkish 12  AFFIXATION case and number endings as in (5) are more of a bor- made for cases where phonological processes interact derline case. with the concatenation, as in (6–8). (5) Ankara ve I˙zmire gideceg˘im ‘Ankara and Izmir I go’, (6) a. active activity i.e. I go to Ankara and Izmir b. intend intention (7) a. a house to house asylum seekers In (5), Ankara does not have any case ending, b. to read a good read whereas I˙zmir is followed by the dative ending -e. (8) a. live life Nevertheless, the dative ending applies to the coordi- b. extend extent nated noun phrase Ankara ve I˙zmir. Therefore, it can be argued that the dative marker is not an inflectional affix In (6), suffixation triggers a phonological change of but a clitic. Similar observations can be made about the the base, stress shift in (6a), and a change of the final genitive marker in the queen of England’s hat. consonant in (6b). In an Item & Arrangement account, Compounding usually combines items that occur as one might say that these phonological changes are trig- independent words, e.g. bookshelf. Words such as phil- gered by the affix. In (7), there is no affix, but the rela- anthropic and anthropomorphic share many properties tionship between the nouns and verbs is very similar to with compounds, but their constituent parts do not what we find in cases of genuine affixation such as occur as independent words (in English). They are encase and entertainment. In an Item & Arrangement often called neoclassical compounds. In fact, anthropo account, one might either say that (7) is not affixation is the Ancient Greek word for human being, a sense it but something else (e.g. conversion), or that (7) also has in English words. Although it is not a word in involves zero affixes, i.e. affixes that do not have a English, it does not behave like an affix either. It has a phonological realization. The examples in (8) show stem-like meaning and can appear in different positions that phonological changes can occur without an affix. in a compound. Other elements of the same general This can be interpreted as phonological rules triggered type, e.g. pseudo, occur almost exclusively as a prefix, by a zero affix or by conversion. The general approach often with a reduced meaning more typical of an affix in Item & Arrangement is that concatenation of base than of a stem, e.g. pseudocultured. It is very difficult and affix is taken as central. Cases such as (6–8) are to draw a clear borderline between stems and affixes treated as exceptions to be fitted in. among such neoclassical elements. The term affixoid is In an Item & Process approach, as argued for by sometimes used to refer to items for which it is difficult Stephen Anderson (1992), the process of affixation takes to determine whether they are stems or affixes. An priority over the affix. In (7), the word formation process example from outside neoclassical word formation is changes the syntactic properties of the input without German reich (‘rich’) in ertragreich (‘productive, prof- affecting its form. In (8), the process modifies the form itable’). In English, -ful in successful is similar. of the input by changing the quality of the final conso- nant and, in (8a), of the vowel. In (6), the process affects the form of the input by changing the stress position or Affix or Process? the voicedness of the final consonant and appending One of the central questions in morphology is whether some further phonological material, the suffix. Affixation affixation should be seen as a rule applying to two ele- of the type illustrated by fondness, the prototypical case ments, a base and an affix, or as a process applying to in Item & Arrangement, is a special case in Item & a base. Charles Hockett (1954) calls the first position Process where there is a perfect match between the input Item & Arrangement, and the second Item & Process. and a part of the output of the affixation process. Although all morphological processes can be It is difficult to find conclusive arguments for either described in either framework, some decriptions are Item & Arrangement or Item & Process approaches to more natural than others. affixation. Proponents of Item & Arrangement usually In a strict Item & Arrangement position, such as claim that their approach is theoretically more restrictive defended by Rochelle Lieber (1992), stems and affix- than Item & Process. Proponents of Item & Process typ- es have separate entries in the mental lexicon. The ically argue that processes such as infixation, reduplica- suffix -ness is described as a noun that requires an tion, vowel change, consonant change, and conversion adjective to its left. Given an appropriate concept of are too frequent, especially in non-European languages, headedness (cf. below), this explains why fondness is to treat them as exceptional compared to ‘pure’ affixa- a noun. The Item & Arrangement approach works par- tion. Linguistic data alone cannot determine as to which ticularly well as long as the form of the resulting word of the two approaches is correct, because every Item & is the concatenation of the forms of the stem and the Arrangement account can be reformulated into Item & affix, as in fondness. Special provisions have to be Process and vice versa. 13 AFFIXATION Headedness The final -o and -a in (10) are not a part of the affix, but so-called word markers, whose form is determined The concept of head of a word only makes sense in an in this case by the grammatical gender of the word (-o Item & Arrangement-based account of morphology. for masculine, -a for feminine). Although the suffix The head determines the syntactic category of the -sit is the rightmost element and the only affix, it is the resulting word as well as other properties such as gen- base that determines the gender as indicated by the der in languages that distinguish them. Thus, fondness article. In an Item & Process approach, there is no is a noun because -ness is the head, and French activité need for a head, because its function is subsumed in (‘activity’) is a feminine noun because -ité is the head. the affixation process. Different methods have been proposed to determine the head. Anna Maria Di Sciullo and Edwin Williams (1987) propose that the head is always the rightmost References element. This concurs with the observation that suffixes Anderson, Stephen R. 1992. A-morphous morphology. often change the syntactic category of their base, but pre- Cambridge: Cambridge University Press. fixes, as in (1b), usually do not. Exceptions include cases Aronoff, Mark H. 1994. Morphology by itself: stems and such as enrich and debug, where the prefixes make verbs inflectional classes. Cambridge, MA: MIT Press. out of an adjective and a noun. An alternative proposed Booij, Geert, Christian Lehmann, and Joachim Mugdan (eds.) by Lieber (1990) is that the last affix to attach is the 2000. Morphologie—Morphology: Ein Internationales Handbuch zur Flexion und Wortbildung—An international head. This generalization makes re- the head in (1b) and handbook on inflection and word-formation, Vol. 1. Berlin: accounts correctly for enrich and debug, but has prob- Walter de Gruyter. lems with prefixes such as counter- in (9). Di Sciullo, Anna Maria, and Edwin Williams. 1987. On the def- inition of word. Cambridge, MA: MIT Press. (9) a. intuitive counterintuitive ten Hacken, Pius. 1994. Defining morphology: a principled approach to determining the boundaries of compounding, b. example counterexample derivation, and inflection. Hildesheim: Olms. c. sign countersign Harris, James W. 1991. The exponence of gender in Spanish. Linguistic Inquiry 22. 27–62. Hockett, Charles F. 1954. Two models of grammatical descrip- In (9), counter- is the only (hence last) affix tion. Word 10. 210–31. attached, but the syntactic category depends on the Lieber, Rochelle. 1990. On the organization of the lexicon. New word it attaches to. The Spanish diminutive, illustrated York: Garland. in (10), is problematic for both. Lieber, Rochelle. 1992. Deconstructing morphology: word for- mation in syntactic theory. Chicago: University of Chicago Press. (10) a. el pintor ‘the painter’ PIUS TEN HACKEN el pintorsito ‘the little painter’ b. la cancyón ‘the song’ See also Compounding; Inflection and Derivation; la cancyonsita ‘the little song’ Word African American Vernacular English African American Vernacular English (AAVE) has The terms have been used to designate nonstandard been known by various names in linguistics, including English varieties spoken by less educated African Negro Nonstandard English, Black English Vernacular, Americans other than those in coastal Georgia and and more recently Ebonics. Except for the last, each South Carolina, who speak Gullah, a creole. The aver- name corresponds to a particular time in American his- age speakers of both vernaculars do not have specific tory when it was more fashionable to refer to African names other than English for the ways they speak; Americans as Negroes or Blacks, among other names. scholars have coined the names to distinguish their 14  AFRICAN AMERICAN VERNACULAR ENGLISH ethnolects (ethnic dialects) from standard English and Americans of standard English structural features, from European-American vernaculars. supposedly because of the mass media, mass educa- Ebonics, which was first used in print in the work of tion, and socioeconomic mobility since the abolition Robert Williams (1975), is probably the most confus- of slavery.) (3) How different is AAVE’s system from ing of the terms. It subsumes Gullah and thus corre- those of standard English and European American ver- sponds to what may be identified broadly as African naculars such as Appalachian and Ozark English? (4) American English. However, in the way Williams How does variation work in its system and what is its defines it, it also refers to other language varieties of significance from an evolutionary point of view? the black diaspora, including English and non-English Overall, a very large proportion of the research has creoles of the Atlantic (e.g. Jamaican and Haitian cre- been devoted to a small subset of structural features oles and Nigerian Pidgin English). The name becomes that makes AAVE peculiar, at least by its statistical quite useless when it is also extended to cover Niger- distribution. This includes the following: Congo languages. The average AAVE speaker might be 1. AAVE allows the loss of consonant clusters at the puzzled as to why their English is lumped with a mul- end of words, and this means that the final con- titude of other languages they hardly understand, if at sonant is often dropped in words such as desk, all, and why the fact that it is first of all an American passed, and old (pronounced des’, pass’, and ol’). linguistic phenomenon is virtually overlooked. 2. Diphthongs (double vowel sounds) such as On the basis of literary sources, the predecessors of [aw] and [ai], as in south side, become long AAVE must have become distinct from other American monophthongs (simple vowels): [a:]. English vernaculars already in the early eighteenth 3. The ‘r’ is dropped in front of consonants and at century (Brasch 1981). Literary representations of the the word end, as in Lord and car (pronounced time are similar to those about the Caribbean, where Lo’d, ca’); the ‘r’ is pronounced before vowels, larger plantations with African majority populations as in the words arrive and road. were already booming. It is not clear whether the liter- 4. The copula ‘be’ is often absent in sentences ary representations (mostly by British, rather than such as She cute and He gon come. American, authors) reflect the speech of African 5. Be is used to denote invariant or repeated states Americans of the time or, instead, generalized stereo- of affairs, as in He be tired/reading every time types that were based on what was developing in the I visit him. Caribbean. Newspaper advertisements about runaway 6. AAVE uses multiple negation, as in He ain go slaves reveal that slaves born in the colonies (a large nowhere (nohow). proportion by the early eighteenth century) or who had 7. The negative auxiliary (e.g. didn’t) is moved to lived there for a long time did not speak differently the front of the sentence, as in Didn’t nobody from the European indentured servants with whom tell me ‘hush yoh mouth’, ‘Nobody told me, they interacted on a regular basis, before the institu- ‘hush your mouth’.’ tionalization of segregation in tobacco and cotton plan- 8. AAVE differs from standard English in various tation ex-colonies in the late nineteenth century. This aspects of time reference, as in I been there, ‘I history suggests that, unlike Gullah, AAVE could not have been there’, and They BEEN married (with have developed as a distinct variety before the late emphasis on been), ‘They have been married for nineteenth century. Aside from letter-writing evidence, a long time and still are’, and He done lef when I more reliable literary representations of AAVE proba- come, ‘He had already left when I came’. bly also date from the nineteenth century, when most of 9. Plural is expressed with dem (in alternation with the authors were also American born. the presence or absence of the standard English AAVE has become central to research in American plural marker -s), as in dem boys, and an dem quantitative sociolinguistics and historical dialectol- signals an ‘associative plural’, as in Yolanda an ogy since the 1960s. William Labov, William Stewart, dem, ‘Yolanda and her friends/relatives’. and J.L. Dillard, together with Walt Wolfram and 10. Many words and word meanings are peculiar to Ralph Fasold, focused on subsets of the following African American speech, as in You are bad questions: (1) Is AAVE a rule-governed vernacular, (with bad pronounced emphatically and with and is it a full-fledged language? (2) Is it an American an extra-long vowel), ‘you are impressive/ English dialect or is it, instead, an erstwhile creole something else’, and you look clean, ‘you look (similar to those of the Caribbean), and therefore a sharp(ly dressed)’. separate language, which has acquired English-like features by ‘decreolization’? (Putatively, this evolution Interest is now growing in describing not only iso- would be caused by approximations by African lated features but also integrated subsystems of the 15 AFRICAN AMERICAN VERNACULAR ENGLISH vernacular, highlighting, on the one hand, similarities Dillard, J. L. 1972. Black English: its history and usage in the with and differences from other American English United States. New York: Vintage Books. Labov, William. 1972. Language in the inner city: studies in the varieties and, on the other, how subsystems such as Black English vernacular. Philadelphia: University of time reference or the noun phrase are internally organ- Pennsylvania Press. ized. This is a corollary of the fact that today, studies Mufwene, Salikoko S., et al. (ed.) 1998. African-American of AAVE’s structure are no longer almost exclusively English: structure, history, and use. London: Routledge. in the quantitative sociolinguistics paradigm, but also Poplack, Shana (ed.) 1999. The English history of African American English. Malden, MA: Blackwell. in theoretical linguistics. Poplack, Shana, and Sali Tagliamonte. 2001. African American There has also been research in African American English in the diaspora: tense and aspect. Oxford: discourse differences, especially on discourse genres, Blackwell. the structure and contents of some narratives, the form Rickford, John R. 1999. African American vernacular English. of language used in them, and whether or not the mean- Malden, MA: Blackwell. Rickford, John Russell, and Russell John Rickford. 2000. ings of particular utterances are transparent. Particularly Spoken soul: the story of Black English. New York: John outstanding in this subarea are playing the dozens, in Wiley & Sons, Inc. which male participants exchange witty, fictional insults Smitherman, Geneva. 2000. Talkin that talk: language, culture directed especially at female relatives and tall tales or and education in African America. London and New York: toasts, rhyming epics in which street culture and its Routledge. Williams, Robert L. (ed.) 1975. Ebonics: the true language of heroes are celebrated in a street verbal style. Black folks. St. Louis, MO: Robert L. Williams & Associates, Inc. References SALIKOKO MUFWENE Brasch, Walter M. 1981. Black English and the mass media. See also Caribbean; Ethnicity and Language; Lanham, MD: University Press of America. Gullah Afroasiatic The Afroasiatic (formerly called Hamito-Semitic, also early 1820s) and Akkadian (in its two main varieties, Semito-Hamitic, more recently Afrasian) phylum unites Assyrian and Babylonian, in the late 1840s). Certain five or six language families that are clearly related, but peculiarities of the Semitic language structure, when the relations among the families are far from clear. The noticed at an early stage in the decipherment of confusion—and the profusion of names—result from Akkadian by Edward Hincks, proved to be very help- the history of the study of these languages. Known from ful in making further progress, and there was soon no antiquity, of course, were three of the principal Semitic doubt that another major Semitic language had been languages: Hebrew, Arabic, and Aramaic in its discovered. With Egyptian, although Jean-François Christian (Syriac) and Jewish literary forms. The Champollion already suggested a Hebrew comparison Ethiopic literary language Ge‘ez became known to in his initial announcement of the decipherment, and European scholars during the Renaissance and was Hincks used Egyptian loanwords in the Hebrew Bible immediately recognized as a close relative of the other in his analyses of Egyptian, the resemblances were three. Beginning in the mid-eighteenth century, a num- less obvious, and the actual kinship was suggested by ber of ancient languages that survived only in inscrip- the pioneering Egyptologist and linguist Richard tions were deciphered, starting with Palmyrene, a close Lepsius by 1880. relative of Syriac, and Phoenician, which is close to Also during the mid-nineteenth century, explorers, Hebrew. In 1781, the name ‘Semitic’ was suggested for notably Lepsius, traveling through Abyssinia (modern these languages as a family, on the basis of the ‘Table of Ethiopia, biblical Cush) gathered information on the Nations’ in Genesis 10:21–4, where Asshur, Aram, and languages they encountered there. Some (notably Eber are among the descendants of Shem. Amharic, the language of the ruling dynasty) proved to The ninteenth century brought two supremely be Semitic languages, others (notably Oromo, which important decipherments, those of Egyptian (in the was then called by the insulting name Galla, and 16  AFROASIATIC Somali) resembled each other fairly closely and tral to any of the modern ones), Egyptian (a single lan- Semitic only remotely; this group was soon dubbed guage attested over thousands of years), Cushitic and Cushitic. Only in the past generation has it been sug- Omotic (whose common ancestry is disputed), Berber gested that ‘West Cushitic’ is actually a separate (a close-knit group of dialects with some ancient mate- branch of Afroasiatic (renamed Omotic), perhaps very rials), and Chadic (a vastly differentiated family with a distant from Cushitic; this was widely accepted for widely used, politically dominant variety). (The Indo- some time, but opinion may be shifting back to the European analogy can also remind us that we have no earlier view. (The principal Omotic language is way of knowing how many languages or entire branch- Wolaytta of Ethiopia, with some 2 million speakers, es of Afroasiatic have vanished without a trace—what but very little is known about most of its relatives.) As ‘Avestan’ or ‘Umbrian’, what ‘Tajik’ or ‘Albanian’ Africa was partitioned among various colonial powers, there might have been.) Thus, it is no wonder that there naturally the investigation of Africa’s indigenous lan- are almost as many views of Afroasiatic relationships guages was carried out largely by scholars from the as there are Afroasiaticists. One view divides the north- respective metropolitan nations. Thus, much of the ini- ern groups—Berber, Egyptian, Semitic—from the tial work on Cushitic was done by German and Italian southern ones—Chadic, Cushitic, Omotic. Another linguists, while French investigators provided most of groups Semitic and Cushitic, plus Berber, against the descriptions of Berber tongues from North Africa Egyptian and Chadic, plus Omotic. (Morocco to Libya). Some characteristics that are widely shared among The last of the five main branches of Afroasiatic to Afroasiatic languages are as follows. Consonants tend receive attention is called Chadic, from its location to be fairly numerous and vowels very few. Stops and near Lake Chad; the principal language of Nigeria, sibilants appear in three series, voiced (b d g, z), voice- Hausa, is its main representative, and what little inves- less (p t k, s), and ‘emphatic’, customarily marked with tigation was done was largely carried out by British a dot (t., .s). The ‘emphasis’ appears differently in linguists. The large number of Chadic-speaking com- different languages: in Semitic, it is mostly represent- munities, their small populations, and their remote ed by velarization or pharyngealization (where con- locations, mostly in northern Nigeria, are factors that striction at the back of the mouth colors the consonant have made Chadic the least known of these groups and adjacent vowels), but elsewhere it appears as glot- overall. talized or even as implosive varieties of these sounds. Returning to the ‘Table of Nations’, we find that There are also a fairly large number of consonants Ham is listed as the father of Cush and Egypt, so not made in the back of the mouth (including the notori- unreasonably these languages that were clearly, ous Arabic ‘ayn /ʕ/). Another peculiarity is a tendency although distantly, related to Semitic came to be called to laterals (sounds produced with air flowing past the ‘Hamitic’. But this proved to be a suboptimal choice sides of the tongue) and interdentals (with the tongue for two reasons: linguistically, it implies that all the tip between the front teeth as in English th [
]). other families form a unit as against Semitic (implying Conversely, Afroasiatic languages tend to have the that the Semitic family was the first to branch off from near-minimal vowel set /a i u/ (both long and short). the others), which is not the case; and ethnographical- The best evidence that the Afroasiatic languages are ly or anthropologically, it caused the classification of related (rather than reflecting much borrowing of languages to become enmeshed with the spurious vocabulary and grammatical features) is the shared notion of a ‘Hamitic race’ of dark-skinned but inflectional paradigms: the personal prefixes and suf- European-featured Africans. (In fact, of course, there fixes self-evidently have a common origin throughout. is no necessary connection between language family The elements used to derive one verb from another relationship and ‘race’ or ethnic group.) For this rea- also appear throughout the phylum (Table 1). It is son, the name ‘Afroasiatic’ is now widely used for the noteworthy that both the earliest and the latest com- phylum, since it simply names the two continents parative grammars of Semitic, those of Heinrich where it is found. This can be contracted to the con- Zimmern (1898) and Burkhart Kienast (2001), sys- venient ‘Afrasian’. tematically include Afroasiatic comparisons. What, then, are the relationships among the branch- (Meanwhile, Carl Brockelmann, who compiled the es of Afroasiatic? The very nature of the data makes it definitive collection of comparative grammatical difficult to achieve consensus. It is as if scholars of information on Semitic as then known (1908–1913), Indo-European were faced with only Gothic and the apologizes in the preface for excluding ‘Hamitic’ other Germanic languages, Greek, modern Romance materials, as the data were inadequate at the time. and Celtic, Armenian, and the modern Indic lan- Other writers on Semitic have followed his lead.) The guages—analogous, respectively, to Akkadian and the well-known ‘triconsonantal root’ structure of the other Semitic languages (an ancient tongue not ances- Semitic languages, however, does not appear to be 17 AFROASIATIC TABLE 1 Some Afroasiatic Comparisons Common Ancient Tuareg Afar Hausa Semitic Egyptian (Berber) (Cushitic) (Chadic) Prefix-conjugation markers (Hausa: aspect pronouns) 1 sing. ’a- — ø-…g´ ø náa 2 m. sing. ta- — t-…-d t- káa 2 f. sing. ta-…¯¹ — t-…-d t- kín 3 m. sing ya- — y- y- yáa 3 f. sing ta- — t- t- táa 1 pl. na- — n- n- mún 2 m. pl. ta-…u¯ — t-…-m t-…-n kún 2 f. pl. ta-…a¯ — t-…-mt t-…-n kún 3 m. pl. ya-…u¯ — -n y-…-n sún 3 f. pl. ya-…a¯ — -nt y-…-n sún Suffix-conjugation or pronominal affixes 1 sing. -ku¯ -w’¹ -i ø níi 2 m. sing. -ta¯ -k -k -t kái 2 f. sing. -t ¯¹ -J -m -t kée 3 m. sing. ø -f -s ø -šíi 3 f. sing. -at -s′ -s -t ’ítá 1 pl. -nu¯ -n -näg´ -tVn múu 2 m. pl. -tunu¯ -Jn -wän -tVn kúu 2 f. pl. -tina¯ -Jn -kmät -tVn kúu 3 m. pl. -u¯ -s′n -sän -Vn súu 3 f. pl. -a¯ -s′n -snät -Vn súu Verb derivation markers causative š — s s — passive n — t m — reciprocal t — m t — shared throughout Afroasiatic; elsewhere in the phy- from the first centuries AD. Proto-Semitic may be con- lum, ‘roots’ tend to have just two consonants. sidered quite reliable for the same reasons—the large The only attempts at description of the grammar amount of data and the small time depth. Contrast the of the hypothetical ancestral language ‘Proto- situation in Afroasiatic: one branch (Semitic) well- Afroasiatic’ are the two small volumes by I. M. known, one branch (Egyptian) known in considerable Diakonoff (1965, 1988), and he changed his mind detail over a long time—but with no vowels known, considerably between the two. A summary, quite close and one branch (Berber) with a scattering of meager to the earlier version, may be found in his article in the inscriptions from Roman times. Thus, four of the six Encyclopædia Britannica (1974). Diakonoff must be branches are known only or primarily from modern considered quite brave for even attempting to recon- data—and the time depth of Chadic alone is estimated struct a proto-language, given the nature of the avail- as about the same as that of Indo-European. able materials: contrast the situation in Indo-European It is no wonder, then, that among the three presen- (the real situation, not the truncated version offered tations of comparative Afroasiatic vocabulary—those above as a parallel to that in Afroasiatic). Proto-Indo- of Marcel Cohen (1947; omitting Chadic, with an European can be reconstructed in considerable detail occasional Hausa comparison), Christopher Ehret (and with considerable agreement among scholars) (1995; omitting Berber), and Vladimir Orel and Olga because the hypothetical community that would have Stolbova (1995)—there is precious little agreement. spoken a language similar to what is reconstructed (Nor has the preliminary requirement of comparative thrived only a couple of thousand years or so before dictionaries of each of the constituent families been the earliest available evidence (Hittite, Vedic, Avestan, accomplished, even for Semitic.) Part of the difficulty Mycenean). We can even verify the accuracy of com- is the shortness of words in the languages, with their parative and reconstructive methodology over such a roots consisting of just two consonants. Thus, it is all time span because we have the precious case of the the more difficult to proceed to the other main consid- Romance languages—reconstructed Proto-Romance eration regarding an ancestral language: the homeland is pleasingly similar to Vulgar Latin as it is known of its speakers. 18  AGE AND LANGUAGE There are three ways of determining where a lan- member of Nostratic, but is coordinate with the lan- guage family may have originated. One, which is hypo- guages that make up ‘Eurasiatic’, is slightly more thetical, suggests that the area with the greatest tenable. An independent suggestion, made by archae- concentration of diverse, related languages is likely to ologists who investigated Indo-European origins, that be where the family originated. (It is not likely that sev- ‘Afroasiatic’—from which they seem to be aware only eral related language groups would all happen to move of Egyptian and Semitic—originated on the southern to the same locality!) For Afroasiatic, this points to coast of the Black Sea, is ipso facto untenable. somewhere in northeast Africa, near where Chadic and The careful study of Afroasiatic languages goes back Cushitic (and Omotic?) are found, say Ethiopia/Sudan. to antiquity. The study of the Afroasiatic phylum is Further support for this suggestion might be found in nearly two centuries old. The tasks that remain for the evidence that the Egyptians entered the Nile Valley researchers are largely the same as those faced by from the south and settled gradually northward. There specialists in other languages: intensive fieldwork, espe- is simply nothing but speculation as to why and cially in Chadic, Cushitic, and Omotic, to record ample whence successive populations of Semitic-speakers information about these languages, and then to bring to appeared on the scene in southwest Asia. A second bear all the analytic techniques of linguistic theory and technique is to examine the vocabulary that can be comparative method. A difference with Afroasiatic is traced back to the protolanguage. A dishearteningly that it also includes some of the oldest written records limited number of words appear to have survived from on earth, providing a unique cross-section of human lin- so long ago, but they seem to point to a preagricultur- guistic behavior across 5,000 years. al, nonmaritime environment. The third source of information is the accidents of discovery owing to archaeology. Terrain, climate, and politics conspire to References make expeditions unlikely; but there is reason to sup- Brockelmann, Carl. 1908–1913. Grundriss der vergleichenden pose that an Afroasiatic homeland is to be sought in Grammatik der semitischen Sprachen. Berlin: Reuther & regions that are now the eastern Sahara Desert, but Reichard. Cohen, Marcel. 1947. Essai comparatif sur le vocabulaire et la were hospitably temperate over 10,000 years ago. An phonétique du chamito-sémitique. Paris: H. Champion. archaeological horizon in the western Arabian peninsu- Diakonoff, I. M. 1965. Semito-Hamitic languages. Moscow: la has been suggested as coordinating with Proto- Nauka. Semitic speakers, but in the absence of written records, Diakonoff, I. M. 1988. Afrasian languages. Moscow: Nauka. no assignment of languages to archaeological sites can Ehret, Christopher. 1995. Reconstructing Proto-Afroasiatic (Proto-Afrasian): vowels, tone, consonants, and vocabulary. be considered certain. Berkeley and Los Angeles: University of California Press. What of wider connections? From time to time, Kienast, Burkhart. 2001. Historische semitische Afroasiatic is claimed to be part of a still older, vastly Sprachwissenschaft. Wiesbaden: Harrassowitz. extended superstock called ‘Nostratic’, which is said Orel, Vladimir, and Olga Stolbova. 1995. Hamito-Semitic ety- to incorporate Indo-European, Uralic, Altaic, and mological dictionary: materials for a reconstruction. Leiden: Brill. various other phyla of Asia. Any perceptible reflection Zimmern, Heinrich. 1898. Vergleichende Grammatik der semi- of a Proto-Nostratic must be uninterpretably dim; but tischen Sprachen. Berlin: Reuther & Reichard. a question that seems not to be addressed by PETER T. DANIELS Nostraticists is the location of the Proto-Nostratic- speaking community—a location from which all the See also Ancient Egyptian; Hausa and Chadic Lan- postulated homelands must be reachable. From this guages; Semitic Languages; Tuareg and Berber point of view, the suggestion that Afroasiatic is not a Languages Age and Language Comprehension and production impose many simulta- and pragmatic. It has been generally assumed that neous demands on the reader to process information these simultaneous processing demands are handled on a number of levels, including syntactic, semantic, by a human’s working memory and that working 19 AGE AND LANGUAGE memory capacity declines with age, affecting lan- examining how to improve older adults’ comprehen- guage processing. Support for this hypothesis is large- sion of language. Task demands, lapses in attention, ly correlational. Older adults are typically found to response biases, or other cognitive deficits contribute have smaller working memory spans than young to the syntactic comprehension problems of adults adults, and such span measures are found to correlate with dementia. with language-processing measures. A particular Semantic and lexical processes are particularly dis- source of difficulty for older adults is the production rupted by Alzheimer’s dementia. Many researchers and comprehension of complex syntactic structures have found that tests of implicit word knowledge involving multiple levels of sentence embeddings that reveal few differences between dementing adults and impose high processing loads on working memory. normal older adults, and lexical knowledge is com- The spontaneous use of such constructions declines monly found to be preserved. Other researchers have with advancing age, and their use contributes to the concluded that the structure of semantic knowledge is breakdown of reading and listening comprehension. destroyed by Alzheimer’s dementia. As a result, per- Text processing is particularly vulnerable to the formance on verbal-fluency tasks, such as generating effects of aging; both cross-sectional and longitudinal examples of categories or words beginning with a spe- research shows that older adults’ reading comprehen- cific letter, is impaired, picture and object naming is sion is impaired, as is their listening comprehension. hindered, and word associations are destroyed. The Further, word and name retrieval is disrupted by aging, semantic network of adults with Alzheimer’s dementia contributing to an increase in tip-of-the-tongue may be intact but inaccessible, perhaps due to the gen- difficulty in finding words. It is unclear as to whether eral, task-independent slowing of cognitive processes, working memory affects the immediate processing of including lexical access. General slowing may words and sentences, or whether working memory increase progressively with disease severity. A break- exerts its affects on the processes involved with the down of inhibitions may also contribute to many of the storage, recall, and recognition of linguistic informa- word retrieval problems experienced by adults with tion. The general slowing of cognitive processes with Alzheimer’s disease if they are unable to inhibit inap- age may also contribute to older adults’ processing propriate word associations. problems by affecting the speed of lexical access and A special speech register, sometimes termed ‘elder- other component processes. Neural inhibitory mecha- speak’, has been described as an accommodation in nisms also appear to weaken with age and permit the communicating with older adults, especially those intrusion of irrelevant thoughts, personal preoccupa- with dementia. This accommodation is characterized tions, and idiosyncratic associations during language as involving a simplified speech register with exagger- processing. These irrelevant thoughts compete for ated pitch and intonation, simplified grammar, limited processing resources, such as working memory capac- vocabulary, and a slow rate of delivery. Many of the ity, and impair older adults’ comprehension and recall. characteristics of elderspeak, such as its slow rate, A controversial issue in the study of aging and lan- exaggerated prosody, and simplified syntax and guage is ‘off-target verbosity’. A minority of older vocabulary, resemble the characteristics of other adults have been observed not only to talk a great deal speech registers such as ‘motherese’, or the speech but also to drift from topic to topic, weaving into their used by adults when communicating with small conversations many unrelated and irrelevant topics. children. Elderspeak may be evoked by the actual Off-target verbosity appears to be related to poor per- communicative needs of the elderly individual as well formance on tests of the function of particular brain as by (culturally based) negative stereotypes of older regions, as well as to psychosocial stress, extraversion, adults, and it is often viewed by older adults as insult- limited social support, and decreased social interac- ing and patronizing. Older adults, especially those in tion. Whereas off-target verbosity has been cited as nursing homes, may adapt to situational demands by providing strong support for inhibitiory deficit theory, becoming more accepting of elderspeak. Some form others have argued that this speech style is limited to of elderspeak may enhance communication with older social settings in which older adults construe their task adults, especially those with Alzheimer’s dementia, differently than do young adults—as a monolog, but little systematic research has been undertaken to responsive to an internal chain of associations. assess intervention and training programs. The debates over working memory capacity limita- In sum, the study of language and aging has focused tions and inhibitory deficits are central to the general on the effects of processing limitations on older adults’ study of gerontology. These core research issues have comprehension and production. Nonetheless, it is been supplemented by two emerging topics: research important not to lose sight of three general points. First, examining the effects of Alzheimer’s disease on lan- many aspects of language processing do not change guage production and comprehension, and research with age, particularly lexical access and semantic 20  AGREEMENT memory, but are vulnerable to the ravages of psychology of learning and motivation, Vol. 22, ed. by Alzheimer’s disease. Other aspects of language, partic- Gordon H. Bower. New York. Academic Press. Kemper, Susan. 1992. Language and aging. The handbook of ularly syntax, may be more resistent to Alzheimer’s aging and cognition, ed. by Fergus I. M. Craik and Timothy disease, yet susceptible to age-related decline due to A. Salthouse. Hillsdale, NJ: Lawrence Erlbaum Associates. working memory limitations or inhibitory deficits. Kemper, Susan, and Tamara Harden. 1999. Disentangling what Second, there has been a gradual shift toward the use of is beneficial about elderspeak from what is not. Psychology sophisticated methodologies such as the study of self- and Aging 14. 656–70. Kemper, Susan, and Reinhold Kliegl (eds.) 1999. Constraints paced reading times, eye movement patterns, and brain on language: aging, grammar, and memory. Norwell, MA: activation patterns to study how aging and dementia Kluwer Academic. affect language processing. Third, the shift toward Light, Leah L., and Deborah M. Burke (eds.) 1988. Language, developing and evaluating practical applications, as memory, and aging. New York: Cambridge University Press. seen in the growing body of research on elderspeak, Park, Denise C., Roger W. Morrell, and Kim Shifren (eds.) 1999. Processing of medical information in aging patients: continues. The linkage of basic research on language cognitive and human factors perspectives. Mahwah, NJ: processing with everyday practicalities has led to the Lawrence Erlbaum Associates. development of consumer standards and guidelines for Ryan, Ellen B., Howard Giles, Giampiero Bartolucci, and presenting medical information, and for electronics and Karen Henwood. 1986. Psycholinguistic and social psycho- telecommunications. logical components of communication by and with the eld- erly. Language and Communication 6. 144–66. Snowdon, David A., Susan J. Kemper, James A. Mortimer, Lydia. H. Greiner, David. R. Wekstein, and William R. Markesbery. References 1986. Cognitive ability in early life and cognitive function and Arbuckle, Tannis, and Deloris Pushkar Gold. 1993. Aging, inhi- Alzheimer’s disease in late life: findings from the Nun study. bition, and verbosity. Journal of Gerontology: Psychological Journal of the American Medical Association 275. 103–13. Sciences 48. 225–32. Wingfield, Arthur, and Elizabeth A. L. Stine-Morrow. 2000. Caplan, David, and Gloria Waters. 1999. Verbal working mem- Language and speech. The Handbook of Aging and ory and sentence comprehension. Behavioral and Brain Cognition, 2nd edition, ed. by Fergus I. M. Craik and Sciences 22. 114–26. Timothy A. Salthouse. Mahwah, NJ: Lawrence Erlbaum Charness, Neil, Denise C. Parks, and Barbara A. Sabel (eds.) Associates. 2000. Communication, technology, and aging. Doylestown, SUSAN KEMPER Pennsylvania: Springer. Hasher, Lynn, and Rose T. Zacks. 1988. Working memory, See also Aphasia; Psycholinguistics; Working comprehension, and aging: a review and a new view. The Memory Agreement Agreement is a means by which languages signal the 1PL nous parl-ons ‘We speak’ presence of a grammatical relation—usually the sub- 2PL vous parl-ez ‘You speak’ ject of a sentence, but sometimes also the object and/or 3PL ils parl-ent ‘They speak’ the indirect object. Typically, it manifests itself as a prefix or a suffix on the verb, as with the English third- The features most often associated with agreement person singular marker -s (e.g. John lives in are person, number, and gender. Given the logistics of Maryland). Some languages have so-called rich agree- speech, there are at least three persons: those referring ment—that is, separate forms for nearly every per- to the speaker (first), the hearer (second), and every- son/number combination. French represents such a thing else (third). Some languages have a fourth case: person—obviative—reserved for entities further removed from the speaker/hearer’s point of view. (1) Rich agreement (French) Inuktitut (Eskimo) and Algonquian are known for this. 1SG je parl-e ‘I speak’ Number usually breaks down into singular and plu- 2SG tu parl-es ‘You speak’ ral, but ‘dual’ represents another possibility. English 3SG il parl-e ‘He speaks’ recognizes two numbers in its pronominal system, but 21 AGREEMENT as far as agreement is concerned, only third-person In traditional grammar, the case of the subject subjects are marked: cf. He understands but We is called nominative, while that of the direct object is understand. accusative. Since only subjects are associated In addition to person and number, agreement may with agreement in English, this may be taken as a sign also involve the gender features of a subject and/or of nominative case. Objects are not overtly marked objects. French nouns are divided into two grammati- with accusative case, but their grammatical function cal genders—masculine and feminine. When noun can be determined by their position following the verb. phrases headed by these nouns appear in subject Another way in which agreement appears function- position, agreement can be discerned on predicate ally similar to case is through its interaction with adjectives and particles (M = masculine, F = feminine): transitivity. In languages like Japanese, transitive and intransitive subjects take the same case-marker, and in (2) Gender agreement (French) opposition to objects. This is known as a nomina- tive/accusative case-marking system. English is essen- a. Le soleil est monté ‘The sun (M) has tially the same with respect to agreement—that is, it risen’ pertains to both types of subject. In some languages, b. La lune est montée ‘The moon (F) has however, intransitive subjects and transitive objects risen’ pattern in the same manner, in contrast to transitive [PARTICIPLE] subjects. Dyirbal (Australian) has a case-marking system of this type, known as ergative/absolutive c. Le soleil est brillant ‘The sun (M) is (transitive subjects are marked with ergative case). bright’ Languages that employ agreement, rather than case to d. La lune est brillante ‘The moon (F) is signal grammatical functions can exhibit the same pat- bright’ tern, as in Mam (Mayan) below (ABS = absolutive; [ADJECTIVE] ASP = aspect; ERG = ergative): Feminine forms are marked with an extra vowel (4) Ergative/absolutive agreement (Mam) that—although not pronounced today—were presum- ably audible at an earlier stage of the language. Of a. ma chiin-x-a course, grammatical gender is not to be confused with ASP 1SG/ABS-go biological gender; the division of nouns into gender ‘I went’ classes should be considered as an aid to reference [INTRANSITIVE] and/or acquisition (German has three genders, for example). b. ma chin-ok t-tzeeq’an-a ASP 1SG/ABS 2/ERG-hit Agreement and Case ‘You hit me!’ [TRANSITIVE] As a grammatical category, agreement appears to function like case-markers—special morphemes that Why is it important to indicate grammatical func- attach to noun-phrases (NPs) (rather than verbs) that tions? Many linguists believe that languages use these signal their grammatical function within the sentence devices to identify the participants in the event speci- (e.g. as subject, object, etc.). In Japanese, for instance, fied by the verb. In a sentence like ‘The President hires NPs marked with [-ga] are interpreted as subjects, young assistants’, for example, the verb calls for two while those marked with [-o] are understood as objects participants: someone who initiates the action (the (TOP = topic; COMP = complementizer): President), and someone/something affected by it (young assistants). The first is sometimes called an (3) Case-marking and grammatical relations Actor, the second (in this sentence) a Patient or (Japanese) Undergoer. Grammatical functions signaled by subject agreement and accusative case (the latter by virtue of Ichiro-wa Taro-ga hirugohan-o its position following the verb) ensure that each par- Ichiro-TOP [Taro-SUBJ lunch-OBJ ticipant is interpreted correctly, or that the sentence does not mean ‘Young assistants hire the President’. tabeta-to iita Languages that employ neither case-marking nor ate]-COMP said agreement must obviously find some other way of ‘Ichiro said that Taro ate lunch’ interpreting such potentially ambiguous sentences. 22  AGREEMENT Agreement and Inflection intricate, and poses a major challenge for linguists attempting to explain it. Agreement is often called an inflectional category, as opposed to lexical ones like nouns, verbs, and adjec- tives. Other inflectional categories include tense, Nonstandard Cases mood, and aspect. Sometimes, agreement is fused with While subjects normally trigger agreement in English- one or more of these to form a single entity. This is the tensed clauses, there are a number of cases in which it case with the English suffix -s: not only does it signal is suppressed and/or the status of the subject itself is a third-person singular subject but also present tense called into question. One case involves ‘subjunctive’ and indicative mood. Agreement disappears in the past mood, following verbs like suggest and demand: cf. tense and/or indicative mood: cf. John lives in I demanded that he be on time; We quickly suggested Maryland now, but last year he lived in Maine; Mary that she leave (compare: …that she should leave, wants to stay, but it is necessary that she leave (NOT: which is also acceptable but does not show signs of leaves). Still, there is some debate as to whether agree- agreement). Other languages show reduced forms ment is truly absent from these forms or simply of agreement under negation, or when one grammati- covert—that is, ‘present but invisible’. In German, for cal function or another is questioned: example, agreement is perfectly compatible with the past tense, and in Palauan (Austronesian) with the (8) Reduction of agreement (Chamorro/ irrealis mood (similar to nonindicative): Austronesian) (5) Agreement and past tense (German) a. Ha-fa’gasi si Juan i kareta 3SG-wash Juan car 1SG ich glaub-t-e ‘I believed’ ‘Juan washed the car’ 2SG du glaub-t-est ‘You believed’ [NORMAL AGREEMENT] 1SG wir glaub-t-en ‘We believed’ b. *Hayi ha-fa’gasi i kareta? who 3SG-wash car (6) Agreement and irrealis mood (Palauan) ‘Who washed the car?’ [WH-QUESTION/NORMAL AGR] REALIS IRREALIS 1SG ak- ku- c. Hayi fuma’gasi i kareta? 2SG ke- omo- who wash+UM car 3SG ng- le- ‘Who washed the car?’ [WH-QUESTION/SPECIAL AGR] Object agreement is often associated with the inflectional category aspect. Again, in Palauan, objects The existential there construction in English of completed events trigger ‘perfective’ agreement on arguably has two subjects: one occurring in clause- the verb, whereas objects of ongoing ones do not initial position (filled by there), the other post-verbal- (IMPF = imperfective; NM = noun-marker; P-preposi- ly and signaling agreement: cf. There’s a man tion; PF = perfective; R = realis): standing on the corner vs. There are three men stand- ing on the corner. Nevertheless, speakers sometimes (7) Agreement based on aspect (Palauan) take the clause-initial subject as the agreeing one, resulting in sentences like There’s four things that can a. ng-chillebed-ii a bilis be said about him. Other nonstandard (but occurring) R3SG-hit-PF3SG NM dog breakdowns involve plural complements of singular ‘S/he hit the dog’ subjects (Neither one of you guys are going to win), [PERFECTIVE] and coordinate (hence plural) subjects with singular marking on the verb: The fact that we were bombed, b. ng-milengelebed er a bilis and that the markets reopened so quickly suggests that R3SG-hit (IMPF) P NM dog the foundations of our country are quite strong. ‘S\he was hitting the dog’ [IMPERFECTIVE] Exotic Forms of Agreement The relationship between agreement—which basi- Agreement in natural language can deviate significant- cally pertains to nouns—and tense, mood, and aspect ly from single forms that encode features of person (a property of verbs) is therefore sometimes quite and number. In Potawatomi (Algonquian), each of 23 AGREEMENT these features can be associated with a separate mor- (11) Portmanteau (Inuktitut) pheme (OBV = obviative): a. tuni-jara nutarar-mut (9) Split agreement (Potawatomi) give-PRT/1SG/3SG child-AL 1SG n-wapm-a ‘I see him’ arna-up taku-jaa-nut 2SG k-wapm-a ‘You see him’ woman-REL see-PRT/3SG-AL 3SG w-wapm-a-n ‘He sees him ‘I gave it to the child that the woman saw’ (OBV)’ 1PL n-wapm-a-mun ‘We see them/him b. arna-up pilauti-milk (OBV)’ woman-REL knife-CMT 2PL k-wapm-a-wa ‘You see him’ 3PL w-wapm-a-wa-n ‘They see them/him angut tuni-jaa (OBV)’ man (ABS) give-PRT/3SG/3SG ‘The woman gave the man the knife’ Here, the prefixes [n-], [k-], and [w-] represent first, second-, and third-person subjects, regardless of their In (11a), the form [-jara] signals a first-person sub- number. The suffixes [-wa] and [-mun] indicate plural- ject and third-person object, while in (11b) [-jaa] indi- ity (the latter also encodes features of the object, i.e. cates two third persons. plurality and/or obviation). Interestingly, the same pre- fixes and suffixes also appear on nouns to indicate pos- session and number: Evolution of Agreement One might well ask why agreement is so pervasive in (10) Nominal agreement language, or where it comes from. Many morphemes appear to originate as independent pronouns that slow- (Potawatomi; c = alveopalatal affricate) ly undergo a process of phonological reduction and 1SG n-ciman ‘my canoe’ subsequent attachment. A comparison of pronouns and 2SG k-ciman ‘your canoe’ agreement from Udi (Caucasian) demonstrates this 3SG w-ciman ‘his/her canoe’ most clearly: 1PL n-cimanwa-nan ‘our canoe’ (12) Pronouns and agreement (Udi) 2PL k-ciman-wa ‘your canoe’ 3PL w-ciman-wa ‘their canoe’ AGREEMENT PRONOUNS 1SG -zu, -z zu In the larger picture then, agreement must be regard- 2SG -nu, -n, -ru, -lu un ed as more than just a property of verbs. ‘Subjects’ of 3SG -ne, -le, -re, -n meno, kano, seno NPs (possessors) corrolate with sentential subjects: it is no accident that the possessive marker [-s] in English 1PL -yan yan has the shape it does (cf. John’s dog). In Irish and other 2PL -nan, -ran, -lan van, efan Celtic languages, prepositions inflect for the 3PL -q’un met’ovon, kat’ovon, person/number features of their objects; this grammat- set’ovon ical relation roughly correlates with sentential objects. To fully appreciate how complex agreement can be, Although the match-up is far from complete, there is consider that objects of transitive verbs in Potawatomi enough similarity of form to allow for a theory of lan- can also be associated with separate morphemes of per- guage change. Independent/emphatic pronouns may son and number. In (9) above, the suffix [-a] signals that pass through an intermediate stage of cliticization on the object is a third person, while [-n] registers its plu- their way to becoming agreement morphemes. rality and/or obviation; also notice how the subject plu- Pronominal clitics also share the fundamental features ral marker [-wa] intervenes between these two suffixes. of person, number, and gender. French has a series of Inuktitut (Eskimo) takes a different tack in the dis- subject-markers traditionally regarded as pronominal: tribution of agreement features, using single forms je, tu, il, etc. Nevertheless, these forms cannot be sepa- called portmanteau to encode the person and number rated from the (auxiliary) verb like their English coun- of both subject and direct object (ABS = absolutive; terparts; cf. *Il vraiment est retourné vs. He really has AL = allative; CMT = comitative; PRT = participle; gone home. This suggests that je, tu, and il are clitics, REL = relative): and in the frequent presence of emphatic pronouns may 24  AINU someday evolve into a new series of agreement mor- to kill, for instance, can involve essential information phemes (cf. Moi, je vais retourné). Also, note that that is separate from the person who initiates the (except for first- and second-person plurals) the tradi- action (cf. kill an hour, kill a motion, kill the mood, kill tional agreement suffixes of modern French are no the light—in addition to the standard meaning ‘cause longer pronounced. to die’). Languages typically recognize this distinction through the mechanism of case-marking, and non- Actor participants are often made visible through Theory of Agreement cases selected by the verb. Actors, on the other hand, How do linguists analyze agreement? Usually, a local are less likely to be singled out by case. Instead, they relationship is assumed to hold between the element must rely on some other mechanism to be rendered that triggers it (e.g. the subject NP), and an inflection- visible, such as agreement. Inflection then—the head al head, which may be comprised of tense, mood, or of every clause—provides a specifier position that can aspect. Agreement is the visible sign of this relation- accommodate one non-case-marked participant for ship. This explains why only verbs in tensed clauses every verb. show agreement—at least in English. In structural terms, the subject NP is called the specifier of the References clause, attached to but outside the immediate domain of inflection: [NP [verb+inflection, etc.]]. This also Anderson, Steven. 1992. A-morphous morphology. New York: Cambridge University Press. parallels the internal structure of the NP, where articles Aronoff, Mark. 1994. Morphology by itself: stems and inflec- like a, the, and some are attached to but outside the tional classes. Cambridge, MA: MIT Press. immediate domain of the noun: [the [dog]]. In fact, Comrie, Bernard. 1989. Language universals and linguistic possessed NPs (e.g. John’s dog) can be assigned typology, 2nd edition. Chicago: University of Chicago Press. essentially the same structure as a clause if the posses- Croft, William. 1990. Typology and universals. New York: Cambridge University Press. sor (John) is regarded as a specifier, and the sign of Givon, Talmy. 1984. Syntax: a functional–typological introduc- possession ([-’s]) as a kind of inflection. It is probably tion. Amsterdam: John Benjamins Press. no accident that English uses the same sound to Marantz, Alec. 1992. On the nature of grammatical relations. express both specifier–head relationships. Cambridge, MA: MIT Press. The function of agreement—that is, as a means of Pinker, Stephen. 1984. Language learnability and language development. Cambridge, MA: Harvard University Press. highlighting one participant in the event described by MARK CAMPANA the verb—can be explained in the following way. First, a verb’s meaning is more closely comprised of its non- See also Austronesian; English; French Language; Actor participants than its Actor participants. The verb German; Grammar, Traditional; Japanese Ainu The Ainu are an indigenous group found mainly on the remaining Ainu are situated on the island of Hokkaido. northern Japanese island of Hokkaido. Physically and It should be noted that the Ainu language found in linguistically distinct from the neighboring Japanese, northern Japan is unrelated to the Turkic language at one time the territorial range of the Ainu extended bearing the same name that can be found in western throughout the northern Japanese island of Hokkaido China. (Yezo), including the Kurile islands to the northeast The genetic and linguistic history of the Ainu has and southern Sakhalin (Karafuto) to the north. been the topic of much debate. Physical features such Evidence from place names, however, suggests that at as increased body hair (when compared to the one time the Ainu had also settled in northern Tohoku, Japanese) and more distinct facial features make the on the main Japanese island of Honshu, and on Ainu appear Caucasian to many people and as such the Russia’s Kamchatka Peninsula as well. Due to reset- Ainu have sometimes been considered Caucasian tlement of the Sakhalin Ainu (1875) and the Kurile (Caucasoid) in ancestry. Other scholars have sought to Ainu (1884) today, however, the majority of the draw a relationship between the Ainu and native 25 AINU American groups. Recent studies conducted with for all forms of epic story in Ainu but as Masayoshi DNA testing, however, suggest the possibility that the Shibatani notes, Ainu are Mongoloid in origin. Numerous attempts have been made to classify In the strict sense of the term yukar refers only to the heroic verse, mythic epics being more specifically Ainu linguistically, including theories that place Ainu referred to as kamuy yukar, mat yukar, or oyna. There in the Malayo-Polynesian, Paleo-Asiatic, and Ural- are as well prose-style old stories and folktales. Altaic language families. At present, there are several competing theories on the relationship of Ainu with As noted above, Ainu is a polysynthetic language, other languages. The first and most widely offered where words tend to be composed of multiple mor- explanation is that Ainu is a language isolate, with at phemes. This is particularly the case in classical Ainu. present no apparent relationship with other languages. In contrast, colloquially spoken Ainu tends to be more The second theory rests on the assumption that Ainu is analytical, with more of the morphemes appearing as related to the Altaic language family, and languages independent words, suggesting the possibility of a such as Japanese and Korean. Some studies arguing long-term change in progress. the latter position suggest that if Ainu is at all con- In terms of its grammar, Ainu is a S(ubject)– nected to other Altaic languages, then the splitting of O(bject)–V(erb) language. But, as seen in the follow- Ainu from Japanese and other Altaic languages must ing example, word order is important in determining have occurred at least 10,000 years ago. However, the grammatical relationship in Ainu. Alexander Vovin (1993), in his work on reconstruction of Proto-Ainu (the hypothesized ancestor language), Kamuy aynu rayke. suggests a connection to the Austro-Asiatic languages bear person kill of the south. Further complicating matters has been ‘The bear killed the man.’ the historical lack of a writing system, and as a conse- quence, a written record of Ainu over time. The rela- tively rapid transition from a viable, functioning Aynu kamuy rayke. language with monolingual native speakers to one that person bear kill is, at present, nearly extinct has further confounded ‘The man killed the bear.’ attempts to conclusively place Ainu among the world’s languages. Phonologically, Ainu has a five-vowel system (/a/, The Ainu have had a continuous presence in /i/, /e/, /o/, /u/) with syllable-initial vowels preceded by Hokkaido for at least the past 1300 years. The first a glottal stop. The consonants of Ainu, 12 in all, record of the Ainu on Yezo occurs in the Japanese include the stops of /p/, /t/, /k/, and the glottal stop, Kojiki (A.D. 720) calling them ‘Emishi’, or those out- sibilant /s/, glottal fricative /h/, affricate /c/, semi- side of the Japanese law. The Ainu refer to themselves vowels /w/,/y/, nasals /m/ and /n/, and the liquid /r/. as Utari or ‘compatriot’ in Ainu, and this term is in Ainu has a pitch accent system in which syllables are more frequent use at present because of the sometimes pronounced with high or low pitch. Its syllable struc- negative connotations that have been leveled on the ture is C(onsonant)V(owel) or CVC. The exception to term ‘Ainu’ by the Japanese. The word Ainu simply this, however, is the Sakhalin dialect, which contained means ‘human’. long and short vowel contrasts, also allowing for CVV While no written form of Ainu existed until recent- to occur. Vowel sequences tend to be avoided in other ly (Ainu is now written in Romanized form or with the forms of Ainu. Vocalic euphony or vowel harmony, the Japanese katakana script), the Ainu culture has assimilation of a vowel to another vowel in the word developed a rich oral tradition most well known for or phrase, is one of the more notable aspects of the their epic poems called Yukar. Yukar follow traditional sound system of Ainu. patterns in their subject matter with some of the Yukar The Ainu vocabulary is rich with references to told from the perspective of an animal, others based on nature, reflecting the close relationship held between activities of ancestors, and still others are hero tales. the people and natural phenomena. Most Ainu villages The Yukar differ from the spoken language, tending to (called kotan) were located alongside rivers and coasts exhibit more complex (polysynthetic) word structures and former Ainu settlements are easily recognizable in and more conservative features than spoken Ainu. place names today with endings such as -nai or -pet These stories have served as a rich source of linguistic (-betsu is the Japanese representation), both meaning material as they maintain archaic forms, providing a ‘river’, as found in the place names of Wakkanai, rich source of data attesting to earlier forms of the Noboribetsu, and Monbetsu. The importance of Ainu language. Yukar has come to be a collective term salmon and bears to Ainu culture is represented by the 26  AINU abundance of lexical items related to these two ani- By 1871, many Ainu customs, including the tattoo- mals. Particularly, the role of the bear was an impor- ing of women’s mouths and the burning of a familial tant part of Ainu culture, and they were often kept in household upon the death of a family member, were villages for ceremonial events and as food. banned. The Ainu language was also banned as the Within Ainu, three distinct dialect groups are gen- Japanese government set out to assimilate the Ainu into erally recognized, the Kurile, Sakhalin, and Hokkaido becoming Japanese and officially classifying them as dialects, each representing the distinct regions where ‘former aborigines’. In 1899, the Hokkaido Aborigine the Ainu once inhabited. The last speaker of the Protection Act was created; yet, despite its name, the act Sakhalin dialect is believed to have died in 1994. did little more than increase the efforts to assimilate the Within these dialect groups, further distinctions can be Ainu. The Japanese government encouraged the Ainu to made, such as that between southern and eastern shift from their traditional role as hunter/gatherers to Hokkaido Ainu. At least 19 dialects of Ainu have been farming. At the same time, Ainu lands were ceased. identified, most of which are now extinct. When land was redistributed to the Ainu, they were It is estimated that the number of Ainu descendants given the least fertile lands. In 1901, Ainu children were (including those of mixed Ainu-Japanese heritage) forced to attend Japanese primary schools and Ainu number around 24,000, with those of solely Ainu children began learning Japanese as their first language. descent being very small. Postwar census figures clas- The assimilationist policies of the Japanese government sify the Ainu as Japanese for census purposes; resulted in a near elimination of the Ainu language and therefore, only estimates of the number of remaining culture by the mid-twentieth century. Today, there are Ainu in Hokkaido are possible. Discrimination aimed no monolingual native speakers left. at the Ainu by Japanese as well as a Japanese desire to Today, the Ainu are fluent Japanese speakers. In view Japan as a single homogeneous society have fur- spite of the past Japanese efforts to eliminate the Ainu ther caused many Ainu to conceal their heritage. It is language, bilingual speakers of Ainu and Japanese do believed that the actual number of Ainu in Hokkaido exist, however, and recently modest efforts have begun may be double or triple the figure of the government to continue the teaching of the Ainu language and tra- estimate of 24,000 (1995). The Ainu face regular dis- ditions to those willing to learn. Ainu language use crimination in employment and marriage opportuni- also continues in ceremonial and ritualistic use and in ties. Such discrimination and fear of disclosing their events performed for tourists. A preserved Ainu vil- Ainu ethnicity have, over time, contributed signifi- lage at Shiraoi is maintained for educational and cantly to the loss of the Ainu language. tourism purposes. The Ainu language and culture have been impacted In 1946, the Ainu Association of Hokkaido was dramatically by the influx of the Japanese into founded, seeking to bring increased attention to the Hokkaido, which began as early as the 1300s. The fight for Ainu rights, the preservation of the language peak of the Ainu culture occurred in the thirteenth and and culture, and a reevaluation of the Hokkaido fourteenth centuries, prior to the large-scale arrival of Aborigine Protection Act. the Japanese (wajin), which began in the fifteenth cen- In 1994, Shigeru Kayano became the first tury. From this point onward, battles with the Japanese Ainu member of the Japanese Diet (parliament). in 1457, 1669, and 1789, as well as diseases such as Subsequently, in 1997, the Japanese government final- smallpox, measles, cholera, and tuberculosis all took ly and officially recognized the Ainu as the indigenous their toll on the Ainu. people of Hokkaido and created the Ainu rights law. In 1550, an agreement was reached between the This law was designed to replace the Hokkaido Ainu and the Japanese to allow a Japanese settlement Former Aborigine Protection Act. It requires local area on Yezo. This would be a turning point in Ainu- governments in traditional Ainu areas to support pro- Japanese relations. Over the next three centuries, motion of the Ainu culture, including promotion of the Japanese control of Hokkaido increased and finally in Ainu language. As a result, government-subsidized 1869 the official government-sanctioned colonization weekly Ainu language radio broadcasts are offered in of Hokkaido began. The Japanese population of Hokkaido with free texts that are widely available. Hokkaido quickly grew to over 1 million, placing the There are also 13 state-supported language classrooms Ainu in a minority. From this point onward, the Ainu in traditional Ainu villages throughout Hokkaido that were systematically stripped of their cultural and lin- seek to provide opportunities for students to learn guistic identity, land, and hunting rights through a Ainu. A bi-weekly Ainu language newsletter has also forced assimilation policy of the Japanese govern- been produced. ment. Also during this period, disease again struck the The 1997 Ainu rights law provided official recogni- Ainu, leading to a decrease in their population. tion to the Ainu and a degree of legimaticy to the 27 AINU efforts aimed at preservation and revitalization of the Hattori, Shiro. 1959. Nihongo no Keitou. Tokyo: Iwanami. language. Whether the Ainu language will undergo a Kayano, Shigeru. 1994. Our land was a forest: an Ainu memoir. Boulder, CO: Westview Press. successful revitalization is dependent upon the ability Maher, John, and Kyoko Yashiro. 1995. Multilingual Japan: an of the Ainu to increase awareness among the Japanese introduction. Journal of Multilingual and Multicultural of what might be lost, should the Ainu language and Development 16(1). 1–17. culture disappear. Patrie, James. 1982. The genetic relationship of the Ainu lan- guage. Oceanic Linguistics Special Publication No. 17. Honolulu: University of Hawai’i Press. References Shibatani, Masayoshi. 1990. The languages of Japan. Batchelor, John. 1889. An Ainu–English–Japanese dictionary. Cambridge: Cambridge University Press. Tokyo: Y. Kumata; Iwanami-Syoten, 4th edition, 1938. Vovin, Alexander. 1993. A reconstruction of Proto-Ainu. Brochure on the Ainu. Sapporo, Japan: The Ainu Association of Leiden: E. J. Brill. Hokkaido, n.d. SHAWN M. CLANKIE Chiri, Mashiho. 1952. Ainugo ni okeru boin chouwa. Hokkaido Daigaku Bungaku Kiyou 1. 103–18. See also Altaic; Japanese Akan and Nyo Languages Nyo, like Left Bank, belongs to the (New) Kwa libr ‘red’ - ebr ‘reds’ subfamily of the Niger–Congo group. Nyo has more lufu ‘white’ - efu ‘whites’ languages than the Left Bank. There are five daughter lorn ‘snake’ - morŋn ‘snakes’ languages in Nyo: Agnéby, Attié, Avikam–Alladian, Potou–Tano (which accounts for over half of New Attié: Attié is spoken in Côte d’Ivoire by about half a Kwa in terms of both the speakers and geographical million people living in Abidjan, Anyama, Alep, area), and Ga–Dangme. Potou–Tano is composed up Adzope, Affery, Agou, Akoupe, and Yakasse– of six language groups: Ega, Potou, Tano (previously Attobrou. The three main dialects of Attié are called Volta–Comoé), Lelemi (and its related lan- Anaindin (Nindin), Ketin, and Bodin (the most presti- guages), Logba, and Basila–Adele. Westermann clas- gious and populous). Most Attié speakers are bilingual sified Avikam–Alladian, Agnéby, Attié, and Potou in Abbey, Anyin, Baoule, Ebrie, and Jula. under the Lagoon subfamily; and Lelemi (and its related languages), Logba, and Basila Adele as Avikam–Alladian: It is a Western Nyo language spo- Togo–Remnant languages. ken in the Western Ivory Coast. It has two languages: Agnéby: The Agnéby languages, a Central Nyo Avikam and Alladian. Avikam is spoken in the basin of group, are spoken in the Ivory Coast in the Agnéby the Bandama River by 12,000 speakers, whereas and Bandama river basins by 230,000 people. Alladian is spoken by about 15,000 people on the east Adioukrou, Abiji, and Abbey are the three languages of Avikam. in this group. Adioukrou (Adjukru) is spoken in the north of the Ebrié Lagoon westward of the Agnéby Potou–Tano: The six language groups identified under river by about 70,000 people. Abidji is spoken by Potou–Tano are Ega, Potou, Tano, Lelemi (and its about 80,000 people in the Abidjan District and in the related languages), Logba, and Basila–Adele. subdistricts of Sikensi and Dabou. Abidji has two main dialects: Enyembe and Ogbru. Speakers of Abiji Ega: Ega was originally classified as Kru, but tend to be bilingual in Adioukrou, Baoule, and Jula. Richards (1982) proved that it is Kwa. It is located Important linguistic features include tone and inside the Kru area. Like other Kwa languages, Ega nasalization on syllables. Abbey, spoken by about has a reflex of the cognate (words maintained by lan- 80,000 people, is spoken around Agboville and to the guages of common ancestry) nyɔ /υ ˜ / ‘two’. north of the Abiji and Krobou areas. Examples of words and their plurals taken from Adioukrou are as Potou: It is spoken in the Eastern Ivory Coast Lagoon follows: area by about 75,000 people. Potou has two languages: 28  AKAN AND NYO LANGUAGES Ebiré (spoken by 60,000 people in the north of the Ebiré (single sounds with two different places of articulation, Lagoon from the Agnéby River to the Potou Lagoon in e.g. /kp, tp, gb/ . For example, in Chakosi, the words the east) and Mbatto (spoken by about 15,000 between kpiε ‘old man’ and gblaki ‘to faint’ have /kp/ and /gb/, the Potou Lagoon and the Comoé River). respectively). Bia languages also have tone terracing, where pitch is lowered toward the word end. Plural is Tano: Tano, previously called Volta–Comoé, has more formed through affixation. For example, in Baule, the speakers than any of the subfamilies of the Potou–Tano plural marker {-mu ˜} can be added to any count noun. group. Tano has four subfamilies: Krobou, Western For example, wa ‘child’ wamu ˜ ‘children,’ ti ‘head’ timu ˜ Tano (Abouré and Eotilé), Central Tano (Bia, Akan), ‘heads.’ There are also independent nasalized vowels and Guang (Southern and Northern). Krobou has about (phonemic nasalization of vowels). For example, in 6000 speakers and is spoken within the Agnéby lan- Baule, wu means ‘see’, whereas wu ˜ means ‘die’. guage area in the Ivory Coast. Abouré and Eotilé are also spoken in the Ivory Coast. Abouré is spoken by Akan about 40,000 people in the Comoé River basin near the sea, whereas Eotilé is spoken by 3000 people in Vitre Akan is the largest of the Central Tano group of lan- #I and Vitre #II on the Ebiré Lagoon to the north of guages. Until the 1950s, these clusters of mutually intel- Grand Bassam. ligible varieties did not have a single name. Some of the dialects (such as Fante, Bono, and Wasa) were viewed Bia: The Bia languages, sometimes referred to as as different languages. However, a unified Akan orthog- Nzema–Anyi–Baule, are spoken by nearly two and a raphy was designed by the Akan Orthography half million people in the Bia, Tano, Nzi, Bandama, Committee in the 1950s and since then, the name Akan and Comoé River basins and along the south coast of has been adopted as the name of the language. Ghana and the Ivory Coast. The two main subgroups Akan is spoken by over ten million people in Ghana are: Northern Bia (comprising Anyi, Baule, Chakosi and in the Ivory Coast. In Ghana, Akan is spoken (Anufo)); and Southern Bia (Nzema and Ahanta). between the rivers Tano and Volta, stretching from the Anyi and Nzema are spoken in Ghana and in the Ivory coast to the inland. Abron (Bono) or Dormaa is spoken Coast, Baule in the Ivory Coast, Ahanta in Ghana, and in Ghana and in Côte d’Ivoire by about a million peo- Chakosi in Ghana and Togo. Nzema has two dialects: ple. In Côte d’Ivoire, Akan (Abron) is spoken in the Nzema and Evalue. Anyi also has two dialect groups: northwest by about 60,000 people. Speakers of Abron Anyi (Aowin) and Sehwi (Sanvi). Anyi has other speak and understand Asante Twi. Wasa is spoken by dialects like Indene, Moronou, Bona, Djoablin, Ano, nearly 300,000 people in Southwestern Ghana. It has and Bini (all in the Ivory Coast), and Afema and two main dialects called Amenfi and Fianse. Wasa and Brossa in Ghana. Baule, with nearly one and a half Abron are mutually intelligible. million speakers, is the largest language group in the In Ghana, Akan is spoken by nearly half of the pop- Bia subfamily. Chakosi is found in an enclave in the ulation as their mother tongue and by two-thirds of the Gur territory on both sides of the Ghana–Togo border. population as a second language. Akan has 11 dialects: In Togo, Chakosi is found in Sansanne–Mango and in Fante (Mfantse), Akuapem, Asante, Agona, Dankyira, Ghana, in the Chereponi area. Most speakers of Asen, Akyem, Kwawu, Ahafo, Bono (Abron), and Nzema (about 120,000) are found in Ghana; in the Wasa. The Akuapem, Fante, and Asante dialects have Ivory Coast, there are about 50,000 speakers. different officially recognized orthographies and are Important linguistic features in the Bia group include used on radio and TV. Quite recently, a unified orthog- consonant mutation (in which consonants change some raphy for all the Akan languages has been adopted. of their phonetic qualities in certain phonetic or gram- The Twi (Asante–Akyem–Kwawu) dialects are the matical contexts. For example, in Nzema /k/ changes to largest dialects, with a total population of over seven /h/ between vowels. /k/ in kɔ ‘go’ becomes /h/ in jihɔ million. Fante, spoken along the coast, has over two ‘she’s gone’) and vowel harmony (cooccurrence restric- million speakers and four important dialects: Agona, tions in the distribution of vowels, i.e. the vowels are Gomua, Abura, and Anomabu. Akuapem is spoken on grouped into two sets and in words of more than one the Akwapim Hills between the Akyem and Ga areas. syllable; only vowels from one set occur. For example, Important linguistic features include a Subject– in Nzema, words otosu mean ‘he still cries’ and ɔtɔkυ Verb–Object word order, example: ‘he still fights’; all the vowels in otosu are produced with an advanced tongue root position, whereas those in S V O ɔtɔkυ are produced with a retracted tongue root). Other Kofi nom nkwan ‘Kofi drinks soup’ linguistic features include double articulated sounds Ama dɔ Kwame ‘Ama loves Kwame’ 29 AKAN AND NYO LANGUAGES Akan languages also have two types of vowel har- of Benin is referred to as Chombulon or Tshummbuli mony: cross-height and rounding. With respect to (similar to a Ghanaian Central Guang variety called cross-height harmony, we observe that the vowels are Nchumburu). grouped into two sets: advanced /i, e, æ, o, u/ and Southern Guang subdivides into Coastal Guang (a unadvanced /¹, ε, a, ɔ, υ/. In any word of two or more dialect cluster—Awutu, Efutu, and Senya). Awutu is syllables, only vowels of one harmonic set occur. For spoken in Winneba, Senya Breku, Obutu, and example, obewu, ‘She’ll die’, has vowels from the Bawjiase in Ghana, spoken by about 200,000 people, advanced set, whereas ɔbεwυ, ‘She’ll give birth’, has and Hill Guang is spoken by about 150,000 people on vowels from the unadvanced set of vowels. With the Akwapim Hills and on the left bank of the Volta respect to rounding harmony, in some of the Akan lan- Lake. Important dialects include Cherepong, Larteh, guages, especially in Fante, vowels of verbal affixes and Anum (Gwa). agree with those of a verb root by being advanced or Other Guang subfamilies include Lelemi-Lefana unadvanced, rounded or unrounded. For example, (Buem), Akpafu (Siwu), and Santrokofi spoken in muruwu, ‘I’m dying’, has only rounded advanced Ghana and Togo by about 45,000 people; Logba spo- vowels, whereas in miridzi, ‘I’m eating’, all the vow- ken by about 4,000 people on the Ghana–Togo border; els are both advanced and unrounded. Rounding har- and Basila–Adele, comprising the Basila and Adele mony is also found between vowels of nominal roots dialects, is spoken by about 10,000 in both Ghana and and affixes in the Asante and Akyem dialects. For Togo. Basila has three subdialects: Gokolodjya, example, the vowel of the noun suffix /o/ in owuo Gilempla, and Giseda. Adele has more speakers in ‘death’ agrees in rounding with the vowel /u/ of the Togo (7,000) than in Ghana (3,000). noun root, owu ‘die’. The vowel /e/ suffix of the noun Like the Akan group of languages, the Guang lan- root esie ‘hill’ harmonizes with those of the root /e, i/ guages have advanced tongue root vowel harmony. For by being unrounded. example, in the following Efutu words, all the vowels The Akan languages have uncommon sounds like are either advanced oi ‘man’or unadvanced ɔ`kυ´ɔ` the alveolo-palatal semivowel /ɥ/, labio palatalized ‘again’. The harmony may operate over entire sen- alveolo-palatal affricate /t ɥ/, occurring in words like tences. For example, muudi nu ‘She is eating (food)’; we /ɥi/’chew’ and Twi /t ɥi:/ ‘name of some Akan mυ´n`wɔ´ ‘She did not go’. dialects’. Negation is formed through affixation and conso- Another important feature of the Akan languages is nant alternation. For example, in Siwu, in negating tone terracing (especially downdrifting, where, during commands, voiceless stops change into voiced stops the production of long utterances, there is a gradual and voiced stops become nasals. Also, the morphemic drop in pitch heights if high tones are interspersed particle daa is preposed to the main verb. Example, with low tones. The drop in pitch could be such that an kε`lε` ‘go’ dàáε´ lε` ‘don’t go’; kpε` ‘weed’; dàábε´ utterance final high tone could be lower in pitch than ‘don’t weed’. an utterance initial low tone. The final drop in pitch could also reach the bottom of a speaker’s pitch range. Ga–Dangme: Ga–Dangme has two languages: Ga and For example, in: Pàpá Kòfí dìdí wìé kòtìé ‘Father/Mr. Dangme. The Ga–Dangme subfamily is spoken in Kofi finishes eating and goes to listen’, the initial low Ghana on the southeastern coast around Accra and tone on /pà/ of Pàpá is higher in pitch than the final inland. Ga is spoken in Ghana along the coast in and high tone on /é/ of kòtìé. around Accra (Ghana’s capital), and in the Densu and Kpone basins by about 500,000 speakers. Dangme Guang: The Guang subfamily, which is spoken by (Adangme) is spoken along the coast (to the east of about half a million people in Ghana, The Ivory Ga) and at the mouth of the Volta and also in the Coast, Togo, and Benin, has more languages than the Akwapim Hills. Like Ga, Dangme has over half a mil- Central Comoé group. However, the languages in this lion speakers. Dangme has six dialects: Ada, Ningo, subfamily are for the most part spoken by small num- Prampram, Shai, Krobo, and Osudoku. Krobo is the bers of people. This group of languages subclassify largest dialect, with about 300,000 speakers. into: North, North-East, Central, and South. North Structurally, Ga has two tones, High and Low, and a Guang languages include Gonja, Nkonya, Nawuri, downstepped high tone (a high tone whose pitch has and Choroba. Gonja (Ngbanito) is spoken in a large been lowered. Downstep is marked with an exclama- area from Bole to Kamabuie near Salago and between tion mark placed before the syllable). For example, the White Volta, Volta, and Daka rivers in Ghana. In ékòmé ‘one’, gbé!kε´ ‘boy’. Plurals are formed through the Ivory Coast, Gonja is spoken in the Bondoukou affixation, (suffixation). For example, tsò ‘tree’, tsèi and in Togo it is spoken at Semere and is called ‘trees’; tʃu`‘ house’ tʃu` i ‘houses’. There are three Bazantcé. The Gonja variety spoken in the Republic tones, High, Mid, and Low, in Dangme. 30  ALBANIAN References Stewart, John. 1989. Kwa. The Niger–Congo languages, ed. by John Bendor-Samuel. New York: University Press of Dakubu, Mary Esther Kropp. 1988. The languages of Ghana. America. London: Kegan Paul International. Greenberg, Joseph. 1955. Studies in African linguistic classifi- SAMUEL OBENG cation. New Haven: Compass Publishing Company. See also Niger-Congo Albanian Albanian is the name English speakers use to refer to the two million in Kosovo, some 300,000 in Macedonia, language of Albania (Albanians call the language and at least 700,000 more living mainly in ‘shqipe’), as well as for the branch of the Indo-European Montenegro, Greece, Turkey, Italy, and the United family to which it belongs. As a branch of Indo- States. In recent years, both voluntary and forced emi- European, Albanian is a sister to many language groups gration of Albanians has greatly increased the pres- from Europe to South Asia: Germanic (including ence of speakers of the language outside their English, Swedish, and German, for example), Romance homelands. (including French, Spanish, and Italian), Slavic (includ- During its known history since the fifteenth century ing Russian, Serbian, and Polish), Indic (including CE, Albanian has been associated with two main Hindi, Gujarati, and Urdu), Hellenic (Greek), Armenian, dialects: Tosk in the south and Gheg in the north. After and Celtic (including Irish and Scottish Gaelic). World War II, the government made increasingly suc- Persuasive evidence for the kinship of Albanian with cessful efforts in Albania to adopt a unified written other Indo-European languages was first advanced in the form of the language, called Standard Literary middle of the nineteenth century and continues to be Albanian, and in 1968 this form of Albanian was also amassed. In addition to regular correspondence of adopted for schools and mass media by Albanians in sounds between particular words and parts of words in Kosovo. However, for ordinary conversation, few Albanian with those in several already established Indo- Albanians in Albania and Kosovo completely aban- European languages, the evidence lies in features of doned their own dialects. In rural areas, ‘unified Albanian verb conjugation and noun declension that Albanian’ may be scarcely used and known only from reflect a common origin with those languages: for exam- television and radio, and even well-educated, urban ple, the irregular forms of certain verbs (e.g. Albanian Albanian speakers exhibit regional and social differ- ështëEnglish isLatin estSanskrit asti, while ences in their language, which appear in pronuncia- Albanian jamEnglish amLatin sumSanskrit asmi). tion, vocabulary, and grammar. The speech of Albanian linguists claim that the predecessor of Albanian speakers living outside Albania reflects their modern Albanian is the ancient Illyrian language, but Gheg or Tosk origins. since little is left of Illyrian, verification for this claim is A modern alphabet was devised for Albanian in elusive. The earliest texts clearly identifiable as 1908. This 36-letter alphabet includes six digraph let- Albanian are scarce and go back only to the fifteenth ters (th, dh, sh, zh, gj, nj) and two special characters (ë century; we have no record of the characteristics of the and ç), and was adopted quite readily by speakers. For language during earlier periods, although extensive doc- Gheg words, a carat (^) is used to mark nasalized vow- umentary evidence exists for many other Indo- els, but the vowel length so distinctive to Gheg dialects European languages. To complicate the task of tracing remains unmarked in written Albanian. its origins, Albanian, like English, has adopted and In addition to its genetic membership in the Indo- adapted a great substance of the vocabulary from other European language family, Albanian shares certain languages throughout its history—in earlier times, basic specific features with other Indo-European languages vocabulary from Latin and Turkish and more recently, in the Balkan area. For example, like Romanian, technical vocabulary from Italian, Russian, and English. Albanian marks ‘definiteness’ (in English, marked Today, Albanian is spoken by more than six million by the word the) by endings on nouns: Albanian people—about three million in Albania itself, another gjysh, ‘grandfather’ becomes gjyshi, ‘the grandfather’; 31 ALBANIAN similarly, Romanian bunic, ‘grandfather’ becomes Another set of endings indicate whether the subject bunicul, ‘the grandfather’. is itself the object of the verb. These ‘medio-passive’ Particular features of Albanian seem particularly endings are used for a range of functions that are quite striking to non-Albanians. Already mentioned is the different in English. For example, e rruan means ‘they vast number of loan words from the international shave him’, but rruhen can mean, ‘they are shaved’ vocabulary—such as heroik, burokrat, standardi- (passive), ‘they shave themselves’ (reflexive), or ‘they zon—as well as many words from Turkish (during the shave one another’ (reciprocal). Still other sets of verb 400-year occupation of Albania under the Ottoman endings can express the speaker’s surprise or dismay, Empire), such as penxhere (window), fukara (poor or and yet others may express a wish. pauper), and çoban (shepherd). Foreign names are Pronouns also take different forms for varying generally spelled in Albanian to approximate their grammatical functions. For example, as subjects of sound—to an Albanian ear—in the language from verbs, objects of verbs, and objects of certain preposi- which they derive, so that Heather Lockwood would tions, first- and second-person pronouns have the become Hedhër Lakvud. accented forms shown in Table 1. As in many other languages, Albanian indicates the Albanian lacks corresponding forms for third- grammatical function, or case, of a noun with noun person pronouns, but instead uses a complex system of endings. Thus, gjyshi, ‘the grandfather’, is the subject ‘pointing’ words that convey the perceived distance of of a verb, but as the direct object of a verb it takes the the referent, as well as the referent’s number, gender, form gjyshin, and as the indirect object of the verb, it and case (see Table 2). takes the form gjyshit. Unaccented pronoun forms indicate that the verbs The verb system in Albanian is quite complex. immediately following have an object (direct or Simple verbs not only have different endings to indi- indirect) (see Table 3). cate whether their subject is the person speaking (‘first These forms appear whether or not an accented person’), the person spoken to (‘second person’), or object is also present in the sentence, so that ‘Agim someone or something else (‘third person’), but also told me’ is më tha Agimi, and ‘Agim told me’ is më tha whether that subject includes the person speaking Agimi mua; similarly, ‘Agim told her’ is i tha Agimi, (‘first person plural’), the person addressed (‘second and ‘Agim told her’ is i tha Agimi asaj. Unaccented person plural’), or is entirely composed of others third-person direct objects may occur in combination (‘third person plural’). Those endings also indicate with any unaccented indirect object, yielding the whether the time is present or past; if past, the ending forms shown in Table 4. indicates whether the verb applies over time or at a The position of the unaccented pronoun (or combi- particular moment. nation) in Albanian is unique among Indo-European TABLE 1 English Subject Direct Indirect Prepositional equivalent object object object First person Singular I, me unë mua mua meje Plural we, us ne ne neve nesh Second person Singular you (thou) ti ty ty teje Plural you all (you) ju ju juve jush TABLE 2 English equivalent Subject Direct Indirect Prepositional object object object Third person Far Near Far Near Far Near Far Near Singular Masculine that one: he, it, him ai ky atë këtë atij këtij atij këtij Feminine that one: she, it, her ajo kjo atë këtë asaj kësaj asaj kësaj Plural Masculine they, them ata ky ata këta atyre këtyre asosh kësosh Feminine they, them ato kjo ato këto atyre këtyre asosh kësosh 32  ALBANIAN TABLE 3 much like English of to indicate possession—djali i Person Number English Direct Indirect Gjonit (the son of John, or John’s son)—but more equivalent object object often to indicate some looser relationship in which the noun after the attributive article narrows the scope of First Singular me më më its referent. For example, such a construction may be Plural us na na used to identify: Second Singular you (thou) të të ● the substance of which something is made: buka Plural you all (you) ju ju e grurit (the wheat bread, or literally, ‘the bread Third of the wheat’) Singular him, her, it e i ● a quality that characterizes the referent: aktet e Plural them, ‘em i u trimërisë (acts of bravery) ● the object of an action: shkelja e rregullit (the TABLE 4 violation of the rule) ● the subject of an action: ardhja e miqve (the Indirect Direct Combination Example Meaning arrival of friends) object object ● the whole of which the referent is part: gjysma e më e = ma ma dha Agim gave bukës (half of the loaf) Agimi it to me ● the identifying element of the referent: muaji i më i = m’i m’i dha Agim gave Agimi them to me korrikut (the month of July) të e = ta ta dha Agim gave ● a characterizing quality after the verb ‘to be’: Agimi it to you është i gojës (he is articulate, or literally, ‘is of të i = t’i t’i dha Agim gave the mouth’) Agimi them to you ● the universe of a superlative: më i madhi i na e = na e na e dha Agim gave it Agimi to us djemve (the biggest of the boys) na i = na i na i dha Agim gave Agimi them to us Among the small national languages of Europe, ju e = jua jua dha Agim gave it Albanian can claim a special pride in the consistency Agimi to you of its standard orthography and grammar, and the rich- ju i = jua jua dha Agim gave Agimi them to you ness of its vocabulary. i e = ia ia dha Agim gave it Agimi to him/her i i = ia ia dha Agim gave References Agimi them to him/her Camaj, Martin. 1984. Albanian grammar. Wiesbaden: Otto u e = ua ua dha Agim gave it Harrassowitz. Agimi to them Hamp, Eric. 1972. Albanian. Linguistics in Western Europe. u i = ua ua dha Agim gave Current Trends in Linguistics, Vol. 9, ed. by Thomas A. Agimi them to them Sebeok. The Hague: Mouton. Hubbard, Phillip L. 1985. The syntax of the Albanian verb com- plex. New York & London: Garland Publishing. languages in one particular aspect: in positive com- Huld, Martin E. 1984. Basic Albanian etymologies. Columbus, mands, it may come directly after the verb, but if the OH: Slavica Publishers. command is addressed to a plural ‘you’, then it comes Newmark, Leonard. 1957. Structural grammar of Albanian. before the verb ending. Bloomington: Indiana University Publications of the Research Center in Anthropology, Folklore, and Linguistics. Finally, Albanian is unusual among languages in its Newmark, Leonard. 1980. Spoken Albanian. Ithaca, NY: use of a set of ‘attributive articles’, which indicate that Spoken Language Services; revised edition, 1997. the word after it refers to a noun or pronoun, its ‘ref- Newmark, Leonard. 1998. Oxford Albanian–English diction- erent’, which may or may not be present in the sen- ary. Oxford: Oxford University Press. tence. The choice of which of the four members of this Newmark, Leonard, Philip Hubbard, and Peter Prifti. 1982. Standard Albanian: a reference grammar for students. set (i, e, të, së) is to be used is determined by the gram- Stanford: Stanford University Press. matical form of the referent. Pure adjectives must Newmark, Leonard, and Vladimir Dervishi. 1999. Albanian always be preceded by one of these articles: këndim i handbook with English and Albanian glossaries. mirë ‘good singing’ with the adjective i mirë ‘good’ Kensington, MD: Dunwoody Press. —contrast the adverb mirë ‘well’ as in këndon mirë Zymberi, Isa. 1991. Colloquial Albanian. London and New York: Routledge. ‘He or she sings well’. LEONARD NEWMARK Any noun with a ‘genitive’ case ending can be pre- ceded by one of these attributive articles, functioning See also Indo-European 1: Overview 33 ALGERIA Algeria Algeria, officially the Democratic and Popular The linguistic situation in Algeria is complicated, Republic of Algeria, is the second largest country in and is a matter of intense discussion within the coun- Africa with 919,595 square miles after Sudan, stretch- try itself. Algeria’s population diversity is reflected in ing from the shores of the Mediterranean to the Sahara the large array of languages spoken. Seventeen lan- Desert. It has a population of 31,736,053 (2001 esti- guages are believed to be spoken in Algeria, Arabic mate) and is divided into 48 provinces called wilayat and, most recently, Tamazight, being the official lan- and 700 local communes. The population consists of guages. The majority of Algerians speak Arabic, fol- Arabs, Berbers (Kabyles being the largest group fol- lowed by French and Berber. Modern Standard Arabic, lowed by Chaouia) and people of mixed Arab and a simpler version of Classical Arabic, is taught at Berber ancestry. Arabs form 80–83%, Berbers schools and is used in formal meetings and in the 16–20%, and others 1%. media, but is not used for ordinary conversation. The Linguistic independence came rather late to Algerian dialect, known as Darja, is used in everyday Algeria. During the Muslim expansion of the seventh life. And increasingly, the vernacular is being used in and eighth centuries, the Arabic language spread theater and in novels because it is believed to reflect from the Arabian Peninsula to North Africa, sup- the Algerian culture. planting the local languages. During the Ottoman After gaining independence, Algeria proceeded to Empire, the Arabic language was suppressed, as else- ‘arabize’ education in order to rid the country of where in the Arab World. Later, the French conquest French influence. However, the fact that the country of Algeria in 1830 further complicated the linguistic was trying to rediscover its roots while forcing its pop- situation. The authorities introduced French as the ulation into ‘modernity’ has pushed it to maintain sole official language. Their main goal was to replace French as a language of science and technology. Thus the Arabic culture by a more ‘civilized’ culture. The French became associated with high social and educa- use of Arabic was generally limited to a small minor- tional status. Arabization on the other hand was pur- ity of religious scholars, and to the rural regions, sued for ideological reasons related to nationalism and where Arabic religious schools were set up to teach religion. the Qur’an. Despite the efforts to introduce Arabization to rid After gaining its independence from France in the country of the cultural traces left by the long 1962, the Algerian government declared Arabic as the French colonialism, Algerians continue to use French official language, which was a difficult task as very for formal and informal conversations. In fact, French few Algerians were actually literate in Arabic. is considered by many the ‘unofficial’ official lan- Teachers were imported from Arab countries such as guage, as it is used in most formal administrative Egypt and Syria. And in the 1970s, the Arabization meetings, gatherings, and various other functions. policy required that street and shop signs be in Arabic, Modern Standard Arabic is used only in formal educa- despite the fact that 60 % of the population could not tion, by the media, and for written purposes. write or read Arabic. Over the years, Algeria struggled The policy that the Algerian government has adopt- between the pressure to eliminate any legacy from the ed toward minority language groups has resulted in French colonialism, to meet the demands of the tension among the different language users. Kabyles to adopt Tamazight as a second official lan- Arabization, the language policy aimed at expanding guage and to teach in schools, and to bear the costs of the role of Modern Standard Arabic to the detriment of rapid Arabization. Arabic is held in high regard French and local dialects and languages, led to dissat- because it is the language of the Qur’an. With the isfaction and resentment among the Berber ethnic spread of Islam, Arabic has become the most used lan- minority and the French educated population. guage among the living Semitic languages. It is relat- Opponents of this policy argued against making ed to Hebrew, Aramaic, and Amharic. Subgroups of Arabic the exclusive language of instruction in public Arabic are Classical Arabic, Eastern Arabic, Western schools when French remains the language of science Arabic, and Maltese. The modernized version of instruction at the universities. Classical Arabic, which is referred to as Modern Most Algerians speak Algerian Arabic, with educat- Standard Arabic, is the language that Arab countries ed and clerical masses speaking Modern Standard use as their official language. Arabic in formal situations. Arabs usually do not 34  ALGERIA speak Berber because it is not taught in schools. Berber Number of Geographical However, most Berbers speak Arabic and French, Language Speakers Areas especially the younger generation. Currently, there is Kabyle Up to 6,000,000 In the Djudjura moun an active movement to regain the Berber language and (1998) tain range, and along teach it in schools. As of recently, Tamazight is being the northern central taught in the Kabyle major cities, but not in the non- and eastern coastal Berber areas, although there are pockets of Berbers in region, east of Algiers Main cities: Tizi- every major city. Ouzou, Dellys, Bejaia Even though the Algerian census does not gather information about household languages, the number of Chaouia 1,400,000 (1993) South and southeast of Berber speakers has been estimated to be as high as the Grand Kabylie 6,000,000 speakers. The Berbers are considered to be region and south of Constantine, in the the original inhabitants of North Africa. They have Aurès Mountains lived in Algeria since 3000 BC, or, according to some Main cities: Batna, historians, as early as 8000 BC. The word Berber Ain el Baidha, Ain comes from the Greeks, who used this term to refer to Mlila the people of North Africa. Although Berber uses an Chenoua 15,000–75,000 Small towns east of ancient script called tifhagh, it is primarily a spoken (1996) Algiers language. Currently, Roman script is used to write the Berber language. Most Berbers, especially the Tachelhit Unknown Southern Algeria near Kabyles, maintain their own language and a sense of the Moroccan border ethnic identity. The Chaouia, however, tend to be more Tahaggart 25,000 (1987) Southern Algeria in assimilated to Arab culture than the Kabyles. The the Hoggar region Kabyles refer to themselves as Imazighen, and to their Major cities: Djanet language as Tamazight. The linguistic division and Tamanrasset between Arabic speakers and Tamazight speakers stems mainly from the government policies that favor Taznatit 40,000 (1995) Around the city of Timimoun the Arabic language over the other local languages such as Tamazight. For years, the Algerian govern- Tumzabit 70,000 (1995) Mzab region, 330 ment suppressed the Berber languages, depriving miles south of Algiers those who did not speak, write, or read Arabic from Main city: Ghardaia basic human rights such as reading the newspaper and watching television. The Berbers, also known as Amazigh in North Because the spoken Algerian Arabic is a simplifica- Africa, are the original inhabitants of North Africa. tion of Modern Standard Arabic, there is less distinc- Because the Berbers have mingled with other ethnic tion regarding number and gender. For example, the groups over the centuries, especially the Arabs, their dual form and the feminine plural do not exist in the identification is purely linguistic and not racial. The Algerian dialect. The Modern Standard Arabic mascu- Berber language (Tamazight) is primarily oral. Its line plural form is used in Algerian Arabic for dual, written form is little known and rarely used. Like feminine and masculine, and for the feminine plural. Arabic, the Berber language is a branch of the Afro- Algerian Arabic consists of three major dialects spoken Asiatic language family. Many languages derive from in three major cities: Algiers (center), Constantine the Berber language family. Some of the Berber lan- (east), and Oran (west). These major dialects differ guages, the number of speakers, and the geographical considerably from the dialects spoken by nomadic peo- areas where they are mostly spoken are given in table. ple in the Sahara. As of 2003, 41 years after independence, French is still taught as a second language and is introduced at Algerian Arabic vs. Modern Standard Arabic the second-grade level. French also remains the lan- guage of science instruction in all universities after a Phonological Differences semi-unsuccessful attempt to arabize the educational Modern Standard Arabic is written from right to left system, including higher education. Attempts to and uses a root system, usually consisting of three substitute English for French as the second national consonants. For example, the English word ‘merge’ language in the mid-to-late 1990s also failed. However, would be made up of the root ‘m-r-g’. In Arabic, since 2003, English is being taught at the seventh-grade the word ‘islam’ is made up of the root ‘s-l-m’. The level, a year earlier than in previous years. Arabic script uses symbols for only consonants and 35 ALGERIA long vowels. The short vowels are not written and are pronunciation. For example, French has nasal vow- represented by diacritics. However, although morpho- els, but Arabic does not. Therefore, the nasal vowels logical and grammatical meaning is indicated by vow- in borrowed French words are denasalized (1). Also, els, the vowel diacritics are usually omitted. The French sounds that do not exist in Arabic are substi- reader has to infer its particular meaning and pronun- tuted by Arabic sounds (2). As the Algerian dialect ciation from the context. relies heavily on emphatism, loan words are no Even though the Algerian dialect is a variant of exception (3). Arabic has three short vowels, a, i, Arabic, its vocabulary differs considerably from and u, and three corresponding long vowels, a:, i:, Modern Standard Arabic. Many words originate from and u:. As a result, French vowels that do not exist sources other than Modern Standard Arabic, such as in Arabic are replaced by one of the three vowels Turkish, Spanish, Italian, and most of all Tamazight (short or long) (4). French influence extends to syntax (Berber) and French. However, most of the vocabulary as well (5). comes from Modern Standard Arabic with phonetic, phonological, and semantic changes. The examples French Algerian Arabic below show Modern Standard Arabic words with (1) bouchon bushu:n (cork) Algerian Arabic correspondences. The elements in (2) il roule yru:li (he wanders) The which the Algerian Arabic words differ from the French r is replaced by the Algerian Arabic ones are indicated in bold. Arabic r. une serviette serbita (towel) The sound v is Modern Standard Arabic Algerian Arabic replaced by b. Assimilation: une savate Sabba:t (a shoe) taji:’u dji (she comes) (3) il sonne ySu:ni (he rings) The French s tazu:ru dzu:r (she visits) is replaced by the emphatic S. tadu:mu ddu:m (it lasts—fem.) une place bla:Sa (a place) Sadr zdar (chest) (4) un bureau bi:ru (a desk, an office) The niSf nuSS (half) French ü is replaced by i:. Metathesis: (5) une table Tabla (a table) The Arabic yartaεid yatrεid (he shivers) feminine indicator a is added. laεana nεal (he cursed) The linguistic situation in Algeria depends on the Dissimilation: future political situation. The Algerian people believe finja:n finja:l that if the Muslim fundamentalists come to power, the ba:dhanja:n badhanja:l trend will be toward arabization of all sectors of life; on the other hand, if democratization is successful, Phone Substitution: bilingualism will be stressed with the French language shajara sajra (a tree) used for the technical and hard sciences, and Arabic Monophthongization: for social sciences and civil services. After years of zawj zu:dj (two) political unrest, the political and cultural situations are Sayf Si:f (summer) far from being resolved. While a solution is being zayt zi:t (oil) sought for the linguistic conflict, Modern Standard Arabic remains a foreign language for some of the Semantic Changes: older Algerians, especially the Berbers. However, yudi:ru ydi:r recent attempts to make Tamazight a second national (he directs) (he does) language have been somewhat successful. In 2002, ‘ajlis ‘uqεud President Bouteflika announced that the Tamazight (sit down—when someone is standing up) language would become an official language in ‘uqεud tgaεεad Algeria alongside Arabic. (sit down—when someone is lying down) References The French Influence Benremouga, Karima. 1987. A survey of a typical eastern dialect of Algerian Arabic: phonology and morphology. The Algerian dialect contains a good number of Master’s Thesis. Lawrence, KS: University of Kansas. borrowed French words. However, the borrowed Benremouga, Karima. 1997. Spoken Algerian ‘Arabic’: vocabulary is subject to Arabic rules of grammar and An overview. [Ed. by In Xingzhong Li, Luis, López, 36  ALTAIC and Tom, Stroik, Papers from the 1997 Mid-America Saad, Zohra. 1993. Language planning and policy attitudes: a Linguistics Conference.] Columbia, MO: University of case study of Arabization in Algeria. Doctoral Dissertation. Missouri. New York: Columbia University. Grimes, Barbara F. (ed.) 1992. Ethnologue, languages of the Talmoudi, Fathi. 1984. The diglossic situation in North Africa: world. Dallas, TX: Summer Institute of Linguistics. a study of classical Arabic/dialectal Arabic diglossia with Harik, Elsa M., and Donald G. Schilling. 1984. The political of sample text in mixed Arabic. Göteborg, Sweden: ACTA education in colonial Algeria and Kenya. Athens, OH: Ohio Universitas Gothoburgensis. University Center for International Studies. KARIMA BENREMOUGA Ruhlen, M. 1987. A guide to the world’s languages. Vol. 1: clas- sification. London: Edward Arnold. See also Arabic; French Language Altaic Altaic, in its broadest conception, is a family of lan- lion people worldwide. The second branch, called guages composed of four subfamilies (Japanese- Okinawan or Ryukyuan, consists of a set of ten lan- Okinawan, Mongolic, Tungusic, and Turkic) and two guages, most of which are mutually intelligible to language isolates (Ainu and Korean). However, the some degree. There are roughly 900,000 speakers status of this family remains a matter of controversy. of Okinawan languages, although very few of the On the one hand, some scholars who accept the genet- languages are being transmitted to children at this ic association of Mongolic, Tungusic, and Turkic point. Japanese and the Okinawan languages share (hereafter, Altaic proper) do not agree that there is a roughly 70% of their vocabulary; they are not mutual- demonstrable relationship between Altaic proper and ly intelligible. Japanese-Okinawan, Korean, or Ainu. On the other The earliest written records of Japanese date back hand, some scholars question even the notion of Altaic to the early eighth century and are composed in proper. The skepticism often rests on purely method- Chinese characters. Since then, two different syllabic ological grounds; that is, although they accept the like- writing systems derived from Chinese characters have lihood that Mongolic, Tungusic, and Turkic languages been developed for Japanese: Hiragana and Katakana. all stem from a common ancestor, some linguists Both are syllabic; i.e. each symbol stands for a sylla- doubt that enough evidence exists to prove this con- ble. Additionally, there is now a phonetic writing sys- nection conclusively. In other instances, linguists tem, Romaji, which uses the Roman alphabet. Here, reject even the likelihood of Altaic proper; they argue the symbols stand for individual sounds. that the similarities among these language families are Japanese is unlike other Altaic languages in that it better explained by borrowing and contact-induced possesses a very simple syllable structure; most syllables changes than by a genetic association. consist of just a vowel or a consonant followed by a Ainu is spoken on Hokkaido, the northernmost vowel. Also, somewhat unusually among the Altaic lan- island of Japan, and on Sakhalin Island, which is part guages, Japanese has only five vowels and a small num- of the Russian Federation. Ainu is spoken only as a sec- ber of consonants. As in other Altaic languages, the basic ond language, primarily by older people. Only about word order is subject–object–verb, and it uses suffixes to 1% of the 18,000 to 25,000 ethnic Ainus claim an abil- indicate the grammatical function of a noun, i.e. whether ity to speak the language, and the number of fluent the noun is the subject (-ga) or an object (-o, -ni). speakers is probably a fraction of this small number. Perhaps the most careful attempt to link Japanese to Ainu, as is typical of Altaic languages, has a basic Altaic proper and to Korean was that presented by Roy word order of subject–object–verb, but the connection Andrew Miller (1971). Despite his efforts, few schol- with other Altaic languages is tenuous at best. Such a ars readily accept the Japanese connection with Altaic, connection was proposed by John Street (1962) and although the affiliation between Japanese and Korean was argued more forcefully by James Patrie (1982), has gained a wider consensus. but few accept Ainu as being part of the Altaic family. Korean is spoken by 65 million people worldwide, Japanese-Okinawan, a subfamily of Altaic, consists the vast majority of them living in North and South of two branches. The first branch consists solely of Korea. The earliest examples of written Korean used Japanese, which is spoken by approximately 130 mil- Chinese characters, a practice that continued into the 37 ALTAIC twentieth century. However, a Korean alphabet, called Like the other Altaic proper families, Mongolic lan- Hangul, was invented in the fifteenth century. Hangul guages typically use vowel harmony. The basic word is now the primary means by which Korean is written. order tends to be subject–object–verb, the grammati- Chinese has had an obvious effect on Korean. In cal function of nouns is signaled by suffixes, and there fact, more than half the Korean words have been bor- are no definite articles (“the”). rowed from Chinese. Even so, the structure of Korean Tungusic languages are spoken over a geographic remains quite distinct from Chinese, and superficially expanse that goes all the way from the Arctic regions at least, it shares many properties with Altaic proper. of the Russian Federation to China’s Xinjiang For example, the basic constituent order is subject– Province. Most of these languages were traditionally object–verb, it has no definite articles, and it exhibits spoken by small nomadic groups. As nomadism has ‘vowel harmony’; i.e. vowels within a word tend to come to an end in China and Russia, Tungusic speak- assimilate to one another. ers have Russified and Sinicized rather rapidly, and A family relationship between Korean and Altaic most Tungusic languages will probably start to die out proper was advanced by two leading Altaicists in the in the near future. twentieth century: G.J. Ramstedt (1952–1957) and Tungusic has traditionally been divided into two Nicolas Poppe (1965). The connection between Korean branches: Northern Tungusic and Southern Tungusic. and Japanese has received more attention in the inter- The best-known member of the Southern branch is vening years, sometimes being examined within the Manchu, which was the official language of China’s context of Altaic (for example, by Roy Andrew Miller last dynastic empire. Although the language was used [1972]) and sometimes independently of Altaic (for for official purposes well into the twentieth century, example, by Samuel Martin [1966, 1991]). very few fluent speakers remain. Sibe, another Mongolic languages are spoken in the central Asian Southern Tungusic language, which, in fact, might be regions of the People’s Republic of China, the Russian better labeled a dialect of Manchu, has 27,000 speakers Federation, and the Republic of Mongolia. Although in Xinjiang Province, making it the largest Tungusic there has been much debate regarding the internal language. The Northern branch of Tungusic consists of structure of the family, it is perhaps most often depict- Even and a cluster of very closely related languages— ed as having two primary branches. Western Mongolic Evenki, Negidal, Solon, and Orochen. Many scholars consists of a single language called Moghul. The lan- suggest that some of the Southern languages should be guage, which is mutually unintelligible with all other set off into a third branch of the family, Central Mongolic languages, is spoken by fewer than 200 peo- Tungusic. The Central Tungusic languages, for exam- ple found in two villages in Afghanistan. Eastern ple, Nanai, tend to be similar to Southern Tungusic in Mongolic is broken down further into three sub- their pronunciation but similar to Northern Tungusic in branches: Dagur (which consists solely of the Dagur the ways in which words are formed. language, spoken by around 60,000 people in China), Only two Tungusic languages have written records Monghuor (a set of closely related languages spoken that extend back before the twentieth century: Manchu in China, all of which have been deeply influenced by and Jurchen. The Jurchen data date to the twelfth cen- Chinese and, in some cases, Tibetan languages), and tury, but its writing system mixes logographic sym- Oirat-Khalka (the subbranch containing most bols, which stand for concepts, with symbols for Mongolic languages). pronunciation, and it is still not completely deciphered. Within the Oirat-Khalka branch, the three most The Manchu script was derived from Mongolic script widely spoken languages are Oirat, spoken by approx- at the end of the sixteenth century. Today, we have a imately 300,000 people in the Kalmyk Autonomous huge body of Manchu literature spanning the seven- Region in Russia; Buriat, spoken by 350,000 people in teenth to the twentieth centuries. the Buriat-Mongol Autonomous Region in Russia; and Tungusic languages are characteristically Altaic in Khalka, the official language of the Mongolian having vowel harmony, case suffixes, and possessive Republic, with more than 2.3 million speakers. It is suffixes. Also in line with other Altaic languages, they this last language that most people refer to when they lack definite articles and have a weak distinction talk about Mongolian. between nouns and adjectives. The oldest written records of Mongolian date back The Turkic language family covers a large area from to the early thirteenth century and take the form of the Balkans and Asia Minor in the southeast to the inscriptions, but the literary text known as The Secret Central Asian regions of both the Russian Federation History of the Mongols had appeared already in 1240. and the People’s Republic of China and up into The script used for this early writing was borrowed the Arctic region of Russia. Turkic is split into two from the Uighurs, a neighboring Turkic group that had main branches: Bulgar and Common Turkic. The main been using the script for many centuries. language in the Bulgar branch is Chuvash, which is 38  ALTAIC spoken by about 1.5 million people in the Chuvash Certainly, there is plenty of evidence that Altaic lan- Autonomous Region of Russia. Chuvash, although guages have been in contact for many centuries, and grammatically similar to the rest of the Turkic lan- everyone involved in the Altaic debate agrees that guages, is marked by a series of historical changes in much borrowing has taken place. The point of con- pronunciation that are hard to interpret in the context of tention is whether these cognates should be dismissed Turkic and Altaic. Therefore, some scholars have sug- en masse as borrowings or whether a subset of the cog- gested that Chuvash is actually a sister to proto-Turkic nates does in fact reflect a family association. rather than a daughter. That is, instead of a branch of Adding fuel to the notion that putative Altaic cog- the Turkic family per se, Chuvash and the Turkic fam- nates are not good indicators of a family relationship ily may both be branches of a larger family unit. is the fact that among the basic vocabulary in the lan- Common Turkic is generally divided into four or guages, for example, the lower numerals, body parts, five subbranches, most of which have languages spo- and natural phenomena, the cognate evidence is not ken by more than a million people. The Eastern sub- compelling. Typically, one expects better cognate evi- branch includes, among other languages, Uzbek (the dence in these vocabulary realms. national language of Uzbekistan, 15 million speakers) One of the strongest pieces of evidence in favor of and Uighur (6.7 million speakers, mostly in China). Altaic, or at least Altaic proper, is the first-person sin- The Southern subbranch includes Turkish (50 million gular pronoun (‘I’). In Tungusic languages, the nomina- speakers), Azeri (14 million speakers, half in Iran and tive form of the pronoun is bi, but for all other cases, the half in Azerbaijan), and Turkmen (3 million speakers). stem is min-; e.g. Oroqen min-ŋi, ‘my’, is the genitive The Northern subbranch includes Tuvan (200,000 (possessive) first-person singular. Precisely the same speakers, mostly in Russian but also in Mongolia) and contrast in stems is found widely in Mongolic. Hence, Yakut (300,000 speakers). The Western subbranch Buriat has bi, ‘I’, for the nominative first-person singu- constitutes Bashkir (1 million speakers) and Tatar (5.5 lar and min-ni, ‘my’, for the genitive. Although the cor- million speakers). Many scholars separate out of respondences in Turkic are not perfect, one does find Western Turkic a fifth subbranch called Central similar patterns in some modern Turkic languages (e.g. Turkic. Within Central Turkic, one finds the national Chuvash), and Old Turkic allowed variability in the languages of Kazakhstan, Kazakh (7.6 million speak- pronominal stems; this, if not fully corresponding to the ers), and Kirghiz (2 million speakers). Tungusic and Mongolic patterns, shows enough simi- The earliest written record of a Turkic language was larity to suggest a connection among all three families. found in a set of inscriptions in Mongolia dating to the early eighth century. These inscriptions were written References in Orkhon Turkic and use what is called a runiform Binnick, Robert I. 1987. On the classification of the Mongolian script. Some scholars have argued that the script is an languages. Central Asiatic Journal 42. 178–95. invention of Orkhon Turkic speakers, although there is Johanson, Lars, and Éva Ágnes Csató, (eds.) 1998. The Turkic growing consensus that the script is actually of languages. London and New York: Routledge. Semitic origin. Kornfilt, Jaklin. 1987. Turkish and the Turkic languages. The world’s major languages, ed. by Bernard Comrie. New York Turkic languages are characterized by vowel har- and Oxford: Oxford University Press. mony. Typically, Turkic languages manifest a sub- Martin, Samuel E. 1966. Lexical evidence relating Korean to ject–object–verb word order and heavy use of suffixes. Japanese. Language 42. 185–251. Nouns and adjectives are only weakly distinguished. Martin, Samuel E. 1991. Morphological clues to the relation- Speculation about some kind of genetic relationship ships of Japanese and Korean. Patterns of change, change of patterns: linguistic change and reconstruction methodology, among Mongolic, Tungusic, and Turkic stems back to ed. by Philip Baldi. New York: Mouton de Gruyter. the eighteenth century, and suggestions about the link Miller, Roy Andrew. 1971. Japanese and other Altaic languages. between these languages and Korean and Japanese Chicago: University of Chicago. arose in the early nineteenth century. Despite a steady Patrie, James. 1982. The genetic relationship of the Ainu lan- stream of scholarship with respect to the classification guage. Honolulu: University of Hawaii Press. Poppe, Nicholas. 1965. Introduction to Altaic linguistics. of these languages over the past 200 years, there is still Wiesbaden: Otto Harrasowitz. widespread disagreement about the merits of an Altaic Ramstedt, G.J. [1952–1957.] Einführung in die altaische language family and, even granting its existence, Sprachwissenschaft (Introduction to Altaic linguistics), which subfamilies should be contained therein. 2 vols. Helsinki. With respect to Altaic, no one denies that there are Street, John 1962. Review of Vergleichende Grammatik der altaischen Sprachen by Nicholas Poppe. Language 38. 92–8. a host of putative cognates, i.e. words that are quite LINDSAY WHALEY obviously related. The issue is whether these cognates reflect shared ancestry or whether they have arisen See also Ainu; Chinese (Mandarin); Japanese; through extensive borrowing among the languages. Korean; Tungusic 39 AMBIGUITY Ambiguity ‘Ambiguity’ is the term for the existence of at least two the telescope, the prepositional phrase with the tele- separate, incompatible interpretations of a stretch of scope modifies either the lady (thus, the lady is a lady speech. Two kinds of ambiguity are recognized: if it is with a telescope) or observes (thus, the policeman due to words, it is called ‘lexical’ ambiguity; if it is due observes by means of a telescope). The ambiguity of to sentence structure, it is called ‘structural’ ambiguity. the sentence is an attachment ambiguity—the preposi- Lexical ambiguity can be either ‘polysemy’ or tional phrase may be ‘attached’ to or associated with ‘homonymy’. Homonyms are distinct units in the different elements in the sentence. vocabulary with identical pronunciation (‘homo- Scope ambiguity refers to the possibility of assum- phones’) and/or spelling (‘homographs’). Each of the ing different logical forms of a sentence. An example homonyms in a set has its own meaning, for example, is the sentence Every man loves a woman, which has bank ‘financial institution’ vs. bank ‘slope’. In this two distinct readings: for each man there is ‘his’ example, homophony and homography co-occur. But woman, and he loves her; or alternatively there is a this need not always be the case; compare peak [pi:k] specific woman who is loved by all the men. With the ‘summit’ and peek [pi:k] ‘glance’, which are homo- first reading, every man ‘has scope over’ a woman, i.e. phones but not homographs. From a historical point of the sentence is primarily about ‘every man’. With the view, homonyms cannot be traced back to a common second reading, it is the other way round, i.e. the sen- etymological origin, as with bat: the word for the fly- tence is primarily about ‘a woman’. ing mammal has a Swedish origin, the word for the Ambiguity, whether lexical or structural, must be implement is related to battle. From the contemporary distinguished from vagueness. A vague expression is viewpoint, homonyms have no shared meaning either. imprecise, whereas an ambiguous expression has sev- These two viewpoints are not watertight, however. A eral precise meanings. An example of a vague expres- very competent speaker knows more about the mean- sion is the predicate red in Mary owns a red skirt. A ing of a given pair of homonyms than a less competent dark pink or a dark orange skirt would be borderline speaker—but whose competence should be taken as cases for this sentence, due to the intrinsic vagueness the yardstick? Nor is it clear at which point etymolog- of red. Vagueness is ‘intrinsic’ in the sense that it has ical reconstruction should come to an end. nothing to do with lack of knowledge (we know what Multiple meanings distributed over several words the color red looks like). There is a close connection with the same pronunciation is called ‘homonymy’; between vagueness and context dependence. The exis- multiple meanings within a single word is called tence of borderline cases is not necessarily a bad thing, ‘polysemy’. The different meanings of a polysemous as different usages may be salient in different situa- expression have a base meaning in common. Further- tions; cf. Mary owns an expensive skirt. Depending on more, the meanings of a polysemous term are often the financial situations of Mary and the listeners, the related by metaphor or metonymy, as in point: ‘punctu- vague predicate expensive will have quite different ation mark’, ‘sharp end’, ‘detail, argument’, etc. Here, readings. In sum, vagueness may be defined as fol- we observe several meanings associated with one word, lows: a sentence is vague if—despite the knowledge of the shared base meaning of which could be something all the circumstances in a given situation—one cannot like ‘smallest unit’—either a concrete one (‘punctuation determine with certainty whether it is true or false. An mark’, ‘sharp end’) or a metaphorical one (‘detail, argu- expression is vague if it occurs in a sentence in such a ment’). There are cases of idiosyncratic polysemy, such way that it is responsible for the sentence’s vagueness. as green ‘a certain color’ and ‘inexperienced’, but there Vague sentences violate the principle of classical are also cases of systematic polysemy, as in the actu- logic demanding that a sentence be either true or false. al/dispositional distinction (as with fast in this is a fast Thus, either there is a third truth value (or many) car) or the distinction between a building and the insti- besides ‘true’ and ‘false’—this is assumed in the tution housed in it (as with school). many-valued logic approach (three-valued logics and Structural ambiguity has two subcases as well: fuzzy logics). Or there are gaps in truth-value assign- ‘attachment’ ambiguity and ‘scope’ ambiguity. ment—this is the thesis of supervaluation semantics. Attachment ambiguity refers to the possibility that the In sum, vagueness and ambiguity have different same sequence of words may be assigned different communicative statuses. Vagueness is part of the com- structures. In The policeman observes the lady with mon understanding of the speaker and hearer, accepted 40  AMERICAN SIGN LANGUAGE by both. With ambiguity, the situation is different. Lyons, J. 1977. Semantics. Cambridge: Cambridge University Usually, the speaker has a certain meaning in mind— Press. Pinkal, M. 1991. Vagheit und Ambiguität. Semantics, ed. by there is no vague reading of an ambiguous sentence. A.v. Stechow and Dieter Wunderlich. Berlin and New York: Instead, any vague reading of an ambiguous sentence is de Gruyter. a communicative breakdown, as ambiguous sentences Rescher, N. 1969. Many-valued logic. New York: McGraw- are always in need of disambiguation. Hill. Williamson, T. 1993. Vagueness. London: Routledge. Zwicky, A., and J. Sadock. 1975. Ambiguity tests and how to References fail them. Syntax and semantics, Vol. 4, ed. by J. Kimball. Ballmer, Th.T., and M. Pinkal (eds.) 1983. Approaching vague- New York: Academic Press. ness. Amsterdam: North-Holland. MONIKA RATHERT American Sign Language American Sign Language (ASL) is a naturally occur- France and eventually persuaded Laurent Clerc, one of ring language with a syntax and culture distinct from the young deaf teachers from de l’Epée’s school, to spoken English. Geopolitically, ASL covers a large travel back to America with him. Together, they estab- territory, serving between 100,000 and 500,000 sign- lished the first American School for the Deaf in ers in both Canada and the United States. ASL is not a Hartford, Connecticut. The French language taught by ‘universal sign language’, nor does it have close Clerc mingled with the indigenous dialects brought in etymological ties with British Sign Language, another by his young American students from across the devel- distinct signed language. Instead, French Sign oping country. Mingling with a couple of fledging Language, or langue des signes française (LSF), has signed languages found in several small established historically given ASL its closest linguistic ties. deaf communities such as Martha’s Vineyard in Throughout early history deaf people lived isolated Massachusetts (descendants of British immigrants), lives, possibly communicating with their hearing par- and Henniker, New Hampshire, a new national signed ents and siblings through the use of idiosyncrati- language began to evolve. cally created home signs. As world populations grew, In 1864, a liberal arts college now known as larger metropolitan groups began to provide the Gallaudet University was established in Washington, contact for visual gestural dialects to emerge. In the DC through the efforts of Abraham Lincoln and late 1760s, the French priest Charles-Michel de Edward Miner Gallaudet, Thomas Hopkins’s son. l’Epée, concerned for the religious salvation of deaf Graduating deaf students, many of whom had previ- Parisians, gathered a small group of deaf children ously attended Clerc’s Connecticut school, returned to together and began teaching them through the use of their native homes and frequently established or an invented manual French that he created as an edu- became responsible for establishing schools for deaf cational counterpart to spoken French. This gathering children in their respective states. The unique lan- in Paris is now recognized to be the first public school guage mixture of Clerc’s LSF signs and indigenous for the deaf in the world. The home signs and various signs brought in or naturally created by Gallaudet stu- Parisian dialects that the deaf children brought to their dents gradually evolved into modern ASL. school merged with de l’Epée’s ‘methodological’ Spoken languages can usually trace their history signs and subsequently evolved into the language now back through many centuries. Even including its his- known as LSF. torical French ties, ASL has only 250 years of docu- The reputation of the school and the emerging lan- mented language use. Yet, ASL is already following a guage flourished. In the early nineteenth century, a process of grammatical evolution that spoken lan- Protestant minister Thomas Hopkins Gallaudet was guages follow. Cross-linguistic studies of the historical sent by wealthy American philanthropists to Europe to relationships between LSF and ASL reveal paths of investigate a school for the deaf in London. Unable to lexical and grammatical change taking place in this obtain any assistance from England, Gallaudet visited young signed language. 41 AMERICAN SIGN LANGUAGE One example can be found in the semantic changes information is expressed in the movement, as Ronnie that have occurred between the LSF nominal sign for Wilbur describes: ‘coin’ and a modern ASL discourse marker used to stop arguments or to change the subject (Wilcox …in a multimorphemic verb sign, the movement can 2000). The old LSF classifier noun for money, carry the information regarding the transfer of a theme- object (e.g. GIVE), the starting location can carry the ARGENT (sign words are conventionally capitalized), source-agent-subject, the ending location can carry the semantically evolved into the sign for GIVE (donner), goal-recipient-indirect-object, the handshape can carry a verb in LSF that can also be found in modern ASL. the physical characteristics of the object itself (size and The verb stem phonology of both ‘give’ signs has a shape indicated by ‘handle’ classifier), and each piece common etymology, but diverge functionally. of information is phonologically distinct from the others. Following a path of lexicalization, one GIVE sign (2000:216) became an inflected verb of both literal and metaphor- ical nature; the other GIVE sign grammaticalized and ASL has a large set of signs called classifiers that extended into the word GIVE-IN-ARGUMENT. This can be used to specify something without actually verb is used metaphorically only and functions solely naming the specific referent. Classifiers can represent as a discourse marker. pronouns and describe in detail the size, shape, depth, Typical paths of grammaticalization in spoken lan- or texture of an object, and they can function as a noun guages reveal that language evolves from the mapping and indicate the location of that noun and its actions of literal objects onto nouns (nominalization) to (Baker-Shenk and Cokely 1999). Classifiers are abun- semantic shifts of abstraction. The original concrete dant in ASL and most signed languages. reference of a French money sign became increasing- Another critical component of ASL is the use of ly abstract as it evolved into the metaphorical dis- nonmanual signs. Grammar can be produced and dis- course marker in ASL. This example, plus the tributed across the face, the head, and the shoulders. documented grammaticalization of modals such as Adverbial and adjectival information can be produced WILL and MUST, demonstrate that ASL is following on the mouth, tongue, and cheeks. For example, one a linguistic path of subjectivity taken by many spoken visual gesture, ‘th’, is made with a slack jaw and a pro- languages. truding tongue pushing through slightly separated lips, ASL has an underlying s(ubject)–v(erb)–o(bject) and indicates carelessness or incorrectness. Other non- sentence structure. In addition, it is strongly charac- manual markers such as the eyebrows, the nose, head terized by topic-comment structure. Topics act as dis- movements, and eyegaze can coarticulate with signed course pivots to link previous information with the words or be produced in isolation. These nonmanuals following and also indicates a shift in the point of mark wh-questions, yes/no questions, rhetorical ques- view that the signer takes of the discourse events tions, conditionals, negations, topics, and relative (Janzen et al. 1999). Topic marking in ASL is gener- clauses. Rhetorical questions, uncommonly used in ally a marker with explicit discourse function. The spoken English, are found quite frequently in ASL following example also demonstrates the flexible conversations. A rhetorical question is accompanied word order of ASL: by a brow raise and a tilting of the head and often involves the use of the signs WHY, WHAT, WHO, HOW, and REASON. Nonmanual facial expressions [KIP BUY NEW COMPUTER]-topic, CURIOUS can also indicate mood and affect. ‘I wonder if Kip bought a new computer.’ Fingerspelling makes use of 26 handshape configu- rations that correspond to the English alphabet. Signed words consist of parts and can be broken Fingerspelling is more prevalent in ASL than in other down into components called parameters: the hand- signed languages of the world. It is primarily used to shape, the location, the movement, and the orientation indicate proper nouns and technical terms. of the palm. These components exhibit both sequen- Fingerspelling is more than a sequence of canonical tiality and simultaneity when the signed words are put handshape-alphabet letter correspondence, however, together, differing from spoken languages that are since the articulatory movements of segments within highly sequential. The iconic nature of ASL, with its the fingerspelled word influence each other. Forward strong link between form and meaning, is pervasive (carry-over) and anticipatory assimilation affects the and is documented throughout its phonological and actual shaping of fingerspelled words, creating a fluid morphological analysis. transition between letters that is prosodic and complex The morphology of ASL verbs is especially com- (Wilcox 1992). plex, more so than spoken English. The verb root is Language borrowing involving fingerspelling takes often expressed in the handshape and grammatical place through the physical proximity of English and 42  AMERICAN SIGN LANGUAGE ASL. Fingerspelled English words are often modified so in schools, universities, courtrooms, and social and that they fit better the structure of a sign (Battison 1978). business settings across America. For example, the word A-L-L can be fingerspelled letter The past 200 years have witnessed a strong expan- by letter, or it can serve as a borrowed loan sign and will sion of ASL throughout the world due to religious or be restricted semantically in several ways. Instead of fin- educational agendas. Also, thousands of deaf students gerspelling the three letters in the conical signing space from around the world have attended Gallaudet in front of the body, a signer can hold up an imagined University (the world’s only liberal arts university for list with one hand while using the other hand to finger- deaf people); thus, the signed language can be found spell the letters, literally running down the entire length throughout Asia, South America, Africa, and Europe — of the arm to show ‘all’ of the items on the list. Or a two- almost every country has deaf alumni who use ASL in handed loan sign can fingerspell an abbreviated A-L addition to their own native signed language. Many deaf while sweeping the letters in front of the signer to des- people have a working knowledge of ASL and the lan- ignate everyone in the room including the signer. This guage is used freely at most of the international deaf particular lexicalized loan sign is highly productive and conferences and workshops. However, its pervasive use limits the semantic meaning according to specific mod- of fingerspelling and marked initialization of English ifications that are applied to the three letters. words has thus far prevented it from being accepted as The culture of deaf people endures not through arti- an official signed lingua franca. Nevertheless, ASL users facts such as food or clothing, but through the powerful provide incentive to the world’s remaining deaf popula- use of the language of ASL. Historically, the hearing tions to maintain their cultural and linguistic identities world has not condoned the language choice of deaf as their respective signed languages advance and evolve. people. A defining event in 1880 still exerts influence on deaf people and their use of signed languages. The rep- References resentatives of the World Conference for the Deaf in Armstrong, David F., Michael A. Karchmer, and John Vickrey Milan, Italy, voted unanimously (except the American Van Cleve. 2002. The study of signed languages: essays in delegates) to ban signed languages from all the schools honor of William C. Stokoe. Washington, DC: Gallaudet in Europe and replace it with an ‘oral’ method of instruc- University Press. tion (Lane 1984). Even in America, this edict had a pro- Baker-Shenk, Charlotte, and Dennis Cokely. 1980. American sign language: a teacher’s resource text on grammar and cul- found effect on the educational systems. Deaf teachers ture. Washington, DC: Clerc Books, Gallaudet University were fired or prevented from educating younger deaf Press. (fourth printing, 1999). children, teachers who signed were not promoted or Battison, Robbin. 1978. Lexical borrowing in American sign respected, older students who might ‘contaminate’ the language. Silver Spring, MD: Linstok Press. younger children were hidden in separate quarters, and Janzen, Terry, B. Shaffer, and Sherman Wilcox. 1999. Signed language pragmatics. Handbook of pragmatics, ed. by Jef curricula were changed to speech-based activities. Verschueren, Jan-Ola Ostman, Jan Blommaert, and Chris The most influential leader of the oral method of Bulcaen, 1–20. Amsterdam: John Benjamins Publishing educating deaf children was Alexander Graham Bell, Company. the inventor of the telephone. His great influence and Klima, Edward S., and Ursula Bellugi. 1979. The signs of lan- wealth had a lasting impact upon the movement. guage. Cambridge, MA: Harvard University Press. Lane, Harlan. 1984. When the mind hears. New York, NY: Organizations that were founded by his proponents Random House. still work to oppress the use of ASL in the United Lane, Harlan, Robert Hoffmeister, and Ben Bahan. 1996. A jour- States today. In the early 1900s, with schools no ney into the deaf-world. San Diego, CA: DawnSignPress. longer administered by deaf adults, the language was Padden, Carol, and Tom Humphries. 1988. Deaf in America: kept alive by the use of deaf children born of deaf peo- voices from a culture. Cambridge, MA: Harvard University Press. ple, a heritage unlike that found in spoken languages. Stokoe, William C., Dorothy Casterline, and Carl Croneberg. Only 8–10% of deaf children are born to deaf parents 1965. A dictionary of American sign language on linguistic who sign, which means that around 90% of the ASL- principles. Washington, DC: Gallaudet College Press. using deaf children learn it from their ‘native’ peers. Reprint, Silver Spring, MD: Linstok Press, 1976. The proponents of the oral method were relatively suc- Wilbur, Ronnie. Layering of nonmanuals in ASL. 2000. The signs of language revisited: an anthology to honor Ursula Bellugi cessful in keeping signed language out of the educa- and Edward Klima, ed. by Karen Emmorey and Harlan Lane, tional arena until the early 1960s when William C. 215–44. Mahwah, N J: Lawrence Erlbaum Associates. Stokoe’s (1965) seminal work on linguistic analysis of Wilcox, Phyllis Perrin. 2000. Metaphor in American sign lan- ASL appeared. His pioneering work ignited an explo- guage. Washington, DC: Gallaudet University Press. sion of research on ASL that now permeates linguistic Wilcox, Sherman. 1992. The phonetics of fingerspelling. Philadelphia, PA: John Benjamins Publishing Company. inquiry. Because of the tenacity of ASL users, along PHYLLIS PERRIN WILCOX with the backing of research that provides linguistic legitimacy, ASL is now finding rewarding acceptance See also British Sign Language; Grammaticalization 43 AMHARIC AND ETHIOPIAN SEMITIC LANGUAGES Amharic and Ethiopian Semitic Languages There are perhaps 11 Ethiopian and Eritrean Semitic Some 76 languages are spoken in Ethiopia and (EES) languages now spoken, although the presence of Eritrea, in addition to the 11 Semitic languages, some dialect continua and lack of information on intelligibil- 21 Cushitic languages, 23 Omotic languages, 20 Nilo- ity between varieties makes a fully confident count Saharan languages, and one as yet unclassified, difficult. Two other languages are known, but are now Ongata. The presence of so many Semitic, Cushitic, extinct. These 13 languages are (spelling of names of and Omotic languages, three of the six subgroups of lesser known varieties varies in the literature): the Afroasiatic family (the other three Egyptian, Berber, and Chadic), suggests that the region is one of 1. Tigre, of western Eritrea, with dialects on the very early Afroasiatic settlement. Red Sea coast and Dahlak islands; Amharic, with over 16 million mother tongue 2. Tigrinya, spoken in Eritrea and northern speakers and, according to the 1994 Ethiopian census, Ethiopia, but especially in Ethiopia’s Tigray over five million second-language speakers, is by far region; the most important Ethiopian language, and the sec- 3. Ge‘ez, the ancient language known in epigra- ond most populous Semitic language after Arabic. phy since about 500 BC, the language of the Amharic is spoken throughout Ethiopia but particular- ancient kingdom of Axum centered on the mod- ly in urban areas, and is taught in all public schools. ern Tigrayan city of the same name, unused as Regional varieties or dialects of Amharic are recog- a mother tongue for many centuries, but contin- nized in regions of Begemder, Gojjam, Wello, and uing in regular use as the liturgical language of Shewa. The Ethiopian capital city of Addis Ababa (in the Ethiopian Orthodox Christian Church; Shewa) is nowadays the focus of Ethiopian economic 4. Gafat, of the Gojjam region of central Ethiopia, and cultural life, and Addis Ababa Amharic has natu- extinct now for some 50 years; rally become the prestige dialect. 5. Amharic, the world’s most populous Semitic The numbers of speakers of most Ethiopian lan- language after Arabic, with over 16 million guages may be estimated from counts of the 1994 speakers in Ethiopia, and long that nation’s Census of Ethiopia. The Census, however, like much official language; of the linguistic literature, grouped as ‘Gurage’ the six 6. Argobba, near mutually intelligible neighbor to languages numbered 7–12 in the list, and provided Amharic, in the northwestern Shewa region; separate figures for only the Soddo and Silte varieties. 7. Soddo (also known as Kestane ‘Christian’), Amharic has three times more speakers than Tigrinya, spoken just west of Addis Ababa, with a dialect with perhaps five and a half million including two mil- Dobbi (or Gogot); lion in Eritrea. Tigre may have a million speakers, all 8. Mesqan, south of Soddo, whose linguistic sep- in Eritrea, followed by Ethiopian Silte (800,000), arateness from neighboring varieties is least Soddo (200,000), and Chaha (perhaps 100,000). confident; Speakers of the others are much fewer in number. The 9. Chaha and mutually intelligible varieties census reported Harari to have 22,000 speakers, and including Ezha, Gomara, Gura, and Muher, Argobba only 10,000. spoken south of Mesqan territory; The region is naturally characterized by bilingual- 10. Inor and mutually intelligible varieties includ- ism and multilingualism, often involving Amharic or ing Endegenya, Enner, Gyeta, Magar, and Oromo, a Cushitic language (known in the old litera- Mesmes (the latter probably now extinct), ture as ‘Galla’) perhaps as populous as Amharic, and southern neighbors of the Chaha group; more populous when Oromo speakers of Kenya are 11. Silte and mutually intelligible varieties includ- added. Research in the late 1960s showed, for exam- ing Enneqor, Ulbareg, and Walane, southeast- ple, that in Jimma town of western Ethiopia some 90% ern neighbors of the Inor group; of speakers of languages other than Amharic knew 12. Zay, about Lake Zway some 70 km south of Amharic, and 42% of Amharic speakers knew Oromo. Addis Ababa; and Today, perhaps the majority of town and city-dwelling 13. Harari (Adare), spoken in and about the ancient Ethiopians, except in largely Tigrinya-speaking Tigre eastern Ethiopian city of Harar. province, are at least second-language speakers of 44  AMHARIC AND ETHIOPIAN SEMITIC LANGUAGES Amharic, and knowledge of the language certainly ‘all’ ‘die’ ‘eye’ ‘house’ ‘name’ continues to spread throughout the region. Hebrew ko¯l ma¯t ayin bayit še¯ m After a largely Tigrayan army overthrew the 1991 Arabic kull ma¯ta ayn bayt ism derg (‘committee’) communist government in Addis Amharic hullu motə ayn bet s m Ababa, whose predecessors had deposed the Amhara emperor Haile Sellassie in 1974, Ethiopian political While a different selection of words might show no power lay in the hands of Tigrinya speakers, who have similarities, and many similarities that exist are far encouraged considerable linguistic freedom of expres- from obvious, the systematic prevalence of such com- sion and identity including, rather controversially, the parisons as these confirms that Hebrew, Arabic, and use of local-majority languages in public education. Amharic are all descendant languages of that spoken Oromo, backed by a nationalist movement, became a by the first Semitic peoples. growing rival to Amharic-language dominance. With its Beyond lexical comparisons, Semitic languages, many second-language speakers and long literary devel- including EES, are particularly characterized by three- opment, however, Amharic has largely continued its consonant roots filled out by vowels and affixes in dominance in literary culture, education, and public life. word derivation (so-called ‘root-and-pattern’ morphol- Written in Ge‘ez, royal chronicles and religious ogy), and other characteristics including the presence writings are known from the fourteenth century, and of a set of subject-agreeing verb prefixes including y- shortly after this time writings in Amharic also began for third-person masculine ‘he’, and t- for both sec- to appear, but especially in the seventeenth century. ond-person ‘you’ and third-person (‘she’) feminine: Publication in Amharic has increased steadily, and Ethiopian publications in Amharic today include writ- 3m.sg. 3f.sg. 2m.sg. ings of all kinds: newspapers, literary and news maga- Hebrew yi-kbad ti-kbad ti-kbad zines, drama, novels, history, textbooks, and poetry. Arabic ya-kbud ta-kbud ta-kbud Amharic-language magazines are published in Europe Amharic y -kbəd t -kbəd t -kbəd and the United States to serve the Ethiopian expatriate populations there. Publication in Tigrinya as well as in The root (kbd heavy’) and the prefixes are cognate; Eritrea and Ethiopia’s Tigray region has flourished. a difference is that the Amharic verbs are subjunctives, Amharic and Tigrinya, and often nowadays other vs. the present tense of the Hebrew and Arabic. EES languages, are written in the Ge‘ez or ‘Ethiopic’ Within the Semitic family of languages, Amharic writing system, an adaptation of the ancient South and its kin in Eritrea and Ethiopia are typically Arabian writing system, itself the southern adaptation grouped as ‘South Semitic’ with the ancient and mod- of evolved Egyptian writing sometimes termed ern Semitic languages of South Arabia, such as epi- ‘Sinaitic’, whose northern adaptations include graphic Sabaen of Yemen and modern Soqotri of Aramaic, Arabic, and Hebrew writing (and Phoenician, Soqotra island. The unity of EES is suggested by sev- perhaps that transformed by the Greeks into their eral features which they tend to share, including a non- alphabet). The epigraphic record suggests that during finite verb conjugation for all but the last of verbs in the fifth century reign of the Axumite king Ezana, sequence (Amh. səbro ‘he having broken’), verbs Ethiopic writing was significantly modified to include expressed as compounds of particle + ‘say’ (z mm alə representation of vowels, as regular extensions or shap- ‘he was quiet’), and a special verb of presence (Amh. ings of its consonantal characters, approximately as alləhu ‘I am present’). early as in Indic Kharosthi writing. EES languages are divided into two groups, North Probably since ancient times, Amharic has spread in (Tigre, Tigrinya, and Ge‘ez) and South (Amharic and territory earlier populated by speakers of other, espe- the rest), which differ in a number of characteristic fea- cially Cushitic, languages, and as a result Amharic has tures, of which two are the presence of so-called ‘bro- come to share a number of features with these other ken’ noun plurals in the North (for example, Tigrinya languages. Amharic has long borrowed from Ge‘ez, a sor ‘ox’, pl. ’aswar), formed by modification of root favored practice for the modern satisfaction of needs structure instead of affixation, and present-tense dou- for philosophical, technical, and other new vocabulary. bling of the second root consonant of triconsonantal Borrowings from Italian entered the language in the verbs (compare Tigrinya y səbb r ‘he breaks’ vs. Amh. Italian Colonial era, but today the principal source of y səb r). The South EES languages are bifurcated as an new words is English, the language of Ethiopian and eastern group of Amharic and Argobba plus the Silte- Eritrean secondary and higher education. group, Zay, and Harari, against a western group of the The obvious Semiticness of Amharic may be sug- Chaha-and Inor-group languages plus Mesqan and gested by a comparison of a selection of five basic Soddo. The eastern group typically augments present- words of Hebrew, Classical Arabic, and Amharic. tense main verbs with a suffixed auxiliary verb, as in 45 AMHARIC AND ETHIOPIAN SEMITIC LANGUAGES Amharic main verb y səbr-al ‘he breaks’ vs. subordi- uted to the influence of Cushitic-language neighbors, nate y səbr. which are mainly of the consistent SOV type, with Argobba differs from Amharic in small but numer- postpositions and case suffixing absent in EES. ous ways, probably resulting in at least slight mutual Amharic and most other EES languages usually case- nonintelligibility. The divergence of Amharic and mark only definite objects, Amharic with -n: w ša-w Argobba may reasonably have resulted from the sepa- l K -u-n nəkkəsə ‘the dog bit the boy’. ration of these peoples upon the adoption of Islam by Amharic and other EES languages have the phono- the Argobbas in about the sixteenth century. The typi- logical characteristic of a series of glottalized ejective cal divergence between pairs of EES languages is consonants, as in Amharic k’utt’a ‘anger’ (with ejec- much greater, on the order suggested by 50% or so of tives k’ and t’). These differ from the so-called cognates in a hundred words of core vocabulary ‘emphatic’ velarized/pharyngealized cognate conso- between North and South EES languages, and about nants of e.g. Arabic. Some EES languages preserve the 60% between South EES languages. Semitic pharyngeals  and ʕ. As in the other Semitic Ge‘ez, because it is the only anciently known EES languages, the voiceless labial stops p and p’ are mar- language, has sometimes been considered the ancestor ginal, or limited to loan words such as ityopp’ya language from which all the others descend, as Latin ‘Ethiopia’, from Greek. Like most of the EES lan- is ancestor to the Romance or Italic languages, but this guages, Amharic has the seven-vowel system i, e, a, , does not seem to be the case, because there are old lin- ə, u, o; others have a five-vowel system with length. guistic features of the other languages unreasonably absent in Ge‘ez, pervasive Ge‘ez features unreason- References ably absent in the others, and because the diversity of the other languages seems too great to date only from Bender, M. L., J. D. Bowen, R. L. Cooper, and C. A. Ferguson (eds.) 1976. Language in Ethiopia. London: Oxford the early Axumite times of likely Ge‘ez dispersal. University Press. Amharic, like other EES languages, constructs typi- Hetzron, Robert. 1972. Ethiopian Semitic: studies in classifica- cal transitive sentences with the main constitutent order tion (Journal of Semitic Studies monograph 2). Manchester: subject–object–verb (SOV), in contrast to the VSO typ- Manchester University Press. icality of Semitic Arabic and Hebrew. An Amharic top- Hetzron, Robert (ed.) 1997. The Semitic languages. London: Routledge. icalized object can precede a subject, but the verb is Leslau, Wolf. 1978. English-Amharic context dictionary. rigidly final. Amharic, however, has a number of word- Wiesbaden: Otto Harrassowitz. order characteristics inconsistent with the basic SOV Leslau, Wolf. 1987. Comparative dictionary of Ge‘ez. type, including prepositions instead of postpositions Wiesbaden: Otto Harrassowitz. (bə-bet ‘in house’, bə- ‘in’) and verb prefixal clause Leslau, Wolf. 1995. Reference grammar of Amharic. Wiesbaden: Otto Harrassowitz. subordinators (s -tsəbr ‘when you break’, s - ‘when’). GROVER HUDSON Typical SOV characteristics of Amharic and EES lan- guages are the linear precedence of adjectives, geni- See also Afroasiatic; Arabic; Hebrew: Biblical; tives, and relative clauses before their modified nouns. Hebrew: Modern; Semitic Languages; Writing The syntactic peculiarities of EES are typically attrib- Systems Analogical Change In a general sense, analogy is the correspondence of are both plural nouns but differ in the way the plural is two or more entities with respect to certain properties. formed: -s vs. -n + vowel change. The form pigs— In linguistics, the term ‘analogy’ is most often used to respectively horses, stones, and the like—is considered describe processes in which a new linguistic object is the model here and the plural of cow is aligned with the created by aligning an existing linguistic object A with plural of pig and thus, by analogy, kine becomes cows. another existing linguistic object B, which serves as a Sound change and analogical change have already model. A and B usually have some properties in com- been considered by Neogrammarians as the fundamen- mon. For instance, Early Modern English pigs and kine tal processes of language change. The two important 46  ANALOGICAL CHANGE types of analogical change are analogical leveling and Other less productive forms of proportional analo- proportional analogy. Analogical leveling results in the gy are backformation and hypercorrection. With back- reduction of allomorphy within paradigms. The para- formation, it is the morphologically less complex item digms of the Old English and Old High German verb that is created. In Early Modern English, pease was a meaning ‘to choose’ contain forms with a fricative [z] mass noun, which ended in a sequence that resembled or [s], where other forms have an [r]. In both lan- a plural morpheme. Its reanalysis led to the creation of guages, the paradigms were leveled out by adopting the a singular form pea by dropping the final fricative. same consonant throughout the whole paradigm. While Backformation also occurs as a word-formation Modern English chose [z], Modern High German process. dropped the fricative in favor of [r]. Old Engl. Mod. Engl. Old High Germ. Mod.H.G. ce-o[z]an > choo[z]e PRES kiu[s]an > kü[r]en ce-a[s] > cho[z]e PAST-(SG) ko[s] > ko[r] cu[r]on > cho[z]e PAST-(PL) ku[r]un > ko[r]en (ge-)co[r]en > cho[z]en PAST-PART (gi-)ko[r]an > geko[r]en The above example also illustrates the interaction of bean : beans [-z] = X : pease [-z] sound change and analogical change. The alternation of (X = pea) [s] and [z] results from a sound change referred to as depression : depress = aggression : X ‘Verner’s Law’ and [r] is the result of a subsequent (X = aggress) change from [z] to [r] in certain positions (‘rhotacism’). Sound change is conditioned by phonetic factors and is Hypercorrection consists in imposing a pattern that not sensitive to the needs of grammatical and semantic relates a colloquial low-prestige variant and a standard functioning. It often obliviates the principle ‘one form— higher-prestige variant onto another word presumed to one meaning’ (‘Humboldt’s universal’). Analogical lev- be low prestige. Thereby, an allegedly high-prestige eling is always conditioned by nonphonetic factors and variant of the word is created: often ‘fixes the damage’ caused by sound change. In contrast to sound change, analogy has been con- fella’ : fellow = umbrella : X (X = umbrellow) sidered to be an unsystematic process. This has given rise to the formulation of ‘Sturtevant’s Paradox’: sound Other phenomena that have been discussed under change is regular but creates irregularity, whereas anal- the heading of ‘analogy’ are blending, contamination, ogy is irregular but creates regularity. However, in some reanalysis, and folk etymology. Blending or ‘portman- cases analogy is fairly systematic. For instance, [z]–[r] teau formation’ consists in combining parts of existing alternations of the above type have been leveled out in words where these parts are usually not morphemes. English quite systematically, as within the paradigms of Examples are brunch from breakfast and lunch or info- loose, freeze, and rise. There are only very few cases tainment from information and entertainment. where the [r] has been retained (e.g. was vs. were). In the case of contamination, a similar fusion of two By proportional analogy—also called ‘analogical words occurs but the resulting expression retains the extension’ or ‘four-part analogy’—an already existing meaning of one of the source words. For example, the morpheme or relation becomes generalized to other [t] of Proto-Indo-European pəte-r should have turned linguistic forms. In most cases, proportional analogy up as [d] by Verner’s Law and a subsequent sound creates or assimilates derived forms on the basis of change, yielding fader. Instead, contaminated by the another morphological pattern. Proportional analogy regularly derived brother, we get father. proceeds like the solution of an equation of the form: In reanalysis, a deviant structure is assigned to an expression. The words napkin and apron both derive A : B = C : X solve for X from French nape ‘cloth’. In the second case, the stone : stones = cow : X X = cows (< kine) Middle English sequence that results from the combi- nation of the word with an indefinite article anapron The solution of this equation with respect to the phe- was reanalyzed as an apron, interpreting n as part of nomenon mentioned above yields the replacement of the article. A related phenomenon is folk etymology. kine by cows as the plural form of cow. As with ana- Sometimes, words change on the basis of what a logical leveling, proportional analogy is not fully sys- speaker assumes is the etymological origin of the tematic: the analogy did not carry over to words like word. For instance, French crévisse should occur in foot: feet / *foots. English as something like crevisse, but appears as 47 ANALOGICAL CHANGE crayfish because the second part got misinterpreted as language acquisition, syntax, and orthography: (i) referring to some kind of fish or aquatic creature. Analogy is assumed to be at work when children over- Theoretical approaches to analogical change mani- regularize morphological forms, like goed instead of fest themselves in attempts to construct a theory of went. (ii) An example of syntactic analogy is the change analogical processes, which is explicit and restrictive of impersonal constructions like me hungreth, me thinks and unifies the different forms of analogy (e.g. within to the more widespread personal constructions, i.e. to I generative phonology; cf. Kiparsky 1982). Some hunger, I think. (iii) Modern High German lieb ‘kind / researchers formulated a number of principles (some- nice’ orthographically reflects a Middle High German times called ‘laws’) in order to explain why and how diphthong, which has been monophthongized to [i:] in analogical changes occur (e.g. KuryAowicz 1949, Modern High German. Words like Wiese ‘lawn’ adopt- Man´czak 1980). These principles are usually subject ed this spelling by proportional analogy, although the to counterexamples and should therefore rather be Middle High German spelling and pronouncation wise understood as tendencies (cf. McMahon 1994). showed a monophthong. Among these principles were the following: (i) Analogy proceeds from a basic to a derived form or from a shorter to a longer form. Counterexample: References backformations. (ii) More overt and complex markers Antilla, Raimo, and Warren A. Brewer. 1977. Analogy: A basic are favored by analogy. For example, the Old High bibliography. Amsterdam: Benjamins. German plural forms of Gast ‘guest’ and Topf ‘pan / Hock, Hans Heinrich. 1991. Principles of historical linguistics, 2nd revised and updated edition. Berlin, New York: de pot’ are Gästi and Topfa. By analogy Topfa adapted to Gruyter. the form Gästi, which has the more complex plural Hock, Hans Heinrich, and Brian D. Joseph. 1996. Language marking (umlaut and suffix) yielding Gäste and Töpfe history, language change, and language relationship. An in Modern High German. Counterexample: Old introduction to historical and comparative linguistics. English lang : leng-ra turned into long : long-er anal- Berlin, New York: Mouton de Gruyter. Kiparsky, Paul. 1982. Explanations in phonology. Holland: ogous to warm : warm-er, not into long : *leng-er. (iii) Dordrecht / Foris: Cinnaminson. In processes of proportional analogy, it is usually the KuryAowicz, J. 1949. La nature des procès dits >analogiques<. more productive pattern that serves as a model. Acta Linguistica Hajniensia 5. 15–37 [Reprint in: Hamp, Counterexample: extension of strong verb forms to Eric P., F.W. Householder and Robert Austerlitz, editors, weak ones as in the change from dived to dove as past Readings in linguistics, 2nd edition. Chicago: Chicago University Press, 1966]. tense of dive analogous to ride : rode. (iv) Allomorphy Man´czak , W. 1980. Laws of analogy. Historical morphology, gets reduced by analogy (in particular by analogical ed. by Jacek Fisiak. The Hague, Paris, New York: Mouton. leveling). Counterexample: the change of the stem Marle, J. van. 1985. On the paradigmatic dimension of mor- Topf- to Töpf- in German mentioned above. (v) When phological creativity. Foris: Dordrecht. a new analogical formation becomes accepted (e.g. McMahon, April M. S. 1994. Understanding language change. Cambridge: Cambridge University Press. brother) and the older form also survives (brethren), STEFAN ENGELBERG the analogical form takes over the basic meaning. Counterexample: blendings. See also Comparative Method; Historical Linguis- Analogy has also been considered an important tics; Inflection and Derivation; Language Change; tool in other branches of linguistics like theories on Morphology Anaphora Linguistic expressions may have their semantic inter- ing examples include phrases like the tall woman, pretation determined solely by their lexical content bought a car, and no tall man. Examples of the (and the way these lexical items are syntactically second type include herself (see (1)), she/her (on arranged), or they may have all or part of their inter- the construal whereby she/her refers to the same pretation fixed by a relation to another linguistic person as the tall woman—see (2) and (3)), his (on expression. Examples of the first type in the follow- the construal whereby its meaning is determined by 48  ANAPHORA no tall man—see (4)), and the unpronounced verb Chomsky’s Binding Theory is based on three con- phrase in (5). straints: Condition A: an anaphor must be c-commanded by (1) The tall woman hurt herself. an antecedent within its binding domain. (2) The tall woman realized that she was tired. Condition B: a pronoun must not be c-commanded (3) The tall woman ran the risk of bumping her by an antecedent within its binding domain. head on the low lintel. Condition C: an R-expression must not be c-com- (4) No short man runs the risk of bumping his manded by any nominal of which it is the antecedent. head on a high lintel. In (the majority dialect of) English, the binding (5) Mary bought a car, and Stanley did ... too. domain for an anaphor is the smallest argument phrase The study of the syntactic configurations in which containing it and another c-commanding N(oun) the antecedent–dependent pair may occur, and of the P(hrase)—usually the clause. The binding domain for a types of semantic relation they may bear to each other, pronominal is the smallest argument phrase containing is called anaphora. it. Grammaticality judgments below (with ungrammat- The most widely studied cases of anaphora in gener- icality indicated by an asterisk) are given for the coref- ative grammar are those involving nominal expressions erence reading of the antecedent–dependent pair; thus like the reflexive in (1) and the pronouns in (2)–(4). (10) is ungrammatical if her and Mary corefer). Semantically, anaphoric dependencies between nom- (9) Mary hurt herself. inals divide into at least two types. When the antecedent (10) * Mary hurt her. refers to an entity in the world (as in (1)–(3)), the rela- (11) * Mary said that Dan admires herself. tion is that of coreference—the dependent refers to (12) Mary said that Dan admired her. whatever the antecedent does, and so they corefer. When the antecedent is quantificational, as in (4), the In (9), herself is a reflexive, i.e. a syntactic anaphor. Its relation is that of bound-variable anaphora. No short antecedent Mary is the c-commanding subject of the man does not, strictly speaking, refer directly to an enti- clause; hence, Condition A is satisfied. In (10), the ty. Rather, it quantifies over the things that are men, and pronoun her is contained in the same clause as the c- his acts as a bound variable. The sentence essentially commanding antecedent Mary, which violates means ‘no thing x which is a man is such that x runs the Condition B and makes the sentence ungrammatical. risk of bumping x’s head… .’ In (11), the reflexive herself does not have an Nominals divide into three types with respect to the antecedent within the binding domain (the embedded syntax of anaphora. Anaphors are nominals that must clause introduced by that), violating Condition A. (12) have an antecedent; these include both reflexives (see is grammatical, because the antecedent Mary of the (1)) and reciprocals, e.g. each other, one another in pronoun her is not located within the same clause. English. Pronominals, like she, her, and his, may have Thus, Conditions A and B are basically locality con- antecedents (as on the most salient interpretations of straints on anaphors and pronominals: anaphors must (2)–(4)), but they need not. Each of (2)–(4) may be be close to their antecedents (the antecedent must be interpreted with the pronoun referring to some unmen- local), and pronominals cannot be (the antecedent tioned person. Such interpretations typically involve must not be local). either some kind of ostension (e.g. pointing at the No locality effects are obtained for Condition C. intended referent), or prominence of the referent (e.g. Coreference is blocked both in the local case (15) and in the speaker at a conference). The third type of nominal the long-distance case (16), where the pronoun and R- is those that do not have antecedents at all, like the tall expression are separated by several clause boundaries: woman, a low lintel, a car, and no tall man. These are (15) * She admires Mary. termed R-expressions (from ‘referring expression’). (16) *She thinks that I said that Dan admires Mary. Each of these three types of nominal has a distinct syntactic distribution with respect to other nominals In both cases, the main clause subject she c-commands with which it has an anaphoric relation. These syntac- Mary. If Mary is interpreted as the antecedent for tic conditions form the center of Noam Chomsky’s she, Condition C is violated—the sentence is ungram- Binding Theory (1980s). Each is critically dependent matical. on a structural relationship between two elements Since c-command plays a critical role, it appears called c-command. Without going into technical that Chomsky’s binding constraints, although they details, one can roughly say that one element c-com- govern a semantic relation, are crucially syntactic in mands another if it is (a) in a structurally more promi- nature. Thus, the binding conditions are a prototypical nent position and (b) not contained within the same example of the tight interplay between syntax and structural unit as the other. semantics. Examination of the parallel conditions in 49 ANAPHORA other languages reveals that the definition of binding anaphoric interpretation of the pronoun would thus be domain varies substantially across languages. The ungrammatical. Notice that both sentences are gram- study of such a variation has led to detailed examina- matical when every man is replaced by John, because tion of dozens of languages, and has also been the sub- he and John would simply corefer, i.e. he would not be ject of a number of studies in the first- and a bound variable. second-language acquisition literature. Returning to bound-variable anaphora, this has a References more restricted syntactic occurrence than coreference. Barss, Andrew (ed.) 2002. Anaphora: a reference guide. (17) John/every man said that he was elegant. Oxford: Blackwell. (18) [The woman who danced with John/*every Fiengo, Robert, and Robert May. 1994. Indices and identity. man] said that he was elegant. Cambridge, MA: MIT Press. Higginbotham, James. 1983. Logical form, binding, and nomi- Coreference may occur where the antecedent c- nals. Linguistic Inquiry 14(3). 395–420. commands the pronoun, as in (17), where Every man May, Robert. 1984. Logical form. Cambridge, MA: MIT Press. is itself the subject of the sentence. When there is no Reinhart, Tanya, and Eric Reuland. 1993. Reflexivity. Linguistic Inquiry 24. 657–720. c-command relation as in (18), where every man is ANDREW BARSS deeply embedded in the subject phrase, the pronoun cannot have an anaphoric relation to every man. An See also Chomsky, Noam Anatomy of the Articulatory System Speech sounds are produced by moving air in and bronchi in turn join to form the trachea (windpipe). out of the body. To accomplish this, a number of The top of the trachea is just behind the notch at the organs in the chest, neck, and head are used. An top of the breastbone. Lying behind the trachea, the understanding of these anatomical organs and their esophagus is a tube that carries food to the stomach. functions is a prerequisite for phonetics, the study of Various muscles of the chest expand and contract speech sounds. the ribcage. The external intercostal muscles lift the The air pathway (Figure 1) consists of the lungs, lar- ribcage, inflating the lungs, and the internal inter- ynx, pharynx (throat), oral cavity (mouth), and nasal costals pull the ribcage down, deflating the lungs. The cavity. The lungs start the process of speech production abdominal muscles also help to deflate the lungs. The by pushing air upward. The vocal folds, which are diaphragm is a large dome-shaped muscle lying just located in the larynx behind the Adam’s apple, may below the lungs; by contracting, the diaphragm moves vibrate, causing the air that flows through them to downward, increasing the capacity of the lungs. The vibrate as well. The airstream is then modified by the elasticity of the lungs resists expansion and contrac- vocal tract (Figure 4)—the pharynx and the oral and tion, attempting to return the lungs to their resting size. nasal cavities. By altering the shape of the vocal tract, a The larynx (Figures 2 and 3) is a complex structure, large variety of sounds can be produced. cylindrical in shape, composed of several cartilages. It The lungs sit inside the rib cage. Expanding the is found above the trachea. Within the larynx, the ring- rib cage causes the lungs to expand, and air is drawn shaped cricoid cartilage sits immediately atop the tra- in; when the rib cage and lungs are compressed, air chea. Above the cricoid cartilage is the plough-shaped is expelled. The lungs are elastic, like a sponge; if thyroid cartilage. The forward external point of the their size is changed, this elastic property tends to thyroid cartilage is easily identified as the Adam’s return the lungs to their resting size. The two lungs apple. The back of the thyroid cartilage has two horns are independent. If one is injured, the other one still pointing down and two pointing up. The two lower functions. horns attach to the cricoid cartilage. The thyroid carti- Tiny tubes within the lungs, called alveoli, combine lage provides a shield for the vocal folds, which are repeatedly to form larger tubes, ultimately resulting in attached behind the Adam’s apple. The two arytenoid a single tube for each lung, called a bronchus. The two cartilages are small pyramid-shaped cartilages sitting 50  ANATOMY OF THE ARTICULATORY SYSTEM Oral cavity Nasal cavity Velum Tongue Pharynx Adam's Apple Larynx Trachea Bronchi Lungs Figure 1. Vocal Organs. on top of the cricoid cartilage at the back on either side Hyoid bone (see Figure 3). Upper horn Crucial to speech are the vocal folds. These are two Epiglottis of thyroid cartilage horizontal shelves of muscle and ligament. The front Lower horn part of both vocal folds is attached to the thyroid car- Thyroid of thyroid cartilage tilage. Each vocal fold is attached at the back to one of cartilage the two arytenoid cartilages. The vocal folds have a tri- Arytenoid cartilages angular space between them called the glottis, through which the air passes upward from the trachea. The Criciod cartilage vocal folds assume a variety of positions to affect the airstream coming from the lungs. The different posi- Trachea tions create the glottal stop as well as different vocal qualities—voiced, voiceless, creaky, breathy, falsetto. In the past, the vocal folds were often called vocal Figure 2. Larynx. cords; they are, however, solid structures and not cord shaped. The attachment of the horns of the thyroid car- tilage to the cricoid cartilage allows the thyroid carti- lage to rock backward and forward, changing the Glottis Arytenoid tension of the vocal folds; this is the basic mechanism cartilages for varying the pitch of the voice (see Figure 3). The epiglottis (Figure 2) is a spoon-shaped carti- lage in the pharynx that forms something of a hood Adam's apple over the main part of the larynx. The horseshoe- Cricoid cartilage shaped hyoid bone (Figure 2) resides behind the chin; it supports the muscles of the tongue above it, and it Thyroid cartilage also stabilizes the muscles of the larynx from above (see Figure 2). Vocal folds The pharynx is the name for the throat. It is a verti- cal tube connecting the larynx with the oral and nasal Figure 3. Larynx Overhead. cavities. The rear wall of the pharynx is the pharyngeal 51 ANATOMY OF THE ARTICULATORY SYSTEM Palate Velum Nasal cavity Uvula Alveolar ridge Oral cavity Velic surface Lips Teeth Blade Oral pharynx Tip Back Pharyngeal Tongue wall Epiglottis Vocal folds Esophagus Adam's apple Trachea Figure 4. Vocal Tract. wall. The root of the tongue forms the forward wall of Blade Back the pharynx. The oral cavity, or mouth, is extremely important in the production of speech sounds. The rear of the Tip Root oral cavity connects to the pharynx. The upper artic- ulators are the upper lip, the upper teeth, the upper surface of the mouth, and the back pharyngeal wall Front (see Figure 4). The lower articulators are the lower lip, lower teeth, and tongue. Sounds are formed by moving a lower articulator toward an upper articu- lator to form a complete or partial obstruction to the Figure 5. Tongue. airflow. The lips and teeth are familiar. A bumpy area ous, without clear landmarks, phoneticians divide it known as the alveolar ridge lies just behind the upper into five parts (see Figure 5). teeth. The hard palate is a smooth bony structure The tip, or apex, of the tongue is its foremost part. behind the alveolar ridge. In phonetics, the hard palate Just behind the tip of the tongue is a small surface is normally referred to simply as the palate. Behind called the blade, or lamina. The middle portion of the palate lies the velum, or soft palate. The back of the the tongue is known as the front. It is not literally the velum narrows at the back to a thin structure called the front of the tongue, but the surface behind the blade. uvula. The pharyngeal wall is considered to be an The rear portion of the horizontal surface of the tongue upper articulator (see Figure 4). is called the back or dorsum. The root of the tongue is The lower articulators are the lower lip, the lower the rear vertical surface facing the pharyngeal wall teeth, and the tongue. The tongue is the most impor- (see Figure 5). tant of the lower articulators; it is a large, complex The top of the pharynx opens into the nasal cavity. organ. Although the surface of the tongue is continu- The upper surface of the velum is known as the velic 52  ANATOMY OF THE AUDITORY SYSTEM surface. The velum can move up or down to close off Ladefoged, Peter. 1993. A course in phonetics. Ft. Worth, TX: the passageway to the nasal cavity. Harcourt Brace Jovanovich. Ladefoged, Peter. 2000. Vowels and consonants: an introduc- tion to the sounds of languages. Oxford: Blackwell. References Ladefoged, Peter, and Ian Maddieson. 1996. Sounds of the world’s languages. Oxford: Blackwell. Catford, J.C. 1977. Fundamental problems in phonetics. Laver, John. 1994. Principles of phonetics. Cambridge: Bloomington, IN: Indiana University Press. Cambridge University Press. Catford, J.C. 1988. A practical introduction to phonetics. Minifie, Fred, Thomas J. Hixon, and Frederick Hixon. 1973. Oxford: Oxford University Press. Normal aspects of speech, hearing, and language. Gimson, A.C. 1980. An introduction to the pronunciation of Englewood Cliffs, NJ: Prentice-Hall. English. London: Edward Arnold. Pullum, G.K., and W.A. Ladusaw. 1996. Phonetic symbol Hardcastle, William J., and John Laver (eds.) 1997. The hand- guide. Chicago: University of Chicago Press. book of phonetic sciences. Oxford: Blackwell. Rogers, Henry. 2000. The sounds of language: an introduction International Phonetic Association. 1999. Handbook of the to phonetics. Harlow, England: Pearson. International Phonetic Association. Cambridge: Cambridge Zemlin, W.R. 1988. Speech and hearing science. Englewood University Press. Cliffs, NJ: Prentice-Hall. Kahane, Joel C., and John W. Folkins. 1984. Atlas of HENRY ROGERS speech and hearing anatomy. Columbus, OH: Charles E. Merrill. See also Speech Production Anatomy of the Auditory System The auditory system consists of three main sections: eustachian tube provides for the aeration and drainage the outer ear, the middle ear, and the inner ear. The of the middle ear system and makes it possible for air structure and function of each of these systems will be pressure to be the same on both sides of the eardrum discussed briefly in the next sections. (also while yawning or swallowing). While we are awake, the eustachian tube opens approximately once per minute; during sleep it opens, on average, once Outer Ear every 5 min. The eardrum is coupled to the oval win- The purpose of the outer ear is to gather sounds from dow, an entrance to the inner ear, through a chain of the environment and to funnel them into the auditory three tiny bones, the hammer (malleus), anvil (incus), system. The outer ear begins at the auricle (also pinna) and stirrup (stapes). Sound waves traveling in the and ends at the eardrum (also tympanic membrane). external auditory canal cause the eardrum to vibrate. The auricle is mainly composed of elastic cartilage This vibration is passed on to the middle ear, where a (except for the earlobe) and aids front-to-back locali- pressure increase is needed because of the mismatch in zation. It leads to the opening of the ear canal (or density and compressibility between air at the eardrum external auditory canal). This canal is, approximately, and fluid at the oval window. As the effective surface 9 mm high, 6.5 mm wide, and 2.5 cm long, and pro- area of the eardrum is 55 mm2 and that of the stapes vides some protection of the eardrum against foreign footplate connecting to the oval window is about bodies. The eardrum is a smooth, nearly transparent, 3.2 mm2, the vibratory pressure at the eardrum is fun- membrane of about 80 mm2 in area. Although it is neled and increased by a factor of 17 (55/3.2) at the extremely thin (about 0.07 mm), it is constructed of oval window. Pressure at the oval window is further- three layers. It is attached to the handle of the hammer more increased by the three bones acting as a lever (malleus), the first of three tiny bones (ossicles) in the system and by the conical shape of the eardrum. middle ear. Inner Ear Middle Ear Although the inner ear is as tiny as a bean, it con- The middle ear is an air-filled cavity that is connected tains thousands of moving parts and is responsible for to the nose cavity via the eustachian tube. The sending information to the brain regarding hearing and 53 ANATOMY OF THE AUDITORY SYSTEM balance. The primary auditory organ of the inner ear is on the membrane from base to apex. Traveling waves the cochlea (latin for snail). It is about 1 cm wide, and produced by high-frequency sounds (shorter wave- 5 mm from base to apex in man, makes about two and length) show maximum displacement near the base of a half coils, and is filled with fluid. The cochlear coil the cochlea, while those produced by a low-frequency is divided into three channels: the scala vestibuli, scala sound (longer wavelength) come to a peak near the media, and scala tympani. The scala vestibuli and scala apex. Arching over the hair cells is the tectorial mem- tympani are filled with perilymph and communicate brane. The differential motion of the basilar and tecto- via an opening at the apex known as the helicotrema. rial membranes results in a shearing motion of the The central duct, the scala media is filled with stereocilia of the hair cells. This bending of the hair endolymph and bounded by two membranes, cells produces an electrical discharge in the cochlear Reissner’s membrane (top) and the basilar membrane portion of the VIIIth nerve. A chemical is released at (bottom), that form a wedge-shaped partition. On the the base of the hair cell when the stereocilia are basilar membrane (about 35 mm long), along the full sheared. Although the mechanics of the organ of Corti length of the scala media, lies the organ of Corti, are very complex as a result of different motions of the which contains the actual receptor cells that are basilar membrane (up and down, side to side, etc.), the responsible for hearing. On the inner side of the tunnel aforementioned briefly describes how mechanical of Corti, a single row of inner hair cells, approximate- motion is converted into neural activity in the organ of ly 3,400 in total, are lined up side by side along the Corti. The size of the electrical response of the cochlea entire length of the organ of Corti (35 mm). On the is directly related to the extent to which the hair cells outer side of the tunnel of Corti are three rows of outer are sheared. hair cells, approximately 13,400 in total in humans. Each hair cell has ‘hairlike’ projections, called stereo- cilia, projecting from their apical surface. About References 90–95% of the fiber of the auditory nerve connect to Durrant, J.D., and J.H. Lovrinic. 1984. Bases of hearing sci- the inner hair cells. Many outer hair cells converge to ence, 2nd edition, 276pp. Baltimore, MD: Williams & a single auditory nerve fiber, while each inner hair cell Wilkins. Lim, D.J. 1980. Cochlear anatomy related to cochlear micro- may excite up to 20 auditory nerve fibers. The outer mechanics. A review. Journal of the Acoustical Society of hair cells act to enhance the selectivity by changing the America 67. 1686–95. mechanical properties of the basilar membrane. If the Lim, D.J. 1986. Functional structure of the organ of Corti: a outer hair cells are selectively damaged, there is a loss review. Hearing Research 22. 117–46. of sensitivity. The hair cells and nerve fibers are held Lipscomb, D.M. 1976. Mechanisms of the middle ear. Hearing disorders, ed. by J.L. Northern, 78–88. Boston: Little, in place by supporting cells. Brown. The motion of the footplate of the stapes in the oval Maue-Dickson, W. 1981. The inner ear—prenatal development. window moves the fluid of the inner ear. This air Medical audiology, ed. by F.N. Martin, 220–35. Englewood vibration, magnified by the middle ear, creates a pres- Cliffs, NJ: Prentice-Hall. sure wave in the fluid, which distorts the basilar mem- Pickles, J.O. 1988. An introduction to the physiology of hear- ing, 2nd edition, 367pp. New York: Academic Press. brane in the scala media. Since the fluids of the Yost, W.A., and D.W. Nielsen. 1985. Fundamentals of hearing: cochlea are incompressible, the displaced fluid, in an introduction, 2nd edition, 269pp. Orlando, FL: Holt, turn, causes an outward displacement of the round Rinehart & Winston, Inc. window. This produces a progressive traveling wave ASTRID VAN WIERINGEN Ancient Egyptian Ancient Egyptian is a branch of the Afroasiatic family main stages in the development of the Egyptian lan- (also called Hamito-Semitic in traditional comparative guage (cf. Loprieno 1995): linguistics) attested in Egypt from 3000 BCE to 1300 CE. Within its family, it shows the closest relation- (1) Earlier Egyptian (3000–1300 BCE), further ships to Cushitic, Semitic and Berber. There are two subdivided into: 54  ANCIENT EGYPTIAN (a) Old Egyptian (3000–2000 BCE: Old (2) A semagram, named determinative in Kingdom/First Intermediate Period); Egyptological tradition, usually accompanies (b) Middle or Classical Egyptian (2000–1300 the sequence of phonograms spelling out the BCE: Middle Kingdom/XVIII Dynasty; semantic domain of the word. used in religious texts until the Greco- Some words are written phonologically, combining Roman period); (mainly mono-)consonantal signs, many others are writ- (2) Later Egyptian (1300 BCE–1300 CE), further ten with logograms (referred to as ideograms in the subdivided into: Egyptological literature), i.e. signs either representing (a) Late Egyptian (1300–700 BCE: Dynasty the object or another entity phonologically similar in XIX/Third Intermediate Period); structure to the intended word (as in a rebus), or even an (b) Demotic (seventh century BCE to fifth object metonymically or metaphorically connected to the century CE: originally the language of intended object. There are 24 monoconsonantal signs. administration during the Late Period, Hieratic and Demotic writing systems derive from used also in literary texts); the hieroglyphic one: Hieratic is an italicization of the (c) Coptic (fourth to fourteenth century CE: latter, whereas Demotic stems from the late Hieratic the language of Christian Egypt); writing. The Coptic writing uses Greek alphabet (with Earlier and Later Egyptian differ from one another in seven additional signs derived from Demotic and rep- the following respect: the former displays a preference resenting sounds absent from the Greek alphabet). for synthetic grammatical structures (suffixes for gen- der, number, etc.), whereas the latter shows a tendency Phonology and Pronunciation towards analytic grammatical devices (prefixation, ver- bal periphrases). Moreover, Demotic and Coptic differ The study of Egyptian phonology is made difficult by from previous stages with respect to the graphic system. the writing system used. As noticed above, vowels are never represented in the hieroglyphic alphabet, semi- vocalic phonemes only seldom. Things are complicat- Alphabet ed by the fact that the phonological reality of Coptic Egyptian hyeroglyphs are a complex set of graphemes, (vocalic) characters, often invoked as a comparison for variable in number according to the period considered. the purpose of phonetic reconstruction, is far from They are pictographic signs representing any kind of being clear, and the Coptic is chronologically far away persons, animals and objects. The two components of from the earliest stages of Egyptian language. an Egyptian word are: Therefore, the pronunciation of Egyptian words is often a matter of convention. Other heuristic criteria (1) A sequence of phonograms that may be mono-, resorted to by egyptologists in order to reconstruct the bi- or triconsonantal (according to the number of phonological reality of Ancient Egyptian are mainly consonantal phonemes represented). The vocalic the contemporary transcriptions and renderings of phonemes are never graphically expressed. Bi- Egyptian words in Akkadian (archive of el-‘Amarna, or triconsonantal signs are usually accompanied fifteenth to fourteenth century BCE) and in Greek by the so-called ‘phonetic complements’, i.e. texts (starting from the fourth century BCE). Also, the monoconsonantal phonograms which specify Egyptian rendering of foreign words may be taken as one or more phonemes of the bi- and triconso- a reconstructing criterion. nantal sign, thus allowing for a more simplified The consonantal inventory of Ancient Egyptian is reading of the intended word. given in Table 1 (adapted from Loprieno 1995; TABLE 1 The Consonantal Phonemes of Ancient Egyptian Consonants Bilabial Dental Alveo-palatal Palatal Velar Uvular Pharyngeal Glottal PLOSIVE: Voiceless p /p / [p(h)] t /t/ [t(h)] t /c/ [c(h)] k/ k / [k(h)] k. /q/ [q’] Voiced b/b / d /d / [t’] d /Ô / [c’] g/g / [k’] FRICATIVE: Voiceless f/ f/ s(s´ ) /s / [s j] š / š/ h /ç x / h
/h
/ h / h/ Voiced z(s)/z/ [s’] /ʕ/ NASAL m /m / n /n/ VIBRANT r /r / [ɾ] A/R/ GLIDE w /w/ j / j/ 55 ANCIENT EGYPTIAN phonemes are given both in the common transcription Early Egyptian has no articles: the form sn may system of Egyptian texts and in their supposed pho- refer both to ‘a brother’ and to ‘the brother’. Later, netic value). articles arose from the deictic pA /tA /nA ‘the said’. Historically, the opposition between voiceless When used attributively, adjectives usually follow and voiced phonemes tends to be neutralized into the noun and agree with it in gender and number (sn=j the voiceless variant. The vowel inventory is fairly nfr [brother-my beautiful] ‘my beautiful brother’/ limited (as usual in Semitic languages), includ- sn.t=j nfr.t [sister-my beautiful] ‘my beautiful sister’/, ing /i/, /a/ and /u/ and their long counterparts. etc.). When used predicatively, they precede the noun The phonemes /e/ and /o/ are attested in Later and do not agree with it (the so-called adjective verbs: Egyptian. nfr sn=j [beautiful brother-my] ‘my brother is beauti- ful’/ nfr sn.t=j [beautiful sister-my] ‘my sister is beau- tiful’/ etc.). Morphosyntax Personal pronouns are either independent or affixal; affixal pronouns are further divided into dependent Ancient Egyptian has inherited the preference of lan- and suffixed pronouns (see Table 3, adapted from guages of its family for bi/triconsonantal roots. The Loprieno 1995): the former are used mainly as objects number of consonants in a root may range from one to of transitive verbs, and as subject of adjectival and four, but the vast majority of roots is biconsonantal (sn adverbial phrases, the latter are used to indicate: (i) the ‘brother’) or triconsonantal (hpr ‘become’). Entire ˘ from a given root, possessor in nominal construct state (sn=j [brother-I] lexical families may be formed ‘my brother’), (ii) the subject of verbal phrases (sd_m=j either by varying the internal vowels or by means of [hear-I] ‘I hear’), (iii) the prepositional complement suffixation. In the transcription system, suffixes are (n=j ‘to me (DAT)’). Independent pronouns, on the indicated by periods preceding them. As for nominal other hand, are used as subjects in nominal phrases morphology, there are suffixes marking gender (mas- (i nk sn=k [I brother-you] ‘I’m your brother’) and in culine and feminine) and number (singular, plural, verbal phrases. In both these cases, the pronoun is in dual [attested only in Early and Middle Egyptian]). In focus and conveys contrastive information (as in cleft Table 2, the morphemes for gender and number are sentences). illustrated for the noun sn ‘brother’. Possessive pronouns (‘mine’, ‘yours’, etc.) are formed by a demonstrative pronoun (pA j /tA j /nA j ‘this TABLE 2 Gender and Number Markers for sn (masc./fem./plur.)’ agreeing with the possessee) and a ‘brother’ suffix pronoun agreeing with the possessor. Possessive pronouns are often used as possessive adjectives in singular plural dual nonofficial texts instead of the suffix pronouns (pA j=j sn [the-my [mine] brother] ‘my brother’ instead of masculine sn.Ø sn.w sn.wj sn=j [brother-I] ‘my brother’). This strategy is the feminine sn.t sn.wt sn.tj most widespread in Demotic. TABLE 3 Egyptian Personal Pronouns Independent Dependent Suffixed SINGULAR: 1st person jnk -wj (later twj-) =j 2nd person masculine T wt (later ntk) -kw (later twk-) =k 2nd person feminine T mt (later ntT ) -T m (later twt-) =T 3rd person masculine swt (later ntf) -sw(later sw- > ef-) =f 3rd person feminine stt (later nts) -sj /-st (later st- > es-) =s DUAL 1st person =nj 2nd person -T nj =T nj 3rd person ntsnj -snj =snj PLURAL 1st person jnn -n (later twn-) =n 2nd person ntT n -T n (later twtn-) =T n 3rd person ntsn (later ntw) -sn/-st (later st-) =sn (later =w) 56  ANCIENT EGYPTIAN TABLE 4 Relative Pronouns TABLE 5 Personal Endings of the Pseudo-Participle positive (who) negative Singular Dual Plural (who … not) 1st person .kj > .kw .wjn Masculine ntj jwtj 2nd person .tj .twnj .twnj Feminine ntt jwtt 3rd person masculine .j > .w .wj .w Plural ntj.w/nt.w jwtj.w 3rd person feminine .tj .tj .tj Relative pronouns are ‘positive’ and ‘negative’ (see Demotic Table 4). The possession relation between two nouns is The stage of development termed Demotic is charac- expressed either (i) with the so-called nominal terized by the evolution from a synthetic to an analyt- ‘construct state’ (i.e. the nouns are juxtaposed with ic morphology. In the field of nominal morphology, for no overt marking of either the possessor or the instance, the development of a definite and indefinite possessee; the word order of this construction is article from the deictic adjective pA /tA /nA , ‘the said’, possessee–possessor: sA Ra [son Ø Ra] ‘son of Ra’) pairs the progressive loss of the nominal endings (see or (ii) with the determinative pronoun n(j)[masc.sg.]/ Table 2). nw[masc.plur.] /nt[femin. sg. and plur.] ‘that of, Singular Plural Singular Plural belonging to’, agreeing with the possessee: sA .t nt R c definite definite indefinite indefinite [daughter DET+AGR Ra] ‘daughter of Ra’. In the Masculine pA sn nA sn(.w) wa sn hjn sn(.w) development of the Egyptian language, this determi- Feminine tA sn(.t) nA sn(.wt) wa.t sn(.t) hjn sn(.wt) native pronoun grammaticalized into an invariable genitive marker n: sA .t nRa [daughter DET-AGR Ra] ‘daughter of Ra’. References The verbal system of Classical Egyptian is synthet- Bresciani, Edda. 2002. Nozioni elementari di grammatica ic: T-A-M morphemes (n past tense; t perfective; w demotica, con Liste grafiche e letture demotiche (a c. di prospective aspect and passive voice; tw passive voice) Angiolo Menchetti). Pisa: ETS. and suffix pronouns are attached directly to the verbal C˘erny´, Jan and Saram Israelit Groll. 1984. A late Egyptian root. T-A-M morphemes, if present, precede the suffix grammar, 3rd edition. Studia Pohl, Series Maior IV. Rome: pronouns. Some verbal roots (‘weak verbs’) show a Pontifical Biblical Institute. Erichsen, Wilhelm. 1954. Demotisches Glossar. Kopenhagen: reduplication of the last radical phoneme in non-finite Ejnar Munksgaard. verb forms. The imperative corresponds to the basic Erman, Adolf. 1933. Neuägyptische Grammatik, 2nd edition. form of the root (with no suffix pronouns or redupli- Leipzig: Wilhelm Engelmann. cation); some verbs, however, display irregular forms Erman, Adolf, and Hermann Grapow. 1926–1953. Wörterbuch of imperative (rdi ‘to give’, imp.: i m; i j ‘to go’, imp.: der Ägyptischen Sprache, Vols. 1–4. Berlin: Akademie Verlag. mi ). A form usually called pseudo-participle, old per- Gardiner, Alan. 1927. Egyptian grammar, being an introduction fective, or stative (cf. Table 5, adapted from Loprieno to the study of hyeroglyphs. Oxford, UK: Oxford University 1995) is characterized by a series of suffix pronouns Press, 3rd edition, 1957. different from that sketched in Table 3. This form con- Helck, Wolfgang, Eberhard Otto, and Wolfhart Westendorf veys perfective meaning. (eds.) 1975–1992. Lexicon der Ägyptologie (LÄ), 7 vols. Wiesbaden: Harrassowitz. Infinitives bear a Ø-morpheme in the regular Johnson, Janet H. 1991. Thus Wrote Onchsheshonqy. An intro- verbs and a .t feminine ending in the class of ‘weak ductory grammar of Demotic, 2nd edition. Chicago, IL: verbs’. Oriental Institute. Sentences are classified according to the kind Loprieno, Antonio. 1995. Ancient Egyptian. A linguistic intro- of predicate; compare the following nominal (1), duction. Cambridge, UK: Cambridge University Press. Osing, Jürgen, 1976. Die Nominalbildung des Ägyptischen, 2 adjectival (2), adverbial (3) and verbal (4) sentences: vols. Mainz: Verlag Philipp von Zabern. Polotsky, Hans-Jakob. 1971. Collected Papers. Jerusalem: (1) i nk sn=k (‘I’m your brother’) Magnes Press. (2) nfr pr=k (‘your house [pr=k] is beautiful Spiegelberg, Wilhelm. 1925 Demotische Grammatik. [nfr]’) Heidelberg: Carl Winters Universitätsbuchhandlung. (3) sA .t=k m pr=k (‘your daughter [sA .t=k] is in [m] ANGIOLO MENCHETTI your house [pr=k]’) (4) sD m=j i t=f (‘I hear [sD m=j] his father [i t=f]’) See also Afroasiatic; Coptic Egyptian 57 ANCIENT GREEK Ancient Greek Overview Doric features in the choruses, because the invention of choral poetry attributed to the Dorians). A(ncient) G(reek) belongs to the Indo-European fami- The two most important literary dialects were Ionic ly, of which it constitutes a separate branch. Its pri- and later Attic. Since these two varieties are closely mary relevance to Indo-European studies and studies related to each other, and literary Attic gave up a num- on language change owes to several factors. In the first ber of vernacular features under the influence of the place, AG is the second earliest attested Indo- more prestigious Ionic, the literary dialect is common- European language after Hittite. It is considered to be ly known as “Attic-Ionic”. The preservation of dialec- pretty conservative under different respects (phonolo- tal variation was favored by political fragmentation. gy, inventory and realization of inflectional categories, After the unification of Greece under the Macedonian word order, among others). Besides, its development kings in the fourth century BCE, linguistic unification continued after antiquity into the Middle Ages also started, resulting in the development of a common (Byzantine and Medieval Greek) up to the present variety, the koiné, which was spoken during the cen- (Modern Greek), which makes Greek the language turies of the Roman (later Byzantine) Empire, and with a longest recorded history together with Chinese. served as the basis for further developments. Unfortunately, this long time span is not evenly docu- mented: there are breaks, one of which concerns the present discussion, covering the period that goes from Mycenean Greek the first written attestations (13th century BCE) to the M(ycenean) G(reek) is known since the 1950s, when redaction of the Homeric poems (eighth century BCE). its decipherment was accomplished by Ventries and Speakers of some AG varieties are thought to have Chadwick. The Mycenean script, commonly known as entered mainland Greece in the second half of the third Linear B, is a syllabary consisting of around 90 signs millennium BCE, then spreading to the Aegean islands, for CV syllables, which makes it unsuitable for writ- Crete and Cyprus, and slightly later to the Aegean ing consonant clusters and word final consonants. The coast of Anatolia, and to Southern Italy and Sicily. The clay tablets that document MG were not meant for first written sources date back to the second millenni- long term preservation, as shown by the fact that they um BCE, and were found in Pylos (in the were not baked: some of them reached us because they Peloponnesus) and on the island of Crete. The lan- were accidentally baked in the fires that destroyed the guage of these early records is usually referred to as archives. The tablets mostly contain lists of various Mycenean Greek. items, lists of workers, and some descriptions of land Although they are several centuries older than the properties. In spite of the limitations connected with main bulk of AG written sources, the Mycenean tablets the text types and of poor spelling, a number of inter- do not preserve a sort of “Proto-Greek”: in spite of dis- esting features can be singled out, including the playing a number of archaic features, including preservation of the Indo-European labiovelars, phonemes that later merged with others in all Greek changed into bilabial or dental stops in later dialects, varieties, Mycenean is already characterized as belong- and of velar glides, dropped in the later literary vari- ing to a specific dialectal group. Dialectal variation is eties, a richer case system, which included at least an another noteworthy peculiarity of AG: contrary to most independent instrumental case, and systematic omis- ancient languages, AG had various local literary tradi- sion of the verbal augment. Other peculiarities of MG tions, based on different varieties; furthermore, inscrip- show its dialectal connections with the Arcado- tions preserve local vernaculars, still different from the Cypriot and the Aeolic groups. standardized literary variety chosen for each dialect. Although variation within the literary language is most- ly limited to Pre-Classic time (before the fifth century The Greek dialects BCE), specific literary genres remained connected with The Greeks, who lacked political cohesion, identified the dialect in which they had their earliest development, themselves as opposed to all other peoples on the basis so that classical authors used to give a specific dialectal of what they felt as a common language. They divided color to certain types of texts (for example, in Attic themselves into three dialectal groups, Ionic (also tragedy it was customary to introduce a number of comprising Attic), Aeolic, and Doric, all connected 58  ANCIENT GREEK with prestigious literary traditions. Modern research them, and did not know that earlier Greeks had used has shown that a fourth dialectal group also existed, another writing system centuries past. The Homeric commonly called Arcado-Cypriot, possibly not recog- poems were composed during this illiterate time; their nized by the Greeks because its speakers did not pro- language, basically an Ionic variety, preserves a number duce any noteworthy literary achievement; of archaic features, which makes the poems one of the furthermore, Doric is now considered a member of a most important documents for Indo-European studies. wider group, called Western Greek. Until the deci- Other early literary texts include Aeolic and Doric pherment of MG, Greek dialects were classified as poetry from the seventh and sixth centuries; prose was belonging to two main groups, Eastern Greek, further apparently an Ionic invention, the most significant divided into Attic-Ionic, Aeolic, and Arcado-Cypriot, early Ionian prose writer being Herodotus, who com- and Western Greek, comprising Doric and North posed his Histories in the sixth century BCE. The bulk Western Greek. The similarities of MG to both the of what is commonly considered classical Greek liter- Arcado-Cypriot and the Aeolic groups have led certain ature, from the fifth and early fourth centuries BCE, scholars to set up a three-way classification, including comes from the city of Athens. It is written in literary Attic-Ionic, Achaean (Aeolic, Arcado-Cypriot, and Attic, and includes the tragedies of Aeschilus, Mycenean), and Western Greek (see Figure 1). Sophocles, and Euripides, Aristophane’s comedies, as (NB: Southern Italy and Sicily were occupied by well as works by philosophers as Plato, historians as speakers of Doric). Thucidides, and numerous political orations. Numerous dialectal variants were simply phonetic, The language of later writers increasingly displays although some phonological differences also existed; typical features of the koiné. One of the most impor- besides there were morphological and lexical features tant sources for the knowledge of the latter is consti- peculiar of each dialectal group, but all varieties were tuted by the New Testament, written in the first century by and large mutually intelligible. CE. From the same period we also have a number of non-literary papyri, which help us understand the real extent of changes undergone by the spoken language. Written sources For during the first centuries CE literary Greek started Apart from the Mycenean tablets, the oldest written to move more and more away from the spoken lan- source is constituted by the Homeric poems. The poems guage, first with a puristic reaction to the intrusion of were first written in the eight century BCE (the oldest “vulgar” features (Atticism), and later with the devel- epigraphic inscriptions also date back to the same cen- opment of a peculiar diglossic situation, which was tury), but their oral composition dates back to several still characteristic of Modern Greek until the 1960s. centuries before. Writing had disappeared from Greece AG was written in alphabetic script, with a wide after the fall of the Mycenean civilization; early histori- number of local variants. Table 1 contains the standard ans trying to reconstruct Greek history describe a sort of Greek alphabet, as used in modern editions of ancient “Dark Age” with rather turbulent social habits; they had authors; it basically corresponds to the Modern Greek no memory of the refined civilization that preceded script, although phonetic values are no longer the same. Thessaly Boeotia Attica Athens Arcadia West Greek Rhodes Attic-Ionic Aeolic Arcado-Cypriot Crete Figure 1. The Ancient Greek dialects (from Horrocks, 1997: 8). 59 ANCIENT GREEK Grammatical sketch of AG Morphology AG has a big number of inflectional categories, both for Phonology nouns and for verbs. Its morphology is highly fusional, AG preserved the five vowels system of P(roto)-I(ndo)- with a fairly high degree of allomorphy, partly due to E(uropean), and the distinctiveness of vowel length simplification of consonantal clusters or monophthon- (long vowels are marked by ‘:’) (see Table 2). gization of diphthongs (vowel contraction). For ancient greek consontants, see Table 3. Nouns are inflected for case (nominative, accusa- tive, genitive, dative, and vocative) and number (sin- TABLE 1 The Greek Alphabet (from Horrocks, gular, plural, and dual); they belong to three genders 1997: xix-xx) (masculine, feminine, and neuter). Nominal inflection only employs suffixes. There are three inflectional Greek letter Ancient Modern classes, called first (/a:/ stems), second (/o/ stems), and pronounciation pronounciation third declension (consonant and /i/ and /u/ stems); Aα (alpha) [a, a] [a] assignment of a noun to a declensional class is deter- Bβ (beta) [b] [v] mined by its phonological shape. Already in MG /a:/ Γ γ (gamma) [g] [γ, j] stems, which were mostly feminine, developed a sub- ∆δ (delta) [d] [o] class for masculine nouns, with special endings for the E ∋ (epsilon) [e] [e] nominative and the genitive singular. The third declen- sion, which displayed the highest degree of allomor- Zζ (zeta) [zd] [z] phy, lost its productivity rather early, and nouns of the Hη (eta) ∋ [ ] [i] third declension moved to other inflectional classes Θ θ (theta) [th] [θ] (some doublets with a consonant and a -o- variant are Iι (iota) [j, i] [i, j] already found in Homeric Greek; in the koiné conso- Kκ (kappa) [k] [k] nant stems switched massively to -a¯- stems). Λλ (lambda) [l] [l] Adjectives, demonstrative, and the definite article Mµ (mu) [m] [m] (created after Homer) share the inflectional categories N υ (nu) [n] [n] of nouns; when modifying nouns they agree with their Ξ ξ (xi) [ks] [ks] heads in all inflectional categories. Adjectives are Ο ο (omikron) [o] [o] divided into two classes, according to whether they Π π (pi) [p] [p] have three different forms for the three genders, or P ρ (rho) [r] [r] only two (with masculine and feminine undifferentiat- ed). They also inflect for degrees of comparison (com- Σσ /ζ (sigma) [s] [s] parative and superlative). T τ (tau) [t] [t] Verbs are divided into two classes, thematic and ath- Y υ (upsilon) [y, y] [i] ematic. They are inflected for mood (indicative, opta- Φ φ (phi) [ph] [f] tive, imperative, and subjunctive), tense/aspect (present, X χ (chi) [kh] [x, ] imperfect, aorist, perfect, future, pluperfect, and future Ψ ψ (psi) [ps] [ps] perfect), diathesis (active, middle, and passive), person, Ω ω (omega) [o] [o] and number (singular, plural, and dual). A separate pas- sive is only found in the aorist, perfect, and future; in the other tenses, the middle can have both middle and pas- TABLE 2 Ancient Greek Vowels sive meaning; the imperfect, the perfect and the pluper- /i/ /u/ /i:/ /u:/ fect are found in the indicative only, the future and the /e/ /o/ /e:/ /o:/ future perfect only in the indicative and in the subjunc- /a/ /a:/ tive. Nominal forms of the verb include an array of infinitives and participles and two verbal nouns. Verbal TABLE 3 Ancient Greek Consonants morphology is more varied than nominal morphology: besides suffixes, one also finds prefixes (so-called “aug- voiceless stops p t k ment” for past tenses in the indicative, and reduplication voiced stops b d g aspirated stops ph th kh in the perfect and some presents), and alternation of the fricative s h root vowel (apophony). Below some typical verb forms nasals m n are analyzed for convenience: lateral l vibrant r 1. leíp-o¯, first sg. pres. indic., “I leave” glide w present stem+inflectional ending 60  ANCIENT GREEK 2. leíp-oi-mi, first sg. pres. opt., “may I leave” genitive with verbs that indicate some sort of separa- present stem+suffix+inflectional ending tion and with the prepositions that take the ablative in 3. é-leip-on, first sg. imperfect, “I was leaving” the other Indo-European languages (e.g. ek, “out of”, augment+present stem+inflectional ending apó, “from”, cf. Latin ex, ab, both with the ablative). 4. é-leip-s-a, first sg. aorist indicative, “I left” The dative is the case of the third argument of verbs augment+present stem+suffix+inflectional of exchange and communication; when occurring out- ending side the verbal valency, it can indicate the participant 5. é-lip-on, first sg. aorist indicative, “I left” for whose sake an action is performed (benefactive); augment+aorist stem+inflectional ending often it occurs in the “external possessor construc- 6. le-loíp-a, first sg. perfect, “I left” tion”: epeide¯` dé moi (dative) he¯ me¯´te¯r eteleúte¯se, “after reduplication+perfect stem+inflectional ending my mother died”, “after the mother died on me”. The dative merged with two other cases: the locative (pos- (this verb has two possible aorists, sigmatic, as in (4), sibly already before the Mycenean age), and the with the suffix -s- and the same stem as the present, instrumental (after MG, where the instrumental case is and thematic, as in (5), with apophonic variation of the still preserved). While the locatival function of the root vowel). plain dative is mostly limited to poetry, its instrumen- AG also has a big variety of derivational affixes. tal function was very productive: only around the first Most category changing affixes are suffixes: díke¯, century CE prepositional phrases start to be found in “right (noun)”, dík-aio-s, “just (adjective)”, dik-aio- place of the plain dative to express instrumental. súne¯ , “justice (noun)”, dik-áz-o¯, “I judge (verb)”. Verbal tenses partly have aspectual value, especial- Prefixed derivates normally maintain the same word ly outside the indicative mood. The basic aspectual class of the base: a-díkaios, “unjust (adjective)”, sun- opposition is between the present (imperfective) and dikázo¯, “I have a share in judging (verb)”. Besides der- the aorist (perfective). The perfect originally had a ivation, compounding was also very productive. resultative function, inherited from PIE: pres. baíno¯, “I go”, perfect bébe¯ ka: “I am (in a certain place)”. Syntax Modality is expressed partly by verbal moods, partly AG has nominative-accusative alignment. The subject by the modal particle an, used in some conditional of both transitive and intransitive verbs is inflected in sentences, and partly by verbal tenses (the imperfect the nominative; note, however, that neuter nouns have may be used for contrary-to-fact modality). Besides, nominativeaccusative. A further peculiarity of the optative also has a purely syntactic use when it neuter subjects is constituted by incomplete verbal occurs in subordinate clauses that depend on other agreement: neuter plural take the verb in the singular: subordinate clauses (“oblique optative”). tà zôia trékhei, “the animals (pl.) run (3rd sg.)”. The AG has a complex system of hypotaxis, including: subject need not be overtly expressed when it is recov- (a) complement clauses, which can have the verb in erable from the context; to a lesser extent, the direct the infinitive and the subject in the accusative or in the object, too, may be omitted if recoverable. Transitive nominative, or the conjunction hóti, “that”, and the verbs can have passive form, in which case subject same moods/tenses of independent clauses; (b) rela- function is assigned to the direct object of the corre- tive clauses, with the relative pronoun he¯´, hós, hó; (c) sponding active; the agent may or may not be added as various types of adverbial clauses. A peculiarity of AG an adjunct (prepositional phrase). subordination is the extremely widespread and elabo- The accusative is basically the case of the direct rated use of participles. The wide use of participles can object of transitive verbs; it also has a limited use as at least in part be explained through the need to make allative (mostly in poetry), and a rather peculiar up for the lack of adverbial verb forms (“converbs”, adverbial use, the “accusative of respect”: pódas corresponding to the English gerund); still, the exis- (accusative) o¯ kùs (nominative) Akhilleús (nomina- tence of participles for most aspect/tenses and for all tive), “Achilles swift-footed”. diatheses constitutes a rather infrequent typological The genitive is used for nouns functioning as modi- feature of AG. fiers of other nouns. It also has an adverbal use: certain AG word order is governed by pragmatic factors. In verbs only take the genitive (árkheín tinos, “to rule particular, the finite verb can stand in any position in over sbd. (genitive)”), others allow variation of the the sentence, and, if transitive, precede or follow the accusative and the genitive, which in such cases has direct object, both in poetry and in prose. The subject, partitive meaning: píen tòn oînon / toû oínou, “to drink if expressed, can be placed before or after the verb, but (all the) wine (accusative)/ some wine (genitive)”. At a preverbal position is more frequent. Modifiers (both pre-literary time the genitive merged with the ablative, adjectival and nominal) can precede or follow their and took over its functions; consequently, we find the heads; adpositions mostly precede their complements, 61 ANCIENT GREEK although Homer has some occurrences where they are PIE. Word order variation connected with pragmatic postposed, and relative clauses follow their head, with factors is common to many ancient Indo-European few exceptions. The definite article precedes the noun languages, although AG appears to have moved in the or noun group it determines; if a head noun is followed direction of bigger freedom: in this connection, the by an attribute, the article is repeated (ho ane¯`r ho retention of Wackernagel’s Law for the placement of agathós, “the good man”, lit. “the man the good”). A clitics and other postpositives is a remarkable archaic noteworthy group of words with respect to word feature. The field in which AG has innovated most is order is constituted by so-called “postpositives”. certainly hypotaxis, whose development is a conse- Postpositives are various types of conjunctive parti- quence of the creation of a highly elaborated literary cles, modal particles, and enclitic pronouns; they all standard. share the peculiarity that they cannot be placed in sen- tence initial position (this opposes them to preposi- References tives, as the negation ou, that can never be sentence Allen, W. S. 1987. Vox Graeca. Oxford: Oxford University Press. final). Most frequently, postpositives are placed after Blass, F., and Albert Debrunner. 1984. Grammatik des neutesta- the first word or first constituent in the sentence mentlichen Griechisch. Göttingen: Vandenhoek & Ruprecht. (Wackernagel’s position), thus separating parts of the Buck, Charles Darling. 1955. The Greek dialects. Chicago: same constituents, sometimes even the definite article Chicago University Press. Chantraine, Pierre. 1953. Grammaire homérique. Tome 2: from the noun it determines. Syntaxe. Paris: Klinksiek. Chantraine, Pierre. 1961. Morphologie historique du grec. Greek and Indo-European Paris: Klinksiek. Chantraine, Pierre. 1968. Dictionnaire étymologique de la AG is generally considered a rather conservative Indo- langue grecque. Paris: Klinksiek. European language. Features that make AG particular- Horrocks, Geoffrey. 1997. Greek. A History of the Language ly valuable for the reconstruction of PIE include the and its Speakers. London, New York: Longman. Meillet, Antoine. 1920. Aperçu d’une histoire de la langue vowel system, diphthongs, the three series of stops grecque. Paris: Hachette. (voiceless, voiced, and aspirated), the retention of Palmer, Leonard. 1980. The Greek Language. Atlantic Heights, numerous inflectional categories, which AG shares NJ: Humanities Press. with Indo-Iranian, and many of the inflectional affix- Schwyzer, Eduard. 1939. Griechische Grammatik. Band 1. es, along with the different morphological processes Munich: Beck. Schwyzer, Eduard, and Albert Debrunner. 1950. Griechische (reduplication, root and suffix apophony, pre- and suf- Grammatik. Band 2. Munich: Beck. fixation) and a highly fusional morphological tech- Ventries, Michael, and John Chadwick. 1973. Documents in nique. In spite of case syncretism, AG still preserves Mycenean Greek. Cambridge: Cambridge University Press. the basic uses of cases which can be reconstructed for SILVIA LURAGHI Animals and Human Language 1: Overview How is human language related to animal communica- tinuist’, respectively, have vigorous champions. Both tion systems? While humans are animals with a com- camps agree, however, that human language and ani- mon chimpanzee–human ancestor living as recently as mal communication systems differ significantly. It is five million years ago, neither chimpanzees nor any these differences, as well as similarities, that provide other species naturally use a communication system linguists an opportunity for analysis of the relationship, with anything similar to the complexity of human lan- if any, between human and animal ‘language’. guage. Is human language an enormously elaborated It is tempting to think of language primarily as a development of an original communication system tool for objective communication. However, humans used by the chimpanzee–human ancestor? Or is it a also use language to think and to create a mental rep- purely human innovation, unrelated historically to any resentation of their experience. Even a hermit who has animal ‘language’? There is no consensus on these never spoken for decades will use language for this questions. These two views, ‘continuist’ and ‘discon- purpose. Some ‘discontinuist’ scholars have argued 62  ANIMALS AND HUMAN LANGUAGE 1: OVERVIEW that the use of language for communication is histori- A vocabulary of 10,000 items would be hard to cally secondary, and that its original function was cog- manage if each item were an unanalyzable whole. nitive. This would render the differences between However, the words of human languages can be divid- human language and animal communication less sur- ed into individual contrasting speech sounds. prising. On the other hand, it is hard to show conclu- Approximately 45 (depending on dialect) of such sively that no animal ‘language’ has cognitive as well sounds are used in spoken English, while other lan- as communicative functions. This alleged peculiarity guages may differ in the number of discrete sounds. All of human language is therefore controversial. languages share the characteristic of being structured, While one commonality among animal communica- or patterned, on two levels: a level of meaningful units tions and human languages is the concern with biolog- (words, phrases, sentences) and of meaningless units ical, objective, or factual information, much human (sounds, syllables). For example, in the word pot, no discourse is not primarily factual in content. Many ani- one sound, p, t, or o, carries meaning. Of the word top, mals vocalize to advertise territorial control, or for sex- containing the same sounds, the same is true. Rather, it ual display. Here, it is not ‘what is said’ that matters, is the relationship of the sounds that imparts meaning but merely the fact that the vocalization occurs, and to the words. Animal communication systems lack this perhaps also its vigor and elaborateness. Similar in so-called ‘duality of patterning’. Individually meaning- function to human singing, factual communication is ful calls are not constructed entirely out of meaningless the primary purpose of animal communication. Human parts recurring in other calls. language, however, can also concern abstract thought. In human language, words are not merely strung Every normal human adult human has a vocabulary together, but are combined into larger units (phrases of thousands of items. Among animal communication and sentences) that are systematically interpretable. systems, the total does not exceed a few dozen. This Thus, for example, all speakers of English will agree may ensure that the domains of experience within on the interpretation of John will kill the crocodile, and which communication is possible among animals are will agree in interpreting it differently from The croc- severely limited, extending scarcely beyond food, dan- odile will kill John. Here, it is the relationships among ger, sex, and dominance. words that imparts meaning. This characteristic of lan- The size of human vocabularies is one of the fac- guage is called syntax. The syntax of human language tors that allow us to talk about things outside our is open-ended in that phrases and sentences can be immediate environment. The fact that a person lives in embedded within larger phrases and sentences, as in New Zealand in the twenty-first century does not pre- Mary thinks her father denied that John will kill the vent her from talking about the surface of the planet crocodile. No naturally occurring animal communica- Mars, or about the civilization of Egypt in the second tion system has a systematic syntax, even of the most millennium BCE. This characteristic of language has rudimentary kind. In particular, no such system has the been called ‘displacement’. In principle, displacement open-endedness that human language derives from does not presuppose a large vocabulary. A chim- embedding. panzee, for example, could conceivably use its danger Some apes (mainly chimpanzees) that have been calls to talk about not immediate danger but a possible exposed to American Sign Language (ASL) have danger tomorrow. However, no chimpanzee does this. acquired vocabularies of a hundred or more signs, and The only animal communication system that resem- have used them to communicate with humans and with bles language in respect of displacement is the dance each other. However, claims that these apes have mas- ‘language’ of bees, which encodes information about tered elements of ASL syntax are highly controversial. the direction, distance, and quality of food sources. Controversial, too, is the claim that a bonobo, or pigmy Until the 1980s, it was widely thought that animal chimpanzee, called Kanzi had at the age of eight calls were all subjective or attitudinal, concerned with acquired a systematic mastery of elements of English the animal’s state of body or mind (‘I’m afraid!’, ‘I’m syntax. What apes can achieve in situations of exposure hungry!’, ‘I want sex!’) rather than with identifiable to human language is remarkable, but their achieve- external characteristics of their environment. It has ments still differ considerably from those of humans. now been demonstrated, however, that vervet mon- Perhaps, then, it is not in having language that humans keys use distinct alarm calls for distinct predators are unique, but in having syntax. (eagles, leopards, and snakes). Furthermore, on hear- While sharing some features, animal communica- ing such a call a vervet reacts to the call itself, not to tion differs from human language in terms of size the predator whose presence is signaled. These calls of vocabulary, contextual dependence on the environ- lack displacement, but in other respects they have ment for communication, and syntax. Examining much of the objective semantic character of words in these aspects of both types of communication sys- a human language. tems, and the overlap between them, enables a fuller 63 ANIMALS AND HUMAN LANGUAGE 1: OVERVIEW understanding of language structure and utility. This Deacon, Terrence. 1997. The symbolic species. New York: more complete analysis could aid research in speech W.W. Norton and London: Allen Lane. Hauser, Mark D. 1996. The evolution of communication. and language disorders and cognitive psychology. Cambridge, MA: MIT Press. Hockett, Charles F. 1960. ‘The origin of speech’. Scientific References American 203(3). 88–96. Savage-Rumbaugh, E. Sue, and Roger Lewin. 1994. Kanzi: the Bickerton, Derek. 1990. Language and species. Chicago: ape at the brink of the human mind. New York: Wiley. University of Chicago Press. Calvin, William H., and Derek Bickerton. 2000. Lingua ex ANDREW CARSTAIRS-MCCARTHY Machina. Cambridge, MA: MIT Press. See also Animals and Human Language 2: Carstairs-McCarthy, Andrew. 1999. The origins of complex lan- guage. Oxford: Oxford University Press. Dolphins; Animals and Human Language 3: Par- Cheney, Dorothy, and Robert M. Seyfarth. 1990. How monkeys rots; Animals and Human Language 4: Primates; see the world. Chicago: University of Chicago Press. Syntax Animals and Human Language 2: Dolphins Can a dolphin master language? Scientists cannot animals can only learn easily in one modality; begin to answer this question until human language dolphins, like apes and humans, are cognitively has been fully mapped and its defining features agreed flexible enough to master more. upon. In the meantime, those interested in comparative (2) Symbol comprehension. The dolphins' signal cognition and in the evolution of human language can set, though small, is open, since the dolphins make great strides by asking smaller and more specif- have mastered signals for objects they could ic questions. Can a dolphin master an artificial signal never have encountered in their evolutionary system; that is, using signals that are not innate? Can history, (such as frisbees and surfboards). They it master a discrete–combinatorial system—here, one extend established signals to new examples, in which these artificial signals, each of fixed form, and they can comprehend such signals even can be combined with each other into strings? Can it when they are given for the first time as part of further master sequential rules, in which (functionally a new function (e.g. a question rather than a paralleling grammatical rules) the order of signals command). They can also master displacement, affects interpretation? the use of a signal for an object that is not pres- Since the 1950s, three groups of researchers have ent, indicating that for the dolphin, a signal worked with bottle-nosed dolphins on the production ‘calls to mind’ the referent of that signal. All or comprehension of such artificial, discrete–combina- this suggests that dolphins to this extent gen- torial systems, or ADCs. However, virtually all of uinely comprehend symbols, and are not just what is currently known and accepted about dolphins’ responding instrumentally. ADC abilities comes from the work of the last and cur- (3) The use of signal order. Each of the dolphins rent group, led by Louis Herman, and most of these has been given a different ‘grammar’. (For data come from two wild-caught female dolphins, example, to ask Phoenix to put the small ball Phoenix and Ake. (As is usual in animal language into the white hoop, the signal string is SMALL research, individual animals are reported by name.) BALL IN WHITE HOOP, while for Ake the This work has shown that dolphins can comprehend a same command is WHITE HOOP SMALL small discrete–combinatorial system involving a few BALL IN). Thus, Ake, at least, cannot be mem- sequence-based rules. Dolphin abilities appear to orizing a signal sequence and executing it include the following: mindlessly, one signal at a time, in the same order as it is presented. Furthermore, their (1) Both auditory and visual modality signals. equal success with two different orders is addi- Phoenix obeys computer-generated whistles, tional evidence that they are indeed operating while Ake obeys gestures. Many nonhuman from sequence-based rules. 64  ANIMALS AND HUMAN LANGUAGE 2: DOLPHINS The use of order to assign roles is robust; the dol- BALL TOUCH). Initially offered as evidence for phins are far less prone than small children to use only underlying mental recursive rules, these strings appear the meaning of signals to make sense of strings that to require merely iteration, a much simpler rule struc- violate established sequential rules (e.g. TOSS LEFT ture. The early claims of recursion have not been FRISBEE rather than the correct LEFT FRISBEE repeated, and to date no other exploration of recursion TOSS). Furthermore, if given a ‘grammatically’ pecu- has been reported. liar signal sequence, the dolphins will extract any cor- Sixth, Schusterman and Gisiner have challenged rect subsequence and act on it, whether it is at the the Herman group’s calculations of probability for beginning, middle, or end of the whole string. commands involving goal objects. However, it appears that regardless of which calculation is used, the dol- (4) Categorization of signals into classes (such as phins do reliably ‘pass the test’, albeit with more things, qualities, and actions). Diverse evidence mediocre scores. suggests that, as in human language, it is these Finally, Herman et al. have been criticized for apply- categories rather than the actual signals that are ing linguistic terms such as ‘sentence’, ‘modifier’, or ordered in the dolphins’ inner rules. indeed ‘language’ to the dolphins’ use of the ADC. (5) The use of the sign strings in diverse function- What may even be called ownership of linguistic terms al contexts. is a perennial argument within animal language Most of Herman et al.’s research postdates the ape research. Herman overtly uses linguistic terms as language controversies of the late 1970s, and has prof- hueristic devices, to provoke reassessment of defini- ited extensively from them. Partly by concentrating tions: others insist that such terms can be used only as almost entirely on comprehension rather than produc- defined within one or another established linguistic tion, the program seems to have eliminated most of the approach. The issue seems unlikely to be soon resolved. errors in method that discredited the early ape work, exceeds even current ape research in its methodologi- References cal rigor, and is highly replicable. Herman, Louis M. 1988. The language of animal language However, Herman’s group has been criticized, first, research: reply to Schusterman and Gisiner. Psychological for not pursuing production skills. Transferring symbol Record 38. 349–62. use between comprehension and production can be dif- Herman, Louis M., D.G. Richards, and J. P. Wolz. 1984. ficult for common chimpanzees, and one small study Comprehension of sentences by bottle-nosed dolphins. with Phoenix suggests that it may prove difficult for Cognition 16. 129–219. Herman, Louis M., and Paul H. Forestell. 1985. Reporting pres- dolphins. Perhaps, this is because Herman’s program ence or absence of named objects by a language-trained dol- uses food rewards rather than communicative training to phin. Neuroscience and Biobehavioral Reviews 9. 667–81. shape the dolphins’ behavior, and this is a second point Herman, Louis M., and Robert K. Uyeyama. 1999. The dol- of criticism. In apes, giving the same response—food— phin’s grammatical competency: comments on Kako (1999). to every correct signal use has been shown to inhibit Animal Learning and Behavior 27. 18–23. Kako, Edward. 1999. Elements of syntax in the systems of three development of maximally human-like symbol use. language-trained animals. Animal Learning and Behavior Third, exactly how many mental rules the dolphins 27. 1–14. use to interpret their ADC strings is unknown, but Premack, David. 1986. ‘Gavagai!’ or the future history of the (depending on how elaborate each rule is) it may be as animal language controversy. Cambridge, MA: MIT Press. few as two. This uncertainty is significant only if it is Roitblat, Herbert L., Louis M. Herman, and Paul E. Nachtigall (eds.) 1993. Language and communication: comparative assumed that the goal is to produce an animal with perspectives. Hillsdale, NJ: Lawrence Erlbaum Associates. extensive or full human-like language skills. More Schusterman, Ronald J., and Robert Gisiner. 1988. Artificial lan- realistically, it gives a direction for future research; guage comprehension in dolphins and sea lions: the essential how many, and how diverse, are the sequential rules cognitive skills. Psychological Record 38. 311–48. that this species can master? Schusterman, Ronald J., J.A. Thomas, and F.G. Wood (eds.) 1986. Dolphin cognition and behavior: a comparative Fourth, early claims that dolphins could master dis- approach. Hillsdale, NJ: Lawrence Erlbaum Associates. placement were justly disputed on the grounds that the Shyan, M.R., and L.M. Herman. 1987. Determinants of recog- tests involved were of memory, rather than of ability to nition of gestural signs in an artificial language by Atlantic use a symbol to access a concept. Subsequent tests bottle-nosed dolphins (Tursiops trunctatus) and humans have since given better proof of displacement. (Homo sapiens). Journal of Comparative Psychology 101. 112–25. Fifth, early and theoretically significant claims of DOROTHEA COGILL-KOEZ mastery of recursion were also justly disputed. The evidence in question was the ability of the dolphins to See also Animals and Human Language 1: respond correctly, on the first trial, to a sequence of Overview; Animals and Human Language 3: Par- two command strings (such as FRISBEE TOSS, rots; Animals and Human Language 4: Primates 65 ANIMALS AND HUMAN LANGUAGE 3: PARROTS Animals and Human Language 3: Parrots In an attempt to determine the relationship between uses ‘no’, ‘come here’, ‘wanna go X’, and ‘want Y’ human language and animal communication, (where X and Y are appropriate location or item researchers have often studied the potential for com- labels). He correctly responds to questions asking him munication between humans and animals. Beginning to judge category, relative size, quantity, similarity or in the early 1960s, most studies on animal–human difference in attributes, and to demonstrate label com- communication focused on great apes and a few prehension. When shown a colored wooden block, he cetacean species such as dolphins, as these animals can respond to questions of ‘What color?’, ‘What were assumed to possess the potential for complex shape?’, ‘What matter?’, and ‘What toy?’ Given two cognitive abilities considered as prerequisites for lan- objects, he can respond vocally to ‘What toy?’, ‘How guage development. Birds, with a few exceptions, many?’, ‘What’s same/different?’, ‘What color big- were rarely subjects of such studies as they were pre- ger/smaller?’, ‘What matter bigger/smaller?’ If given sumed to be merely talented mimics. For example, a tray with different numbers of, for example, red and experiments on pigeons using standard laboratory par- blue balls and blocks, he can respond to ‘How many adigms demonstrated capacities far inferior to those of blue blocks?’ with the appropriate quantity label. He mammals. These results were thought to represent the thus not only produces labels in response to simple abilities of all birds, despite evidence suggesting that queries but also comprehends questions that involve species such as jays, crows, and parrots might be capa- multiple labels and recursion. He combines labels to ble of more impressive cognitive and communicative identify, request, comment upon, or refuse more than feats. In particular, parrots, like primates and 100 items, and to alter his environment. He uses a cetaceans, have large brains, long lives, and highly very limited form of syntax, involving sentence social natures; couldn’t parrots also have evolved com- frames (e.g. ‘I want X’) and consistent application of plex cognitive capacities? Could proper training the adjective–noun order construction. He requests enable them to demonstrate language-like abilities objects in their absence and refuses substitutes, show- comparable to those of nonhuman primates and ing that he reacts to the dissonance between the con- cetaceans? cept encoded in his request and the proffered item. Research spanning more than 20 years has provid- Such data suggest that his labels are representational. ed positive answers to these questions. After being He semantically separates labeling from requesting. trained with a technique called the model/rival (M/R) He practices on his own and builds upon previous procedure, Gray parrots (Psittacus erithacus) learned knowledge to learn new sounds. After learning ‘gray’, to use aspects of English speech. In the standard M/R he spontaneously produced ‘grape’, ‘grate’, ‘grain’, protocols, two humans demonstrate the labeling abili- ‘cane’, and ‘chain’. When these words were rein- ty that a bird is to acquire. One human trains the sec- forced with the corresponding object, Alex quickly ond (the model/rival), i.e. presents and asks questions associated the two. about the item (e.g. ‘What’s here?’). The trainer Particularly noteworthy are additional data demon- rewards correct identification with the item, demon- strating that Alex’s labeling goes well beyond simple strating referential and functional use of labels, association of object and word. His labeling is equally respectively, by providing a 1:1 correspondence accurate when he is presented with items that are relat- between label and object, and modeling label use as a ed but not identical to training objects. He also trans- means to obtain the item. The second human not only fers use of utterances learned in one context to another. is a model for a bird’s responses and its rival for the Thus, after learning to state ‘none’ when asked trainer’s attention, but also illustrates aversive conse- ‘What’s different?’ about two identical items, he quences of errors: trainers respond to errors by scold- could, without training, give the same response when ing, and temporarily hide the object. Birds quickly asked ‘What color bigger?’ to two identically sized, acquire referential label use with such training. differently colored blocks. Although Alex lacks all but This two-way communication code was then used a very few verbs, he nevertheless exhibits communica- to test the parrots’ cognitive abilities. The oldest bird tive capacities once presumed limited to humans or in the study, Alex, labels more than 50 objects, seven nonhuman primates. Similar studies with younger par- colors, five shapes, up to 6 quantities, and three cate- rots indicate comparable results, demonstrating that gories (color, shape, material). He also appropriately Alex is not an exceptional Gray parrot. 66  ANIMALS AND HUMAN LANGUAGE 4: PRIMATES Studies on Gray parrots also show that birds given that certain levels of communication and cognitive training that lacks some aspect of input present in M/R development, and the responsible underlying neural protocols (reference, functionality, social interaction) structures, are present in birds and not limited to pri- fail to acquire referential English speech. If input fails mates or mammals. to include, for example, joint attention of the trainer and parrot to the object being labeled, Gray parrots, References like young children, do not learn the label. For similar reasons, Gray parrots fail to learn referential use of Pepperberg, Irene M. 1990. Conceptual abilities of some non- labels presented only through video or audiotapes. primate species, with an emphasis on an African Grey par- rot (Psittacus erithacus). Language and intelligence in Depending upon the organization of M/R input, these monkeys and apes, ed. by Sue T. Parker and Kathleen R. parrots may exhibit a form of mutual exclusivity simi- Gibson. Cambridge and New York: Cambridge University lar to that of young children: they replicate the stage in Press. which a child believes not only that every object has a Pepperberg, Irene M. 1991. A communicative approach to ani- label, but also that each object can have only one label. mal cognition: some conceptual abilities of an African Grey parrot. Cognitive ethology: the minds of other animals, ed. For example, what is called ‘dog’ cannot be called ‘ani- by Carolyn Ristau. Hillsdale, NJ: Erlbaum Associates. mal’. Such research on acquisition patterns may aid our Pepperberg, Irene M. 1998a. The African Grey parrot: how cog- understanding of other forms of exceptional learning: nitive processing might affect allospecific vocal learning. learning unlikely in the normal course of development Animal cognition in nature, ed. by Russell P. Balda, Irene but that can occur under certain conditions. These M. Pepperberg, and Alan C. Kamil. London and New York: Academic Press. results have implications both for teaching psychology Pepperberg, Irene M. 1998b. Cognitive capacities of avian sub- and linguistics students to ‘think outside the box’ and jects. Comparative psychology: a handbook, ed. by Gary for designing intervention programs for children who Greenberg and Maury M. Haraway. New York: Garland fail to acquire normal communication skills. Press. Although exceptionally rich, the behavior of the Pepperberg, Irene M. 1999. The Alex studies: cognitive and communicative abilities of Grey parrots. Cambridge, MA: parrots described here is fully equivalent neither to Harvard University Press. human language nor to the vocal behavior exhibited IRENE M. PEPPERBERG even by young humans. Some behavior patterns of these birds, however, are comparable to those See also Animals and Human Language 1: observed in humans, nonhuman primates, and Overview; Animals and Human Language 2: Dol- cetaceans. The parrots’ behavior thus demonstrates phins; Animals and Human Language 4: Primates Animals and Human Language 4: Primates Apes (here, orangutans, gorillas and chimpanzee In 1969, the Gardners reported the first great suc- species) are our closest animal relatives. Can they too cess with a chimpanzee named Washoe. (As is usual in master language? The earliest researchers to tackle this animal language research, individual animals are question, in the 1930s and 1950s, were the Hayes and reported by name.) They were followed through the the Kelloggs. Each team tried to raise an infant chim- 1970s and 1980s with many more chimpanzee proj- panzee to understand and produce speech. Although ects, led by researchers such as Premack, Rumbaugh, both chimpanzees reportedly learnt to understand some Fouts, and Matsuzawa, and one each with a gorilla, spoken English, only one produced words — a mere Koko, trained by Francine Patterson, and an orang- four, and these with much difficulty and apparently utan, Lyn Miles’ Chantek. with little understanding of their meaning. In these programs, many apes soon learnt to pro- The studies were useful, however, in showing that duce well over a hundred signs or symbols. They also speech is physically impossible for apes. Instead, in seemed to use their signs consciously, as true symbols. the second chapter of ape language research (ALR), Many birds and mammals can learn to press certain apes were taught to use manual signs, to arrange indi- buttons or produce certain behaviors ‘instrumentally’ vidual plastic symbols into symbol strings, or to point merely for rewards, rather like pushing buttons on a to or press symbols on keyboards. food dispenser. However, the ALR apes extended old 67 ANIMALS AND HUMAN LANGUAGE 4: PRIMATES signs to new examples of a referent, applying a sign the apes as meaningful learnt signs. Secondly, for ‘dog’, for example, to a particular dog they had even where clear signs were produced in never seen before. Some apes could explicitly distin- appropriate contexts and with the same two- guish referents (e.g. the actual dog) from the signs for word grammar as young children, critics said referents (here, the word dog), while many appeared that the apes' strings should not be assigned a to spontaneously invent new signs. Evidence like grammar as, unlike children, the apes did not this suggested that these apes were using their signs later develop a fuller grammar to justify such as true symbols in a human-like way, rather than just ‘rich interpretation’. instrumentally. (5) Statistical inadequacy: Those researchers who The researchers of the 1970s also believed that they did undertake careful comprehension tests had found evidence for rudimentary syntax in apes, at often gave such restricted answer choices that a the same level, and with the same patterns, as children high percentage of correct responses could in the two-word stage of language acquisition. Some have arisen from chance alone. apes, such as the Rumbaugh team’s Lana, could arrange (6) Lack of first trial data: When tested with new much longer strings of signs into regular orders. Others, sentences, apes were shown to respond quickly, such as the Premack team’s Sarah, could react appro- However, there is a crucial difference between priately to sequences that differed in sign order alone, ‘quickly’ and ‘on the first trial’. Only the latter such as RED ON GREEN vs. GREEN ON RED. can be used as evidence that the apes have From these circumscribed achievements, many access to grammatical rules, by which to researchers of the 1970s argued broadly that apes decode the novel sentence: as they are known to could master language. quickly master new variations on old routines, Criticism of the ALR accumulated, however, and it is possible that quick (but not instant) com- towards the end of the 1970s it reached overwhelming prehension of novel sentences is merely quick proportions. To varying degrees, different ALR pro- mastery of a new variation on a rote task. The grams up to 1980 suffered from the following: research of the 1970s typically did not identify this crucial first trial data. (1) Redundancy and formulae: Some apes could (7) Indeed, methodology and testing in general produce and respond correctly to long strings of were underreported in the work of this time. signs, but these strings commonly took on a few formulaic patterns. Rather than manipulating The turning point, however, was the work of a sin- grammatical rules, the apes could have been gle research group, led by Terrace and working with simply learning a few arbitrary sign sequences one chimpanzee, Nim. Careful analysis of certain in which one or two elements could be varied videos of Nim’s signing revealed the following: depending on the context. The apes, it was (8) His ‘grammatical’ regularities could be ascribed argued, were merely pressing rote sequences of to imitation. Nim certainly tended to order signs buttons on a machine, without seeing the but- as his teachers did, but the videos showed that tons as symbols or the sequences as grammar- his sign strings were commonly repetitions or governed: it was the researchers who were minor variations on those his teachers had just assigning words to buttons and thus translating, produced. His ‘grammar’ could therefore have for example, the button strings 21593 versus been the result of simply parroting the true 21598 into ‘Please machine give me M&M’ ver- grammar of his human companions. sus ‘Please machine give me milk’. (9) His ‘sentences’ could be unusually long, but (2) Contextual cues: Training and testing of signs these long strings consisted of empty repetition. and strings often took place in sets of many, vir- After entering the ‘two-word stage’, Nim’s sign tually identical repetitions. How much of the strings never increased in grammatical or apes’ success was due to grasping the signs semantic complexity. Children, conversely, themselves, and how much was due to their only produce longer sentences on demonstrably picking up the general pattern of the exercise? acquiring further grammatical structures. (3) Paralinguistic cues: It was often argued that the apes could have been reacting to cues in the Terrace then analyzed shorter video segments of body language and gaze of the researchers, apes in other programs, and asserted that the same pat- rather than to the signs themselves. tern of imitation could be seen in them. (4) Overgenerous interpretation: Researchers, it Further, symbol use came under scrutiny once was claimed, were like doting parents, inter- more. Were the apes consciously using symbols to preting the random or the natural gestures of refer, or were they just ‘button-pushing’ — behaving 68  ANIMALS AND HUMAN LANGUAGE 4: PRIMATES instrumentally? Savage-Rumbaugh and colleagues Savage-Rumbaugh and colleagues developed new then demonstrated that: methodologies to address this problem, finding that a more human-like social context was a key to human- (10) Their apes, which used signs appropriately in like symbol use. Their two chimpanzees, Austin and familiar contexts, were often unable to use the Sherman, repeatedly failed to transfer sign use to new same signs in new tasks. If the apes were gen- contexts. These failures were repeatedly overcome, not uinely processing the signs as symbols, this by yet more training for more rewards, but by adjusting failure is unexpected. It suggested that these the social interaction of the two apes from being ape- apes were indeed simply producing signs like to more human-like. Fuller, more human-like sym- instrumentally. bol use then flowed spontaneously from this. Finally, apart from all these methodological flaws, The Savage-Rumbaugh group also opened a new critics could always point to the enormous gap chapter in ALR research with another chimpanzee, between the small vocabulary, two-word combinations species, the bonobo. In the wild, this species exhibits, and food-focused conversations of the apes, and the some more human-like social features than the com- massive vocabulary, complex grammar, and wide- mon chimpanzee. This may be why captive bonobos ranging topics of an adult human. For many critics, appear capable of spontaneously acquiring human-like these differences alone were enough to justify the symbol use. Furthermore, they were found to acquire broad conclusion: ‘no ape has shown any abilities rel- symbols (including comprehension of spoken words) evant to language’. in the same way as is natural for humans — simply by Thus, in the late 1970s/early 1980s, the second exposure as an infant to meaningful symbol use. (The chapter of ALR closed in great acrimony. Critics latter discovery also showed that chimpanzees, like accused the ALR researchers of naivity and anthropo- humans, have an early critical period for effortless morphic bias; researchers in turn accused their critics ‘language’ acquisition.) of hysterical denial of ape–human continuity, and In many of Savage-Rumbaugh’s critical tests, the of constantly ‘shifting the goal-posts’ by hastily methodological problems identified in the 1970s were redefining language each time an ape passed a test of eliminated. The weight of evidence is now in favor of language. the idea that, under appropriate conditions, bonobos Throughout the 1980s, the third chapter of ALR and common chimpanzees can indeed acquire sym- began. Firstly, the Nim project itself came under fire. bols, rather than just instrumentally manipulate signs. The video segments Terrace analyzed were argued to Savage-Rumbaugh’s team have also presented evi- be atypical of Nim’s signing. These video segments, dence that bonobos can master grammatical cues such although long, were all taken from a single, socially as sign order, and even clause structure. Thus, Kanzi, and physically restrictive test situation (and one that their first successful bonobo, could respond appropri- the ape apparently found aversive). The tasks were ately and differentially to commands such as ‘Put the limited to those in which the chances of instrumental tomato in the microwave’ versus ‘take the tomato that’s signing were arguably maximized. Others who worked in the microwave outdoors'; an achievement that, in with this ape claimed that under better conditions, children, is held to indicate a grasp of clause structure. Nim signed differently. Nevertheless, as ALR moves into this new territory, Nim’s signing was also argued to be unrepresenta- new issues arise. Does Kanzi also exploit clause struc- tive of the species. The infant Nim had a chaotic social ture and symbol order to decide, for example, what environment; in less than four years, he had a turnover tomato to take, or is his grasp of semantics so good of nearly 60 trainers, almost none with any prior expe- that he is able to deduce much ‘grammatical’ informa- rience of apes or signing. So stressful was this that the tion just from the meaning of the symbols? chimpanzee became intractable at an unusually young Finally, brain imaging is beginning to enter ALR. age. If Nim was behaving abnormally in the social At the time of writing, published reports are limited to context, it is possible that his ‘linguistic’ behavior was only one common chimpanzee, from the Rumbaugh also abnormal. program, showing bilateral brain activity, rather than Most importantly though, Nim, like many other the typical left-hemisphere cortical activity of modern ALR apes of that time, was taught signs through drills humans. Little can be built as yet from this solitary and food rewards. Thus, ‘symbol’ learning and use result, but the chapter that opens here is likely to be an was in fact taught to these apes as an instrumental task. exciting one. The natural consequence of this is that their sign use ‘Can an ape learn a language?’ The weight of should indeed and predictably be instrumental. If so, evidence now indicates that under certain condi- such programs cannot reliably tell us much about an tions some species can master symbol use. Beyond ape’s ability to acquire symbols. this, rather than many firm conclusions, ALR has 69 ANIMALS AND HUMAN LANGUAGE 4: PRIMATES developed more appropriate methodologies and Premack, David, and A.J. Premack. 1983. The mind of an ape. focused more specific and answerable research ques- New York: Norton. Ristau, Carolyn A., and Donald Robbins. 1982. Language in the tions; 'Which aspects of human language can an ape Great Apes: a critical review. Advances in the study of master? Under what social and training conditions? behavior, Vol. 12, ed. by J.F. Rosenblatt, R.B. Hinde, C. How, within the brain, does it process its language- Beer, and M.-C. Busnel. New York: Academic Press. like behavior?’ This is not surprising and should not Rumbaugh, Duane M. (ed.) 1977. Language learning by a be disappointing. In any rapidly developing field, chimpanzee: the LANA Project. New York: Academic Press. Savage-Rumbaugh, E. Sue. 1986. Ape language: from condi- most past research, from the perspective of the pres- tioned response to symbol. Oxford: Oxford University Press ent, consists of demonstrations of what does not work and New York: Columbia University Press. and what is not true. ALR is in its infancy, it is to be Savage-Rumbaugh, E. Sue, Jeannine Murphy, Rose A. Sevcik, hoped that the field has decades, perhaps centuries, of Karen E. Brakke, Shelly L. Williams, and Duane M. research ahead. Rumbaugh. 1993. Language comprehension in ape and child. Chicago, IL: University of Chicago Press. Sebeok, Thomas A., and R. Rosenthal (eds.) 1981. The Clever References Hans Phenomenon: communication with horses, whales, de Luce, Judith, and Hugh T. Wilder (eds.) 1983. Language in apes and people. New York: New York Academy of Sciences. primates. New York: Springer-Verlag. Terrace, Herbert S. 1979. Nim. New York: Knopf and London: Gardner, R. Allen, and Beatrix T. Gardner. 1969. Teaching sign Eyre Methuen Ltd., 1980. language to a chimpanzee. Science 165. 664–72. Thompson, Claudia R., and Russell M. Church. 1980. An Lyn, Heidi, and E. Sue Savage-Rumbaugh. 2000. Observational explanation of the language of a chimpanzee. Science 208. word learning in two bonobos (Pan paniscus): ostensive and 313–14. non-ostensive contexts. Language and Communication DOROTHEA COGILL-KOEZ 20(3). 255–73. Peng, F.C. (ed.) 1978. Sign language and language acquisition See also Acquisition; Animals and Human Language in man and ape: new dimensions in comparative pedolin- 1: Overview; Animals and Human Language 2: Dol- guistics. Boulder, CO: Westview. phins; Animals and Human Language 3: Parrots Aphasia Aphasia is an acquired language disorder that occurs ‘focal’ means that the cerebral damage is restricted to after language skills have been completely developed, rather clearly circumscribed brain lesions, as opposed due to focal brain damage in the hemisphere that is to diffuse brain damage, that is, damage spread over dominant for language (usually the left hemisphere). the entire hemisphere. This diffuse cerebral damage In this definition, several other language and speech may cause language problems (for example, in demen- disorders are excluded. The term ‘acquired’ indicates tia), but these are quite different, and hence distin- that aphasia cannot be congenital. Of course, perinatal guishable from aphasia. brain damage may cause speech and language prob- Aphasia is the result of damage to the hemisphere lems, but these are not referred to by the term ‘apha- that is dominant for language, most often the left sia’. ‘Language disorder’ distinguishes aphasia from hemisphere. The right hemisphere is involved in com- pure articulatory and fluency disorders, which may or munication, however. Although damage in this hemi- may not be the result of brain damage, but are not cen- sphere does not affect grammatical or semantic tral to the language faculty. The fact that aphasia is a processing, communication may be impaired, because language disorder implies that (usually) the entire lan- of problems at the discourse level, often referred to as guage faculty is affected, that is, speaking, compre- ‘pragmatic’ problems. In some left-handers, damage to hension of spoken language, reading, and writing, the right hemisphere may result in aphasia. because the same language system serves these four modalities. Aphasia is by definition due to brain dam- Typology age. In some psychiatric diseases (e.g. schizophrenia), the production deficits may appear similar, but as there The symptoms of aphasia are rather diverse and vary, is no focal brain damage, this is not aphasia. The term depending on the site and extension of the lesion. In 70  APHASIA general, large lesions cause more severe deficits than Reading and writing are also impaired, comparable to small lesions and anterior lesions (in the frontal parts oral comprehension and production (see Figure 1). of the left hemisphere) cause an impairment in the ability to construct sentences, whereas lesions to the (2) Wernicke’s Aphasia more posterior parts (the temporal (and parietal) lobes) This aphasia type has been named after Carl Wernicke cause problems in word finding. These relatively (1848–1905), a German neurologist who was the first to selective impairments to the different language levels describe this syndrome. He located the lesion which can best be explained by the so-called ‘classic aphasia causes this aphasia type in the posterior parts of the left syndromes’. These are more or less accepted in the temporal lobe (see Figure 1). Speech production is flu- aphasiological world, although pure syndromes do not ent and well-articulated, but the aphasic speaker pro- occur regularly (estimated proportions in the entire duces sound and word substitutions, so-called aphasic population are between 15 and 100%). The ‘paraphasias’, for example ‘spool’ for ‘spoon’ or ‘moth- main types are the following. er’ for ‘wife’. Distortions of words may be further away from the target words than in these examples, and as a (1) Broca’s Aphasia consequence the listener may not recognize existing This aphasia type has been named after Paul Broca words; these sound strings are called ‘neologisms’, for (1824–1880), a French neurologist who was the first to example, ‘stroofel’. When many of these words are pro- mention that impairments to the faculty of speech duced, the text will no longer be understandable, which were caused by a lesion in the left frontal lobe, more is referred to as ‘jargon’ or ‘word salad’. specifically to the foot of the third frontal gyrus, now Comprehension of language is moderately to known as Broca’s area (see Figure 1). (N.B. he did not severely impaired. The combination of a severe com- describe what is now called Broca’s aphasia). This prehension and production disorder results in poor aphasia type is characterized by slow, elaborated communication abilities. Writing and reading are par- speech in which mainly content words (nouns, verbs, allelly affected. and adjectives) are produced; function words (deter- miners, prepositions, pronouns) are often omitted. This speech style is called telegraphic speech. (3) Anomic Aphasia In most of the speakers with Broca’s aphasia, artic- There is no clear lesion site in the brain that is respon- ulation is affected as well, because the brain site sible for this aphasia type, but the lesions are usually responsible for articulation is so close to the area that, found in the temporal and/or parietal lobe. Anomic if damaged, it causes Broca’s aphasia. Comprehension aphasia is characterized by word-finding problems. of spoken language is relatively good: in conversa- The oral output of a speaker with anomic aphasia is tions, speakers with Broca’s aphasia understand what basically fluent, but due to the pauses that arise from is said. When tested for the comprehension of complex the word-finding problems, speech rate may be low. syntactic structures, the performance drops dramati- The word-finding problems are not only recognizable cally. It is therefore assumed that in Broca’s aphasia by pauses, but also by the use of less specific words, the grammatical component of language is affected. such as ‘thing’, ‘do’, ‘here’, ‘there’. The latter is called ‘empty speech’. Comprehension of spoken language is relatively good, although complex commands are Wernicke's difficult to understand. Reading is usually spared as Area well, but in writing the word-finding problems show up again. (4) Conduction Aphasia Conduction aphasia was first described by, again, Carl Wernicke. The name of the syndrome derives from the fact that it is caused by a lesion in the tract that connects Wernicke’s and Broca’s area. It is characterized by the production of sound errors, for example, ‘scrample’ for ‘scrabble’. The patient is well aware of his errors and Broca's this results in many self-corrections, for example, Area ‘scramble, no strapple, no stample …’. The errors are particularly prominent when the patient is asked to repeat longer words and sentences. Comprehension of Figure 1. Left hemisphere of the human brain. spoken and written language is relatively good, but 71 APHASIA impaired for complex and long sentences. In writing, a References similar error pattern is shown as in speaking. Overviews Albyn Davis, G. 2000. Aphasiology: disorders and clinical (5) Global Aphasia practice. Needham Heights: Allyn & Bacon. Nadeau, S.E., L.J. Gonzalez Rothi, and B. Crosson. 2000. The lesion responsible for global aphasia is often large Aphasia and language: theory to practice. New York: The and extends to both the frontal, temporal, and parietal Guilford Press. lobes. All language modalities are severely impaired. Sarno, M.T. (ed.) 1991. Acquired aphasia. San Diego: Spoken output is often restricted to one- or two-word Academic Press. utterances, sometimes only ‘yes’ and ‘no’, not always Visch-Brink, E.G., and R. Bastiaanse (eds.) 1998. Linguistic levels in aphasia. San Diego: Singular Publishing Group. used correctly. Comprehension is severely impaired and reading and writing are hardly possible. This, of course, results in very poor communication abilities. Clinical Aspects Code, C., and D. Müller (eds.) 1995. Treatment of aphasia: Again, it should be stressed that although this clas- from theory to practice. London: Whurr Publishers. sification has contributed considerably to the under- Goodglass, H., and E. Kaplan. 1983. The assessment of aphasia standing of aphasia, most patients suffer from a and related disorders. Philadelphia: Lea & Febiger. ‘mixed’ form of aphasia, which cannot be labeled with Helm-Estabrooks, N., and A. Holland (eds.) 1998. Approaches any of the types mentioned above. to the treatment of aphasia. San Diego: Singular Publishing Group. Prognosis Linguistic Aspects Aphasia is caused by brain damage and this is normal- Grammar ly irreversible. This does not mean that no improve- Bastiaanse, R., and Y. Grodzinsky (eds.) 2000. Grammatical ment is possible. During the first few months and disorders in aphasia: a neurolinguistic perspective. London: especially the first few weeks following brain damage, Whurr Publishers. spontaneous improvement, if not recovery, is certainly Grodzinsky, Y. 1998. Theoretical perspectives on language deficits. Cambridge, MA: MIT Press. possible. If, however, the symptoms are persistent and are still present after 3–6 months, improvement is questionable. In most countries, aphasic patients are Lexical semantics Ellis, A.W., and A. W. Young. 1996. Human cognitive neu- treated by speech therapists. The goal of speech thera- ropsychology. Hove: Psychology Press. py is to teach the patient to deal with his language problems, through improvement of communicative strategies as well as language competence. Phonology Whitaker, H.A. (ed.) 1988. Phonological processes and brain mechanisms. New York: Springer-Verlag. Psychosocial Consequences Above, the linguistic aspects of aphasia were empha- Pragmatics Joanette, Y. and H. Brownell (eds.) 1989. Discourse ability and sized, since aphasia is a language disorder. One should brain damage. New York: Springer-Verlag. realize, however, that a language problem has severe consequences for communicative abilities and therefore ROELIEN BASTIAANSE has a severe impact on the daily life of the patient and his surroundings. Nowadays, part of the speech therapy See also Brain Organization and Auditory Pathways is focused on this psychosocial aspect of aphasia. Applied Linguistics: Overview Applied Linguistics, in its present form, is an emerg- use, more specifically, into verbal communication ing multidiscipline, or trans-discipline, developed on within a given social and/or institutional context. a multidisciplinary basis of inquiry into language in Theoretical and methodological problem-oriented 72  APPLIED LINGUISTICS: OVERVIEW approaches fundamentally define applied research of knowledge that served as the basis of much of the objects and require the development of multidiscipli- work done in the field since its beginnings, in the nary models and methods to fit the process-oriented 1950s, as a branch of linguistics. In fact, as an (rather than product-oriented) and multifaceted con- emerging scientific area in its own right, Applied textualized uses of language that researchers in this Linguistics is increasingly interested in research ques- field of study are facing. tions that differ qualitatively from the language-relat- Such a multidisciplinary basis of inquiry involves a ed research carried out in linguistics, as well as from combination of approaches from linguistics and several the traditional nontheoretical practical applications of other disciplines within the domain of human sciences, mainstream linguistic findings, which developed when such as psychology, sociology, pedagogy, anthropology, the term ‘applied linguistics’ was used to designate a ethnography, philosophy, politics, and psychoanalysis, subfield of inquiry within the mainstream field of as well as newer areas of investigation such as, for Linguistics. The primary concern of early applied lin- instance, computer sciences and cybernetics. This guists was to develop empirical material for foreign inquiry leads to a number of new insights and issues that language teaching, which was inspired and oriented by are extremely relevant in response to the challenge of theories, methods and purposes, and, based on linguis- penetrating deeper into epistemological and ethical tic studies, mainly structural linguistics. questions imposed by the complexity of the real world of Despite the fact that this traditional role assigned to human interactions, as well as by the dynamism of lan- applied linguists is still emphasized by several scholars guage in use, i.e. of language as parole, in Saussurean working on language learning methods and models, the terms, or performance, in Chomskyan terms. fundamental reevaluation of traditional linguistic con- Thus, in applied research, real human interactions ceptions of language, subject, and context by different are seen as a complex object of inquiry with a crucial areas of the social sciences has begun to change the social and institutional, as well as a psycholinguistic understanding of these ideas and seriously challenges dimension, and not simply as an individual or speak- the predominant modes and forms of knowledge of er/hearer subject matter. According to this view, there applied research tradition. Over the past two decades, are also a number of political and ideological aspects of much of the relevant discussions about epistemological language and discourse to be accounted for, since they and ethical issues within the subfields of bilingualism, shape and are shaped by everyday current linguistic language planning, language teaching and learning, and interactions. Similarly, in applied research, language in second and foreign language acquisition, for example, use is viewed as a complex processual object com- involve critical rethinking of traditional approaches. posed of observed units compiled into a system, which A good example of this increasingly changing is characterized not simply by a static range of appro- process in action in the applied research tradition is the priate rules to constrain the shape of words, sentences well-established subfield of second and foreign lan- and texts, but by continuous change and variability. guage teaching and learning. A large body of recent Although this growing research perspective began work seeks to understand the role of teachers’ and to emerge in the 1980s and covers a wide range of learners’ reasons and accounts in the teaching and issues that go beyond traditional linguistic concerns, it learning process, as well as the role of the historical, has not rendered irrelevant the significant and ongoing social, and politicoideological aspects of the macroso- role of the theoretical and methodological contribu- cial context of society as a whole in the microinterac- tions of linguistics in applied studies. Nevertheless, by tional context of the school as an institution, which adopting a view of language as a complex and dynam- also shapes teacher/learner exchanges in the class- ic object, applied studies have challenged various room. Another relevant contribution to this subfield aspects of linguistic tradition and have attempted to has to do with language and discourse as important offer alternative approaches and issues. A classic sites of subjectivity construction on the one hand, and example of such a result of applied inquiry is the rele- as important moves in the reproduction of power rela- vant theoretical and methodological discussions and tionships on the other. Indeed, the purpose of much of findings that led to and emerged from the sociolin- the contemporary research on teacher training and guistic notion of ‘communicative competence’ as the instructional materials and methods focuses on explor- ability to use language appropriately in varying social ing these new concerns. Research that explores ethno- contexts, i.e. in communicative practices that are cul- graphic insights and methodologies has been crucial turally and historically situated. for most of the work involving classroom linguistic As many different authors have pointed out over the interactions, as well as that of collaborative researcher/ last two decades, through multidisciplinary research teacher work in teacher training projects. efforts, Applied Linguistics is also arguing for and In general terms, a great deal of recent rethinking responding to critical accounts of the received theories about traditional modes and forms of knowledge in 73 APPLIED LINGUISTICS: OVERVIEW applied research has drawn on a critical view of posi- Graddol, David (ed.) 2001. Applied linguistics for the 21st cen- tivist epistemologies inherited from mainstream scien- tury. AILA Review, Vol. 14. Association Internationale de Linguistique Appliquée. tific tradition in order to face the new challenges Hymes, Dell. 1972. On communicative competence. outlined above. In response to the deficiencies of pos- Sociolinguistics, ed. by J.B. Pride and J. Holmes. itivist models of understanding language, subject, and Harmondsworth: Penguin Books. context, exploratory analyses of the complementary Hymes, Dell. 1992. The concept of communicative competence roles ascribed by other epistemologies to inductive revisited. Thirty years of linguistic evolution, ed. by M. Putz. Philadelphia: Benjamins. and abductive thinking and to qualitative and quantita- Kaplan, Robert B. (ed.) 1980. On the scope of applied linguis- tive data and methods, for instance, ought to be essen- tics. Rowley, MA: Newbury House. tial. The attempts to model the dynamism and Larsen-Freeman, Diane. 1991. Second language acquisition variations of language as an open and changeable sys- research: staking out the territory. TESOL Quarterly 25. tem have, likewise, been of fundamental importance. 315–50. Mauranen, Anna and Kari Sajavaara (eds.) 1995. Applied lin- guistics across disciplines. AILA Review, Vol. 12, Oxford: References Catchline/AILA. Adger, Carolyn and Jeffrey Connor-Linton. 1997. Ethical issues Pennycook, Alastair. 2001. Critical applied linguistics: a critical for applying linguistics: introduction and prologue. Issues in introduction. Mahwah, NJ: Lawrence Erlbaum. applied linguistics 4(2). 169–77. INÊS SIGNORINI Cristal, David. 1981. Directions in applied linguistics. London: Academic Press. See also Chomsky; Noam; Saussure; Ferdinand de Arabic Arabic is spoken natively by approximately 170 mil- exhibits innovations not found in other Semitic lan- lion people (Arabs) living in dozens of countries, most guages, such as a definite article with l. in what is known as ‘the Arab world’, which stretches The origins of the term “Arab” are unknown. from Iraq in Southwest Asia to Morocco in Northwest Assyrian, Babylonian, and Hebrew use of the word Africa. It is the language of prayer and scripture for refers to nomads coming from the desert. The earliest approximately one billion Muslims. Originally the clearly attested use of Arabic (written in the Nabataean language of some tribes of the Arabian Peninsula, it script) dates to a 328 CE tomb inscription in the Syrian came to be a major world language as a result of the desert. The oldest samples of writing (graffiti and Islamic conquests that began in the seventh century inscriptions) in a script that can be seen as Arabic are CE. In 1974, Arabic became one of the six official lan- also found in Syria and date to the fourth and sixth guages of the United Nations. centuries. Arabic is a Semitic language, one branch of the The Arabic alphabet developed from that of the Afroasiatic language family. Scholars long considered Nabataeans. Unlike Nabataean, Arabic is a cursive Arabic as a relatively pure and unchanged ancient script. It consists of 28 letters and is written from right Semitic language, practically identical to a proposed to left. As is typical of Semitic alphabets, the Arabic Proto-Semitic, but this position is now widely viewed alphabet contains no letters to indicate short vowels. For as a romantic notion that saw the deserts of Arabia as example, a word written as qbl could be read in a num- the guardian of a relatively untouched language. ber of ways, including qabila ‘he accepted’ or qabla Arabic is conservative in many respects, but also inno- ‘before’. Like Nabataean, early forms of the Arabic vative. It alone preserves the full ancient Semitic case- alphabet failed to distinguish between some phonemes. marking system found in Old Akkadian. It participated Typically, a letter could be read in one of two ways, for in or was considerably influenced by some develop- example, /z/ or /r/. Given the absence of short vowels ments found only in South Semitic languages such as and the consonantal ambiguity, deciphering the mean- South Arabian and Ethiopic, as well as some found ing of a given text could pose problems. With the need only in the Northwest Semitic language complex, for more precision, a system that did away with this which includes Aramaic and Hebrew. Arabic also ambiguity became standard. For example, /z/ is marked 74  ARABIC with a dot above the letter in order to distinguish it from speaking a local variety of Arabic that differs consid- /r/. A method for marking short vowels and other erably from MSA and that is officially denounced but phonological details was also devised, but these mark- nevertheless thrives. For example, it is the vehicle of ings are used mostly in scripture and in materials for most popular music and television programs. beginners, such as children’s stories. In practice, the MSA is the modern realization of at least the formal absence of short vowels in most texts is not a problem; register of Arabic used in parts of the Arabian context almost always indicates the correct reading (and Peninsula in the sixth and seventh centuries. This vari- where this is not the case, a marker is sometimes sup- ety of Old Arabic was apparently the language of an plied to assist the reader). That the Arabic alphabet is extensive pre-Islamic oral poetry and it is the language relatively easy to learn comes as a surprise to foreigners of the Qur’an (‘reading, recitation’), the book of scrip- accustomed to seeing the ornate and often complicated ture that Muslims believe the Prophet Muhammad calligraphic styles of Arabic that are used in decorating received over a 23-year period, beginning in 610 CE. mosques or book covers. Calligraphy plays a central Due to a power vacuum, within a hundred years of role in Islamic art due to the general avoidance of Muhammad’s death, a relatively small number of Arabs human and animal figures in art and due to the promi- and their allies came to control a vast empire stretching nent role of scripture in Islam. from present-day Spain and North Africa to the borders The domination of consonants in Arabic orthogra- of India. In time, the religion of Islam (and, therefore, phy is indicative of a defining characteristic of Semitic Arabic for religious purposes) spread to such far-flung languages known as the root and pattern system. Most places as China and what is now Indonesia (the latter words are based on a root consisting of three conso- being the most populous Muslim country). nants. Arabic, more than any other Semitic language, The language of the Qur’an is commonly held by excels in exploiting this morphological resource. Each Muslims to be that of God. Its sound and content are root is associated with a core meaning. The systemat- felt to be so interwoven that translation is considered ic manipulation of these roots is the basis of Semitic impossible. As a result, there are no official transla- derivational morphology. For example, the root d-r-s is tions; however, translations that attempt to impart its associated with studying. The verb darasa ‘he studied’ meaning are available. All Muslims are expected to is based on the template or pattern C1aC2aC3a, the recite scripture and pray in Arabic even if they do not final vowel being the third-person masculine singular understand what they are saying or reading. perfect conjugation marker; the corresponding The nature of spoken Arabic before the Islamic con- yaC1C2uC3u yields yadrusu ‘he studies’ (the prefix ya quests is debated. Early sources mention dialectal vari- marks it as third-person masculine imperfect and the - ation in the Arabian Peninsula but give only glimpses u suffix indicates indicative mood). Doubling the mid- of it. Personal letters and other documents dating to dle consonant (gemination) of the root—one method the early Islamic centuries provide much more insight for forming causative verbs—produces the verb pat- into the nature of early spoken Arabic. From these, it tern C1aC2C2aC3a and results in darrasa ‘he taught’. is clear that the essential characteristics of the modern The corresponding imperfect, based on the pattern Arabic dialects (collectively called New Arabic), such yuC1aC2C2iC3u, is yudarrisu ‘he teaches’. The corre- as the absence of case marking in nouns and adjec- sponding active participle, based on the imperfect verb tives, go back to at least the first centuries of the stem, is mudarris ‘teacher/teaching’ (muC1aC2C2iC3). Islamic era. In other words, the language situation of The ‘noun of place’ pattern maC1C2aC3a(t)—the t is the modern Arab world—where a variety of Old not pronounced in some contexts—yields madrasa(t) Arabic is used for formal purposes and a variety of ‘school’. New Arabic (or something akin to it) is the common ‘Arabic’ is in fact a cover term for a number of lan- language—is over a thousand years old. There is no guage varieties that could easily be designated distinct evidence that New Arabic developed later than Old languages, given the right social and political circum- Arabic. On the other hand, there is some evidence that stances. The Arab world is united by a standardized points to Old Arabic containing innovations not known variety of Arabic used mainly for writing but also for in New Arabic or other Semitic languages. some formal oral communication. Regional differ- New Arabic was possibly formed or at the very ences exist but this is mostly restricted to usage and least significantly affected by the language contact sit- pronunciation. Linguists often refer to this officially uation that arose from the Islamic conquests. Arabic sanctioned variety as Modern Standard Arabic (MSA). speakers mixed and intermarried with non-Arabic MSA is typically learned at school and most Arabs, speakers and Arabic began to gain hold in places far even those who obtain advanced degrees in an Arab from the Arabian Peninsula. In a gradual process that educational system, lack confidence (and typically took centuries, various populations came to adopt experience) in using it. In contrast, Arabs grow up Arabic and Islam. Subsequent migrations from the 75 ARABIC Arabian Peninsula aided in the spread of Arabic. In the of Arabic speech communities, it is now taking place case of places like Persia (modern Iran) and Turkey, on a far wider scale. There are few ‘pure dialect’ the populations embraced Islam but the relatively few speakers. MSA elements (mostly lexical) are common Arabs who settled there eventually assimilated the lan- even in informal speech. The importation of more guage of the local inhabitants. Nevertheless, both informal rhetorical styles (perhaps connected with Persian and Turkish borrowed a great deal from Western-style democratic values or populist rhetorical Arabic, including its alphabet. This is true of other approaches) may have contributed to New Arabic Asian and African languages. spreading to domains and functions formerly associat- Although the pre-Islamic Arabs had a highly devel- ed with Old Arabic. Whatever the causes, new Arabic oped poetic tradition, they did not have a prose or aca- is commonly used—with a substantial admixture of demic literary tradition. This changed soon after the Old Arabic lexical material in particular—in contexts advent of Islam. Classical Arabic developed in where it would not have been expected in the past, response to the new demands placed on Old Arabic. It such as in sermons and university lectures. is theoretically based on the language of the Qur’an. In A distinctive characteristic of the phonology of fact, there are distinct differences (orthographic, gram- nearly all varieties of Arabic is the presence of matical, stylistic…). The vocabulary of Classical emphatic consonants paired with nonemphatic coun- Arabic expanded considerably to meet the needs of the terparts, for example: /t/ and /t· /, /d/ and /d· /, /s/ and /s./. empire by deriving new terms from Arabic roots. Emphatic consonants are formed by raising the back Classical Arabic soon became the vehicle of an of the tongue, which simultaneously results in the tip impressive corpus of literary and scientific works, of the tongue being drawn back somewhat from the spurred on by the Arabs’ contacts with the advanced point of articulation of the corresponding nonemphat- cultures that they had conquered. For example, the ic consonant. Old Arabic has only three vowels, /a/, /i/, Abbasid dynasty in Baghdad aggressively pursued the and /u/—as in the case of consonants, these may be translation of Greek philosophical and scientific texts lengthened with a corresponding difference in mean- into Arabic. Scholars writing mostly in Arabic pro- ing. There is considerable allophonic variation in the ceeded to considerably build on these works. These vowels, which is conditioned by the consonants in a scholarly works, in turn, were passed on to Western word. Pronunciation and word stress in MSA are typ- Europe mostly via Spain and Sicily and significantly ically influenced to a great degree by one’s native contributed to European cultural and scientific devel- dialect. New Arabic realizations of the phoneme cor- opment. Particularly in the 12th century, and for some responding to Old Arabic /q/ have become a special time after, learning Arabic was popular in Europe in regional and social marker. The extreme case is order to read the scientific works of the Greeks (many Jordan, where in some rural dialects it is realized as of which were preserved only in Arabic) as well as to [k], but for Bedouins it is [g]. Its urban distribution is read the contributions of the Arabs themselves. complicated: men not of Palestinian extraction typi- Through such contact, European languages came to cally use [g], perhaps for its association with Bedouin borrow numerous words of Arabic origin, such as toughness, whereas men from Palestinian urban fami- alcohol, asthma, and zenith. lies might use [ʔ], unless they want to downplay their MSA is the modern manifestation of Classical Palestinian origins. Urban and increasingly nonurban Arabic. However, Classical Arabic’s ornate styles and women favor [ʔ], which is seen as sophisticated. some of its complex grammatical constructions are Much of the morphology of Arabic is a model of rarely used. Again, a major difference is found in logic and order. This is not true, however, of the rela- MSA’s new vocabulary and phrasing. Like Classical tionship between singular and plural forms of nouns Arabic, MSA arose in response to a language contact and adjectives. There are two types of plurals, ‘sound’ situation—with modern Western Europe, which is usu- and ‘broken’ (which designate whether the stem is left ally dated to the French invasion of Egypt in 1798. The intact or is modified to form the plural). Sound plurals importation of Western institutions, like newspapers, are formed by suffixation, for example: mudarris/ has had a profound impact on Arabic. Whereas docu- mudarrisuuna ‘teacher/teachers’. The plural of most ments composed before the modern era in Classical nouns is of the broken type and numerous patterns are Arabic typically targeted a limited and elite audience, used, for example: safiir/sufaraaʔ ‘ambassador/s’, the spread of education—particularly in the second half madiina(t)/mudun ‘cities’, jariida(t)/jaraaʔid ‘news- of the twentieth century—has resulted in MSA, a ver- paper/s’. Some plural patterns are predictable from the sion of Old Arabic streamlined for mass consumption. pattern of the singular, most are not. Some nouns have The spread of education (among other factors) has two or more plurals. also resulted in a blurring of the lines between New The basic word order of Old Arabic is considered and Old Arabic. While mixing has long been a feature to be Verb–Subject–Object, whereas New Arabic is 76  ARABIC classified as being of the Subject–Verb–Object turally and scientifically superior to the West. It is a type. Both, however, exhibit a good deal of flexibility. symbol of Muslim unity, as well as the cornerstone of The head noun of a noun phrase precedes adjectival Arab nationalism. As such, it is jealously guarded. modifiers, which agree with the head in number, gen- The realities of the modern world undercut this lin- der, and definiteness (as well as grammatical case in guistic idealism. In their professional lives, many edu- Old Arabic). Old and New Arabic exhibit intricate pat- cated Arabs are more adept or at least more terns of agreement variation. Numeral constructions comfortable in expressing themselves in English or and the negation of verbs are particuarly complex in French than in MSA. Some influential members of Old Arabic. Arab society criticize the Arabic abilities of recent Arabic has played an important role in Western lin- high school graduates while simultaneously sending guistics on a number of fronts. The root-and-pattern their own children to English or French private system has led to new approaches to morphology. schools. Economic forces similar to those that resulted Arabic has played a particularly prominent role in soci- in their ancestors’ learning and eventually adopting olinguistics. Charles Ferguson, a founding figure in the Arabic are now drawing young Arabs to write French field, was profoundly influenced by his contact with email messages to Arab friends or to study medicine in Arabic. Some mark the publication of his paper English at Arab universities. In short, bilingualism is ‘Diglossia’ in 1959 as the beginning of sociolinguistics, widespread and growing. in which he undertook to describe the particular lan- The Arab world is, nevertheless, far from being on guage situation typical of Arabic speech communities the verge of abandoning MSA. In many ways, the in which two closely related but substantially different modernization of Arabic (its adaptation to the needs of varieties of a language (a ‘High’ and a ‘Low’ variety) a changing world) over the past two centuries has been exist side by side in a single speech community in a sta- quite successful. Moreover, a far greater percentage of ble and functionally complementary relationship. This Arabs are literate in Arabic than at any time in history. article has resulted in some 3,000 studies of the topic. Nevertheless, one wonders what the future holds. The dynamics of Arabic speech communities have Arabic changed a great deal as a result of the contact also played an important role in language attitude stud- situation that arose with the Islamic conquests. The ies, and in challenging theories of language and gender. present is also a time of intensive language contact. In The fact that some modern spoken varieties of Arabic addition to contact with the West, the very process of are mutually unintelligible underscores the difficulty of educating the masses has resulted in a new contact sit- defining precisely what a language is (as opposed to a uation. Millions of Arabs with limited proficiency in dialect). For example, Maltese, the official language of MSA are using it on a day-to-day basis. While it is not the island nation of Malta located near Sicily, is con- likely that they will wrench the language from the sidered by scholars to be a dialect of Arabic, whereas hands of the traditional gatekeepers, it nevertheless its speakers typically deny this relationship, prefering seems unlikely that it could acquire so many users to see it as a Semitic language whose roots go back to without their leaving their mark on it. ancient Phoenician. Maltese is the only variety of New Arabic to become a national language. It is written with References Latin characters, with some modifications. Arabic provides a fascinating case study in lan- Bakalla, Mohammad H. (ed.) 1983. Arabic linguistics: an intro- guage standardization and adaptation. Few languages duction and bibliography. London: Mansell Publishing. Belnap, R. Kirk, and Niloofar Haeri (eds.) 1997. Structuralist provide 1,400 years of documented continuous use. In studies in Arabic linguistics: Charles A. Ferguson’s papers, many ways, it has changed remarkably little over the 1954–1994. Leiden: Brill. centuries. This applies to both Old and New Arabic, Ferguson, Charles A. 1959. Diglossia. Word 15: 325–40. which share far more in common than is genuinely Fischer, Wolfdietrich, and Otto Jastrow. 1980. Handbuch der admitted. The modern dialects are so conservative that arabischen Dialekte. Wiesbaden: Otto Harrassowitz. Haeri, Niloofar. 1996. The sociolinguistic market of Cairo: gen- in at least some ways they are more like Pre-Islamic der, class, and education. London/New York: Kegan Paul. Arabic and the language of the Qur’an than is the stan- Holes, Clive. 1995. Modern Arabic: structures, functions, and dardized MSA. varieties. London/New York: Longman. The differences between MSA and an Arab’s day- Versteegh, Kees. 1997. The Arabic language. New York: to-day speech are real and do present challenges. Some Columbia University Press and Edinburgh: Edinburgh University Press. have called for teaching Arab children basic literacy R. KIRK BELNAP skills in their native dialect, but there is little chance of this idea taking hold in the Arab world in the near See also Afroasiatic; Algeria; Diglossia; Ferguson, future. Old Arabic holds tremendous symbolic value. It Charles Albert; Language: Contact—Overview; binds Arabs to their golden age, when they were cul- Semitic Languages 77 ARABIC TRADITIONAL GRAMMAR Arabic Traditional Grammar The essential quality of Arabic linguistics is its prag- (we would now say ‘function’) appropriate to matism, that is to say, its concern with what can be their status, just as the moral value of an action done with words in terms of the speaker’s intentions, is determined by the place and time of its the listener’s expectations, and the context of the utter- occurrence. ance. The Arabic word for grammar itself is nah.w, lit. ‘way, manner’, i.e. way of speaking, using the same It is a principle of Arabic grammatical analysis that metaphor as other key Islamic terms for kinds of every element operates (or not, according to its status) behavior, such as sunna ‘way [of orthodoxy]’, sB ar¯‘a on the following element, this operation being termed ‘path [of the law]’, t.ar¯qa ‘way [of the mystic]’, ‘amal, lit. ‘effect.’ Although this very much resembles madhab ‘way [of thinking]’, i.e. ‘school of thought’, the western notion of ‘government,’ there is no histor- ¯ ical or technical connection; ‘government’ is hierar- and it is consequently not surprising that Arabic gram- matical analysis should deal primarily with the exter- chical, based on a vertical (paradigmatic) relationship nal features of language, nowadays generally called between words, while the Arabic ‘amal is purely linear the surface structure, rather than the implicit or hidden (syntagmatic), denoting the immediate effect of an features, viz. the underlying or deep structure. ‘operator’ (‘a¯ mil) on another element ‘operated on’ ‘Meaning’, therefore, is literally’ what you have in (ma‘mu¯ l f ¯hi) sequentially. mind’ (ma‘na¯ ) or ‘what you intend’ (mura¯ d), as mani- Arabic linguistics also recognizes analogy (qiya¯ s), fested in and only accessible through what you actual- both as the inductive device for extracting general ly say (kala¯ m ‘speech’). principles from the data and as the means by which Accordingly, the criteria for correct speech are the speakers create new utterances by extrapolating from same as those for ethically correct behavior. A wholly speech patterns already known. successful utterance must be both ‘good’ (h.asan, lit. (2) Morphological categories are very few, being ‘beautiful’), i.e. well-formed, consisting of the proper effectively limited to three classes of words: words in the proper order, and semantically ‘right’ Noun, Verb, and Particle. This division is based (mustaq¯ m, lit. ‘morally straight’), i.e. conveying to on a combination of formal and semantic prop- the listener exactly the meaning intended by the speak- erties: the Noun (ism, lit. ‘name’) and Verb er. Ill-formed utterances (qab¯h., lit. ‘ugly,’ hence ‘eth- (f ¯‘l, lit. ‘act, action’) are distinguished by their ically bad’) may still be ‘right’ if they succeed in meaning and by certain unique features (only conveying the speaker’s intended meaning, just as a Nouns have Definite Articles, only Verbs have well-formed utterance may fail to do so, and will thus Tense, for example), while the Particle (h· arf, be termed g·ayr mustaq¯m ‘not right’. Finally, some lit. ‘bit, piece’) is classified negatively by the utterances convey no meaning at all, because of inner absence of a generic meaning and lack of the structural and semantic contradictions: these are formal markers of Nouns or Verbs. The full def- labeled muh.a¯ l (‘wrong’, lit. ‘perverted, twisted’), i.e. inition of the ‘Particle’ usually includes the incoherent. Significantly, in the earliest grammar at qualification ‘which occurs for some [syntac- least, before the influence of Greek logic appeared, the tic] meaning’ (ja¯ ’a li-ma‘na¯ ) to separate it from truth or falsehood of an utterance were linguistically other linguistic units also called h· arf, which irrelevant. would nowadays be labeled ‘phoneme’, For convenience, Arabic linguistic theory will be ‘grapheme’, ‘lexeme’, and other nonsyntactic dealt with in three terminological groups: (1) system- units. atic, (2) categorical, and (3) functional. (3) The functional categories reveal the ultimate (1) The systematic terminology includes the crite- preoccupation of Arabic linguistics with speech ria of correctness mentioned above, to which as behavior. There are more than 70 terms for two further terms of similar ethical origin can the various ways in which words can be used, be added: each item in an utterance has a ranging from the decisions of the speaker at the given ‘status’ (manzila) and a corresponding highest level (e.g. ibtida¯ ’ ‘starting an utterance’ ‘place’ (mawd·i‘). The formal correctness of an [specifically a statement of the form ‘x is y’, utterance depends on using words in the place where Arabic has no verb for ‘is’], istifha¯ m 78  ARAMAIC ‘asking a question’) to the specific events occur- substantial changes to the theory as Islamic culture ring within the utterance as a result of grammat- absorbed the methods of Greek science, leading in one ical rules (e.g. nas·b ‘making dependent’ [scil. direction to a marked pedagogical and scholastic char- ‘accusative’ or ‘subjunctive’], tatniya ‘making acter and in another to increased abstraction and grow- ¯ dual’). The precise boundary between the ing interest in underlying features. What might be speaker’s autonomy and submission to gram- called the metaphysical aspects of linguistics, such as matical rules is never very clear, but this is the origins of language itself, the nature of meaning, intentional, as both the speaker and speech ele- and the question of grammatical causality, were enthu- ments can be ‘operators’ (‘a¯ mil, see above), i.e. siastically and revealingly explored but, as might be affect the form of words. It is noteworthy that expected, they had to be dealt with in a manner that the 70 or more names of speech acts are all did not conflict with Islamic doctrines. expressed in Arabic by Verbal Nouns (‘inflect- ing’, ‘making diminutive’, ‘describing [adjecti- vally]’, ‘eliding’, ‘doubling [a sound]’ etc.), so References that Arabic grammar, like an Islamic legal trea- Auroux, Sylvain, E.F.K[onrad] Koerner, Hans-Josef Niederehe, tise, is nothing but a list of actions and condi- and Kees Versteegh (eds.) 2000. Handbook of the language tions for their permissibility. sciences, Vol. 1. Berlin, New York: Walter de Gruyter. Bakalla, Muhammad H. 1983. Bibliography of Arabic linguis- The many functions enable words of the same form tics, 2nd edition. London: Mansell. class to be subcategorized: thus, Adjectives and Bohas, George, Jean-Patrick Guillaume, and Djamel E. Adverbs are simply Nouns used in a certain ‘way’ Kouloughli. 1990. The Arabic linguistic tradition. London (nah·w) or ‘place’ (mawd· i‘), and conversely and New York: Routledge. Carter, Michael G. (ed.) 1981. Arab linguistics, an introductory Demonstrative and Relative Pronouns, which do not classical text with translation and notes. Amsterdam/ have the form of Nouns, nevertheless have the ‘status’ Philadelphia: John Benjamins. (manzila) of Nouns and are thus labeled ‘noun of Carter, Michael G. 1991. ‘Lexicography’ and ‘grammar’. The pointing’ (ism ’ia¯ ra) and ‘noun of the attached Cambridge history of Arabic literature, Vol. 1, Religion, [clause]’ (ism al-maws·u¯l), respectively. The western learning and science in the Abbasid Period, ed. by Michael J. Young, et al. Cambridge: Cambridge Univerity Press. Prepositions, Interjections, Conjunctions, and Articles Carter, Michael G. 2004. Sι¯ bawayhi. Oxford: Oxford Univer- are all considered Particles (h·arf), and are categorized sity Press/1. B. Tauris. by the function they express; hence, a Vocative Particle Gully, Adrian. 1994. Grammar and semantics in medieval is called a h·arf nida¯ ’, ‘a bit for calling’, a Negative Arabic. Richmond: Curzon. Particle is a h·arf nafy, ‘a bit for negating’, a [coordi- Sezgin, Fuat M., Geschichte des arabischen Schrifttums, Vol. 8, ‘Lexikographie’, Vol. 9, ‘Grammatik’. Leiden: E.J. Brill, nating] Conjunction is a h·arf ‘at· f, ‘a bit for joining’, 1981, 1984. and so on. In this way, every speech element, from Versteegh, Kees. 1995. The explanation of linguistic causes. Al- sound to sentence, is identified and its behavior is Zag g a¯ gι¯’s theory of grammar. Introduction, translation, accounted for. commentary. Amsterdam/Philadelphia: John Benjamins. The above sketch is based on the description of Versteegh, Kees. 1997. The Arabic linguistic tradition. Landmarks in linguistic thought III. London and New York: Arabic by the first grammarian, S¯bawayhi (who died Routledge. in late eighth century AD), which remains unsurpassed MICHAEL G. CARTER for its completeness and scientific adequacy. Later grammar added almost nothing to the data but made See also Arabic Aramaic Aramaic is a northwest Semitic language, related to The nomenclature of Aramaic is quite confusing. Hebrew and more distantly to the now extinct Aramaic is sometimes called Chaldee, Chaldaic, or Akkadian, which it replaced around the third century Assyrian. The latter appellation is also used for one of BCE as the lingua franca of the middle East, and to the dialects of Akkadian, and it should not be confused Arabic, South Arabian, and Ethiopic. with that different language. In Israel, Aramaic is 79 ARAMAIC sometimes called Kurdish, because Aramaic speakers is a literal translation into Aramaic. Additionally, two living there hail from Kurdistan, but the term Kurdish of the four so-called Ancient Versions of scripture are is properly applied to a quite different western Iranian in Aramaic: the Targum (‘translation’), which is a language. An important eastern dialect of Aramaic is Jewish translation of the Hebrew Scriptures into usually called Syriac. Syriac has three native scripts Aramaic, and the Peshitta (‘simple version’), a that have a one-to-one relationship with the standard Christian translation of the Bible into the dialect of Hebrew alphabet, which was itself originally used for Aramaic called Syriac. These versions give important Aramaic and pushed out the native Hebrew alphabet evidence of variant readings and show how obscure during the Babylonian exile. Jewish dialects of passages were understood in olden times. The Talmud, Aramaic are normally written in Hebrew characters. originally oral in nature, is written predominantly in Aramaic is indigenous to Syria and is known from the Aramaic, with an admixture of Hebrew. The city of tenth century BCE, but it has been in use at different Edessa, now called Sanliurfa, in southeastern Turkey, times from Egypt to India. It has remained a living lan- was once a center of Christian religious schools that guage in a few areas up to now. Thus, in 1930, in the produced a vast literature in the Syriac dialect, includ- town of Zakho in Iraqi Kurdistan, there were approxi- ing philosophy, theology, and poetry, much of which mately 1,500 Aramaic-speaking Jews. All of these peo- exists in unpublished manuscripts in various scholarly ple and their descendants have moved to Israel, where libraries. The scriptures of the Mandaean religion, a their language lives on tenuously. There are other small dualistic faith that still exists in Iraq and Iran, are writ- pockets, particularly in Maalula, near Damascus, ten in Mandaic, which is an Aramaic dialect, and adds Syria, where the language is spoken mostly by Greek yet another name to the language. Catholics. Other related dialects are spoken by Various Aramaic inscriptions exist on stone, clay, Christians in the Tur Abdin area of southern Turkey. and other writing materials from the period before 700 Aramaic must nevertheless be considered an endan- BCE over a wide area from Asia Minor to northern gered language, constantly pressed by Arabic, Hebrew, Arabia. The importance and international character of Turkish, and other languages, and it seems unlikely the language may be gauged from the fact that when that there are any monolingual speakers now. In the the Akkadian-speaking Assyrians besieged Jerusalem year 2000 election in the United States, the city of at this time, the officials on the wall begged the emis- Chicago published instructions for voter registration in sary to say what he had to say in Aramaic, because various languages, including Aramaic, and offered a they understood it, but did not wish the common peo- telephone hotline where Aramaic speakers could call ple to understand. The emissary refused and spoke his for information in their own language. There is a sec- threats as to the horrors of the siege in Hebrew, to dis- tion in that city where the shops have Aramaic signs courage the defenders (II Kings 18:26). A particularly and advertisements. Aramaic is also quite widely used important discovery from a later period is a collection as a liturgical language by ancient Christian churches, of documents, all in Aramaic and written on papyrus, including the Chaldean Catholic church, and that of found at the ancient city of Yeb in Egypt, whose ruins four groups of Malabar Christians in India, known as are on an island called Elephantine opposite the mod- the Christians of St. Thomas. The latter group has been ern city of Aswan. It seems that there was a Jewish replacing its Aramaic liturgy with its vernacular temple in Yeb from the seventh century BCE, and there Malayalam language. The Samaritans, a Jewish sect was a community of Aramaic-speaking Jewish merce- that has existed from biblical time until the present day, naries who left this substantial evidence of their exis- seem to have spoken Aramaic until about the tenth cen- tence. The documents contain official and private tury CE and possessed texts in the language, but now letters, legal documents, and some literary works. Of they speak Arabic. A number of important prayers in particular historical importance is a copy of a decree the Jewish liturgy are still recited in Aramaic, particu- by the Achaemenid king Darius II Nothus, who larly the Kaddish (Doxology) and the Kol nidre (annul- reigned in the second half of the fifth century BCE, ment of vows). instructing the Jews of Elephantine to observe the Aramaic is of crucial importance for religious stud- Passover. This lent credence to the Aramaic decrees ies. The biblical books of Daniel and Ezra have sub- found in the book of Ezra, which some had thought to stantial portions written in Aramaic, and there is one be forgeries. Another celebrated document is a request verse in Jeremiah [10:11], as well as two words in the for permission to rebuild the magnificent Jewish tem- book of Genesis [31:47.] The last item displays the dis- ple of Yeb, which had been destroyed around 410 BCE parity in the lexicon between Hebrew and Aramaic: by the priests as well as various legal documents, that Jacob calls a monument Gal-Ed [‘heap of witness’: the shed a flood of light on the history of the time. story explains the name of the area east of Jordan called Aramaic has a large number of dialects, but all have Gilead], whereas Laban calls it Jegar-Sahadutha, which the features characteristic of Semitic languages. Words 80  ARAWAK of the core vocabulary contain a root, usually of two, and sentences are linked together by conjunctions. The three, or four consonants, to which brief prefixes and result is a language that is mostly regular and straight- suffixes are added, and they also undergo internal forward, with a minimum of verb tenses, no case end- vowel change; many of these roots have equivalents in ings on nouns, and a native script that serves it well. other Semitic languages. Nouns have two genders, masculine and feminine, and two numbers, singular and plural. The verb is inflected with suffixes in the References perfect tense and with prefixes and suffixes in the Brown, Leslie. 1982. The Indian Christians of St Thomas. imperfect tense, like other Semitic languages. Aramaic Cambridge: Cambridge University Press. numerals above two share the common Semitic pecu- Driver, G.R. 1957. Aramaic documents of the fifth century B.C. Oxford: Clarendon Press. liarity of using the feminine form with masculine Heinrichs, Wolfhart. 1990. Studies in Neo-Aramaic. Atlanta: nouns, and vice versa. Scholars Press. Literary western Syriac is a clearly defined standard Jastrow, Marcus. A dictionary of the Targumim. New York: language. The 22 letters of its alphabet represent the Putnam, 1886–1903 frequently reprinted. consonants of the language. The vowel system distin- Johns, Alger F. 1966. A short grammar of Biblical Aramaic. Berrien Springs, ml: Andrews University Press. guishes /a/,/a:/,/e/, /i/ and /u/; /u/ and /o/ were not dis- Lindenberger, James M. 1994. Ancient Aramaic and Hebrew tinctive in this dialect. For an illustration of word letters. Atlanta: Scholars Press. formation, consider the word [ri:ʃ] ‘head.’ Pronominal Maclean, Arthur J. 1971. Grammar of the dialects of vernacular suffixes may be added to nouns, of which this is a Syriac. Amsterdam: Philo Press. selection: [ri:ʃe:h] ‘his head’, [ri:ʃa:n] ‘our head’, Macuch, R. 1965. Handbook of classical and modern Mandaic. Berlin: De Gruyter. [ri:ʃkun] ‘your (plural) head’. Various nouns have Marcus, David. 1981. A manual of Babylonian Jewish Aramaic. irregularities; thus, the plural of [bra:] ‘the son’ is Lanham: University Press of America. [bnaya:], the plural of [a:ta:] ‘the sister’ is [?awa:ta:], McNamara, Martin (project director). 1986. The Aramaic and the plural of [?atta:] ‘woman, wife’ is [neʃe:]. Bible.Wilmington, NC: Glazier. The fact that the participle in Aramaic is not used to Robinson, Theodore H.1962. Paradigms and exercises in Syriac grammar. Oxford: Clarendon Press. construct complex sentences with frequent subordi- Stevenson, William B. 1962. Grammar of Palestinian Jewish nate clauses, as occurs, say, in Germanic, Slavic, or Aramaic. Oxford: Clarendon Press. Dravidian, means that Aramaic syntax is very simple Wright, William. 1966. A short history of Syriac literature. compared with that of those languages. The verb usu- Amsterdam: Philo Press. ally comes first, followed by the subject and object, ALAN D. CORRÉ Arawak The Arawak language family is the South American such as hammock, tobacco, potato, guava, and many language family with the largest number of languages. other names for flora and fauna. Geographically, it spans four countries of Central The creation of a ‘mixed’ language of Arawak/Carib America—Belize, Honduras, Guatemala, Nicaragua— origin in the Lesser Antilles is one of the most interest- and eight of South America—Bolivia, Guyana, French ing pieces of evidence for the history of languages in Guiana, Suriname, Venezuela, Colombia, Peru, and pre-Conquest times. Speakers of Iñeri, a dialect of the Brazil (and also formerly Argentina and Paraguay). Arawak language now (misleadingly) called Island There are about 40 living Arawak languages. The Carib, were conquered by Carib speakers. They devel- first Native American peoples encountered by oped a ‘mixed’ Carib/Arawak pidgin that survived until Columbus—in the Bahamas, Hispaniola, and Puerto the seventeenth century (Hoff 1994). ‘Men’s speech’ Rico—were Arawak-speaking Tainos. Their language and ‘women’s speech’ were distinguished in the fol- became extinct within a hundred years of the lowing way. Women used words of Arawak origin, Columbian invasion. Spanish—and many other lan- while men used words of Carib origin and grammatical guages—inherited a number of loanwords from elements mostly of Arawak origin. The pidgin coexist- Arawak languages. These include widely used words ed with Carib used by men and Iñeri used by women 81 ARAWAK and children; it belonged to both parties and served as TABLE 1 Pronominal Prefixes and Suffixes in Proto- a bridge between them. This diglossia gradually died Arawak out with the spread of Island Carib to both men and Prefixes Suffixes women. As a result, Island Carib, an Arawak language, had a strong influence of Carib with respect to both Person Singular Plural singular plural 1 nu- or ta- wa- -na, -te -wa vocabulary and, possibly, grammar. 2 (p)i- (h)i- -pi -hi The languages in areas settled by the European 3nf ɾi-, i- na- -ɾi, -i -na invaders soon became extinct. Those on the north coast 3f thu-, ru- na- -thu, -ru, -u -na of South America perished first, before 1700. When the ‘Impersonal’ pa- — — — search for gold and rubber extended up the Amazon and its tributary the Rio Negro, additional languages succumbed, a process continuing from the eighteenth name Arawak (Aruák) to refer to the group of unques- century to the present day. Sometimes, the Indians tionably related languages easily recognizable by retaliated, attacking settlements and missions; but the shared pronominal prefixes such as nu- or ta- ‘1sg’, pi- invaders always returned. Indian rebellions often pro- ‘you’, and prefixes ka- meaning ‘have’ and ma- mean- voked forced migrations that sometimes ended up in ing ‘not’. A number of scholars, mainly North the creation of a new dialect or even language. For Americans, prefer to use the term Arawak(-an) to refer instance, in 1797 the British authorities removed the to a much more doubtful higher-level grouping and rebellious inhabitants of St. Vincent (an island in the reserve the term Maipuran (or Maipurean) for the Lesser Antilles) to Belize on the mainland. These were group of undoubtedly related languages, which was racially a mixture of black slaves and Indians, who claimed to be one branch of ‘Arawakan’. Here, I fol- spoke Island Carib. This resulted in the creation of a low South American practice and use the name new dialect of Island Carib—known as Central Arawak for the family of definitely related languages. American Island Carib, Kariff, Black Carib, or The limits of the family were established by the Garifuna—which by the twentieth century had devel- early twentieth century. Problems still exist concern- oped into a separate language, now one of the two ing internal relationships within the family and possi- Arawak languages with the largest number of speakers. ble relationships with other groups. Internal The overwhelming majority of Arawak languages classification and subgrouping of Arawak languages are now endangered. Even in the few communities remains a matter of debate; further detailed work is with over 1,000 speakers, a national language needed on both the descriptive and comparative fronts. (Portuguese or Spanish) or a local lingua franca The putative studies of ‘Arawakan’ by Ester (Lingua Geral Amazônica, Quechua, or Tucano) is Matteson, G. Kingsley Noble, and others are deeply gradually gaining ground among younger people. The flawed. Unfortunately, these have been adopted as the few healthy Arawak languages include Guajiro in standard reference for the classification of Arawak lan- Venezuela and Colombia (probably up to 300,000 guages, especially among some anthropologists, speakers) and Garifuna in Central America (with up to archeologists, and geneticists, influencing ideas on a 190,000 speakers). The Campa languages (total esti- putative homeland and migration routes for proto- mate 40–50,000) form one of the largest groups of Arawakan. The classification found in Campbell indigenous population in Peru. (1997) contains a number of factual mistakes and Most of the materials on Arawak languages collect- omissions and has to be treated with extreme caution. ed during the second half of the twentieth century are Little is known about a homeland for the Arawak by missionary linguists. Their quality and quantity family. The linguistic argument in favor of an Arawak vary. A full description is available for only three or homeland located between the Rio Negro and the four languages. Orinoco rivers—or on the Upper Amazon—is based on The fact that Arawak languages were related was the fact that there is a higher concentration of struc- first recognized by Father Gilij as early as in 1783. The turally divergent languages found in this region. This recognition of the family was based on a comparison area has also been suggested as one of the places where of pronominal prefixes in Maipure, a now extinct lan- agriculture developed. This is highly suggestive and guage from the Orinoco Valley, and Moxo, from corroborated by a few mythical traditions of northern Bolivia. Gilij named the family Maipure. Later, it was origin found among Arawak-speaking peoples south of renamed Arawak by Daniel Brinton after one of the the Amazon. The origin myths of the Tariana, in north- most important languages of the family, Arawak (or west Amazonia, suggest that they could have come Lokono), spoken in the Guianas. This name gained from the north coast of South America. wide acceptance during the following decades. The Arawak languages are complicated in many majority of native South American scholars use the ways. Words can be differentiated by stress in some 82  ARAWAK languages, such as Baure and Waurá (south of For instance, in Baniwa, one says nu-kapa ‘I see’ and Amazonas), and Tariana, Achagua, and Warekena nu-watsa ‘I jump’, but nu-kapa-ni ‘I see him’ and (north of Amazonas). At least two use tones, i.e. pitch hape-ni ‘he is cold’ (nu- refers to ‘I’ and -ni to ‘him’). differences, to distinguish words—Terêna in the south, And ‘my hand’ is nu-kapi. and Resígaro spoken in the far northeast of Peru. Some languages have lost the pronominal suffixes; Each Arawak language has a few prefixes and these include Yawalapiti (Xingú park, Brazil) and numerous suffixes. Prefixes are typically monosyllab- Chamicuro (Peru) to the south of the Amazon, and ic, while suffixes can consist of one or more syllables. Bare, Resígaro, Maipure, and Tariana, to the north. Roots usually contain two syllables. Prefixes are rather The form of the first person pronoun ‘I’ is ta- in the uniform across the family, while suffixes are not. What Caribbean (Lokono, Guajiro, Añun, Taino) and nu- in is an independent word in one language can be a gram- other languages. This is the basis for classification of matical suffix in another language. An Apurinã noun Arawak languages into Nu-Arawak and Ta-Arawak. maka means ‘clothing’—this is where the English Most Arawak languages distinguish two genders— word hammock comes from. In Tariana, -maka is a masculine and feminine, e.g. Palikur amepi-yo ‘thief classifier for clothing, as in pa:-maka ‘one piece of (woman)’ (literally ‘thief-MASCULINE’), amepi-ye clothing’ (literally ‘one-classifier: clothing’). ‘thief (man)’ (literally ‘thief-FEMININE’), Tariana nu- Most grammatical categories in Arawak languages phe-ri ‘my elder brother’ (literally ‘I-sibling-MASCU- are expressed on the verb. Cases for marking subjects LINE’), and nu-phe-ru ‘my elder sister’ (literally and objects are atypical. Tariana, spoken in northwest ‘I-sibling-FEMININE’). Genders are not distinguished in Brazil, has developed cases for such basic grammati- the plural. Some languages also have complicated sys- cal relations to match the pattern in nearby East tems of classifiers—these characterize the noun in Tucano languages; cases are also reported for Apurinã. terms of its shape, size, or function. Tariana and Arawak languages spoken south of the Amazon Baniwa have over 40 classifiers, e.g. Tariana pa:-da (‘South Arawak’) have a more complex verb structure (one-classifier: round) ‘one round thing’, hanu-da than those north of the Amazon (‘North Arawak’). (big-classifier: round) ‘big round thing’, i.e. the suffix South Arawak languages such as Amuesha or Campa -da is a classifier for ‘round things’ and can be have up to 30 suffix positions. North Arawak languages attached to numerals (‘one’), adjectives (‘big’), etc. such as Tariana or Palikur do not have more than a All Arawak languages distinguish singular and plu- dozen. Suffixes express meanings that are realized by ral. Plural is only obligatory with human nouns. Plural independent words in familiar Indo-European lan- markers are *-na/-ni ‘animate/human plural’, *-pe guages, e.g. ‘be about to do something’, ‘want to do ‘inanimate/animate non-human plural’. something’, ‘do late at night’, ‘do early in the morning’, Throughout the Arawak language family, nouns ‘do all along the way’, ‘in vain’, ‘each other’. A typical divide into those that must have a possessor (inalien- example from Amuesha, spoken in Peru, shows the ably possessed) and those that do not require a posses- complexity of meaning that verbal suffixes can express sor as a must (alienably possessed). Inalienably ø-omaz-amy-eʔ-ampy-es-y-e.s-n-e.n-a (3person.singu- possessed nouns include body parts, kinship terms, lar-go.downriver-distributive-epenthetic-dative- and a few others, e.g. ‘house’ or ‘name’. Inalienably epenthetic-plural-epenthetic-late-progressive-reflexive) possessed nouns have an ‘unpossessed’ form marked ‘They are going downriver by canoe in the late after- with a suffix, e.g. Pareci no-tiho ‘my face’ (literally ‘I- noon stopping often along the way’. face’), tiho-ti ‘(someone’s) face’ (literally ‘face- Arawak languages spoken north of the Amazon have UNPOSSESSED’). Alienably possessed nouns take a evidentials, i.e. forms with which speakers specify how suffix when possessed, e.g. Baniwa nu-inu-ni ‘my they acquired the information. Baniwa and Piapoco dog’ (literally ‘I-dog-POSSESSED’). have just a ‘reported’ evidential, which is used for sec- The overwhelming majority of Arawak languages ond-hand information. Tariana has four evidentials: have a negative prefix ma- and its positive counterpart visual (if one saw the event happen), nonvisual (if one ka-, e.g. Piro ka-yhi ‘having teeth’ (literally ‘POSITIVE- heard or smelt it), inferred (if one can judge what hap- tooth’), ma-yhi ‘toothless’ (‘NEGATIVE-tooth’); Bare pened by some indirect evidence), and reported (if one ka-witi-w ‘a woman with good eyes’ (‘POSITIVE-eye- acquired the information from another person). FEMININE’), ma-witi-w ‘a woman with bad eyes, a blind Verbs are divided into active verbs (‘hit’ or ‘jump’) woman’ (‘NEGATIVE-eye-FEMININE’). Most languages and stative verbs (‘be cold’). All Arawak languages have just the numbers ‘one’ (pa-; also meaning ‘some- share pronominal affixes and personal pronouns. one, another’) and ‘two’ ((a)pi or yama). Pronominal suffixes (e.g. -ni ‘he’) are used for sub- The common Arawak vocabulary consists mostly of jects of stative verbs and for direct objects. Subjects of nouns, with quite a few body parts, fauna, flora, and active verbs are expressed by prefixes (e.g. nu- ‘I’). artifacts. Only a few verbs are retained in most Arawak 83 ARAWAK languages, e.g. ikau ‘arrive’, *p (da) ‘sweep’, *po or Campbell, Lyle. 1997. American Indian languages. New York: *da ‘give’, *(i)ya ‘cry’, *kama ‘be sick, die’, *itha Oxford University Press. Dixon, R.M.W., and Alexandra Y. Aikhenvald. 1999. ‘drink’. Introduction. In Dixon and Aikhenvald, pp. 1–21. Dixon, R.M.W., and Alexandra Y. Aikhenvald (eds.) 1999. The References Amazonian languages. Cambridge: Cambridge University Press. Aikhenvald, Alexandra Y. 1995. Person-marking and discourse Facundes, Sidney da Silva. 2000. The language of the Apurinã in North-Arawak languages. Studia Linguistica 49. 152–95. people of Brazil. Ph.D. Thesis. SUNY at Buffalo. ———. 1996. Areal diffusion in North-West Amazonia: the Hoff, Berend. 1994. Island Carib, an Arawakan language which case of Tariana. Anthropological Linguistics 38. 73–116. incorporated a lexical register of Cariban origin, used to ———. 1998. Warekena. Handbook of Amazonian languages, address men. Mixed languages: 15 case studies in language Vol. 4, ed. by D.C. Derbyshire and G.K. Pullum, 215–439. intertwining, ed. by P. Bakker and M. Mous, 161–8. Berlin: Mouton de Gruyter. Amsterdam: IFOTT. ———. 1999a. The Arawak language family. In Dixon and Payne, David L. 1991. A classification of Maipuran (Arawakan) Aikhenvald, pp. 65–105. languages based on shared lexical retentions. Handbook of ———. 1999b. Areal diffusion and language contact in the Amazonian languages, Vol. 3, ed. by D.C. Derbyshire and Içana-Vaupés basin, North West Amazonia. In Dixon and G.K. Pullum, 355–499. Berlin: Mouton de Gruyter. Aikhenvald, pp. 385–415. Taylor, Douglas M. 1977. Languages of the West Indies. ———. 2001. Areal diffusion, genetic inheritance and prob- Baltimore, MD: Johns Hopkins University Press. lems of subgrouping: a North Arawak case study. Areal dif- Tovar, Antonio, and Consuelo Larrucea De Tovar. 1984. fusion and genetic inheritance: problems in comparative Catálogo de las lenguas de América del Sur. Madrid: linguistics, ed. by Alexandra Y. Aikhenvald and R.M.W. Editorial Gredos. Dixon, 167–94. Oxford: Oxford University Press. Wise, Mary Ruth. 1986. Grammatical characteristics of Aikhenvald, Alexandra Y. 2002. Language contact in Preandine Arawakan languages of Peru. Handbook of Amazonia. Oxford: Oxford University Press. Amazonian languages, Vol. 1, ed. by D.C. Derbyshire and Aikhenvald, Alexandra Y. 2003. A grammar of Tariana, from G.K. Pullum, 567–642. Berlin: Mouton de Gruyter. northwest Amazonia. Cambridge: Cambridge University Press. ALEXANDRA AIKHENVALD ———, and Diana Green. 1998. Palikur and the typology of classifiers. Anthropological Linguistics 40. 429–80. See also Carib and Cariban Languages Archeology and Language The history of related languages is at the same time a European language family, at some time in its prehis- history of societies to which those languages toric existence, has occupied a territory far more con- belonged. When we reconstruct prehistoric relation- fined than that of its earliest historically attested ships among a group of languages, we simultaneously branches; Indo-European languages are thus supposed postulate the historical existence of the societies that to have spread from somewhere to their historical spoke those languages. Since the very beginnings of location in prehistory, invading or occupying areas linguistics as a science in the nineteenth century, the where languages not belonging to the Indo-European problem of the archeological correlates of language family were spoken; some of these relic non-Indo- change and dispersal has thus been crucial to the joint European languages are attested throughout Eurasia efforts of linguists and archeologists. (Basque, Tartessian, Iberian, Etruscan, Dravidian lan- It was within the field of Indo-European linguistics guages), and are surrounded by Indo-European lan- that archeological evidence has been more and more guages. To some extent, the archeological record can widely resorted to determine the location of the Indo- suggest the sociohistorical conditions under which the European homeland, the area in which a population Indo-Europeanization of Eurasia may have occurred. spoke Proto-Indo-European, a broadly defined group From the great number of archeological solutions to of dialects that later developed into the various Indo- the homeland problem, one can select three scenarios European languages. The homeland problem can be that enjoy wide currency: stated as follows: there are several reasons for suppos- (1) The continuity hypothesis. Archeologists have ing that the historical distribution of Indo-European found that there is a major cultural border languages is not the original one and that the Indo- between the steppe cultures and those of 84  ARMENIAN Northern/Central Europe, which was not seri- ward diffusion of some linguistic traits. In Africa, sig- ously transgressed at any period from the nificant steps have been made toward uncovering the Neolithic period onward. Thus, a Proto-Indo- early historical developments among the speakers of European linguistic continuum should already each of the major language families of the continent: have existed during the Mesolithic (8500–5000 Nilo-Saharan, Niger-Congo, Afro-Asiatic, and BC) between the Baltic and the Black/Caspian Khoisan. The Proto-Nilo-Saharan speech territory most seas. The Neolithic cultures that emerged probably lay in the southern fringe of the central throughout this region are then to be considered Sahara desert. The early periods of Khoisan history as ancestral to their respective regional Indo- have been played out in eastern Africa: Khoisan speak- European languages. ers were makers of the Eastern African Microlithic (2) The Anatolian hypothesis. According to this Tradition; as early as the sixth millennium BC, hypothesis, the dispersal of Indo-Europeans is Khoisan speakers began to spread south across south- connected with the spread of agriculture from ern Africa. Proto-Afroasiatics are to be identified with Anatolia into Europe and Asia: Indo-European the early wild-grass-collecting cultures of northeastern farming colonists absorbed non-Indo-European Africa. The overall history of the dispersal of Niger- cultures expanding in a wave of advance. Congo languages is exceedingly complex, but an initial (3) The Kurgan model. According to this model, stage (between 8000 and 6000 BC) can be clearly iden- the Indo-European homeland was located tified in which Niger-Congo peoples, starting as far among pastoralist tribes of south Ukraine and east as the Nuba Mountains of Sudan, expanded south south Russia that expanded both to the east and across the woodland savannas of West Africa. west (the kurgan or tumulus being a typical marker of this expansion). This model implies a References process of progressive acculturation of non- Dixon, R.M.W. 2002. Australian languages. Their nature and Indo-European peoples across Europe and development. Cambridge: Cambridge University Press. Asia. Ehret, Christopher. 2000. Language and history. African lan- Archeological models of dispersal have also been guages. An introduction, ed. by Bernd Heine and Derek applied to other areas. In Australia, patterns of Nurse. Cambridge: Cambridge University Press. Gimbutas, Marija. 1991. The civilization of the goddess. San linguistic diffusion show considerable similarities to Francisco: Harper. the patterns of diffusion of nonlinguistic traits. For Renfrew, Colin. 1987. Archeology and language. London: instance, both the use of dug-out canoes and pearl shell Jonathan Cape. ornaments appear to have permeated down from the ANDREA SANSÒ coastal regions of northern Australia as far as South Australia, a pattern that closely resembles the south- See also Indo-European 1: Overview Armenian Armenian is a cover term for a number of distinct lan- Abkhazia and Russia; and dozens of other mutually guages, the best known of which are the literary forms unintelligible variants of Armenian originally spoken Standard Eastern and Western Armenian, Middle in Turkey, Armenia, Azerbaijan, Iran, Georgia, (/Medieval/Cilician) Armenian, and Classical Abkhazia, Russia, and Israel. Lomavren, the language Armenian. The label ‘Armenian’ also encompasses the of the Bosha gypsies of Turkey and Armenia, draws its closely related languages Zok, formerly spoken by the phonology, morphology, and syntax from the Erzerum Armenian inhabitants of southeastern Nakhichevan; dialect of Armenian but its lexicon is mostly of Indic Kistinək, spoken by the Armenian inhabitants of origin; for this reason, it is not clear whether or not Musaler, Turkey; Kesbənuək, spoken by the Armenian the language should be classified as a member of inhabitants of Kesab, Syria; Homshetsma, spoken by the Armenian family. All but Homshetsma employ the the Hemshinli of northeast Turkey and the Hamshen Armenian alphabet created by Mesrob at the begin- Armenians of the Black Sea coastal regions of ning of the fifth century. 85 ARMENIAN Armenian is an Indo-European language, generally primarily from Parthian, and to a lesser extent from believed to be most closely related to the Greek and Greek and Syriac; several hundred and perhaps as Indo-Iranian subgroups. (For instance, all three share a many as several thousand words are of unknown ori- particle *me: used in expressing prohibitions (Greek gin, most likely having come from Urartian, Hurrian, me:, Sanskrit ma:, Armenian mi) and the imperfect and other now-extinct autochthonous languages. third-person singular particle *e- (as in Greek e-pher-e, Armenian also incorporated large numbers of Arabic Sanskrit a-bhar-a-t, Armenian e-ber ‘(s)he/it carried’).) words following the expansion of the Arabs in the Because of the large number of loan words from vari- Middle East in the seventh century, and thousands of ous Middle Iranian languages, especially Parthian, Turkish words following the arrival of the Seljuks and Armenian was thought to be an Iranian dialect until other Turkic tribes in Anatolia beginning in the Heinrich Hübschmann demonstrated in 1875 that it eleventh century. was a distinct branch of the Indo-European family. Although there are dozens of mutually unintelligi- Scholars disagree on how the Armenians came to his- ble varieties of Armenian, all share certain features. torical Armenia, the eastern half of present-day Turkey Proto-Armenian had four verbal conjugations, charac- centered around Lake Van and Mount Ararat. Some terized by theme vowels -e-,-i-,-ɑ-, and -u- (beɾ-e-m ‘I believe they came southward from the Russian steppe, carry’, χaws-i-m ‘I speak’, χnd-a-m ‘I rejoice’, and others believe they and the Hittites came eastward zgen-u-m ‘I wear’); most modern dialects (including from Greece, and others suggest they moved only a the Western and Eastern literary languages) have com- short distance from an original Indo-European home- pletely or partially lost the -u- conjugation, and stan- land in the Transcaucasus. It is most likely that this dard Eastern Armenian has merged the -i- conjugation settlement occurred in the second millennium BC. The into the -e- conjugation. There were originally three earliest mentions of the Armenians occur in the morphologically distinct sets of personal endings for inscriptions of the Achaemenid Persian king Darius verbs—present, imperfect, and aorist—which were (sixth century BC) and the Greek historian Herodotus used in combination with additional tense and aspect (fifth century BC). markers to form the various tenses and moods. The The earliest written records of the Armenian lan- system of nominal morphology in Proto- and Classical guage date from the fifth century AD, shortly after the Armenian was rich, preserving the Indo-European conversion of the Armenians to Christianity in the nominative, accusative, genitive, dative, instrumental, fourth century led to the creation of an Armenian ablative, and locative cases in both singular and plural alphabet by Mesrob in around 401 and a systematic (but the Indo-European dual was lost); there were at program of translating the books of the Bible. The lan- least eight different declensions, distinguished prima- guage of the earliest translations was Classical rily by different theme vowels. This system was sig- Armenian (also called grabar, ‘written (language)’), nificantly reduced by the medieval period; Middle which continued as the preferred literary form of Armenian and the modern varieties now use the singu- Armenian until the nineteenth century, when it was lar endings for the plural as well, and have only one supplanted by the two modern literary dialects. productive declension, formed from parts of the origi- In linguistic terms, Armenian is notable for its sig- nal -i- and -o- declensions. With the exception of pro- nificant divergences from Proto-Indo-European (the nouns, the inventory of cases has significantly reduced reconstructed ancestral language of the Indo-European as well: the accusative has merged with the nomina- family), particularly in terms of pronunciation and tive, and the genitive with the dative. Proto-Armenian vocabulary. Some of the more striking phonological had several participial forms, but only two of these changes are the development of a rich set of affricates survive into the modern period: the original past par- (ts, dz, etc.), the loss of final syllable rimes (e.g. ticiple -eal is now -el in the Eastern dialects, and the Proto-Indo-European *worgjom ‘work’ > Classical original present participle -oʁ is now used as a present Armenian gorts), the change of initial *dw to erk- (e.g. participle and for relativizing subjects of subordinate Proto-Indo-European *dwo: ‘2’ > Classical Armenian clauses. The Western dialects have replaced -eal with erku), and the change of original *w to g. Most strik- -ats (> -adz in most varieties) for past participles; all ing in the vocabulary of Armenian is the rarity of modern dialects also use the -ats participle to relativize words inherited from Indo-European, and the over- nonsubjects of subordinate clauses. whelming predominance of words of unknown origin. Most of the changes between Classical and Modern Not surprisingly, native Indo-European words survive Armenian first appeared in the medieval period in primarily in the core vocabulary: mayr ‘mother’ < Middle Armenian documents, associated with the *ma:ter, hayr ‘father’ < *pater, khoyr ‘sister’ < *swe- Armenian kingdom of Cilicia, which flourished from sor, kov ‘cow’ < *gwows, tun ‘house’ < *domos, em ‘I the eleventh to fifteenth centuries AD in what is now am’ < *esmi. The remainder of the lexicon is drawn south-central Turkey. Middle Armenian is generally 86  ARMENIAN Western in character, although it shares many features ers of Armenian in 1996, but all varieties of the language with Eastern dialects as well. It inverts the pronuncia- except for Standard Eastern Armenian are in immediate tion of the Classical Armenian plain voiced and voice- danger of extinction as very few diaspora Armenians less stops (e.g. berem ‘I carry’ > perem, pat ‘wall’ > under the age of 30 speak the language fluently. bad), a feature that is preserved in the modern Cilician Whereas Classical Armenian was relatively Indo- dialects of Zeytun and Hadjin but differs from the European in its syntactic and morphological structure, Western and Eastern literary varieties (Eastern pre- all varieties of Modern Armenian are typologically serves the Classical system (bεɾεm); Western devoic- much closer to Turkish and the Balkan languages. es and aspirates the original voiced consonants Compare, for instance, the formation of relative claus- (phεɾεm)). The Cilician kingdom was in close contact es, exemplified by ‘I saw the bird that was singing in with several Crusader kingdoms, as a result of which the tree’: Classical tesi ´z-tH´rtSHun-´n oR eRgEr i it borrowed a significant number of words from veRAj tsAr-oj-n (I.saw specific-bird-definite that Crusader French, most famously what comes out as was.singing in on tree-genitive-definite), Western the standard Western form for ‘mister,’ baron. dzAr-i-n v´RA jERkHç“ tH´rtSHun-´ dEsA (tree-geni- In the nineteenth century, Armenian nationalists tive-definite on singing bird-definite I.saw). Western became interested in developing a literary form of the Armenian has been additionally influenced by Turkish modern language. This was brought about by excising and Greek (cf. ´sdEp“in ‘carrot’, istAkHçz ‘lobster’, most Turkish forms from the regional dialects and bAntHçg ‘hotel’), whereas Eastern Armenian has been replacing them with new borrowings from the classical heavily influenced by Russian (e.g. the standard form language. The intellectual center around which the new for ‘potatoes’ is kHARtHçfli, and the word for ‘gay’ is Western literary language was organized was gAlubçj, from the Russian word originally meaning Constantinople, although many features of the standard ‘sky blue’; the native word for ‘blue,’ kApujt, cannot dialect (including the pronunciation of the consonants) be used in this sense). do not come from the Armenian dialect originally spo- ken there. The same holds for Eastern Armenian with respect to Erevan. The relationship between the two The Armenian Alphabet, with IPA Equivalents for modern literary dialects is somewhat complicated; Eastern Pronunciation there are many grammatical differences (e.g. Western a A c ts À dZ g´ siREm vs. Eastern siRum Em ‘I love’, Western bidi b b k k ® r g g h h s s siREm vs. Eastern k´siREm ‘I will love’) and lexical d d j dz v v differences (e.g. Western dεɾmɑ vs. Eastern spitAk e E @ “ t t ‘white’; Western hçs vs. Eastern εstεʁ ‘here’, Western z z ç tS r R bEdkHARAn vs. Eastern zukHARAn ‘bathroom’, Western ê E m m … tsH hɑvith vs. Eastern dzu ‘egg’), and most Western å ´ y j u u † tH n n § pH speakers have difficulty in understanding Eastern; ± Z š S Ï kH however Eastern speakers are relatively comfortable i i o ç ó ç with the Western dialect. This asymmetry in mutual l l æ tSH f f intelligibility most likely results from the fact that large x X p p numbers of speakers of Western dialects fled to Eastern Armenia following the Russo-Turkish war in 1828 and The Lord’s Prayer in Different Varieties of the Turkish Genocide in 1915–1920, whereas before Armenian, Rendered in the IPA the fall of the Soviet Union in 1991 most Western Armenians had little or no exposure to Eastern Classical Armenian (E¯ miacin ms. 229, 989 A.D.) Armenian. The fact that there is some mutual intelligi- Hayr mer or yerkins. surb e¬i…i anun !o. eke…ê bility in both directions can also be linked to the fact ar!ayu†iwn !o. e¬i…in kam! !o orpês yerkins ew that the literary dialects tend to borrow the same forms yerkri. zha… mer hanapazord tur mez aysawr. ew from Classical Armenian, and (at least in recent †o¬ mez zpartis mer. orpês ew me! †o¬um! mero… decades) employ the same newly coined words. partapana…. ew mi tanir zmez i §orju†iwn. The destruction of the Armenian homeland and more ay¬’ §rkea zmez i æarê. zi !o ê ar!ayu†iwn ew than a million Armenians by the Ottoman government in zawru†iwn ew §a®! yawiteans amên: 1915–1920 rendered most nonstandard varieties of mod- hAjR meR oR jeRkin´s. suRb e“itsHi Anun kHo. ern Armenian moribund; with few exceptions, the eketsHE ARkHAjutHiwn kHo. e“itsHin kAmkH kHo Armenians in the diaspora (primarily Lebanon, France, oRpEs jerkin´s ew jeRkRi. ´zhAtsH meR and the United States) speak only Standard Western hAnApAzoRd tuR mez AjsAwR. ew tHo“ mez Armenian. There were approximately 6.8 million speak- ´zpARtis meR. oRpEs ew mekH tHo“umkH meRotsH 87 ARMENIAN pARtApAnAtsH. ew mi tAniR ´zmez i pHoRdzutHiwn. Zeytun Dialect (Cilicia, South Central Turkey) aj“J pH´RkeA ´zmez i tSHARE. zi kHo E ARkHAjutHiwn çv mEj bob´ çj ijginkHn-is, kHu Anun´t sujp ew zAwRutHiwn ew pHArkH jAwiteAn´s AmEn. tHç“nA. kHu tHEkHEvyytHyn´t tHu“ kç. kHu gçmkH´t tHu“ lA, intSHbEs ijginkH´, indEn El ijgEjin vijç. mij Standard Eastern Armenian AmEn øjvEn hçtsH´ Esøj miz tuj. jEv miz nEjE mij Hayr mer, vor yerkn!um es. surb †o@ lini !o bçjdkH´, tSHçtsH vçR minkH El g´ nEjinkH mij anunå. !o †agavoru†yunå †o@ ga. !o kam!å †o@ bçjdkHi dEjERun. jEv miz pHçjtsutHAn mi dAn´j, lini yerkri vra, inæpes vor yerkn!um ê. mer hAbç tSHçjEn miz AzAdE. tSHunkHi kHin˘ E hanapazorya ha…å tur mez aysór. yev †o@ mez tHEkHEvyjytHyn´ jEv zçjutHyn´ u pHçrkH´. mer part!erå, inæpes yev men! en! †o@num mer hAvidjAn´s hAvidEnitsH. AmEn. partakannerin. yev mi tar mez !orju†yan, ayl §rkir mez æari…. vorovhetev !ónn ê †aga- Kesab (Syria) voru†yunå yev zoru†yunå yev §a®!å havityans. øv mieR byb´, suRp E“ni kHE Qnun, kHE amen: tHEkHEvyRutHyn´ tH´“ kç, kHE iRAdEtH´t ´n˘ç, hAjR mER, vçR jERk´nkHum-Es. suRpH tHç“ lini tSHytsH´R kHi iRgQnkH´ tH´RzEn El ikEdin´, mieR kHç Anun´. kHç tHAgAvçRutHjun´ tHç“ gA. kHç AmEn EvyR ho´tsH´ duR miez Es EvyR El, mieR kAmkH´ tHç“ lini jERkRi v´RA, intSHpEs vçR bçRdkH´ miezi bA“´tSHlAmuS ´Rç tSHytsH´R kHi jERk´nkHum E. mER hAnApAzçRjA hAtsH´ tuR mEz mienkH ginonkH mieRontsH´, vE z´zmiez AjsçR. jEv tHç“ mEz mER pARtkHER´, intSHpEs jEv pHçRtsytHjAn mi dAnç, hAbç XAl´sç i tSHARien, mEnkH EnkH tHç“num mER pARtAkAnnERin. jEv mi tSHynkHi kHE E tHEkHEvyRutHyn´, tSHEREf´, kHuvEtH´, tAR mEz pHçRdzutHjAn, Ajl pH´RkiR mEz tSHARitsH. hAvidiein´s hAvidçnitsH Amçn. vçRçvhEtEv kHçn˘ E tHAgAvçRutHjun´ jEv zçRutHjun´ jEv pHArkH´ hAvitjAn´s. AmEn. References Adjarian, Hrachea. 1909. Classification des dialectes Standard Western Armenian arméniens. Paris: Champion. Ov hayr mer or erkin!n es, !u anund surb ålla. Bardakjian, Kevork, and Bert Vaux. 1998. Eastern Armenian: a !u †agaworu†iwnd ga. !u kam!d ålla inæpês textbook. Ann Arbor: Caravan. erkin!å` noynpês erkri vray. mer amên óruan Clackson, James. 1994. The linguistic relationship between ha…å aysór al mezi tur, mezi nerê mer part!erå Armenian and Greek. Oxford: Blackwell. Dumézil, Georges. 1964. Notes sur le parler d’un Arménien inæpês men! al kå neren! mer partakannerun. musulman de Hems,in. Brussels: Palais des Académies. u mez §orju†ean mi tanir, hapa æarên mez Finck, Franz Nikolaus. 1907. Die Sprache der armenischen azatê. !anzi !ukd ê †agaworu†iwnå ew Zigeuner. St. Petersburg: Imperial Science Academy. zóru†iwnå u §ar!å yawiteans: Amên: Karst, Josef. 1901. Historische Grammatik des Kilikisch- çv hAjR mER vçR jERginkHn-Es, kHu Anun´tH suRpH Armenischen. Strassburg: Trübner. Meillet, Antoine. 1936. Esquisse d’une grammaire comparée de ´l˘A. kHu tHAkHAvçRutHjun´tH kHA. kHu gAmkH´tH l’arménien classique, 2nd edition. Vienna: Imprimerie des ´l˘A intSHbEs jERginkH´, nujnbEs jERgRi v´RA. mER Pères mékhitharistes. AmEn çRvAn hAtsH´ AjsçR Al mEzi duR, mEzi nERE Thomson, Robert. 1975. An introduction to classical Armenian. mER bARdkHER´ intSHbEs mEnkH Al g´ nEREnkH mER Delmar, NY: Caravan. bARdAgAn˘ERun. u mEz pHçRtsHutHjAn mi dAniR, Vaux, Bert. 1998. The phonology of Armenian. Oxford: Oxford University Press. hAbA tSHAREn mEz AzAdE. kHAnzi kHug´tH E BERT VAUX tHAkHAvçRutHjun´ jEv zçRutHjun´ u pHArkH´ hAvidjAn´s. AmEn. See also Indo-European 1: Overview Artificial Intelligence Artificial Intelligence (AI) is the subfield of computer as perception, reasoning, understanding, and learning. science that aims to enhance computers with capabili- However, AI researchers usually give more elaborate ties traditionally attributed to human intelligence, such definitions for AI, focusing either on the problems that 88  ARTIFICIAL INTELLIGENCE AI aims to solve or on the methods used for approach- as the Turing test: a human observer is allowed to inter- ing them. Thus, AI can be alternatively defined as the act via a teletype with another human and a machine effort to make computers capable of solving problems that pretends to be a human. The observer does not that currently only humans can solve, or as a set of know which is which and tries to figure it out by con- methods for confronting problems for which exact versing with both via the teletype. Turing argued that if solution algorithms are not yet known. The latter defi- the machine could successfully fool a knowledgeable nitions also explain why topics such as symbolic cal- observer, then it should certainly be considered intelli- culus and neural networks are not considered pure AI gent. However, many philosophers, such as John Searle topics anymore and why some not-well-defined prob- and Hubert Dreyfus, have argued not only against the lems, such as natural language understanding, are tra- correctness of the Turing test but also against the feasi- ditionally positioned in the domain of AI. bility of ever building a nonbiological intelligent AI has been strongly influenced by the works of machine. This debate led to the formation of two philo- Norbert Wiener, Claude Shannon, John von Neumann, sophical directions: strong AI and weak AI. Supporters Alan Turing, and Alonzo Church. However, the term of strong AI believe that computers, even with current ‘artificial intelligence’ is attributed to John McCarthy, technology, are capable of supporting intelligence, who introduced it on August 31, 1955, in a document given appropriate programming. Supporters of weak titled A Proposal for the Dartmouth Summer Research AI believe that although programs able to imitate intel- Project on Artificial Intelligence. The proposal involved ligent behavior can be made, it is impossible to make a a two-month study ‘on the basis of the conjecture that machine that understands or develops consciousness in every aspect of learning or any other feature of intelli- the way that humans do. gence can in principle be so precisely described that a The distinction between weak and strong AI is machine can be made to simulate it’. It was signed by closely related to another distinction: that between John McCarthy, Marvin L. Minsky, Nathaniel symbolic and subsymbolic AI. The key assumption of Rochester, and Claude E. Shannon. The conference that symbolic AI is that knowledge is represented by struc- was held at Dartmouth College (in Hanover, New tures of semantically meaningful symbols and that a Hampshire) the next summer is considered a milestone set of such symbols is therefore required to support the in the history of AI because it brought together all the reasoning process of an intelligent system. A program researchers interested in the field and established the for proving logical theorems is an example of a sym- birth of AI. After the Dartmouth conference, Minsky bolic AI system. However, the key assumption of sub- and McCarthy founded the AI Laboratory at the symbolic AI is that intelligent behavior can be attained Massachusetts Institute of Technology, and two other without the need for semantically meaningful sym- participants—Herbert Simon and Allen Newell— bols. A neural network is an example of a subsymbol- founded the AI Laboratory at Carnegie Mellon ic AI system. University. McCarthy, Minsky, Simon, and Newell are A generic AI system can be described in terms of considered the founders of modern AI. four main components: perception, reasoning, knowl- The feasibility of the ultimate AI goal (i.e. the devel- edge, and learning. Figure 1 shows their interconnec- opment of intelligent machines) has been questioned tion. Perception is the task of processing the input to repeatedly. A theorem known as the Church–Turing extract structured information needed for the particular thesis has provided strong supporting evidence, assum- task. The reasoning component is responsible for the ing that intelligent behavior is based on some kind of manipulation of this information, according to the sys- computation. Turing has also described a process to tem knowledge contained in the knowledge component. test whether the intelligence of an AI system is compa- The output of reasoning may be the outcome of the AI rable to that of a human being. This process is known system itself, or it may be fed to a learning mechanism Figure 1. Structure of a Generic Artificial Intelligence System. 89 ARTIFICIAL INTELLIGENCE whose function is to update the existing knowledge. The European Coordinating Committee for Artificial structure of Figure 1 should be considered as a descrip- Intelligence (ECCAI), the Society for Artificial tive model, not an implementational one, because some Intelligence and Simulation of Behavior (AISB), and systems may lack some components, and others, espe- the Special Interest Group on Artificial Intelligence cially the subsymbolic ones, may have some of the (SIGART) of the Association for Computing components merged indistinguishably. Machinery (ACM). However, the majority of the topics addressed by AI concern one or more of the above four components. References These topics are discriminated as AI branches, accord- ing to the types of problems, the techniques used, and Barr, Avron, and Edward A. Feigenbaum (eds.) 1981. The hand- book of artificial intelligence. Los Altos, CA: William the classes of applications involved. There are many Kaufmann Inc. AI branches: search methods, symbolic reasoning, Haugeland, John (ed.) 1997. Mind design II. Cambridge, MA: production systems, heuristic techniques, knowledge MIT Press. representation, logic programming, statistical reason- Jackson, Philip C. Jr, 1985. Introduction to artificial intelli- ing, fuzzy logic, game playing, planning, vision, natu- gence, 2nd edition. New York: Dover Publications. Konar, Amit. 2000. Artificial intelligence and soft computing: ral language processing, machine learning, neural behavioral and cognitive modeling of the human brain, Boca networks, genetic algorithms, common sense reason- Raton, FL: CRC Press. ing, expert systems, speech understanding, theorem Nilsson, Nils J. 1998. Artificial intelligence: a new synthesis. proving, machine translation, robotics, perception, Los Altos, CA: Morgan Kaufmann. automatic programming, constraint satisfaction, ontol- Pollock, John L. 1989. How to build a person: a prolegomenon. Cambridge, MA: MIT Press. ogy, and intelligent agents, among others. Their out- Rich, Elaine, and Kevin Knight. 1991. Artificial intelligence, comes are very diverse. Some of them, such as expert 2nd edition. New York: McGraw-Hill. systems and game playing, have produced impressive Russel, Stuart, and Peter Norvig. 1995. Artificial intelligence: a results, whereas others, such as machine translation modern approach. Eaglewood Cliffs, NJ: Prentice–Hall. and speech understanding, are still far from their goal. Turing, Alan M. 1950. Computing machinery and intelligence. Mind 59. 433–60. There are several scientific organizations devoted to KYRIAKOS N. SGARBAS AI research. Among the best known are the American Association for Artificial Intelligence (AAAI), the See also Turing, Alan Artificial Languages The term ‘artificial language’ can refer to several dif- Descartes, Gottfried Leibnitz, and John Wilkins, were ferent kinds of symbolic systems, including program- investigating the possibility of constructing a linguis- ming languages, symbol systems used in various tic system on the basis of scientific taxonomies and branches of science, ‘controlled’ languages for unam- logicomathematical operations. This work inspired biguous technical communication, intermediate lan- many so-called philosophic or a priori language proj- guages in machine translation, highly regularized ects throughout the next two centuries, both written ethnic or literary languages, fictional languages creat- (pasigraphies) and spoken (pasilalies). Despite their ed for literary purposes, and planned languages for influence on the development of mathematics, formal intercultural communication. In linguistics, the term is logic, and general linguistic theory, none of these sys- most often used to designate this last category, but tems acquired a significant body of users. metaphors and concepts from these neighboring fields In the nineteenth century, the rise of historical and are a frequent source of both insight and confusion. evolutionary views of language contributed to a waning The idea of improving existing languages, by mak- of interest in philosophical language projects. At the ing them more regular, less ambiguous, or more same time, improvements in transport and communica- expressive arose in medieval Europe. By the mid-sev- tions turned attention toward the practical benefits of a enteenth century, a number of leading intellectuals, common international language, while the Romantic including Jan Amos Komensky (Comenius), René equation of language with culture and identity led 90  ARTIFICIAL LANGUAGES many to associate the idea of a common language with arbitrary modifications of Western European words, ideals of peace and brotherhood among nations. Today, making new coinages controversial. After its extraor- all three of these orientations—the philosophical-sci- dinarily rapid spread in the 1880s, including a certain entific, the instrumentalist, and the value oriented—can amount of oral use, Volapük underwent an even more be found among users, researchers, and theorists of rapid decline in the 1890s in response to internal con- international planned languages. flicts over the control and development of the lan- The systematic study of such languages, a field now guage. It has not been in active use since the 1920s. referred to as interlinguistics, was inaugurated in 1903 Volapük’s successor, Esperanto (1887), was pub- with the publication of Couturat and Leau’s Histoire de lished by a Jewish physician in Warsaw, Ludovik la Langue Universelle (History of the Universal Zamenhof. Initially slower to spread than its predeces- Language), supplemented four years later by Les sor, Esperanto experienced explosive growth between Nouvelles Langues Universelles (The New Universal 1900 and 1912, the period in which its oral tradition Languages). The authors analyzed a total of 75 lan- became firmly established. Today, it remains by far the guage projects and project outlines, most of them orig- mostly widely used international planned language, inating in the previous half-century. No fewer than 912 and thus it is the most accessible to linguistic study. projects are listed in the most complete cataloge to For much the same reason, it is also the language in date, Aleksandr DuliÍenko’s MeËdunarodnye which the greatest share—approximately 50%—of the Vspomogatel’nye Jazyki (1990: International Auxiliary interlinguistic literature has been written. Research on Languages). However, the vast majority of these have Esperanto is reviewed in greater detail below. seen no practical use, and linguistic research on them is In 1907, Ido, a project derived from Esperanto, was necessarily theoretical. Empirical research requires a published and quickly attracted some prominent community of language users, and the number of proj- adherents from the Esperanto movement; rivalry ects that has achieved this status is quite small. between the two languages continued until after World The most widely used classificatory scheme in inter- War II. It is now generally accepted that Ido was the linguistics places language projects on a continuum creation of Louis Couturat, a prominent French logical between ‘naturalism’ (modification of an existing lan- philosopher and the principal author of Histoire de la guage that leaves most of its features intact) and Langue Universelle (mentioned previously). Some of ‘schematism’ (implying a high degree of rationaliza- Couturat’s modifications tended toward greater tion on a priori lines). All of the more naturalistic proj- Latinization. Others introduced greater schematism. ects that have acquired a community of users are Like the Volapük community before it, the Ido move- Latinate. They include Latino sine flexione (1903), a ment soon became embroiled in linguistic disputes regularized form of Latin invented by the Italian math- that alienated many potential users. Today, only a very ematician Giuseppe Peano; Occidental (1922; later small community of speakers remains. renamed Interlingue), the project of Edgar de Wahl, an The newest language projects to have acquired a Estonian mathematician who had been active in earlier small number of users are Glosa, a derivative of projects; Novial (1928), the creation of Danish linguist Lancelot Hogben’s Interglossa (1943), and Lojban, a Otto Jespersen; and Interlingua (1951), a project pub- derivative of James Cooke Brown’s Loglan (1955). lished under the auspices of the International Auxiliary Glosa combines Greek and Latin words with particles Language Association, but primarily designed by its to denote tense, number, and so forth; phrase and sen- director Alexander Gode. Like the earlier naturalistic tence construction resemble those of English. Lojban’s projects that it replaced, Interlingua was intended pri- syntax is based on symbolic logic; it was originally marily for communication among educated Europeans; proposed, in the tradition of the philosophic language Gode’s idea was to distill a ‘Standard Average projects, as a means of exploring the relationship European’ (a concept inspired by Benjamin Whorf) between language and thought, and attempts to use it from lexical and grammatical elements common to for oral communication are fairly recent. Neither proj- English, French, Italian, and Spanish/Portuguese. ect has acquired a speech community. Interlingua acquired perhaps a few thousand users in All but the last two projects are discussed in greater the 1950s and 1960s, primarily in the United States and detail in Detlev Blanke’s seminal work Internationale Europe, and retains a small following today. Plansprachen (1985; International Planned Lan- The remaining language projects to have acquired a guages), the standard reference work on the subject. body of users display both naturalistic and schematic This work introduced an alternative classificatory traits. They are also quite diverse. The earliest of them, scheme, listing 18 stages through which a language Volapük (1879), was developed by Johann Martin project might be expected to pass on its way to full Schleyer, a Catholic priest in Baden, southern socialization. On this scale, Ido and Interlingua were Germany. The vocabulary of Volapük consisted of identified as ‘semilanguages’, having developed beyond 91 ARTIFICIAL LANGUAGES the project stage to acquire an original and translated lit- Fundamento (1904; Foundation of Esperanto), which erature, oral use at international meetings, and other Zamenhof persuaded the first international Esperanto markers of communicative use. Esperanto, however, was congress (1905, Boulogne-sur-Mer) to adopt as an unal- shown to be a qualitatively different phenomenon, being terable paradigm of acceptable usage. It also accounts used on a scale and in a range of situations as extensive for the effort he invested in translations of such works as, although different from, many ethnic languages. as Gogol’s The Tax Inspector, Shakespeare’s Hamlet, Zamenhof’s original blueprint for Esperanto con- the Old Testament (from Hebrew), and Andersen’s sisted of 16 short rules on a variety of topics, a glos- Fables; these influential texts did much to develop sary of 940 words, some samples of poetry and prose, Esperanto literary style and semantics. After World War and a lengthy essay on the key characteristics of his I, literary and scientific writing in Esperanto flourished proposal. Like most planned languages, Esperanto and diversified in both book and periodical form; major uses the Latin alphabet, and pronunciation follows libraries currently hold 10,000 to 20,000 volumes. orthography; in words with several syllables, intona- Development of the oral language has also been tional stress falls on the next to last syllable. Chief marked, particularly since the 1970s, with the expan- among its distinguishing features is its system of sion of international telephony and travel. Today, word-class endings: -o marks nouns, -a adjectives, and although influential dictionaries, grammars, and text- -e adverbs; -i/-is/-as/-os/-us/-u form verbs in the infini- books exist, none can be taken as a definitive or com- tive, past, present, and future indicative, conditional, plete description of Esperanto as it is actually used. and imperative modes, respectively. Apart from certain Valuable research on Esperanto has been carried closed class words (the definite article la, prepositions, out both by professional linguists and philologists and conjunctions, comparatives, correlatives, and a small by brilliant amateurs such as the Hungarian polymath number of adverbs ending in -au˘), all lexical items Kálmán Kalocsay (Plena Gramatiko de Esperanto, must be combined with one of these endings to func- 1932; Complete Esperanto Grammar). Only a small tion in a sentence. Word roots are borrowed intact or sample of this work is available in English. Particular with inflectional endings removed, primarily from interest has been focused on the rules governing word Romance (75%), but also from Germanic (20%), and sentence formation; the etymology and semantics Slavic, and other sources. Direct loans from non- of the core vocabulary; the interplay of prescriptive European languages remain relatively rare. principles (such as the 16 rules), linguistic ideologies, A second key characteristic of Esperanto is the and tacit linguistic models; the relationship between invariance and high combinability of all of its ele- language, culture, and identity; the development of ments. Zamenhof introduced more than 20 affixes to Esperanto literature; and the use of the language in express ideas related to the root word, such as ‘place various scientific and technical fields. In the 1980s, a of’, ‘container for’, ‘opposite of’, and so on; these Dutch project examined Esperanto’s suitability as an have since been supplemented by a handful of others. intermediate language for machine translation. The Word roots may also be freely combined. Many com- Modern Languages Association of America lists 200 mon compounds have acquired conventional meanings to 300 publications in its annual bibliography. over time, so that lern-ej-o, for instance (literally Several million people have studied Esperanto, on ‘learning-site’), usually indicates a school of some the basis of textbook sales and other data; however, the kind. Early compound formation was influenced by population of those with an active command of the lan- Germanic and Slavic models (e.g. el-don-i, to publish, guage is much more difficult to characterize. Taking literally ‘to give out’), but modern usage favors greater into account the reported membership of the Universal transparency and what might be called ‘indigenous’ Esperanto Association (the most representative interna- models, e.g. the direct transformation of adjectives tional organization of Esperanto speakers), annual sales into verbs (bon-a, ‘good’, bon-i, ‘to be good, to mani- of books and periodicals, and economic barriers to par- fest goodness’) or the use of affixes and prepositions ticipation, a reasonable current estimate would be as independent roots (ad-e, ‘continually, repeatedly’, something more than 50,000 active speakers, along per-i, ‘to mediate, broker’, et-a, ‘small’). with a substantially larger number of passively compe- This introduces a third key characteristic of tent, lapsed, or occasional users. Some three quarters of Esperanto: its evolution through communicative use. active speakers are located in Europe (including Throughout 25 years as the central figure of the move- Eastern Europe and Russia), with established commu- ment, Zamenhof consistently emphasized an inductive nities of users in most East Asian and Latin American approach to learning and using the language, with new countries, in West Africa, and in regionally isolated speakers developing their own tacit linguistic models on countries such as Iran and Israel. There are active, self- the basis of a growing body of literary and scientific organizing networks of both young speakers (under 30 texts. This approach is reflected in the exercises of the years) and Esperanto-speaking families, including a 92  ASPECT few hundred first-language speakers. Active specialist Forster, Peter G. 1982. The Esperanto movement. The Hague: organizations exist in many areas, including predomi- Mouton. Gledhill, Christopher. 1998. The grammar of Esperanto: a cor- nantly oral domains such as popular music and radio pus-based description. München/Newcastle: Lincom broadcasting. Increasingly, the World Wide Web con- Europa, 2nd edition, 2000. stitutes a valuable tool for research into the language Janton, Pierre. Espéranto. 1973. Paris: Presses Universitaires de and its domains of use. France; as Esperanto: language, literature, and community, ed. by Humphrey Tonkin, translated by Humphrey Tonkin, Jane Edwards, and Karen Johnson-Weiner. Albany: State References University of New York Press, 1993. Blanke, Detlev. 1985. Internationale Plansprachen: Eine Jordan, David. 1992. Being colloquial in Esperanto: a reference Einführung (International planned languages: an introduc- guide for Americans. Lanham: University Press of America, tion). Berlin: Akademie-Verlag. 2nd edition, El Cerrito: Esperanto League for North Couturat, Louis, and Léopold Leau. 1979. Histoire de la langue America, 1999. universelle (1903), Les nouvelles langues universelles Large, Andrew. 1985. The artificial language movement, (1907). Hildesheim and New York: Olms. Oxford: Blackwell. DuliÍenko Aleksandr. 1990. MeËdunarodnye vspomogatel’nye Schubert, Klaus (ed.) 1989. Interlinguistics: aspects of the sci- jazyki (International auxiliary languages). Tallinn: Valgus. ence of planned languages. Berlin: Mouton de Gruyter. Eco, Umberto. 1993. La ricerca della lingua perfetta. Roma: Tonkin, Humphrey (ed.) 1997. Esperanto, interlinguistics, and Laterza; as The search for the perfect language, translated planned language. Lanham: University Press of America. by James Fentress. Oxford: Blackwell, 1995. MARK FETTES Aspect ‘Aspect’ in its proper sense—also called ‘grammatical event from outside as an undivided whole, often aspect’—is one of the grammatical categories of the focusing on its completion. From the imperfective verb besides tense, mood, and others. The term perspective, the speaker views the event from inside ‘aspect’ is also used as a cover term for grammatical as an ongoing process, a state, or a habitual action aspect and two other closely related phenomena: without directing attention to the completion of the ‘aktionsart’ as a lexical-semantic category of verbs and event. From a textual point of view, verbs in the ‘predicational aspect’ as a semantic property of verb imperfective aspect contribute to the description of phrases. Aspect in its broader sense has to do with the the background of a scene, whereas verbs in the per- internal structure of the event the verb refers to, with fective aspect are used to describe the events that take the perspective the speaker adopts toward the course place on this background in their temporal succes- and structure of the event, or with both of these. sion. Thus, the Russian equivalent for English to write Because the expression of grammatical aspect is has an imperfective form, pisat’, and a perfective one, often closely intertwined with the expression of tense, napisat’. The imperfective one is used in contexts in in traditional grammars, aspectual distinctions are which English would use expressions such as she was often confusingly dealt with under the heading of writing or she used to write. The perfective form ‘tense’. Among the categorical distinctions in tense- occurs in contexts in which the completion or the aspect systems that are of an aspectual nature are those result of writing is emphasized (he wrote/has written between imparfait and passé composé/passé simple in the letter) and in which it occurs in a sequence of French, progressive and nonprogressive in English, or events (he sat down, wrote the letter, and put it in an imperfect and aorist in Greek. Formally, aspect can be envelope). Aspectual semantics is a very intricate expressed by very different means, e.g. by prefixes and matter, and the meaning of the perfective–imperfec- suffixes (Russian), reduplication (Chamorro), inflec- tive opposition involves cross-linguistic differences. tional and periphrastic forms within a tense-aspect Furthermore, some languages do not show a fully system (French, English), verb stem formation developed perfective–imperfective distinction but (Greek), or combinations thereof (Bulgarian). express aspectual meanings that can roughly be clas- The main opposition within the grammatical cate- sified as meaning components of the imperfective gory of aspect is the one between perfective and aspect (progressive vs. nonprogressive aspect, habitu- imperfective aspect. Informally, it is often said that in al vs. nonhabitual aspect) or the perfective aspect using the perfective aspect, the speaker looks onto the (completive vs. noncompletive aspect). 93 ASPECT In contrast to grammatical aspect, the term ‘aktions- events as culminating or delimited) like draw a circle, art’—which translates as ‘kind of action’ from run a mile, write a letter. Despite the popularity of German—is considered to be a lexical-semantic cate- these so-called Vendler classes, they are problematic gory. Approaches to aktionsart differ as to whether in many respects. In particular, the four classes are dis- they consider all verbs to show a certain aktionsart or tinguished by semantic properties belonging to very whether they restrict the term to phenomena in which different domains (length of event, stativity, comple- aktionsart meanings are expressed by certain prefixes tion), and it is often unclear if the domain of the clas- or suffixes. Aktionsart distinctions are mainly used to sification is verbs, verb phrases, or sentences. classify verbs according to their reference to parts of Formal semantics has been especially interested in events, to events of a certain length, and to repetitions the difference between activities (e.g. she ate, she ate of events (affixes and their translations are under- apples, she watched the apples) and accomplishments lined): German loslaufen ‘to start running’, Russian (e.g. she ate the apples). Expressions describing activ- posidet’ ‘sit for a while’, Russian kriknut’ ‘to scream ities can be modified with phrases like for ten minutes, once’, German austrinken ‘to drink up’, Russian pisy- but not phrases like in ten minutes. Because watch the vat’ ‘to write repeatedly’. Because both grammatical apples is an activity, She watched the apples for ten aspect and aktionsart are mainly expressed by means minutes is acceptable, but She watched the apples in of word formation in Slavic languages—which have ten minutes is not. With accomplishments, such as eat played an important role in research on aspect—the the apples, it is the other way around: She ate the proper distinction between the two categories has apples in ten minutes is acceptable, whereas She ate always been a subject of debate. The diversity of the the apples for ten minutes only seems to be so if the phenomena covered by the term ‘aktionsart’ is reflect- context forces eat the apples to be reinterpreted as an ed in the diverse semantic approaches to aktionsart. It activity. can be observed, though, that formal semantic approaches to aktionsart phenomena often involve References notions like ‘part of an event’ (an event referred to by German austrinken ‘to drink up’ as part of an event Binnick, Robert I. 1991. Time and the verb. A guide to tense referred to by trinken ‘to drink’), ‘recurrence of an and aspect. New York, Oxford: Oxford University Press. Chung, Sandra, and Alan Timberlake. 1985. Tense, aspect, and event’ (rewrite, rebuild), and ‘event presupposition’ mood. Language typology and syntactic description. Vol. (he ate up presupposes he ate; she rebuilt the house III: Grammatical categories and the lexicon, ed. by Timothy presupposes somebody built the house before, etc.) Shopen. Cambridge: Cambridge University Press. Sometimes, the term ‘aktionsart’ is also applied to Comrie, Bernard. 1976. Aspect. An introduction to the study of or—better—confused with the phenomenon of ‘predi- verbal aspect and related problems. Cambridge: Cambridge University Press. cational aspect’ (also called ‘lexical aspect’). The most Dahl, Östen. 1985. Tense and aspect systems. Oxford: Basil influential description of predicational aspect was Blackwell. given by Zeno Vendler (1957), who distinguished four Vendler, Zeno. 1957. Verbs and times. The Philosophical ‘time schemata:’ (1) statives like know something, hate Review LXVI. 143–60. somebody; (2) achievements (roughly, expressions Verkuyl, Henk J. 1993. A theory of aspectuality. The interaction between temporal and atemporal structure. Cambridge: referring to punctual events) like reach the summit, Cambridge University Press. win a race, recognize somebody; (3) activities (rough- STEFAN ENGELBERG ly, expressions referring to events as ongoing or non- culminating) like push a cart, run, and swim; and (4) See also Inflection and Derivation; Tense and accomplishments (roughly, expressions referring to Aspect Marking; Time and Tense Assimilation and Coarticulation Assimilation refers to a change in sound due to the anticipation of the next sound (= anticipatory or right- influence of a neighboring sound (aspiration, glottaliza- to-left assimilation), or it can be carry-over (left-to- tion, nasal release, etc.). For instance, the /s/ in ‘hus- right) assimilation, in which an ongoing feature is band’ becomes voiced as a result of the voiced continued to the next sound. Assimilation is similar to phonemes surrounding it. The influence can be either in coarticulation, in that both processes deal with the mod- 94  ASSIMILATION AND COARTICULATION ification of sound due to contextual variability. Voicing However, the standard generative phonology (Chomsky This occurs, for instance, when ‘v’ in ‘have to’ is pro- and Halle, the Sound Pattern of English, 1968) makes a nounced with an [f] instead of [v] due to assimilation clear-cut distinction between assimilation and coarticu- of the voiced fricative to the following voiceless lation. Assimilatory processes are part of a grammar consonant. and are language specific (although they occur in many Not all speech sounds are equally affected by con- languages), and they can be accounted for by phono- textual variability. Bilabial, [b] and [p], tend to be very logical rules. Coarticulation results from the inertia of stable and show no effect of assimilation irrespective of the speech mechanism and is, hence, more universal. the context. Velars such as [k] and [g] are also stable, Both the distribution (universality vs. language speci- although the place of articulation is more fronted in the ficity) and the quality of the contextual change (mere context of front vowels: the closure of the [k] in ‘kick’ articulatory adaptation vs. intentional phonetic modifi- is more palatal, while that of the [k] in ‘cook’ is more cations) distinguish the two processes. However, there velar. Of the different plosives, assimilation occurs most is no simple way of distinguishing between those assim- in alveolars, i.e. in [t] and [d]. The closure of the [t] or ilation effects that are due to the inherent properties or [d] will not be alveolar, but will occur at the place of limitations of speech production and those that are not, articulation of the following segment. For instance, the unless the latter are very obviously language-specific. closure of a [t] or [d] before a dental will also sound Assimilation often appears to be motivated by ease of dental (indicated by a diacritic), e.g. ‘ho[t θ]ing ‘hot articulation, but what seems easy and natural in one lan- thing’, while that before a velar will sound velar, i.e. guage often turns out to be less so in another. ‘ho[k k]ake’, ‘hot cake’ or ‘ba[g g]irl’, ‘bad girl.’ Assimilation effects are not fully understood: it is Preceding a bilabial, the [t] or [d] will also be bilabial, not clear to what extent we can explain them by i.e. ‘hot pie’ becomes ‘ho[p p]ie’ and ‘bad boy’ becomes assuming that high-level commands associated with ‘ba[b b]oy’. specific segments are confounded by biomechanical ‘sloppiness’ or to what extent high-level commands are deliberately planned to optimize transitions Manner of Assimilation between targets and to yield the best possible vocal Alveolar stops are not only influenced by surrounding tract performance in running speech. consonants but also by surrounding vowels. Here, the Assimilation is often mentioned in connection with assimilation involves manner rather than place of historical changes. Many of the sound changes in lan- articulation. In several North American accents, the [t] guages can be described as assimilatory. For example, and [d] are neutralized, i.e. their characteristic physi- English words such as mission, special, or passion were cal properties are replaced by the voiced alveolar flap once pronounced with a medial [sj] but in modern (transcribed as [ɾ]). This sound contains voicing and a English have [ʃ]: by a process of assimilation, the [s] flap (= a very brief contact between the tongue tip and has been retracted in anticipation of the following [j] or the alveolar ridge). Using this flap, the words ‘Adam’ [I]. However, it is important to distinguish between his- and ‘atom’ may sound identical, both words having the torical processes (which are over and done with) and flap for the intervocalic ‘t’ and ‘d’. This flapping processes that are still current or operative in the mod- occurs whenever what would be [t] or [d] in other ern language. They are part of the speaker’s organiza- accents occurs between two vowels, not only within tion of pronunciation and are relevant to our description words (see previous example) but also across word of the language (e.g. prefix un-, which is pronounced as boundaries, as in ‘ge[ɾ] away’ (get away). However, [n] in nonvelar contexts (untidy, unsettled) but which flapping does not occur when the stop precedes a are pronounced as [ñ] before a velar (unkind, ungainly). stressed syllable, as in ‘attract’, where the second syl- lable carries the stress. Place of Articulation This occurs in English, for example, by ‘ratbag’ or Affricates ‘oatmeal’ pronounced with [p] instead of [t] in rapid or There is little assimilation of the affricates in English, informal speech, by assimilation of the alveolar stop to although there may be some variation among speakers. a following bilabial. Labiodental fricatives [f] , [v] do not show a great deal of assimilation, although [v] may become a voice- less word, finally preceding a voiceless obstruent Manner of Articulation (ha[f] to, o[f] course. Here, ‘Indian’ is pronounced as ‘Injun’, where the stop In English, in fast speech, dental fricatives [ð] often [d] and approximant [j] merge to form an affricate assimilates entirely to a preceding alveolar sound ‘I[n (same for soldier). n]epub ‘in the pub’. 95 ASSIMILATION AND COARTICULATION The alveolar [s] and [z] are often assimilated to the able to distinguish phonological (language-specific) following palatal glide [j] or palatoalveolar fricative from phonetic nasalization by manipulating speech as in ‘mi[š j]ou’ (‘miss you’), ‘it wa[ j]ellow’ (‘it was rate. They found a quite different distribution of the yellow’). temporal patterns of nasalization in the two languages, Assimilation is quite a general phenomenon in con- as a function of rate: in American English the extent of nected speech, but it spreads wider in fast speech than nasalization on the vowel preceding the nasal was pro- in slow speech. It involves the change of some seg- portional to the varying vowel duration, while in ment under the influence of another one, and the Spanish it remained constant. They concluded that the change makes the two speech sounds more similar. spread of nasalization as vowel duration increases in The phrase ‘ten books’, for example, is pronounced as AE must be intentional (phonological), while the short [tem buks], where /n/ assimilates to the adjacent /b/ by and constant extent of nasalization in Spanish must be adopting its bilabiality feature. Most forms of assimi- an automatic consequence of the speech mechanism, lation are to be located at the segment spellout level. since it reflects the minimum time necessary for the Assimilation should be distinguished from coartic- lowering gesture of the velum. However, there is no ulation. Adjacent speech sounds interact because of strict dichotomy between universal and language-spe- the physiology and the mechanics of articulation. cific variations, as coarticulation differs in degree These interactions become more intense at higher across languages. speech rates. They depend on the time allotted to the articulation of syllables (coarticulation). References There is a wide range of assimilation processes that take place across a syllable or word boundary. In Borden, G.J., and Harris K.S. 1980. Speech science primer: Swedish, a syllable final /r/ is integrated with any physiology, acoustics and perception of speech. Baltimore: Williams and Wilkens. (phonemically adjacent) apical dental consonant(s) Clark, J., and Yallop C. 1995. An introduction to phonetics and making the latter a retroflex alveolar, and this effect phonology, 2nd edition, Oxford: Blackwell. spreads over the word boundary. Brosnahan, L.F., and Malmberg B. 1970. Introduction to pho- Quite often, context-dependent changes involving netics. Cambridge, England: W. Heffer & Sons Ltd. the same articulatory structures have different acoustic Chomsky, N., and Halle M. 1968. The sound pattern of English. New York: Harper & Row. and perceptual manifestations in different languages, Davenport, M., and Hannahs S.J. 1998. Introducing Phonetics so that it is possible to distinguish what can be consid- and Phonology, 196pp. Great Britain: Arnold. ered universal phonetic behavior from language par- Farnetani, E. 1997. Coarticulation and connected speech ticular rules. A classical example is the difference processes. The handbook of Phonetic Sciences, ed. by W.J. between vowel harmony, an assimilatory process pres- Hardcastle and John Laver, 371–404. Cambridge, MA: Blackwell. ent in a limited number of languages (Hungarian) and Ladefoged, P., and Maddieson I. 1996. The sounds of the the process of vowel to vowel coarticulation, attested world’s languages. Oxford: Blackwell. in many languages and probably present in all. In other Solé, M.J., and Ohala J. 1991. Differentiating between phonet- cases, cross-language differences are not easily inter- ic and phonological processes: the case of nasalization. pretable and inferences about the nature of the under- Proceedings of the 12 International Congress on Phonetic Sciences, Vol. 2, 110–13, Aix-en-Provence. lying processes can be made only by manipulating ASTRID VAN WIERINGEN some of the speech parameters, for example segmental durations. In a study of vowel nasalization in Spanish See also Anatomy of the Articulatory System; and American English, Solé and Ohala (1991) were Assimilation and Dissimilation; Speech Production Assimilation and Dissimilation All words have citation forms, or dictionary pronunci- sound changes when the word is used in connected ations, which occur when they are pronounced in iso- speech. One of the most common processes of sound lation. These dictionary pronunciations may undergo change is assimilation. 96  ASSIMILATION AND DISSIMILATION ‘Assimilation’ is a cover term that includes a variety Assimilation especially affects voicing, manner of of processes and refers to the change in a language articulation, and place of articulation. Voicing assimi- sound that makes it more similar to a neighboring lation is seen in the case of English -s, where the voic- sound. The process by which the Latin word noctem, ing of the first consonant carries over to the voicing of meaning ‘night’, became notte in Italian shows an the suffix. The effect of assimilation on place of assimilation of the [k] sound, represented by <c> in articulation can be seen in the case of English <n>. the spelling, to the same point of articulation as the The overwhelming pattern for this kind of assimilation following [t]. is for the first consonant to assimilate to the place of ‘Dissimilation’ is the opposite process, in which articulation of the second, and not the opposite. In sounds become more unlike neighboring sounds. The English, this occurs commonly with words like input, German word Kartoffel ‘potato’ comes from an earlier pronounced usually as imput, and income, frequently form Tartuffeln. In this case, the [t] at the beginning of pronounced as iŋ come. The same in- prefix in Latin the word became [k], a sound dissimilar to the follow- before the labial sounds [m], [p], and [b] was written ing [t]. Compared to assimilation, dissimilation is an as [m], as in immemorial, impossible, and imbibe. uncommon process in the world’s languages and can Evidence of the manner of articulation assimilation be less obvious to an observer. As an example, in many with the in- prefix can also be seen historically in Latin Bantu languages, prefixes must disagree in voicing in words that were borrowed by French and later with the first consonant of the root. Thus, the conso- English. Before the continuant sounds [l] and [ɺ], in- nant of the prefix tu-/du- for ‘I’ changes, depending on was pronounced and written identically to the follow- whether or not the first consonant of the verb is voiced. ing sounds, so that in+legal and in+regular were writ- Thus, bona ‘to see’ starts with a voiced consonant [b] ten as illegal and irregular. and ‘I see’ is consequently tu-bona, whereas soma ‘to The most common explanation for assimilation read’ starts with a voiceless consonant [s], and the pre- is that speakers adjust the articulation of sounds fix shows up with a voiced one: du-soma ‘I read’. because it is easier than keeping each sound distinct Assimilation can occur both across word bound- in the stream of speech. When speakers are produc- aries and within words, with sounds affecting each ing hundreds of sounds per minute, it is only na- other in three main ways: regressively, progressively, tural that the precise articulation of sounds will be and bidirectionally. affected by other sounds near them. There is also evi- ‘Regressive’ or ‘anticipatory’ assimilation refers to dence that some types of assimilation may have the effect of a later sound on an earlier sound. One of their roots in listener perceptions rather than speaker the most common types of regressive assimilation in articulations. many languages involves nasal consonants, which reg- While assimilation takes place in all languages, the ularly change place of articulation to match the place extent of assimilation is sensitive to speech rate and of articulation of the following consonant. In English, register. All other things being equal, assimilation is the <n> in the prefixes in- and un- is usually pro- more likely to occur in casual styles of speech than in nounced [n] only when the following sound is an alve- more careful, formal styles. The kinds of assimilation olar consonant, as in intention. In input, the following and the extent to which it occurs are also relevant in sound is bilabial [p]; hence, the <n> is actually pro- distinguishing differences in varieties of a language. nounced as bilabial [m]. One clear difference between standard southern ‘Progressive’ assimilation refers to the effect of an British and General American English is in the pres- earlier sound on a later one. The English -s mor- ence or absence of coalescent assimilation in words phemes found in plural nouns, third person present such as gradual or issue. The British pronunciation tense verbs, and possessives are examples of this type generally does not assimilate the sounds in the middle of assimilation, being pronounced [s] after voiceless of words, i.e. the du in gradual is pronounced [dju], sounds (e.g. cats, walks, Pete’s, respectively) and [z] and the su in issue is pronounced [sju]. The American after voiced sounds (e.g. dogs, reads, Mary’s). pronunciation, however, does assimilate, leading to ‘Bidirectional’ or ‘coalescent’ assimilation occurs [dZ u] and [ʃu], respectively. when two sounds influence each other, as in the com- mon American English spoken forms for an alveolar References consonant followed by an unstressed syllable beginning with a [j]. In the following examples, the [t] and [d] Broe, Michael. 1992. An introduction to feature geometry. In Docherty and Robert. sounds join the [j] to create an affricate: did you [did Docherty, Gerald, and D. Robert Ladd (eds.) 1992. Papers in juá]
[dIdZ u], can’t you [kænt juá]
[kænu], laboratory phonology II: gesture, segment, and prosody. graduate [g»æ dZueIt], ritual [»Iu ə]. Cambridge: Cambridge University Press. 97 ASSIMILATION AND DISSIMILATION Katmaba, Francis. 1989. An introduction to phonology. Nolan, F. 1992. The descriptive role of segments: evidence from London: Longman (especially Chapter 5). assimilation. In Docherty and Ladd. Kingston, John, and Mary Beckman (eds.) 1990. Between the Ohala, John. 1990. The phonetics and phonology of aspects of grammar and physics of speech. Cambridge: Cambridge assimilation. In Kingston and Beckman. University Press. JOHN M. LEVIS Lodge, Ken. 1992. Assimilation, deletion paths, and underspec- ification. Journal of Linguistics 28. 13–52. See also Assimilation and Coarticulation Austin, John Langshaw Austin was a leading member of a school of philosophy argued that, contrary to what the weight of philosoph- variously referred to as ‘Oxford philosophy’, ‘linguis- ical and grammatical tradition would have us believe, tic philosophy’, or ‘ordinary language philosophy’ that not all declarative sentences are used to make state- developed in Britain in the years between the two ments. While a sentence such as ‘The cat is on the World Wars and enjoyed its heyday in the late 1940s mat’ may indeed be used to make a statement about a through the early 1960s. Each of these descriptive certain state of the world involving a feline creature terms is, however, arguably inaccurate to some extent. and its whereabouts at the time of speaking, it is not at Not all of the members were, for instance, based at all clear that words such as ‘I hereby pronounce you Oxford University; nor is it true that all the philoso- man and wife’ (as uttered by a priest at a given phers at Oxford were sympathetic to its basic tenets. In moment during a marriage ceremony) or ‘I name this fact, it may even be argued that the group did not prop- ship the HMS Queen Elizabeth’ (as pronounced by, erly constitute a ‘school’ but at best a loosely knit gath- say, the Prime Minister of Britain at the launching cer- ering of philosophers who shared some common views emony of the latest addition to his country’s naval or at least a certain attitude toward philosophy and who fleet) are used to make statements. These latter utter- regularly gathered for an informal chat for what were ances do not describe anything in the world; if any- famously known as the ‘Saturday morning talks’. thing they help alter the world in significant ways by Austin published only seven papers and a book dur- effectively bringing about certain changes. The priest’s ing his lifetime. The book was an English translation words, for instance, literally help ‘create’ a newly of Grundlagen der Arithmetik (Foundations of arith- wedded couple. Austin coined the term ‘performa- metic) by the German mathematician and philosopher tives’ to refer to the utterances of the latter kind that, Gottlob Frege (1848–1925). Austin’s major works despite their declarative form, did not result in state- How to do things with words (HDTW) and Sense and ments being made. The former, run-of-the-mill declar- sensibilia (S&S) were both published posthumously. ative sentences were, by contrast referred to as So too was Philosophical papers, a collection of 12 ‘constatives’. Constatives, being straightforward dec- papers, most of which had been originally presented at larations, could be assessed in terms of truth and falsi- the annual meetings of the Aristotelian Society and ty. The performatives were neither true nor false. They published in the conference proceedings. HDTW and could be considered ‘felicitous’ or otherwise, depend- S&S are the written versions (based on Austin’s own ing upon a number of attendant circumstances—in the annotations as well as notes prepared by members of case of the marriage ceremony, for example, the words the audience) of two series of talks delivered by Austin had to be pronounced by a priest and nobody else, etc. at Harvard (1955) and the Berkeley Campus of the Furthermore, performatives were not descriptive; to University of California (1958), respectively. utter them was to perform (whereof the neologism) HDTW is undoubtedly Austin’s major philosophi- certain acts—later subsumed under the broader cate- cal work and the one responsible for his reputation gory of ‘speech acts’. outside the disciplinary bounds of philosophy, mainly Pursuing his line of inquiry, Austin painstakingly linguistics. In it, Austin develops with great verve a sought formal criteria by which to distinguish perfor- philosophical thesis that, despite its apparent simplici- matives from constatives. Although a number of can- ty and unpretentiousness, turned out to be a ground- didates emerged, such as the use of the first person, breaking discovery in philosophical analysis. Austin present indicative, active voice, etc., counterexamples 98  AUSTIN, JOHN LANGSHAW soon convinced Austin that the very attempt was a can be made explicit by the use of the performative wild goose chase. Even the hypothesis that peforma- verb ‘promise’ as in ‘I promise etc.’ Curiously, there tives were mostly utterances that were culture-specific are no formulas for making explicit the perlocutionary (such as marrying in a Christian church) turned out be nature of utterances; utterances such as ‘I threaten you difficult to sustain. This was confirmed by the discov- that …’ and ‘I persuade you that …’ are bizarre, to say ery that even such ordinary utterances such as ‘I prom- the least. This is but a consequence of the fact that the ise that p’ (where p is any complement denoting a success of a perlocutionary act depends on the hearer future course of action to be performed by the speak- doing his/her (i.e. feeling threatened, persuaded, etc.) er) meet all the features that help identify performa- part rather than the speaker simply intending to tives. A promise is neither true nor false; it may be achieve a certain goal. described as infelicitous if it turns out, for instance, S&S addresses the familiar problem of sensory per- that the speaker had no intention whatsoever to carry it ception and argues against the then mainstream view out. So a promise is as good a speech act as marrying patronized by, among others, A.J. Ayer and H.H. Price, a couple or launching a ship. And so too are such rou- to the effect that what one perceives is a sensory tine acts as ordering, requesting, etc. impression which may or may not be an accurate rep- Austin was also struck by the fact that a ‘p’ sim- resentation of the object perceived. Austin argued that pliciter can just as well do the job as ‘I promise that p’ the correct answer to the question ‘What exactly is one provided the circumstances are appropriate. Thus ‘Do looking at when one gazes a church camouflaged as a it’ as uttered by an army general to a private is readily barn or a straight stick placed a glass full of water’ understood as an order rather than, say, a request, even should be ‘A church-camouflaged-as-a-barn and a in the absence of the ‘performative formula’ ‘I order stick-made-to-look-bent’. In other words, the epis- you to …’. The only difference between them is one of temic state of the observer is part of the objectual real- explicitness. Accordingly, Austin distinguished ity that is perceived. Now, this is indeed a bold thesis between implicit (or ‘primary’, as he called them) and because it flies in the face of the claim made by naïve ‘explicit performatives’. realists that truth is nonepistemic. But then, at this precise stage, the Austin reader is Austin became a household name in disciplines only being prepared for a dramatic turn in Austin’s such as linguistics, psychology, sociology, anthropolo- reasoning. The sentence ‘The cat is on the mat’, he gy, etc., thanks to John Searle who took up the torch argued, is a primary performative whose explicit coun- after Austin’s untimely death in 1960. Searle is wide- terpart would be ‘I state that the cat is on the mat.’ This ly recognized as Austin’s intellectual heir. In fact, means, Austin went on to argue, that there really are no many textbook presentations of the speech act theory constatives but only some very ‘clever’ performatives leave the reader with the impression that there is a masquerading as such—so successfully indeed that smooth continuity between the two. Recently, howev- generations of philosophers were tricked by them. er, the continuity thesis has been challenged by a Linguistic activity is fully taken up by the performance growing number of scholars. Austin’s philosophy con- of speech acts. tinues to attract attention, and many scholars from The distinction between constatives and performa- fields other than philosophy have come up with fasci- tives was thus replaced by a three-way distinction: nating readings of his works. locutionary acts, illocutionary acts, and performative acts. Locutionary acts were acts of producing utter- Biography ances with certain sense, while illocutionary and per- locutionary acts were acts done in saying certain John Langshaw Austin was White’s Professor of things and by saying certain things, respectively. The Moral Philosophy at the University of Oxford, locutionary acts were further subdivided into ‘phonet- England (1952–1960). He was born in Lancaster, ic acts’ (acts of merely uttering certain noises), ‘phat- England on March 26, 1911, and spent his early years ic acts’ (acts of uttering certain vocables or words), in Scotland. At 13, he obtained a scholarship in clas- and ‘rhetic acts’ (acts of using the vocables with a cer- sics and went to Shrewsbury school. Five years later, tain more or less definite sense and reference). Unlike he won another scholarship in classics and moved to the illocutionary acts whose performance depends on Balliol College, Oxford. In 1931, he won the presti- the speaker saying the words with a certain intention, gious Gaisford Prize for Greek Prose. He moved to etc., perlocutionary acts can be said to be felicitous Oxford definitively in 1933, when he was examined just in case the hearer reacted to the utterance in the and elected to a Fellowship at All Souls College. He way intended by the speaker. Thus, promising is an moved to Magdalen as a tutor in 1935. In 1939, he illocutionary act whereas threatening is a perlocution- published his first paper. The following year, he was ary act. The illocutionary force of an act of promise commissioned in the Intelligence Corps of the Army 99 AUSTIN, JOHN LANGSHAW and was posted to the war Office in London. In 1941, ———. 1962. How to do things with words. Oxford: he married Jeane Courts. In 1944, he moved to Clarendon Press. ———. 1962. Sense and sensibilia. Oxford: Clarendon Press. Normandy and from there to Versailles. He left the Berlin, Isaiah, et al. (eds.) 1973. Essays on J.L. Austin. Oxford: Army in 1945 with the rank of a lieut.-col. and the title Clarendon Press. of OBE, and was awarded the Croix de Guerre by the Cavell, Stanley. 1995. Philosophical passages: Wittgenstein, French and appointed Officer of the legion of Merit by Emerson, Austin, Derrida. Oxford, UK: Blackwell. the Americans. In 1950, he published an English trans- Fann, K.T. (ed.) 1969. Symposium on J.L. Austin. London: Routledge & Kegan Paul. lation of Gottlob Frege’s Grundlagen der Arithmetik Felman, Shoshana. 1980. Le Scandale du corps parlant: Don Juan (Foundations of Arithmetic). He visited Harvard in the avec Austin ou La séduction en deux langues. Paris: Seuil. spring of 1955 where he delivered the William James Rajagopalan, K. 2000. On Searle [on Austin] on meaning. lectures (later published as HDTW) and the Berkeley Language & Communication 20. campus of the University of California (where he Searle, John R. 1969. Speech acts: an essay in the philosophy of language. New York: Cambridge University Press. delivered lectures on perception, later published as Shaw, H.E. 1990. With reference to Austin. Diacritics 20. S&S). He died of cancer in February 1960. Vendler, Z. 1967. Review of Austin’s Sense and sensibilia and How to do things with words. Foundations of Language 3. References Warnock, G. J. 1988. J.L. Austin. London: Routledge. KANAVILLIL RAJAGOPALAN Austin, John Langshaw. 1961. Philosophical papers. London: Oxford University Press. See also Searle, John Australia Australia with its overseas territories is somewhat Australians tend to speak of ‘nations’ or ‘peoples’.) smaller than the USA and, despite a population of only Each language thus averaged 600 speakers, although 20 million, is linguistically richer than many other there were much smaller and much larger ones. But countries. Although English is its principal language, even the larger languages never exceeded a few thou- Australia is host to some 250 other languages as well. sand speakers. According to indigenous creation sto- It is an interesting area for studies of languages in con- ries, languages were implanted onto the land that tact: (1) Many languages have been imported, and tribes were granted as caretakers, so that the land, English has undergone independent development. (2) people, and language formed a symbolic unit. Languages and dialects have influenced each other in No secure hypothesis regarding genetic affiliation many ways, with contact languages emerging. (3) with languages outside the continent can be proposed. Many communities, especially indigenous ones, are Although there is a high level of diversity, especially bilingual, although many languages have been lost due among non-Pama-Nyungan languages, there is evi- to overt pressure or a natural shift to English (rarely to dence for the assumption that all languages have another language); however, a number of languages descended from a ‘proto-Australian’ language. With have been reintroduced in the twentieth century, creat- respect to their typological affiliation there are com- ing a multilingual environment. (4) Australia has peting hypotheses. An older one argues that they form developed language and educational policies that have a separate group with two families: the non-Pama- had important effects on the texture of society. Nyungan languages in the central north and northwest Much of Australia’s pre-European human history— and the Pama-Nyungan languages elsewhere. A more the initial immigration, spread, and growth, social recent one maintains that there are clusters of lan- organization, and the relationship of indigenous lan- guages definable in terms of descendency, and others guages with their neighbors—remains obscure. The that are more related in terms of areal (or contact) fea- population was small; most estimates suggest a figure tures. But all make use of grammatical inflections, approximately above 300,000. As for the number of mark semantic roles more than syntactic function, languages, experts agree on about 250 languages with have free word order, and have complicated, subtle 250–300 dialects, which were tied to social units, i.e. distinctions in the semantics of pronouns, kinship rela- tribes, clans, or moieties. (Today, indigenous tions, and so on. 100  AUSTRALIA With 250 languages in use by a small population, from Britain. Speakers of other languages were assim- multilingualism was common and was maintained ilated or integrated over time and shifted to English through social practices: ceremonies, marriage outside without leaving a linguistic trace. the immediate group, and the association of languages Immigration of non-English speakers has had peaks with the land. Intertribal communication was facilitat- and troughs. Nineteenth-century migration was fed ed by the fact that neighboring languages shared lin- mainly by Celtic-speaking Britons, Germans, Chinese, guistic features, by the existence of a shared sign and, from the late nineteenth century, Italians. There language and translators. There is little indication that were small numbers of Jews, Poles, Afghans, languages functioned intertribally, although today Japanese, Malays, and Indians. Asians and Pacific some are used as lingua francas. Islanders were brought in to provide labor, as in The continent’s isolation meant that European households or for camel driving, pearl fishing, or sugar colonization began late. The expansion from early cane growing. Gold fields were multilingual sites, but settlements—in the Sydney region, Port Arthur cities became multilingual as well, while indigenous (Tasmania), and Norfolk Island—was slow, and not diversity was suppressed. The ‘White Australia’ poli- until the twentieth century had the whole continent been cy from the late 1880s to the mid-1960s put pressure explored. Indigenous cultures and languages were thus on non-whites to assimilate or emigrate. The education better maintained in the remote northwest and center. In acts (Victoria’s was the first in 1872), which made areas that were settled early, indigenous languages education free and secular, as well as World War I and ceased to be used decades ago. Thanks to language poli- other developments, sped up the loss of immigrant lan- cies and community efforts, over 50,000 people today guages, so that Australia assumed an Anglo-Celtic use an indigenous language in addition to English or a outlook. English became the only medium of commu- contact language. Efforts are under way to revive or nication in the public domain. document traditional languages and transmit knowledge The period from 1945 to the 1980s witnessed about them to children so as to halt the breaking of the renewed waves of immigration. Immigrants in the chain of transmission and their complete loss. early post-World War II years were predominantly Contact with English has led to borrowing, restruc- European and reintroduced their ‘Languages-Other- turing, and code switching. Thus, the word kangaroo is Than-English’ (LOTEs). The end of the White used in many languages, although it originated in Australia policy in the late 1960s triggered an influx of Queensland. The elaborate kinship terminology is grad- Vietnamese, Chinese, and Middle Eastern immigrants ually being abandoned and English words like father who introduced Vietnamese, Chinese languages, and brother acquire indigenous shades of meaning. This Bahasa Indonesia, Arabic, and other languages. Asian example from Warlpiri (around Alice Springs) shows languages are today among the strongest LOTEs and code switching: September-rla-rnalu meeting wangka- have the youngest speech communities. Most ja, wali natirli yanurnu committee member-patu panu European speech communities are aging; only Greek purda-nyanjaku. ‘During September, we attempted to is holding its ground. Approximately 25% of the pop- hold an AGM [= annual general meeting], but many ulation use a LOTE ‘at home’, according to the 2001 committee members were unable to attend.’ Although a census, which includes the 10% of the 450,000 indige- few languages have the vitality to be used in existing nous Australians who use an indigenous language. domains of communication, most have peripheral status There are regional differences among states and cities. and have lost the association with the land. Sydney has the largest proportion of LOTE speakers The immigrant population reached the 1 million and reflects recent Asian and Middle Eastern immi- mark as late as the 1850s. Growth accelerated as the gration, while Melbourne remains more European. herding, fishing, and manufacturing industries became Language maintenance and shift depend on well-stud- incentives for immigrants. Although immigration has ied factors, but it is worth pointing out differences remained a significant factor, natural growth never fell between different communities speaking particular below the 40% mark per decade from the end of the LOTEs such as Spanish, Arabic, and Chinese lan- nineteenth century, which ensured that English could guages regarding maintenance and shift. A very sig- develop reasonably independently from British nificant effect of recent immigration has been the rise English. But as most immigrants came from Britain of multiculturalism and, more recently, a concept of and Ireland, there was a continuous influx of native pluralism and the persistence of a climate of opinion speakers and an outside model to emulate. Its standing favorable to language policy and LOTE maintenance. was enhanced by the fact that children of well-off par- A significant feature is the existence of contact lan- ents were educated in Britain and that higher-level staff guages. A pidgin was first attested in Sydney as early in the education system, the churches, and later the as the early 1790s. Its spread followed the paths of the Australian Broadcasting Corporation were imported expansion of settlement and exploration. The wider 101 AUSTRALIA context of trade in the Pacific ensured that it was influ- texts. Australian English is known for its informality, enced by the maritime jargon of the Pacific, which nonstandard grammar, and slang expressions; note itself had been influenced by that in the Atlantic, by expressions like bullsh [= bullshit], bloody [as an emo- local languages and, later, by South Seas jargon and tive adjective], to have kangas in your paddock ‘to be Melanesian pidgin. It was not merely on the receiving crazy’. As elsewhere in the English-speaking world, end, but it influenced those pidgins during the nine- nonstandard grammar contains features like double teenth century as well. Kriol and Torres Strait Creole, negation; fairly unique to Australian English is the use two offsprings of Australian pidgins, developed in the of she for inanimate objects as in she’ll be alright ‘it’ll Northern Territory and north Queensland, respectively, be alright’. Today, Australian English is fully stratified and are used by more than 10,000 speakers each— socially but still lacks well-defined regional patterns. more than any traditional language. Depending on For well over a century, Australian English was con- their ‘broadness’, they are more similar either to sidered a low derivation from British English, but the indigenous languages or to Australian English. Due to last 40 years have witnessed the rise of a standard vari- decreolization, they now participate in a gradient that ety, which is the target of instruction and use in the connects them with Australian English via Aboriginal public domain. It is now well covered in dictionaries, English. usage guides, and even grammars. The participation of Aboriginal English is a dialect of Australian English Australian English in international movements, e.g. that straddles the space between pidgin and mainstream nonsexist language, plain English, and the dominance English, depending on a speaker’s background and of American English are putting pressure on it to adopt region. It has replaced indigenous languages as a international English, especially American English, source of influence on Australian English. Words like terms at the expense of local usages. business, sorry (day), and shame reflect indigenous A crucial feature of Australian English is the influ- meanings and relate business to traditional practices ence of indigenous and, to a lesser extent, immigrant and shame to emotions in the presence of strangers. languages. Indigenous loans like boomerang and kan- Aboriginal English is increasingly used in literature. garoo have become international words. The shift in Independent pidgin formation occurred in Western expressions to denote social units is interesting. The Australia and, possibly, in Tasmania. Norfolk Island term Aborigine is being restricted to a generic use, as Creole and (British) Pitcairnese are worth noting, both regional groups have been adopting indigenous names, of which go back to the early period of colonization. e.g. Koori (southeast) or Nyunga (west), which have Immigrant ethnic varieties of English have been become a part of standard Australian English. Most reported for Italian, Greek, and Asian communities. loans are, however, little used. Immigrant languages English, the typical lingua franca in multilingual set- have had little effect, mainly in the sphere of food, tings such as the workplace, has acquired discourse although German has contributed a few additional features that reflect the broad cultural background of words. As the title of Australia’s informal anthem non-English-speaking immigrants. ‘Waltzing Matilda’ shows, it is often impossible to No language can be understood without taking into ascertain whether a word has come into Australian account language contact and the wider situation, and English directly or by some other route. Both waltzing this is especially true of English. Today, Australian and matilda are originally German but are unlikely to English is a distinct center of English, but its historical have been brought in by immigrants. dependency on British English dialects is well known. With about 250 languages in use, the development Its accent and vocabulary have absorbed features from and implementation of a comprehensive language pol- many dialects, especially Cockney and those of south- icy since the late 1980s is a remarkable fact. Its key eastern England, Scotland, and Ireland. Australia’s features are: (1) English is the de facto national lan- demography ensured that dialects from southeastern guage and should be acquired by all citizens; (2) England provided the basis for Australian English. indigenous languages are to be maintained because Vowels in words like lead, late, and loud are similar to they link modern Australia to a distant, indigenous London’s Cockney; those in lid, led are high, as in tra- past; and (3) immigrant languages complement ditional accents from southeastern England. The English for as long as communities wish to maintain vocabulary has retained dialect words, e.g. tucker, and them. Some, such as Chinese, Japanese, German, and words from late eighteenth-century Anglo-American Spanish, are seen as beneficial to exports, tourism, and English, e.g. township, squatter. One can find traces of so on, and are offered more widely in education. While Anglo-American English and the northern type of multilingualism may not continue on the current level, English in the sentence structure of Australian it has widened the consensus on languages for nation- English, as in the definite article in ‘he goes to the uni’ ally important functions and has shifted the emphasis and the avoidance of shall outside very formal con- from ‘high cultural’ European languages to the ones 102  AUSTRIA that are closer to Australia’s geopolitical region. Leitner, Gerhard. 2004a. Australia’s many voices. Australian Although pluralism is still official policy, changes in English—the national languages. Berlin: Mouton-de Gruyter. Leitner, Gerhard. 2004b. Australia’s many voices. Ethnic the political climate and financial constraints have Englishes. Indigenous and migrant languages. Policy and reemphasized the role of English. education. Berlin: Mouton-de Gruyter. Malcolm, Ian. 1995. Language and communication enhance- References ment for two-way communications. Report. Perth: Edith- Cowan University. Clyne, Michael. 1991. Community languages. The Australian Ozolins, Urs. 1993. The politics of language in Australia. experience. Cambridge: Cambridge University Press. Cambridge: Cambridge University Press. Dixon, R.M.W. 2002. Australian languages. Their nature and Wurm, Stephen, Peter Mühlhäusler, and Darrel Tryon, (eds.) development. Cambridge: Cambridge University Press. 1996. Atlas of languages of intercultural communication in Encyclopedia of Aboriginal Australia. 1994. Canberra: the Pacific, Asia, and the Americas. Berlin: Mouton-de Aboriginal Studies Press. Gruyter. Leitner, Gerhard. 1992. Pluricentric English. Pluricentric lan- GERHARD LEITNER guages, ed. by Michael Clyne, 178–237. Berlin: Mouton-de Gruyter. See also English Austria The Republic of Austria is a country situated in sitzen (to sit), stehen (to stand), knien (to kneel), and Central Europe with an area of 83,853.3 square kilo- hocken (to squat/ sit) rather than the use of haben (to meters and an estimated population of approximately have) as prevalent in Germany. A further example is eight million. It is divided politically into nine states the use of different genders with particular nouns— (Bundesländer) and is bordered by the Czech Republic while the word for ‘salary’ is das Gehalt in Germany, to the north, Germany to the northwest, Switzerland it is der Gehalt in Austria. Variation is also found in and Liechtenstein to the west, Italy and Slovenia to the the level of language use (pragmatics). It has been south, and Hungary and Slovakia to the east. established, for instance, that realizations of requests The official language of Austria is German. and complaints in Austrian German are longer and However, the variety of German spoken in Austria is include more subjunctives and more modal verbs than not identical to that spoken in other areas, such as in in German German. There is no variation in the ortho- the Federal Republic of Germany. Rather, variation is graphical level given that the recent spelling reform of found in the phonological, morphological, syntactical, the German language was also adopted by Austria. lexical, and pragmatic levels of language. The most Up until the early 1980s, the German language was obvious differences between the German in Austria seen from a monocentric point of view. In other words, and in Germany are found in the lexicon. Although it the German spoken in the Federal Republic of is estimated that more than 95% of German vocabu- Germany was regarded the standard variety for lary is common to both these varieties, differences are Austria, as, indeed, it was for other countries in which apparent in everyday vocabulary and also in the vocab- German is an official language. Any specific features ulary used in relation to political, social, and econom- of these subvarieties were consequently viewed as ic institutions and structures. Many words specific to deviations from the main variety. The 1980s witnessed Austrian German, so-called Austrianisms, stem from the rejection of such monocentric views and the birth dialects, colloquial speech, or foreign languages. In of pluricentricity, whereby German was regarded as a particular, Austria’s proximity to non-German-speak- language with a number of national varieties, such as, ing countries to the south and east is evident in the for example, Austrian German. This view was based many borrowed words used in Austria. The Hungarian on the fact that the German spoken in different coun- influence is apparent, for example, in the use of the tries evidenced linguistic features particular to these term Schinakel for ‘rowing boat’, as is the French areas, and also on the fact that the inhabitants identi- influence in the use of Volant for ‘steering wheel’. fied with their particular country-specific variety. Differences in the area of syntax include, for instance, Current linguistic debate regarding the status of the use of the auxiliary sein (to be) with the verbs Austrian German concentrates on variants of this 103 AUSTRIA pluricentric thesis. Three particular standpoints are of Austrian standard abroad, particularly in the educa- interest here. Firstly, the pluricentric state-oriented tional context in Eastern Europe. thesis that we have just mentioned suggests that vari- As far as language planning, i.e. the use of deliber- eties of German should be classified according to the ate, often official, measures relating to the status particular nations with which they are associated. This and/or form of a language, is concerned, the codifica- view sees it as irrelevant that features suggested to be tion of Austrian German is yet to be completed. The specific to Austrian German may also be used in Österreichisches Wörterbuch (ÖWB) (Austrian dic- Southern Germany, for example, since such features tionary), published under the auspices of the Austrian are not standard in these latter areas—i.e. they are not Ministry of Education and the Arts, contains the stan- found in newspapers, for example, while they are in dard lexicon of Austrian Standard German (ASG) and Austria. The pluriareal thesis, on the other hand, sug- also includes some grammatical descriptions. gests that the varieties of German be differentiated However, further detailed descriptions of Austrian according to the areas in which particular features are Standard German remain to be conducted. found, irrespective of national boundaries. Proponents It is not the case that Austrian Standard German is of this view argue that dialects of Southern Germany, always spoken in Austria. Rather, a complicated con- such as Bavarian, share many features with Austrian tinuum between the poles ‘dialect’ and ‘standard’ has German—consequently, so-called ‘Austrianisms’ are been identified, with colloquial speech found between not Austrian-specific. In addition, they point out that the two poles. Regional variation is most pronounced many ‘Austrianisms’ are not common to Austria as a between eastern and western varieties. In addition, due whole, an east-west divide being identified. The final to late industrialization, variation in Austrian German standpoint is termed pluricentric-neutral and sees is influenced by social class to a larger extent than is Austrian German as the sum of the differences relative the case in Germany. to other varieties of German. National or regional fac- Approximately 98% of the population of Austria tors are irrelevant. speak German as their mother tongue. Ethnic minori- Although the concept of pluricentrism implies that ty groups account for the remaining figures. In addi- different varieties of a language exist, it does not sug- tion to the ethnic groups of workers and their families gest that all varieties enjoy equal status. The attitudes from Turkey and the former Yugoslavia who are con- of the Austrian people to Austrian Standard German centrated in towns and speak Turkish and Serbo- vary between viewing it as a soft, melodious language Croatian, respectively, there are four established to rejecting it in favor of German Standard German ethnic minority groups in Austria, namely the (GSG). Overall, a general, so-called ‘linguistic cringe’ Slovenes, the Croats, the Hungarians, and the Czechs, can be identified among Austrians relative to German the former two of which are the principal groups. Standard German. This stems from the fact that Austrian law ensures the rights of these groups to learn Germany is a vast country, with a much larger popula- their language in school, to use it in an official capac- tion, and much greater economic and political power ity, and to mark local signposts bilingually. Speakers than Austria. In addition, some semantic confusion are today bilingual, speaking German as well. The exists. Not only does ‘German’ denote the language coexistence of two or more languages within one state but also the people of Germany, leading the ‘German’ where speakers use these different languages alter- language to be associated with ‘the German people’. nately in particular situations (i.e. language contact) Also, there is a tendency for Germans to view nation- has effects both on the development of the languages al varieties of German as regional varieties, given the themselves and also on their status in a particular overlaps that exist between Southern German regional speech community. forms and Austrian standard forms. Furthermore, the Slovenian is the oldest ethnic minority language of majority of German language publishing houses are Austria, dating back over a millennium. It is spoken in based in Germany, and, thus, commonly demand the the southern state of Carinthia (Kärnten) in the Gail, use of German Standard German. Finally, the Duden Rosen and Jaun Valley south of Lake Woerth and the grammar and dictionary, published in Germany, and capital Klagenfurt, and is divided into several dialects. accepted as a norm in Austria, takes features of Language contact has led to changes in the lexical and German Standard German to be standard and marks phonetic levels in both Carinthian Slovenian and Austrianisms in the same way as it does regional vari- Austrian German. Tscherfl, borrowed from the ants. Nevertheless, in recent years, an overall Slovenian Í revlje, is, for example, used in the German increased national awareness has led to a heightened Carinthian dialect to mean ‘old shoes’, and venahti identification with Austrian Standard German. This is from the German Weihnachten (‘Christmas’) is found reflected, not least, in an increased promotion of the in Carinthian Slovenian. Despite various efforts by 104  AUSTRIA interested groups to encourage the use of Slovenian, approximately two or three dozen dialects of the this minority language continues to be used less and language, a number of which are currently spoken in less and the majority language, German, is used more Austria. All but exterminated at the hands of the and more. National Socialists in World War II, it is only in recent The Croatian minority group has lived in a number years that the self-confidence of these speakers has of areas in the eastern Austrian state of Burgenland increased, and organizations and interest groups have since the fifteenth and sixteenth century at which time been established in Austria and elsewhere. In both whole villages fled the approaching Ottomans and Burgenland and Vienna, these groups are responsible were taken in by feudal lords in need of laborers. for the publication of a bilingual newspaper, for exam- These Croatians were speakers of all three Croatian ple, and once a month a program is also produced and dialects. Today’s Burgenland Croatian differs from broadcasted by various groups of Romany. Since Croatian proper to such an extent that mutual intelligi- 1993, the Romany ethnic group has been recognized in bility is difficult. This development stems partly from Austria. Speakers are usually bilingual, using the language contact with German and Hungarian, which minority language predominantly at home and, on has led to interference on the level of the lexicon, occasion, at work also. Dialects of Romany have only phonology and syntax, and partly, unlike the case of been codified to a small extent—the dialect of the Carinthian Slovenians, from efforts of the Austrian Romany spoken in the Austrian state of Burgenland Croatian minority to preserve their own identity rather being one example of a codified variety. than to nurture links to the homeland—a partial by- The principal foreign languages on offer in Austrian product of the fact that whole communities fled secondary schools include English and French, and together. With the goal of an independent identity in occasionally Italian or Russian. Latin and Classical mind, this Croatian minority rejected the adoption of Greek are also taught. the modern written language of their home country in favor of their own variety of written language, based on Îakaver, the Croatian dialect spoken by the major- References ity of Burgenland Croats. There now also exists a dic- Ammon, Ulrich. 1995. Die deutsche Sprache in Deutschland, tionary of Burgenland Croatian. Nonetheless, the trend Österreich und der Schweiz. Das Problem der nationalen from monolingualism in Croatian, to bilingualism, and Varietäten. Berlin and New York: de Gruyter. Barbour, Stephen, and Patrick Stevenson, 1990. Variation in finally to monolingualism in the majority language, German. A critical approach to German sociolinguistics. German, continues. Cambridge: Cambridge University Press. Hungarian has been spoken in the state of Clyne, Michael G. (ed.) 1992. Pluricentric languages. Differing Burgenland since the tenth and eleventh century when norms in different nations. Berlin and New York: de Gruyter. Hungarian guards were sent to the frontier to protect Clyne, Michael G. 1995. The German language in a changing Europe. Cambridge and New York: Cambridge University the Hungarian border from the West. This Hungarian- Press. speaking population is concentrated in a number of Martin, Victoria. 2000. The German language in Austria. areas in Burgenland—the most important of which are National varieties of German outside Germany, ed. by those in Oberwart (Felsöör) and Oberpullendorf Hogan-Brun, Gabrielle, 103–19. Bern: Lang. (Felsöpulya). Despite promotion of the language in Muhr, Rudolf, and Richard Schrodt, (eds.) 1997. Österreichi- sches Deutsch und andere nationale Varietäten plurizen- bilingual schools via the option to learn Hungarian as trischer Sprachen in Europa. Empirische Analysen. Vienna: a noncompulsory subject and the—albeit limited— Hölder-Pichler-Tempsky. transmission of Hungarian on television and radio, Ohnheiser, Ingeborg, Manfred Kienpointner, and Helmut Kalb there is a clear trend toward monolingualism favoring (eds.) 1999. Sprachen in Europa: Sprachsituation und Austrian German. Sprachpolitik in europäischen Ländern. Innsbruck: Institut für Sprachwissenschaft. Finally, Czech is spoken predominantly by former Österreichisches Wörterbuch (ÖWB), 35th edition, 1979. intellectuals of the monarchy in Vienna who, rather Vienna: ÖBV and Jugend & Volk; 36th revised edition, than returning to Czechoslovakia upon its new foun- 1985; 37th revised edition, 1990; 38th revised edition, 1995. dation, remained in Austria. Russ, Charles V.J. (ed.) 1990. The dialects of modern German: Romany, the language of the gypsies, termed Sinti a linguistic survey. London: Routledge. Wiesinger, Peter (ed.) 1988. Das österreichische Deutsch. and Roma by the speakers themselves, originated in Vienna: Böhlau. India around AD 1000, and is now a language without ANNE BARRON a particular homeland, not currently spoken in India but rather in Europe and also further afield—in the See also Germany; Language: Contact—Overview; USA, Australia, and Canada, for example. Language Death; Language Planning; Standard Corresponding to its wide distribution, there exist Language 105 AUSTROASIATIC Austroasiatic The Austroasiatic language family contains languages (spoken mainly in Madhya Pradesh, India), Juang (spo- spoken by about 80 million people, primarily in ken mainly in Orissa, India), Korwa (spoken mainly in Southeast Asia; however, only two of the languages in Bihar and Madhya Pradesh, India), and Mundari (spo- the family, Vietnamese and Khmer, have national sta- ken primarily in Assam and Bihar, India). Mundari are tus. Nowhere do the speakers of major Austroasiatic typically hill peoples, while the related Santali is spoken languages appear to be recent migrants. This is a lan- by lowland farmers. Mundari and other Munda lan- guage family that must have once filled more of the guages have been spoken alongside Indo-Aryan map of Southeast Asia and eastern India, now reduced languages for two millennia. They show the signs of this to scattered islands by encroaching Indo-Aryan, Sino- long coexistence not only in loan words but also in the Tibetan, Tai, and Austronesian languages, including structure of typical sentences. Santali is traditionally an Bengali, Assamese, Burmese, Thai, Lao, and Cham. unwritten language; speakers who learn to read and The Austroasiatic language family is thought to have write do so in Bengali or Oriya, and on the occasions two major branches, which are typologically very when Santali is written, Bengali or Oriya scripts are divergent from each other. The division between the generally used. As with its close relative Mundari, Mon-Khmer group and the Munda languages must Santali and its speakers have little recorded history have taken place many thousands of years ago. before the nineteenth century. The Munda or Western branch of the Austroasiatic Munda languages appear to have been greatly influ- family, spoken entirely in east central India, is distant- enced by their Indo-European and Dravidian neigh- ly related to Mon-Khmer. It shares the South Asian bors. In contrast to these neighbors, Munda languages subcontinent with three other language families. Indo- have traditionally been unwritten, and their peoples European languages, particularly the Indo-Aryan have little recorded history until they came into the branch of Indo-European, dominate in the north, while sights of British administrators in India’s colonial peri- the south is the domain of the Dravidian languages. od. Typologically, Munda languages are agglutinative, The subcontinent’s northern fringe is occupied by and very long sequences of affixes may be found, Sino-Tibetan languages, and Austroasiatic languages especially in verbs. There is even noun incorporation are scattered from central India eastward into Vietnam. (e.g. in Sora), which is usually a feature of polysyn- In India itself, the Austroasiatic language with the thetic/incorporating languages. There are two genders most speakers is Santali, but the major languages of for nouns in most of the Munda tongues: animate and this family, Vietnamese and Khmer (Cambodian), are inanimate. Munda languages use affixes extensively. spoken in Southeast Asia. The major Sino-Tibetan lan- Case relations are signaled by word order, postposi- guages are spoken in East and Southeast Asia (e.g. tions, and pronominal affixes. The basic word order is Chinese, Burmese). In addition to these four families, subject–object–verb. Most Munda languages also have there is one language isolate, Burushaski, spoken in three numbers—singular, dual, and plural. Suffixes northern Pakistan, while the genetic affiliations of the and particles placed after the noun are used to express Andaman Islands remain unclear. such features as number and possession, which are The Munda branch consists of two groups or sub- often indicated in Indo-European tongues by case branches. The Nahali or Nihali sub-branch has about inflection. 5,000 speakers in Madhya Pradesh and Maharashtra, The Mon-Khmer or Eastern are typologically quite northeastern India, and the genetic affiliation of this lan- different from the Munda languages. The Mon-Khmer guage is controversial. The Munda sub-branch is found languages are mostly prefixing, monosyllabic, usually in northeastern India, in the mountains and plateaus of with a large number of vowel contrasts. The most Madhya Pradesh, Bihar, and Orissa. It is geographical- commonly found word order is subject–verb–object. ly surrounded by Indo-European and Dravidian speak- Vietnamese has borrowed extensively from Chinese, ers, whose languages have greater social prestige. It especially during the Tang dynasty, when Vietnam consists of a fairly large number of languages, including came under the strong cultural and political influence Santali (spoken mainly in India, but also in Bangladesh of China, and was even written with Chinese charac- and Nepal), Ho (spoken mainly in Bihar and Orissa, ters until Western Christian missionaries introduced India), Sora (spoken mainly in Orissa, India), Korku the Roman alphabet. Grammatically, the Mon-Khmer 106  AUSTROASIATIC languages make considerable use of affixes (prefixes, Khasi or Khasian is spoken mainly in Assam, northern infixes, and suffixes). They are agglutinative in that India, and Bangladesh, with over 500,000 speakers. different linguistic elements, each of which exists sep- Khasi is the state language of the Indian state of arately and has a fixed meaning, are often joined to Meghalaya, whose capital is Shillong, and it has many form one word. Cambodian and Mon have their own loan words from Bengali, Urdu, and English. scripts, which descended from the alphabets of India. Palaung-Wa, formed by two groups, Wa and Both are written from left to right. Palaung, consists of a large number of minor lan- Scholars have long realized that Khmer and Mon guages spoken in scattered pockets located in were related, and Mon-Khmer is the well-established Thailand, Laos, Burma, and Yunnan. Parauk is one of group of Austroasiatic languages typified by these the more prominent languages in this group. Most of two. Mon language is now in decline, as Burmese and the 650,000 speakers live in Burma and China. Wa Thai expansion supplanted Mon rule. Through Mon, probably has well over 1,000,000 speakers. The Buddhist culture was transmitted to the early speakers majority of speakers live on the mountainous border- of Burmese. By way of Mon script, Burmese and Shan land between Burma and China. The legendary center first became written languages. Mon is now a minori- of the Wa country is Lake Nawngkhio, high in the ty language of southern Burma and central Thailand, mountains on the China–Burma frontier. The older and most of its speakers are bilingual in Burmese or history of the language and its speakers is unknown, Thai. Although Mon is not an official language in but Austroasiatic speech may well have a history of Burma, it is now being used again in traditional several millennia in this region. Apart from their monastic education, which most boys receive. Mon is mountain heartland, Wa speakers are found in scat- not a tonal language but, like Khmer, has a ‘clear’ and tered communities in mountainous parts of a ‘breathy’ register, a feature that now forms part of the Xishuangbanna, the southern Shan State, and north- Burmese sound pattern as well. Modern Mon can be western Thailand. Wa languages are largely monosyl- divided into three dialect groups: Pegu, Martaban and labic; they have no tones, but tend to have rich and Moulmein, and Ye. Mon villages in central Thailand, complicated vowel systems. resulting from settlements in the last three or four hun- The Eastern Mon-Khmer sub-branch is further sub- dred years, speak dialects close to those of Martaban divided into Khmeric, Bahnaric, and Katuic spoken in and Moulmein. Vietnam, Malaysia, Thailand, Cambodia, and Laos. Khmer language survives, and is now the majority Khmeric consists of a single language, Khmer. Most language of Cambodia. Close to Khmer are some of the over 7 million speakers reside in Cambodia, minority and hill languages of Southeast Asia, includ- with others in Vietnam, Malaysia, and Thailand. The ing Sre, Mnong, Stieng, Bahnar, Hre, Sedang, Kuy, Bahnaric sub-branch includes about 35 minor lan- Bru, So, and others. Khmer script, like others of South guages spoken in central and southern Vietnam, south- and Southeast Asia, descends from the Brahmi of ern Laos, and eastern Cambodia. Bahnar, with over ancient India. 85,000 speakers in Vietnam, is one of the major lan- Mon-Khmer consists of six sub-branches: guages. And Katuic sub-branch includes several minor Nicobarese, Northern Mon-Khmer, Eastern Mon- languages scattered in Vietnam, Laos, Thailand, and Khmer, Viet-Muong or Vietic, Monic, and Aslian. Cambodia. The Monic sub-branch includes Mon and Nicobarese consists of six different languages with a Nyahkur, spoken in Burma and Thailand. And, the total of over 22,000 speakers on Nicobar Islands Aslian languages (roughly 40,000 speakers) of inac- (under Indian administration) in the Bay of Bengal. cessible districts in the Malay Peninsula and southern These include Car, Nancowry, Great Nicobarese, and Thailand include Kensiu, Sengoi, Orang Benuja, others. In these languages, taboo leads to word avoid- Temiar or Northern Sakai, and others with even small ance on a considerable scale, and thus to a rapid numbers. turnover of vocabulary. Thus, despite their dissimilar- The Viet-Muong or Vietic sub-branch includes ities among each other, and with other Mon-Khmer Vietnamese and a number of minor languages spoken languages, the relationships are historically closer than in northern Vietnam and northern Laos. Most this would suggest. Vietnamese speakers reside in Vietnam, with the rest The Northern Mon-Khmer sub-branch has many in the United States, Cambodia, and China. Muong is languages of very small communities, including spoken in inland northern Vietnam. It has five tones Mlabri, spoken by 300 hunter-gatherers. This sub- and, like Vietnamese, is a monosyllabic language. It is branch has three sub-branches itself: Khasi or Khasian, likely that the speakers of early Muong and of early Palaungic or Palaung-Wa, and Khmuic. Khmuic con- Vietnamese lived not far apart 2,000 years ago. sists of several languages spoken in northern Laos and The inclusion of Vietnamese and Muong (with northern Thailand, of which the main one is Khmu. some tiny minority languages of Vietnam and Laos) in 107 AUSTROASIATIC the Austroasiatic family continues to be seen as con- accepted by nearly all specialists that Vietnamese troversial by scholars, since Vietnamese has been resembles its Austroasiatic neighbors. under the influence of Chinese for 2,000 years. On the basis of some lexical and typological simi- Whatever its shape at the beginning of this period, larities, some linguists speculate that Austroasiatic Vietnamese is now a tonal language (with six tones languages are very distant relatives of Sino-Tibetan. that frequently help to distinguish homonyms): its Also, Schmidt (1906) first proposed a relationship sound pattern rather resembles that of Chinese. between Austroasiatic and Austronesian (the so-called Vietnamese is basically monosyllabic, but it has many Austric hypothesis) on the basis of certain phonetic, words of two or more syllables. It uses particles but lexical, and grammatical similarities, which was dis- has no prefixes and suffixes. Word order is very impor- cussed further by Benedict (1973) and Headley tant for showing grammatical relationships since there (1973). However, most linguists currently maintain is no inflection. The vocabulary has many loan words that there is very little solid, convincing evidence so from Chinese; Chinese words now make up as much far to support either of these hypotheses. The single as 60% of the vocabulary of written Vietnamese. Many Austric family grouping has not yet been generally of these loans arrived before the tenth century, as is accepted. evident when their Vietnamese pronunciation is com- pared with modern Chinese. An alphabet based on References Roman letters and adapted for Vietnamese, as by Benedict, P.K. 1973. Austro-thai and Austroasiatic, the ‘Austro adding diacriticals, is generally used today in place of Linkage’. Paper presented at the First International the traditional Chinese-type writing of the past. Conference on Austroasiatic Linguistics, Honolulu, January According to another school of thought, the Mon- 2–6, 1973. Khmer branch is actually two branches. In other Headley, R.K., Jr. 1973. Some sources of Chamic vocabulary. Paper presented at the First International Conference on words, the Austroasiatic language family is believed to Austroasiatic Linguistics, Honolulu, January 2–6, 1973. have three branches: the Mon-Khmer languages, the Henderson, E.J.A. 1991. Austroasiatic languages: essays in Munda languages, and the Annamese-Muong branch- honour of H.L. Shorto. London: School of Oriental and es. There is considerable evidence but no definite African Studies. proof as yet that these groups are derived from a sin- Nguyen, D.-H. 1987. Vietnamese. The world’s major lan- guages, ed. by B. Comrie, 777–96. New York: Oxford gle ancestor language, which is the essential require- University Press. ment for classification in the same linguistic family. Schmidt, W. 1906. Die Mon-Khmer Völker. Ein Bindeglied The Annamese-Muong subfamily is composed of zwischen Völkern Zentral-Asiens und Austronesiens. Archiv Muong and Vietnamese (also called Annamese). The für Antrhopologie 5. 59–109. classification of Vietnamese is still disputed; some Zide, N.H. (ed.) 1966. Studies in comparative Austroasiatic lin- guistics. regard it as a Mon-Khmer tongue, others as a Tai (or CAMILLA COCKERTON Thai) language, and still others as a language unrelat- ed to any other known tongue. It is now generally See also Indo-European 1: Overview Austronesian The Austronesian language family (formerly Malayo- Cambodia, Laos, and Taiwan. Although both the Indo- Polynesian) is the most widespread family in the European and Sino-Tibetan families have considerably world and the largest in terms of number of languages, more speakers, Austronesian includes some of the with approximately 1,200 (one fifth of the world’s world’s largest languages in terms of number of speak- total). The languages are spoken by more than 270 ers, as well as some of the smallest. million people in the Malay Peninsula, Madagascar, Indonesia has 13 Austronesian languages with a mil- Indonesia, the Philippines, the island groups of the lion or more speakers, including Indonesian with 17 Central and South Pacific (except for Australia and million, Javanese with 75 million, Madurese with 13 much of the interior of New Guinea), and New million, Sundanese with 27 million, Balinese and Zealand, as well as in scattered areas of Vietnam, Buginese with 4 million, and Acehnese with 3 million. 108  AUSTRONESIAN Javanese alone accounts for about one quarter of all Peninsula, as well as in Brunei on the coast of north- speakers of Austronesian languages and has more western Borneo. Malay was spoken on both sides of speakers than all of the languages in the Oceanic group. the strategic Strait of Malacca between Sumatra and The other large languages are found in the the Malay Peninsula. It became an important trade lan- Philippines, all of whose 160-some languages are guage in the larger region when the India–China trade Austronesian. Eight have more than a million speak- began around the start of the first century because the ers, such as Tagalog, which is the official language but preferred route passed through the Strait of Malacca. also one of many regional languages spoken on hun- When the Dutch East India Company arrived in dreds of islands with 17 million speakers, Cebuano Indonesia at the beginning of the seventeenth century, with 15 million, and Ilocano with 8 million, they found Malay in use as a lingua franca in major Hiligaynon with 7 million, Bikol and Waray with 3 ports throughout the archipelago, a role it has retained million, and Kapampangan and Pangasinan with to some extent today. Malay was chosen as the basis 2 million. for the national language of Malaysia (Bahasa Outside the Philippines and Indonesia, there are Malaysia), Brunei (Bahasa Kebangsaan ‘national lan- only two Austronesian languages with more than a guage’), and Indonesia (Bahasa Indonesia), where million speakers: Malay with 17 million and Malagasy Malay was felt to be a more neutral alternative to on Madagascar, an island off the southeast coast of Javanese, the most widely spoken language. Africa, with 9 million. Malagasy is the westernmost The Austronesian language family is approximate- member, and although it has some influence from ly 6,000 years old and can attribute its present wide mainland African languages, it is classified as distribution to the development of open-ocean travel Austronesian on the basis of its predominantly by the ancestors of the present-day speakers. With its Austronesian vocabulary and structure. multitude of languages on thousands of islands, it has Fourteen of the 21 or 22 Austronesian languages been of key importance in the development of the spoken by the pre-Chinese aboriginal population on comparative method in historical linguistics, as well as Taiwan are still spoken, although they have been large- critical in the study of what clues languages can give ly submerged by the influence of Chinese immigration us about where the homeland of the language family from the mainland. The largest language has only must have been and what kind of culture the speakers 100,000 speakers, and most of the others have only a of the ancestral proto-language must have had. few thousand or fewer. Some languages, such as Scholars have used the spread of Austronesian to look Siraya, are already extinct and are documented only in at issues of importance in human history, such as the historical texts. spread of agriculture, which Peter Bellwood (1991) Other Austronesian languages include the argues to be the driving force behind the spread of this Polynesian languages, such as Tongan, Tahitian, and language family and others. Samoan, all of which have relatively few speakers, Both linguistic and archeological evidence indicate particularly Hawaiian, which is severely endangered. an initial dispersal of Austronesian languages from In Vanuatu, none of the indigenous Austronesian lan- Taiwan several centuries after Neolithic settlers intro- guages has more than 3,000 speakers, and some there duced grain agriculture, pottery making, and domesti- (e.g. Lakona) as well as in New Guinea (e.g. Sirak) cated animals to the island from the adjacent mainland have as few as (or fewer than) several hundred speak- of China, approximately 4000–6000 BCE. The lin- ers. Approximately 600 Austronesian languages are guistic evidence suggests a steady southward and east- spoken in New Guinea and in scattered coastal areas ward movement, with Austronesian speakers moving and on the islands to the east of it, but there are prob- around the northern coast of New Guinea into the ably fewer than 2 million speakers. western Pacific in approximately 2000 BCE. From Most of the well-described languages are the larger New Guinea and the Bismarck Archipelago, settlers ones in the Malay and Indonesian archipelagoes and in spread out very rapidly, crossing the ocean in outrig- the Philippines. Politically, the most important lan- ger canoes. There is, however, a puzzling thousand- guage is probably Malay, native to the Malay year gap before the settlement of central and eastern Peninsula, adjacent portions of southern and central Polynesia, with Hawaii settled only within the past Sumatra, and some smaller neighboring islands. The 1,500–1,700 years and New Zealand within roughly oldest attestations of the language (in fact, the oldest the past millennium. Madagascar was settled by immi- of any Austronesian language) are found on stone grants from southeastern Borneo sometime between inscriptions in the late seventh century in southern the seventh and thirteenth centuries CE. Sumatra. After Islam was introduced at the end of the An early split separated the aboriginal Austronesian thirteenth century, Malay-speaking sultanates were languages of Taiwan (Formosan) and the rest of established in Malay-speaking regions of the Malay the family (Malayo-Polynesian). The designation 109 AUSTRONESIAN ‘Malayo-Polynesian’, formerly applied to the language Samoan, and Tahitian, which are remarkable for their family as a whole, has been restricted since the mid- wide geographic spread yet close relationship. In addi- 1960s to the non-Formosan subgroups of Austronesian. tion, some 18 so-called Polynesian Outliers are found Formosan, however, does not represent a subgroup in Micronesia and Melanesia. Yapese also seems to be defined by exclusively shared linguistic innovations. Oceanic, although its place within this group remains Rather, it is a collective term for a highly diverse group uncertain, partly because it shares little vocabulary of languages, perhaps as many as six, most of which with more typical languages because of a complex his- share broad typological similarities with languages in tory of borrowing. the Philippines and some other areas, such as The list below shows some very basic items of Madagascar. Western Malayo-Polynesian is a cover vocabulary, which reveal the general similarities term for the Austronesian languages of the Philippines, among some of the languages that have allowed lin- western Indonesia (Borneo, Sumatra, Java-Bali- guists to reconstruct the Austronesian family and to Lombok, Sulawesi), mainland Southeast Asia, analyze its internal relationships. Cognates for ‘2’ ‘5’ (also ‘hand’) ‘fish’ ‘fire’ ‘me/I’ Indonesian dua lima ikan api saya/aku Malagasy roa dimy tsondro afo izaho/ahy Tagalog dalawa lima isda apoy ako Fijian rua lima ika buka au/noqu Hawaiian lua lima i’a ahi a’u/(w)au Madagascar, and at least Chamorro and Palauan in The Austronesian languages have some of the western Micronesia. These languages do not exhibit largest and smallest sound inventories, with these any of the innovations characteristic of Central-Eastern inventories showing areal characteristics. Hyenghène Malayo-Polynesian. They include some of the largest in New Caledonia has 30 consonants, whereas and best-known Austronesian languages, including Hawaiian has only eight. Sakao in Vanuatu has 12 Malay, Javanese, Acehnese, Sundanese, Balinese, vowels, whereas several languages of the Philippines Buginese, Ilokano, Tagalog, Cebuano, and Malagasy. and western Indonesia, such as Cebuano Bisayan and The Central Malayo-Polynesian languages are Banjarese, have only three. Rare sounds include pre- found throughout much of eastern Indonesia, includ- nasalized bilabial trills (made by trilling the lips after ing the Lesser Sunda Islands from Sumbawa through [m]) of the languages of Manus Island and apico-labi- Timor and most of the Moluccas. Many of the changes al sounds (made by touching the upper lip with the tip that define this subgroup apply to most of the lan- of the tongue) of some central Vanuatu languages. guages but do not reach the geographic extremes, lead- Some languages, such as Kiput in Borneo, distinguish ing some scholars to question it. The small group of short and long vowels as well as short and long conso- languages comprising the South Halmahera-West New nants. Many Oceanic languages permit no final conso- Guinea subgroup is found in the northern Moluccan nants or consonant clusters. Tonal contrasts are also island of Halmahera and in the Doberai Peninsula found in a few Austronesian languages, e.g. in the (also called Vogelkop or Bird’s Head) of western New Chamic languages as well as in two widely separated Guinea. Most of the languages are known only from parts of New Guinea and in southern New Caledonia. short word lists, although preliminary descriptions Languages with the word order verb–subject– exist for Buli of Halmahera and Numfor-Biak and object (VSO) and verb–object–subject (VOS) are Waropen of western New Guinea. common in Taiwan, the Philippines, and northern The Oceanic subgroup is the largest and best Borneo. Other VSO languages are Malagasy and Old defined of all major subgroups in Austronesian, Javanese. Although the Polynesian languages are also including all the languages of Polynesia, all the lan- VSO, they have developed this order through a prior guages of Micronesia (except Palauan and Chamorro), subject–verb–object (SVO) stage, unlike the other lan- and all the Austronesian languages of Melanesia east guages, which have inherited it from Proto- of the Mamberamo River in Irian Jaya. Some of the Austronesian. SVO languages are widely distributed better-known languages are Motu of southeastern New in western and eastern Indonesia, Melanesia, and Guinea, Tolai of New Britain, Mota of the Banks Micronesia. Subject–object–verb (SOV) languages are Islands in northern Vanuatu, Chuukese of Chuuk in restricted to the New Guinea mainland and some near- Micronesia, Fijian, and the many Polynesian lan- by islands, where they have been in contact with char- guages, including Maori, Hawaiian, Tongan, Niuean, acteristically SOV Papuan languages. 110  AUXILIARIES Most languages of the Philippines and Taiwan, as ples, which has also brought about drastic grammati- well as the languages of northern Borneo, northern cal convergence as well as extensive borrowing of Sulawesi, and Chamorro and Malagasy, mark verbs vocabulary. There is still some dispute over the mem- with affixes indicating the grammatical role that a spe- bership of many languages. The Maisin language of cific noun bears in relation to the verb. One of the most southeastern Papua New Guinea is now generally fundamental distinctions in the verb systems of regarded as an Austronesian language with heavy con- Austronesian languages is the division into stative and tact influence from Papuan languages. Other contro- dynamic verbs. Stative verbs often translate as adjec- versial or aberrant languages include those of Buka tives in English (e.g. Hawaiian make ‘die/to be dead’), and Bougainville, Arove, Lamogai, and Kaulong of and in many Austronesian languages it is doubtful New Britain in Papua New Guinea, Ririo and some whether a category of true adjectives exists. other languages of the western Solomons, Asumboa of Most of the languages distinguish between inclusive the Santa Cruz archipelago, Aneityum and some other first-person plural pronouns (‘we’, listener included) languages of southern Vanuatu, several languages of and an exclusive form (‘we’, listener excluded). Some New Caledonia, and Nengone and Dehu of the Loyalty Oceanic languages distinguish a dual (i.e. ‘we/you/they Islands in New Caledonia. Most of the Chamic lan- two’). The distinction between alienable and inalien- guages distinguish words with different intonations able possession is also characteristic of Oceanic lan- and have also acquired other features characteristic of guages. Nouns referring to inalienably possessed the Mon-Khmer languages with which they have been entities (e.g. body parts, relatives) are typically a small, in contact in Vietnam and southern China. closed set, and the nouns referring to alienable posses- sions can be freely added to. Inalienable and alienable References possession are often manifested grammatically in dif- ferent construction types, and some languages have Bellwood, Peter. 1991. The Austronesian dispersal and the ori- several subclasses of alienable possession, each with its gin of languages. Scientific American. Blust, Robert A. 1995. The prehistory of the Austronesian- own marker. Other languages, including many in speaking peoples. Journal of World Prehistory I: 4. Micronesia, have developed a complex system of clas- Dutton, Tom, and Darrell Tryon (eds.) 1994. Language contact sifiers to mark categories of alienable possession. In and change in the Austronesian world. Berlin: Mouton de Pohnpeian, alienably possessed nouns belong to one of Gruyter. more than 20 different classes, whose members usual- Lynch, John. 1998. Pacific languages. Honolulu: University of Hawaii Press. ly have some distinctive semantic feature. Tryon, Darrell T. 1995. Comparative Austronesian dictionary. Historical connections are complicated by exten- An introduction to Austronesian studies, 4 vols. Berlin: sive long-term contacts and movements between Mouton de Gruyter. Austronesian- and non-Austronesian-speaking peo- SUZANNE ROMAINE Auxiliaries Auxiliaries are a special type of verb typically verbs. They can occur by themselves, without auxil- expressing the grammatical notions of tense (e.g. past iaries, as in they understood that problem. or present), aspect (e.g. finished or continuing), and The number of auxiliary verbs is limited and they mood (i.e. commitment to the truth of the utterance). can comprise a relatively short list (about 13 in Auxiliaries do not occur by themselves but always English). They are often referred to as function words, ‘help’ other verbs, hence their name auxiliary, which which are considered a closed class. On the contrary, can be translated as ‘helping verb’. Examples of the number of lexical verbs is vast, and new verbs are English auxiliaries are have, as in I have eaten, and invented daily, e.g. to bookmark a URL. Hence, they might, as in It might rain. Examples of verbs that are are members of the open class or lexical words. not auxiliaries are surmise, perceive, swim, under- In addition to lacking meaning, auxiliaries possess stand, and eat. These lexical verbs carry more mean- a number of unique characteristics of word formation ing than auxiliary verbs and are referred to as main and syntax. For instance, they tend to be unstressed 111 AUXILIARIES and therefore can be reduced, or contracted, as in I’m In the late 1970s and early 1980s, auxiliaries cause going there and I’ve seen him. They are also posi- a very lively debate as to whether the category of AUX, tioned differently. In English, auxiliaries occur in introduced in Chomsky’s Syntactic Structures, is uni- questions (namely before the subject, as in (1)) and in versal. Since then, some linguists have argued that all negative sentences (namely before not/n’t, as in (2)). If auxiliaries are simply verbs, and others that they are an auxiliary is not present, main verbs need the inclu- special and that this special category exists universally, sion of a form of do, as in (3) and (4), to question or hence the similarities between languages as diverse as negate sentences: Japanese, Arabic, and Lummi (a language spoken in the NW of the US). Yet others, working mainly within (1) Has he eaten yet? cognitive linguistics, see auxiliaries in terms of a con- (2) He hasn’t eaten yet. tinuum from auxiliary to main verb, making it possible (3) Did he eat yet? to think of aspectual verbs, such as continue, go, and (4) He didn’t eat yet. got in She continued writing, She went fishing, and He got going, as somewhat auxiliary-like and somewhat In other languages, auxiliaries also occupy special verb-like. The latter approaches are also interested in positions in the sentence and are marked for negation. what kinds of verbs are the sources of auxiliaries. In English, auxiliaries are used in tag questions, those Auxiliaries form part of a language’s Tense, Mood, tagged onto the end of a statement, such as, He has Aspect (TMA) system. Even though the core auxiliary eaten, hasn’t he. Again, if a sentence only contains a is said to express TMA, they may also express passive main verb, the inclusion of a form of do is needed, as voice and negation. TMA represents a huge collection in He went, didn’t he? of different notions, and not all languages express all. In English, the auxiliary verbs include (a) have, be, English auxiliaries indicate progressive aspect (as in I do, get and (b) may, might, can, could, shall, should, am reading), present relevance (I have lived here a will, would, and must. Those listed under (a) can also long time), voice (It was read), uncertain mood (It be used as main verbs, as in He has a headache, and might rain), polite mood (Could you please do that), the ones listed as (b) are called ‘modals’, since they ability and volition, also considered mood. Other lan- help to express mood and modality. A few auxiliaries guages indicate how the speaker acquires the knowl- are marginal since they have characteristics of both edge (by direct perception or via hearsay). The main verbs and auxiliaries. For instance, dare occurs similarities of these systems in languages that have both with and without do in questions, as in (5) and developed independently of each other has prompted (6), and in negations: much debate as to whether this category is biological- (5) Shall I part my hair behind? Do I dare to eat a ly innate, i.e. part of Universal Grammar. peach? (T.S. Eliot in The Love Song of J.A.P. In short, auxiliaries are different from main verbs in 122) terms of position, unstressed pronunciation, and uses (6) Dare I eat a peach? as TMA markers. This has prompted many linguists to consider them separate from main verbs, and to regard Other marginal auxiliaries (also called semimodals this as a universal property of languages. or semiauxiliaries) include ought (to), need (to), have (got) to, be able to, be going to, be about to, be to, and References used to. Modern English is quite unique in being able Chomsky, Noam. 1957. Syntactic structures. Den Haag: to ‘stack’ the auxiliaries, as in (7): Mouton. (7) She may have been being seen (while commit- Heine, Berndt. 1993. Auxiliaries. Oxford: Oxford University Press. ting that murder). Krug, Manfred. 2000. Emerging English modals. Berlin: Historically, auxiliaries have evolved from main Mouton-de Gruyter. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan verbs. The discovery of these gradual changes inspired Svartvik. 1985. A comprehensive grammar of the English many linguistic theories on the nature of linguistic language. London: Longman. change. Thus, I have to read a book is seen as having Steele, Susan, Adrian Akmajian, Richard Demers, Eloise developed from possessive I have a book (to read) to Jelinek, Chisato Kitagawa, Richard Oehrle, and Thomas purposive I have a book to read to obligation I have to Wasow. 1981. An encyclopedia of AUX: a study of cross- linguistic equivalence. Cambridge; MA: MIT Press. read a book. Originally, have was a main verb but it Warner, Anthony. 1993. English auxiliaries. Cambridge: developed into a modal. Have to is moving toward the Cambridge University Press. reduced form hafta in many dialects, but does not yet ELLY VAN GELDEREN behave like a modal in terms of its position in negative sentences and questions. See also Chomsky, Noam; English 112  B Babylonian Traditional Grammar Ancient Mesopotamia (Iraq) witnessed the first The intricacy of the cuneiform writing system, as attempts at studying the structure of language. These well as Sumerian language and culture were taught by attempts are preserved in grammatical texts dating means of lists of signs, bilingual vocabularies, and from Old Babylonian times (2000–1600 BC), more proper grammatical texts. As education in ancient than a thousand years before Pa¯nini’s grammar of Mesopotamia was orally conducted and it mainly Sanskrit. As in India and Europe, grammatical study in involved memorization, these teaching aids were also Mesopotamia emerged following major concerns with used as material for memorizing and copying practice. the preservation of a classical literature written in a The earliest grammatical texts from the Old language that was disappearing at the time. Babylonian period were studied and developed in the In the Old Babylonian period, two languages were succeeding periods up to the sixth and fifth centuries spoken in the area: Akkadian, a group of Semitic BC. The earliest grammatical training texts (from c. dialects (with Babylonian as a principal division), and 3000 BC, cf. Black 1989:76) consisted of lists of Sumerian, an isolated language among them. words grouped according to their determinatives. They Sumerian, the language of a major part of Ancient were later complemented (from c. 1900 BC) by lists Mesopotamian literature, was dying out as a living containing the signs of the cuneiform writing system. language and was being replaced by Akkadian. They vary from single-column to four-column lists. Nevertheless, Sumerian was to remain the literary lan- They include paradigms, lists of morphemes, word guage used in administrative, legal, scholarly, and reli- lists, as well as other forms of conveying grammatical gious areas. Texts written in cuneiform script appeared information, such as dialogues. The texts were written for the first time around 2900 BC. They continued to on clay tablets generally forming large collections be written until the beginning of the Christian era. known as ‘series’. Since Akkadian coexisted with Sumerian, the The analytical method characterizing Old awareness of linguistic diversity was alive among their Babylonian grammatical texts is the paradigm. The speakers. Sumerian was taught at school and the texts contain bilingual lists (Sumerian on the left and cuneiform script was the first type of writing that stu- Akkadian on the right) of selected forms arranged dents learned. Copying texts at school was a duty, according to morphologically determined patterns. As writing (additional) lines being used as a punishment. there is less variation in the nominal system, much Sumerian grammatical texts were subject matters on attention is devoted to the Sumerian verbal system. exams, together with mathematical problems. It is Sumerian is a highly agglutinative ergative language believed that Sumerian grammar was still being taught and listing all potential forms of a verb would be in southern Mesopotamia under its Persian rulers almost impossible. Thus, the structure of the paradigm about a decade before the battle of Marathon (490 BC) is often governed by Akkadian, showing the appropri- (cf. Black 1989:76). ate correspondences in forms of the Akkadian verb. 113 BABYLONIAN TRADITIONAL GRAMMAR The earliest examples of paradigms date from c. Word lists are the largest number of linguistically 2500 BC and they include short lists of contrasting interesting Sumerian texts. They date from c. 2800 BC, verbal forms inserted in word repertoires used in scrib- being found among the oldest tablets discovered in al training (cf. the following examples from Civil Mesopotamia, and are still found until the end of the 1994:77): third millennium. Originally unilingual, lexical compi- lations turned Sumero-Akkadian bilingual around the Transitive verb: in-na-sum ‘he will give to’ eighteenth to seventeenth centuries BC. ì-na-sum ‘has been given to him’ Other types of Babylonian grammatical texts nu-ì-na-sum ‘has not been given to him’ include dialogues used in scribal schools for training hˆe-na-sum ‘let it be given to’ purposes, often included in parts of the curriculum. They cover various topics related to writing, spelling, Intransitive verb: and vocabulary. Some were intended for scribal train- ba-til ‘has been finished’ ing, and others for instructing administrators and fore- nu-til ‘has not finished’ in-til ‘he finishes’ men on how to give orders and conduct businesses in hˆe-til ‘let him/her/it finish’ Sumerian. The grammatical tradition of Babylon offers the The largest and most complex paradigm is that of oldest instance of linguistic analysis and an invaluable the Sumerian verb gen (Akkadian ala¯ku) ‘to go’, source for the reconstruction of Sumerian. It is inter- which has 318 entries. esting to observe with Black (1989:96) that, even if it In addition to the verbal paradigms, there are also was a closed system, the Babylonian grammatical tra- long paradigms of pronominal forms with different dition spread to other territories as well. It reached the prepositional and conjunctive elements, etc. It is Hittite capital Hattusas and the city-state of Ugarit by believed that one of these paradigms must have origi- the thirteenth or twelfth centuries. It also reached nally had around 1,200 forms (cf. Civil 1994:77). Assyria in the eighteenth or seventh centuries. Besides paradigms, ‘grammatical vocabularies’ are the second type of Babylonian grammatical texts. References Initially, they were probably simple vocabularies of pronominal, adverbial, and prepositional phrases Black, J.A. 1989. The Babylonian grammatical tradition: the first grammars of Sumerian. Transactions of the Philological composed in Middle Babylonian times before the Society 87(1). 75–99. twelfth century BC. Larger systematic lists were Black, J.A. 1990. Recent study of Babylonian grammar. Journal developed much later (fifth to fourth centuries BC) by the Royal Asiatic Society 1. 95–104. the introduction of paradigmatic groups in which one Civil, Miguel. 1994. Sumerian. History of linguistics, Vol. I: or more elements were systematically varied. The eastern traditions of linguistics, ed. by Giulio Lepschy, 76–87. New York: Longman. Although their organization is less rigorous than that Hymes, Dell (ed.) 1974. Studies in the history of linguistics. of paradigms, these lists of morphemes are very help- Traditions and paradigms, Bloomington, London: Indiana ful for present-day linguists, thanks to their tabular University Press. form with translations. LAURA DANILIUC Balkans The term ‘Balkan’ refers to the southeasternmost the Pindus Mountains in Greece, while in the north- peninsula of Europe and, collectively, the countries west there are the Julian Alps and the Carpathians. located there. The Balkan Peninsula was named after a Although there is no sharp separation in the north mountain range, the Stara Planina in northern Bulgaria between the peninsula and Central Europe, the line of whose Turkish name is Balkan (mountain). The relief the Sava and Danube rivers is commonly considered to of the region is shaped by mountain ranges: the Balkan be its northern limit. The geographical area covers Mountains lie east–west across Bulgaria, the Rhodope about 629,000 square kilometers with a population of Mountains extend along the Greek–Bulgarian border, about 70 million people divided into many countries, the Dinaric range extends down the Adriatic coast to including Albania, Bosnia–Herzegovina, Bulgaria, 114  BALKANS Macedonia, the mainland of Greece, and the European Herzegovina, Montenegro, Macedonia, and Serbia). part of Turkey. Geographically, it also includes parts of The use of alphabets is also related to religious influ- Croatia, Slovenia, Serbia and Montenegro (formerly ences that have tended to divide the Balkan peoples cul- Yugoslavia), and SE Romania, which are often con- turally. The Slovenes and Croats are predominantly sidered Balkan countries as a whole because of their Roman Catholic, as are a number of Albanians and close ties with the region in history and politics. Romanians. Islam is adhered to by the Turkish minori- ty, by the majority of Albanians and Gypsies, and by some Bulgarians, while in Bosnia and Herzegovina, and Linguistic Landscape in Macedonia, Montenegro, and Serbia, many indige- In the Balkans, many languages of different genetic nous Slavs converted to Islam following the arrival of affiliation are being spoken belonging to different the Ottoman Turks in the fourteenth century. Elsewhere, groups of the Indo-European (IE) family along with the Eastern Orthodox tradition predominates, organized some non-IE languages. Most numerous are those of the through Bulgarian, Greek, Macedonian, Romanian, Slavic family with more than 23 million speakers, Serbian, and Albanian national churches. including Bulgarian, Macedonian, Bosnian, Croatian, Through the long-standing areal contact and exten- Serbian (the last three were formerly labeled Serbo- sive exchange of language materials, speakers of typo- Croatian), and Slovenian as national languages, in addi- logically distinct languages of the Balkans have tion to Slovak, Rusyn, Czech, Polish, Ukrainian, and adopted features from each other over the centuries, Russian varieties of the respective ethnic minorities. sharing today sets of typological properties termed The languages belonging to the Romance family spo- balkanisms that are not the result of either their com- ken by about 20 million speakers are (Daco-) Romanian mon genetic origin or chance. This specific develop- (the official language of modern Romania), Aromanian ment of the relations between the Balkan languages (also known as Vlach or Macedo-Romanian, the variety gave grounds for the term Sprachbund (from German) of Romanian spoken by minorities in Albania and to be used to denote the particular relations between Greece for at least several centuries), Megleno- them in terms of linguistic union. This unity in diver- Romanian (the variety of Romanian spoken by a minor- sity and similarity without genetic affiliation have ity in Greece), Istro-Romanian, Istro-Romance or Istriot remained outstanding features of the linguistic picture (two archaic languages spoken in Istria on the western of the peninsula, reflecting both the influence of geo- part of the Balkans), Italian and Venetian (as minority graphical factors and the historical processes of ethno- languages in Croatia and Slovenia), and Judeo-Spanish cultural interaction. (in small dispersed communities mainly in Bulgaria, Greece, and Turkey). Geography Other IE languages spoken on the Balkan are Greek, with about 10 million speakers (in its different Geographically, the area is composed of two parts: one dialects and written forms), Albanian, with about 6 southern and Mediterranean, and the other northern million speakers (including its two dialects Tosk and continental. Albanian, the official language of Albania, Gheg The southern part is open to the easily navigable Albanian spoken in Kosovo as well as Arvanitika—the seas with the Adriatic and Ionian seas on the west, the variety of Tosk Albanian spoken in Greece for some Sea of Marmora, the Dardanelle and the Aegean Sea 600 years or more), several idioms of the Indo-Aryan on the south, and the Black Sea and the Bosporus on language Romani spoken by about 3 million speakers the east. Surrounded on three sides by water, the of dispersed communities in all Balkan countries peninsula has been linked through history with deep (Balkan Romani, Vlach Romani, and Sinte Romani), and lasting bonds to Anatolia, and Asia beyond, the and two Germanic languages—German and Yiddish— Italian peninsula to the west, and to the eastern spoken by minorities in Romania. There are also non- Mediterranean and Egypt to the south. Therefore, it IE languages that are spoken by minorities on the has always been open to various influences and pres- Balkans—the Uralic, Finno-Ugric language, sures in a permanent exchange of populations to and Hungarian, and a number of Turkic languages belong- from all these destinations throughout the whole of its ing to the Altaic family. history. This communicative permeability and contin- This diversity of languages is further reflected in the uous intimate contacts of its peripheral parts with usage of two alphabets. In the western part of the neighboring non-Balkan areas are contrasted with the Balkans, the Slavic languages traditionally have been central mountainous part characterized by isolation written in the Latin script (Croatia, Slovenia, and parts and perpetuation of a most conservative way of life. of Bosnia and Herzegovina) while the Cyrillic script is Except for the northern plains, which are quite used in the eastern parts (Bulgaria, parts of Bosnia and open toward Central Europe, the continental part is 115 BALKANS characterized by the mountains that have contributed peninsula, Thracian and Ancient Macedonian to the to the continued fragmentation of human groups in the northeast, Dacian in the north, and Greek to the south- area. They have been physical obstacles, preventing east. With the exception of Greek, very little is known efforts at regional, economic, and cultural integration about them from the fragments conserved in ancient dividing the region into small units, in which distinct Greek manuscripts consisting mainly of toponymic ethnic groups have tended toward distinct national cul- terms, personal names, and some words, aside from tures and languages, local economies, and political the fact that they are all of IE origin. They all seem to autonomy. The mountains have also subdivided the be related and have a set of common traits in syntax whole area into vertical ecological zones, ranging and morphology with Greek dialects and other lan- from lowland farming areas to wooded or rocky guages of the region. The Greeks established colonies uplands acting as places of refuge for ethnic groups outside Greece during the last millennium BC, while expelled from more desirable coast and valley lands during the reign of the Macedonian Alexander the throughout history. Great in the fourth century BC, much of the Balkan Although considerably different, the two parts of the Peninsula came under Greek influence so that the Balkan peninsula are connected by all kinds of histori- Ancient Macedonian language assimilated entirely by cal, cultural, and linguistic links as several rivers cross- Greek disappeared in the third century BC. This ing the almost continuous mountain barrier provided process of Hellenization of the Balkans caused the for- for continental roads, which, although difficult and mation of several common features in all languages dangerous, were always important as lines of commu- that were spoken here later in history. nication and channels of exchange of goods and ideas. The Romans first invaded the peninsula in the third century BC, and by the first century AD it was entire- ly under their control. At the height of Roman power, History the Balkan peoples were united under a common legal Since the beginning of its history, the Balkan system, a single ultimate political and military power, Peninsula has been very complex from the linguistic and were connected through several important com- point of view, an area of great linguistic diversity in mercial routes. From the third century BC until the which a variety of different languages have always time of Slav immigration in the sixth century AD, been in close contact. It has been a crossroads of when the entire region belonged first to the Roman and migration and mixing of peoples for many centuries, later to the Byzantine Empire, radical romanization so that ethnic diversity has come to be one of its most occurred, which is reflected in a number of toponymic prominent features. Many different peoples have traces, while the Latin language of the Roman domi- migrated into the area, absorbing or being absorbed by nation represents a major linguistic substratum of the already existing groups that left traces in the languages subsequent Slavic languages. spoken there in different layers of language contact. Venetic was quickly assimilated by Latin under the The whole area shows the same stratification of lin- Roman rule in the first century; Thracian language dis- guistic elements since early prehistory, which reflects appeared in the fifth or sixth century AD, as Thracians the successive settlement of the region by different were first assimilated by Romans and later by Slavs populations until today. Archeological and historical and Bulgars. The Dacians were assimilated by data report that the Balkan Peninsula had been an Romans and are thought to be ancestors to the present inhabited area even before the immigration of the IE Romanians. Some Illyrian tribes were either roman- tribes. This pre-IE substratum (or substrata) belonged ized or assimilated by later Slavic migrations, but oth- to an old pre-IE language complex of Mediterranean ers moved south into present-day Albania, where they languages, far related to IE (probably linked with the managed to retain their identity, including the Illyrian expansion of agrarian economy along the coast of the language that is said to be the ancestor of the modern Mediterranean around 5000 BC). As they were Albanian language. Some of the romanized Thracian completely assimilated by later IE expansion, little is and Illyrian populations survived by taking refuge in known about them, and all opinions about their origin the isolated mountain areas where they lived like shep- and their relation to the IE family remain purely theo- herds conserving several Balkan Romance varieties. retical. However, there is no doubt that they left traces After the fall of the Roman Empire, this Latin was in all the ancient IE languages of this area.The gradual considerably modified in isolation, developing through Indo-Europeanization of the Balkans was a rather long absorption of different Paleobalkan substrata in what is process, which occurred during the second millennium today called the Vulgar Latin variants. The mixture of BC. It brought to the region tribal groups speaking those tongues with later additions of Slavic and other several IE languages termed collectively Paleobalkan elements created interesting structures and the vocabu- languages: Illyrian and Venetic to the northwest of the lary of several Romance varieties still spoken in the 116  BALKANS Balkans: Daco-Romanian, Aromanian, Istro-Romanian, their customs and terminology of dress, food, and and Istro-Romance or Istriot, in addition to Dalmatian, music. A large number of Turkish loan words can be an extinct language spoken along the Adriatic coast referred to as Balkan Turkisms because of their exis- until the beginning of the twentieth century. tence in all Balkan languages, Bosnian, Macedonian, Later invasions by peoples from the north culmi- Bulgarian, Serbian, Albanian and, to a lesser extent, nated in the arrival of the South Slavs, who occupied Romanian and Greek. much of the region by the sixth century AD. The breakup of the Turkish empire in the nineteenth Consequently, non-Slavic peoples of the Balkans were and twentieth centuries led to the establishment of new influenced by Slavic culture, and their languages con- states that had to find a way of including many nation- tain many words of Slavic origin. The Slav invasions alities. The recent history of the area reflects a similar were followed by the arrival of other smaller groups. fundamental contrast between convergent and diver- In the seventh century, horsemen called Bulgars set- gent tendencies in the development of Balkan cultural tled in the area of present-day Bulgaria, but they and linguistic identity. During the twentieth century, in became absorbed into the more numerous Slavs and the multinational state of Yugoslavia including all ceased to exist as a separate people. In the ninth cen- South Slavs except Bulgarians, Serbo-Croatian as the tury the Hungarians, or Magyars appeared in the official language acted as a converging factor, while Danube Plain, who, unlike most other small groups of the later struggle for political independence has con- invaders, retained their language and culture. tributed to prevailing divergent cultural and linguistic After the fall of the Western Roman Empire in 476, tendencies of various ethnic groups until the present. the region of ancient Illyria became part of its eastern portion, the Byzantine Empire, which continued its Linguistic Union political, cultural, and linguistic influence through Byzantine Greek on the eastern part of the Balkans for This common historical fate of the Balkan peoples, another 1,000 years, until it was defeated by the Turks their coexistence within the borders of the same in 1453. It also introduced Orthodox Christianity to empires (Roman, Byzantine, Ottoman), and exposure the Romanian and Slavic tribes of this part of the to similar cultural influences have contributed consid- Balkans. For this purpose, in the ninth century, Greek erably to the emergence of a shared cultural identity in priests Cyril and Methodius translated the Bible from all fields of human activity including common linguis- Greek into what is now called Old Church Slavonic, tic features, both lexical and morpho-syntactic. which reflected several Common Slavic features and Although the Balkan linguistic phenomena have was at that time widely understood by the Slavs. attracted the attention of numerous linguists since the Translating the Bible, the Byzantine priests introduced third decade of the nineteenth century, in spite of their plenty of Greek, Latin, and Hebrew words into Slavic. prolonged efforts no general agreement has been In the fifteenth century, the Ottoman Empire con- reached as to which particular languages belong to the quered the peninsula, and a long period of its rule has union, the linguistic traits that characterize the area, had a lasting effect on the arrangement and character- and the sources of these areal features. It is true, that istics of the Balkan peoples changing the geographic not all Balkan languages have an equal share in the pattern and cultural characteristics, mainly by forcing Balkan Sprachbund and the common properties are migrations throughout the peninsula. For almost 500 most numerous in those parts of the Balkans where the years, Albania, Bulgaria, Bosnia and Herzegovina, Byzantine and Turkish influence was strong and long Montenegro, Serbia, and areas of present-day term, and where the greatest number of languages are Romania were under the Turkish control. The area of co-territorial. The epicenter of Balkanisms seems to be present-day Croatia and Slovenia in the western part in the southern bordering area of Macedonia, where of the peninsula was throughout this period under the Greek, Albanian, Macedonian, Aromanian, and political, cultural, and linguistic influence of Italy Balkan Romani intersect. Therefore, due to varying (Venice) and later German-speaking Austria-Hungary. distribution of Sprachbund features, many scholars The most notable result of these movements of consider five languages including Bulgarian, small groups of people and even entire small commu- Macedonian, Greek, Albanian and Romanian, and nities from region to region and small-scale language some southeast Serbian dialects to be the core Balkan shifts in certain regions at different times was a radical languages, while others displaying only some of the change in the distribution of various Slavic dialects features are regarded as less typical and peripheral. and linguistic discontinuity in many parts of the A considerable contribution to the study of these Balkans, as well as large-scale multilingualism. phenomena was offered by Kristian Sandfeld in his During this period, all peoples of the Balkans were classic comparative study of the Balkan languages strongly influenced by Turkish culture, as reflected in Linguistique balkanique (1930), in which he recorded 117 BALKANS over one hundred Balkan Sprachbund properties, mak- therefore, has been increasingly seen as a complex of ing a distinction between general concordances and shared features in the respective systems of languages concordances between individual Balkan languages. involved, resulting from an interplay of several factors Since then, however, a number of different analyses including inheritance, innovation, and contact in a mul- pointed out different combinations of Balkan tilingual environment, in which, among the random Sprachbund properties, although most of them agree changes in each language, those were more easily along with a number of similarities in vocabulary on spread that contributed most to direct intertranslatabili- one phonological property—the presence of the schwa ty between them. phoneme, a high or mid central vowel, and six gram- The current linguistic tendencies of the Balkans at matical properties: (1) substitution of case inflections the beginning of the twenty-first century are closely by uninflected function words; (2) grammaticalization related to the issues of ethnic identity and are shaped of the category of definiteness through definite arti- by recent interethnic conflicts resulting in puristic and cles; (3) pronominal doubling of objects; (4) expres- divergent developmental tendencies in all languages sion of future tense using the verb want; (5) perfect spoken in the region. tense with an auxiliary verb corresponding to have; and (6) partial or total loss of the infinitive and its sub- stitution by subjunctive clauses. References In spite of various proposals for source languages Banfi, Emmanuel. 1985. Linguistica Balcanica. Bologna: N. of these common features, the exact nature of the con- Zanichelli. tact that led to the Balkan linguistic phenomena Fraenkel, Eran, and Christina Kramer (eds.) 1993. Language contact–language conflict, Balkan studies, Vol. 1. New York, remains a matter of some controversy. Several early Bern, Berlin, Frankfurt/M., Paris, Wien: Peter Lang scholars tried to explain their origin through the sub- Publishing. stratum hypothesis, suggesting that they developed Joseph, Brian D. 1983. The synchrony and diachrony of the under the influence of an ancient substrate—that of the Balkan infinitive: a study in areal, general, and historical lin- Paleobalkan languages Thracian, Dacian, or Illyrian. guistics. Cambridge: Cambridge University Press. Katiic´, Radoslav. 1976. Ancient languages of the Balkans Other scholars attributed the decisive influence to (Trends in linguistics). The Hague and Paris: Mouton. either Greek or vulgar Balkan Latin. All these Lindstedt, Jouko. 2000. Linguistic Balkanization: contact- hypotheses have been considered to be unrealistic as induced change by mutual reinforcement. Languages in the Balkan Sprachbund properties developed in the contact, ed. by Dicky Gilbers, John Nerbonne, and Jos post-Byzantine period and were not present either in Schaeken. Amsterdam: Rodopi. Rivero, Maria-Luisa, and Angela Ralli (eds.) 2001. Compara- classical Greek or Latin, while the Paleobalkan lin- tive syntax of the Balkan languages. Oxford: Oxford guistic material is too limited to allow any conclusions University Press. of this type. More recent theories suggest multiple Sandfeld, Kristian. 1930. Linguistique balkanique. Problemes causation including the reciprocal influence of the et resultats. Paris: Honore Champion. Balkan languages and Slavic influence through mas- Schaller, Helmut. 1977. Bibliographie zur Balkanphilologie (Bibliography for Balkan linguistics). Heidelberg: Carl sive bi- or multilingualism, sociolinguistic accommo- Winter. dation, or some combination of these factors, which ANITA SUJOLDZIC has led to the convergence of different languages. These contact-induced linguistic changes, however, See also Albanian; Greek; Language: Contact— have not led to a significant, massive restructuring either Overview; Latin; Old Church Slavonic; Romanian; in syntactic structure or in lexicon, even in the most Serbo-Croatian and South Slavic languages; affected Balkan languages. The Balkan Sprachbund, Sprachbund Baltic Languages The Baltic languages, a branch of the Indo-European Lithuanian, spoken on the eastern shores of the Baltic family of languages, are spoken in the area bordering the Sea, and Old Prussian (extinct since the seventeenth Baltic Sea. The principal Baltic languages are Latvian, century). Other extinct Baltic languages are associated 118  BALTIC LANGUAGES with historical regions of northern Europe, such as on the German orthography of that time and was quite Yotvingian, Curonian, Selonian, and Semigallian. inconsistent. Linguists consider that the Baltic languages are Latvian and Lithuanian are the only Baltic lan- more closely related to Slavic, Germanic, and Indo- guages still alive. However, Lithuanian is far more Iranian (in that order) than to the other branches of the archaic than Latvian. family. For this reason, some linguists regard Baltic A frequent false impression about the Baltic coun- and Slavic as branches of a single Balto-Slavic divi- tries (Estonia, Latvia, and Lithuania) is that they all sion of the Indo-European family. speak similar languages. In fact, there are three Baltic The early common ancestor of the Baltic languages countries and only two Baltic languages. Latvian and is traditionally referred to as Proto-Baltic and it is Lithuanian belong to the Baltic subgroup of the Indo- believed that it broke off from the other Indo- European language family, while Estonian belongs to European languages before 1000 BC. A further divi- the Finno-Ugric language family. sion into East Baltic (Latvian and Lithuanian) and West Baltic (Old Prussian) is believed to have taken Lithuanian place before 300 BC. In terms of phonology and grammar, the Baltic lan- The Lithuanian language, the oldest of the two surviv- guages are said to be the Indo-European languages ing Baltic languages, is spoken by approximately 3 closest to Proto-Indo-European. They display a high million people in Lithuania, where it has been the offi- degree of inflection in both the nominal and verbal cial language since 1918, and by an additional half systems. Beside Sanskrit, the Baltic languages are million elsewhere in the world, in Poland and sought to be the closest to the origins of the Indo- Belorussia and in such countries as the United States, European languages. The reason is that the Baltic lan- Canada, and Australia. It is estimated that more than guages, Lithuanian in particular, remained relatively 85% of the population of Lithuania are native unchanged in time, while other Indo-European lan- Lithuanians. guages suffered several alterations. From all Indo- Lithuanian occupies an important place in linguis- European languages spoken nowadays, Lithuanian has tics as it is considered to be the most ancient of the liv- best preserved the ancient sound systems, many mor- ing Indo-European languages. It is also the closest phological and lexical characteristics. Therefore, a language to Proto-Indo-European, the parent language basic knowledge of Lithuanian is indispensable in the from which all the Indo-European languages emerged. study of Indo-European languages. It has inherited many Proto-Indo-European character- Old Prussian is the most important of the extinct istics, such as a pitch-accent system, a rich inflection- Baltic languages and the most archaic of the recorded al case system, and many little-changed Proto- Baltic languages. The speakers of Old Prussian were Indo-European word stems. It is believed that Modern completely assimilated to German territories by the Lithuanian is the oldest living Indo-European lan- seventeenth century. guage; nevertheless, it has one of the most recently The earliest Old Prussian written document is a established standard literary traditions. German–Prussian vocabulary, the so-called Elbing Lithuanian has two main dialect groups, Low (or vocabulary, compiled in 1300 AD and consisting of 802 Western) Lithuanian (along the Baltic coast), with Old Prussian words written in a South Prussian dialect. three subdialects, and High (or Eastern) Lithuanian, The most important Old Prussian written records are with four subdialects. Although the country is rather three catechisms translated from German in the six- small, it is still rich in dialectal variation. The Low teenth century. Every written record in Old Prussian dialect is spoken by the Lowlanders, who live in the has its own specific orthography. west and along the Baltic Sea; High Lithuanian is spo- The Old Prussian nominal system had seven noun ken by the Highlanders, who live in the Eastern (and types and five cases: nominative, genitive, dative, accu- greater) part of Lithuania. sative, and vocative. It did not have the dual number, Even if a literary language had been in use since the but only the singular and plural. However, it displayed sixteenth century, three literary dialects fought for a neuter gender, lost by Lithuanian and Latvian. supremacy in the nineteenth century. The most promi- Old Prussian verbs had three separate forms in the nent person in the formation of the rules of the stan- plural, but not in the singular. The third person was the dard language was Jonas Jablonskis (1861–1930). same in both the singular and the plural. There were Modern Standard Lithuanian is based on the southern three tenses: present, preterite, and future. branch of the West High Lithuanian dialect and com- Old Prussian vocabulary was quite similar to pleted its development after the first period of Lithuanian and Latvian (closer to Lithuanian than Lithuanian political independence (1918–1940), when Latvian). The orthography was almost wholly based it became the country’s official language. Lithuanian 119 BALTIC LANGUAGES experienced a major Russian influence especially after Latvian World War II. Old Lithuanian is known from written documents Latvian, or Lettish, one of the two surviving Baltic dating from the sixteenth century, more precisely from languages, has been the official language of Latvia 1547 when the first book in Lithuanian was published. since 1918, when Latvian independence was won. It is Phonetically, Lithuanian went rather far from Indo- the mother tongue of some 3 million persons living on European. Baltic languages, as well as Slavic lan- the eastern shore of the Baltic Sea. guages, did not preserve aspirated and labiovelar Historically, Latvian has three groups of dialects: consonants, syllabic sounds, and some vowels including Central Latvian, the language of literature; Livonian schwa. However, Baltic languages were not influenced Latvian, greatly influenced by Livonian and Curonian; by any other non-Indo-European sound system, and and High Latvian (or Latgalian), which underwent con- basically no new phonemes appeared in Lithuanian. siderable Slavic influence. Nowadays, there are two The most complicated feature of Lithuanian pho- other major dialects: West Latvian (also called netics is stress: there is a system of three tones that is Livonian or Tahmian) and East Latvian (High Latvian). very important in pronunciation. Acute, circumflex, The earliest printed writings in Latvian are transla- and gravis intonations act for long and short vowels of tions of religious works: a Roman Catholic catechism, Lithuanian, and furthermore the stress is absolutely Catechismus Catholicorum (1585), a Lutheran version free. The Lithuanian has a free stress in contrast to (1586), and a translation of the Bible (1685). In 1638, Latvian fixed stress, which occurs on the first syllable. the first Latvian (–German) dictionary, by Georgius Lithuanian is rich in the use of diphthongs and, like Mancelius, appeared; the first grammar of the Latvian Latvian, in rising and falling intonations. language was published in 1644 by Johann Georg Lithuanian is written in the Latin alphabet (33 let- Rehehausen. The works of the Latvians Juris Alunans ters), with additional diacritical marks. (1832–1864) and Atis Kronvalds (1837–1875) had a Lithuanian is a highly inflected language. In stan- great influence on the development of a standard dard Lithuanian, nouns have seven cases (some Latvian language, which was finally established at the dialects have eight or more), three numbers (in classi- end of the nineteenth and the beginning of the twenti- cal Lithuanian, while in modern colloquial language eth century, based on the Central Latvian dialect. the dual number is seldom used), and two genders, Latvian is closely related to Lithuanian, but it is less masculine and feminine, as neuter disappeared earlier conservative, displaying many sound changes in its his- in Baltic languages and was preserved just in some torical development. Phonetically, the main difference adjective forms. There are twelve inflectional types is that Latvian has its stress always on the first syllable, and five nominal declensions, each of them represent- due to the neighboring Finnish influence. Besides, ing Indo-European nominal stems. The article is not Latvian has short vowels and monophthongs in the used. Lithuanian adjectives have three declensions. final syllables, while Lithuanian has long vowels and The Lithuanian pronominal system is a very rich diphthongs. Latvian is more archaic than Lithuanian in one, with several classes of pronouns: personal, the intonations inherited from Proto-Baltic. demonstrative, interrogative, attributive, negative, def- Based on the fact that Latvian has undergone a num- inite, and indefinite. There is also a reflexive pronoun, ber of linguistic changes that Lithuanian has not, histor- similar to the one in Romance and Slavic languages. ical linguists generally consider that Latvian is the The Lithuanian verbal system, although less com- younger of the two Baltic languages. One of the most plex than its Proto-Indo-European ancestor, has quite important changes is the retraction of main word stress a number of inflected forms. Four simple verb tenses almost always onto the first syllable of a word (an influ- (present, preterite, frequentative past, and future) and ence from Livonian). Because of the change in the several compounds exist in the indicative mood; stress pattern of the language, many inflectional endings imperative, subjunctive, reflexive, infinitive, and par- in Latvian have either shortened or been lost altogether. ticipial forms are also clearly defined. The frequenta- Grammatically, Latvian, like Lithuanian, has a rich tive past tense does not exist in Latvian. There are inflectional morphology, having seven noun cases and three verbal conjugations. The dual number is also six verb declensions. The nominal system lost the dual preserved. The third person in Lithuanian, as well as in number and the neuter gender; the instrumental case all other Baltic languages, does not vary in number. has the same form as the accusative in the singular and Lithuanian, as well as Latvian, has many compound as the dative in the plural. Latvian has six nominal tense forms, based on the forms of the verb buti ‘to be’ declensions with eight inflectional types. and participles. Lithuanian and Latvian word order is The verbal system is generally the same as in quite free, and, in general, the syntax of both lan- Lithuanian. It has three conjugations (genetically differ- guages is quite similar. ent). There are three persons, the third of which is the 120  BAMBARA, MANDENKAN AND THE MANDE LANGUAGES same (apparently from the time of Proto-Indo-European) References in both the singular and the plural (as well as the dual). Andersen, Henning. 1996. Reconstructing prehistorical The verb in Latvian has three tenses (present, dialects: initial vowels in Slavic and Baltic. Berlin, New preterite, and future). Distinct characteristics are the York: Mouton de Gruyter. descriptive and the obligatory moods. Latvian also has Barentsen, A.A., B.M. Groen, and R. Sprenger (eds.) 1991. a relatively free word order. Studies in West Slavic and Baltic linguistics. Amsterdam: Rodopi. The basic vocabulary is originally Baltic, but there Dambriunas, L., A. Klimas, and W.R. Schmalstieg. 1966. are numerous loan words, mainly from Finnish, Introduction to modern Lithuanian. Brooklyn, NY: German, and Russian. Latvian also had close contacts Franciscan Fathers Press. with Swedish and Polish. Fennell, Trevor G., and Henry Gelsen. 1980. A grammar of First written in Gothic script, the Latvian language modern Latvian. The Hague: Mouton. Hjelmslev, Louis. Études baltiques. 1932. Copenhagen: Levin has used the Roman alphabet with diacritical marks to & Munksgaard. indicate the distinctive quality of some vowels and Magner, T.F., and W.R. Schmalstieg (eds.) 1970. Baltic linguis- consonants since 1922. The Latvian alphabet has 33 tics. Pennsylvania: Pennsylvania University Press. letters, 11 with diacritical marks. The orthography has LAURA DANILIUC undergone numerous changes and improvements, and is now quite closely related to pronunciation. See also Indo-European 1: Overview Bambara, Mandenkan and the Mande languages Some 35–40 million people in 15 West African states Early classifications subdivided the Mande lan- make use of Mande languages as mother tongues or guages into a Mande-Tan and a Mande-Fu group, linguae franca (trade languages). These languages are based on differences in the word for ‘10’ in the indi- spoken from Senegal to Nigeria and from Mauritania vidual languages. The current view basically distin- to the Ivory Coast. The name Mande derives from a guishes between Western and Eastern languages. phonological variant of the name held by the former Western Mande is dominated by two-tone lan- kingdom of Mali. The official languages of large king- guages, i.e. languages that use intonational patterns doms in West Africa were often Mande, such as with two pitch levels to express lexical or grammatical Soninke in Gana (seventh to eleventh century), meaning. Languages with three pitch levels are rare Malinke in Mali (thirteenth to fifteenth century), and and found specifically in the Southwestern group. Bambara in the kingdom of Segou (seventeenth to Mani-Bandama languages (Eastern Mande) usually nineteenth century). show very reduced and short word forms. They com- Going back to Joseph Greenberg’s classification, pensate this loss of information by 3–5 different pitch the Mande languages, some 100 in total, have been levels, as seen in the Dan language of the Ivory Coast considered to represent one of the main subgroups of with 5 levels. While other languages often use the syl- Niger-Congo, the largest African language family. The lable as the tone-bearing unit, Mande languages — affiliation of the Mande languages to the Niger-Congo and specifically Western Mande — assign tonal pat- group of languages remains tentative. Unlike other terns to the word. Regardless of the number of sylla- Niger-Congo language groups, Mande shows no signs bles, words in the Bambara language will always be of noun class or concordance systems. Some authors limited to one of two tonal structures covering the have demonstrated interesting links to Songhai, a lan- entire word. guage belonging to the Nilo-Saharan group of lan- Most Mande languages show a very distinct conso- guages. They showed similarities in the areas of even nant–vowel (CV) syllable structure, and word stems of basic vocabulary (e.g. a similar word root for ‘heart’), 2–3 syllables prevail. Specifically in the Mani- of word structure and derivation, as well as in word Bandama languages, reduction processes are responsi- order. Some authors even identified links to such non- ble for many monosyllabic word forms and altered African languages as Basque or Canarian in syntax or syllable structures. In Gban, numerous words display vocabulary. ε¯ ‘Moskito’. a ClV syllable structure, such as kl~ 121 BAMBARA, MANDENKAN AND THE MANDE LANGUAGES The Western Mande languages show 5–7 oral vow- Mande. In comparison to other groups of languages, els and, at times, as many nasal vowels (7 oral and 7 words undergo very little change due to grammatical nasal vowels in Bambara). Many Mani-Bandama and influences. Typically, the position of a word in the sen- Volta-Niger (Eastern Mande) languages are particular- tence fully determines its function. Compare cε` mùso ly complex and make fine distinctions between differ- ‘the man’s wife’ with mùso cε` ‘the woman’s husband’. ent vowel sounds. Languages with many different In most cases, singular and plural nouns have the same vowels often use vowel harmony, which means that all form, provided other logical indications of plurality vowels in a given word must share certain characteris- are present. See Mauka mɔ` ɔ à shyá ‘The persons are tics. For example, the Volta-Niger language Northern numerou’s, literally’ ‘person is numerous’. Samo of Bangasoko only uses either open (long) or Nouns and verbs are indistinguishable in form in closed (short) vowels within the same word stem. The the predominant majority of the Mande languages. only difference of words like zɔ` rɒ´ ‘remove’ and zùró Position alone determines the function of such ‘throw’ is vowel aperture. These words could just as noun–verbs or neutrals. A noun–verb in a nominal simply be transcribed as zùróO and zùróC. position functions as a noun, and as a verb in a verbal Long vowels in Mande languages are historically position. Compare Bambara à bε´ kúma ‘He talks’ with most often the result of compensating for the loss of a kúma dòn ‘It’s a speech’. These functions are distinct consonant between two vowels. The Mauka language in a purely tonal respect in several languages. In in the Ivory Coast lost most of its intervocalic conso- Northern Samo, each nominalized verb receives a high nants, a process resulting in increased vowel length, tone on the last syllable. In Kelinga-Bozo, a Western e.g. sìi ‘to sit’ (Bambara sìgi), sì ‘spend the night’. Mande language, the tonal pattern is inverted in the Mani-Bandama developed many very short words, so case of a change from a nominal to a verbal function. vowel length became an important distinguishing Compare ʃ yâ , ‘laugh’ with ʃ ya ‘laughter’. (phonemic) factor between different words. In dialects Western Mande often uses affixes to express mean- of Northern Samo, a Volta-Niger language, there are ings equivalent to sentential structures in Eastern five different vowel lengths, two of which are Mande. For example, the Northern Samo phrase ‘to phonemic. put somebody into the action of ….’, à ba gònì dà Some special consonants occur only in specific táágó mà ‘he made the man march’, literally ‘he has geographic areas. Doubly articulated stops such as gb put the man in the action of marching’ is expressed as or kp, or ,  and other implosives, are only found in à ye cè lá-taama ‘He made the man march.’ in southern regions, not in the Sahel belt. Dialects of the Bambara, where the prefix la- expresses causation. same language may contain these sounds if they are Mande languages have portemanteau morphemes spread further to the south. that express two grammatical meanings simultaneous- Some Niger-Volta and Mani-Bandama languages ly. For example, the element mu in Susu (Western seem to lack phonemic nasal consonants altogether. If Mande) in a mu fa ‘He has not arrived’ details the nasal consonants occur in these languages, they are negation of action as well as the perfect tense. always followed by nasal vowels, i.e. the nasal charac- In Guro, Northern Samo, and some other lan- teristic is clearly a quality of the vowel. For example, guages, final negated sentence markers do appear in Northern Samo words (Bangasoko dialect), such as addition to the (already negating) portemanteau mor- bε`rε´ ‘crust, scab’ or m~ε`n~ε´ ‘snake’, differ merely in phemes. See Northern Samo à n` dàa kɔ´. ‘He has not vowel nasalization, which means that ‘snake’ should be come.’ Both kɔ´ and n` (portemanteau morpheme) indi- analyzed as bε`rε´ and b~ε`n~ε´. In several Mani-Bandama cate sentence negation. The final negated sentence and Volta-Niger languages — e.g. Northern Samo or markers are therefore highly redundant and easily Dan in the Ivory Coast — nasalization appears to be a prone to loss. Borrowings of such sentence markers characteristic of words as a whole: words are either can frequently be seen, particularly from neighboring completely oral or completely nasalized. Gur languages. In Southwestern Mande languages, the initial con- Mande languages place the possessor in front of the sonant of certain word classes changes in specific possessed. In nearly all these languages, two forms of grammatical constructions. In Mende, the initial con- possession are formally distinguished. The possessor sonant of nouns changes in possessive constructions: may dominate the relation (‘active relation’) or be ndε ndε í mia: ‘That is the canoe’, but Mε ndε unable to determine the nature of the relation with the lε´ndε í mia ‘That is the Mende canoe’. possessum (‘passive relation’). Compare Bambara n Unlike other Niger-Congo language groups, there ká só ‘my house which I bought and which I may sell are absolutely no signs of noun class systems in at any moment’ with n ká bolo ‘my hand (given to me Mande. With the exception of a feminine pronoun in at the moment of birth)’. In some languages, e.g. Jo (Northwestern Mande), gender is nonexistent in Northern Samo, there is a three-fold distinction of 122  BAMBARA, MANDENKAN AND THE MANDE LANGUAGES associations: mutual social relations (e.g. kinship), of Mandenkan to the dominant language in that part of passive nonsocial relations and active relations. Africa can be found in (a) the dynamic Mande traders All Mande languages have subject–object–verb who, over hundreds of years, spread their language word order. In most Mande languages, a sentence is along with their products in multiethnic areas; (b) the interpreted as passive if a normally present direct enormous prestige of Mandenkan as a successor lan- object is missing. Compare Western Mande Jeli guage of the kingdom of Mali, one of the most impor- Lamina-a sibeo kunε ‘Lamini eats the meat’ with tant African civilizations; (c) the use of Bambara by the sibeo-a kunε Lamini munu ‘The meat is eaten by French as a means of communication for their West Lamini’. African troops during World War II; (d) the important The Mande languages often use nouns such as role of Mandekan variants in traditional and modern words for body parts as postpositions, i.e. words with entertainment (griots, music, cinema, etc.); and (e) the grammatical functions. For example, Western Mande accelerated intra-African migration over the past 40 Jeli kɔ ‘belly’ can double as the postposition ‘inside, years that has increased the necessity of intraethnic in’. Thus, Jeli bɔgɔ kɔ may both be analyzed as ‘in means of communication. More than 2 million people the bush’ or ‘the belly of the bush’. from Burkina Faso have migrated to the Ivory Coast — Mande languages often use nonverbal structures while only 11% speak Mandenkan as mother tongue in where other languages may prefer a verbal one: Mani- the Ivory Coast, 61% use it as lingua franca. Bandama Guro e à da-na ‘He comes’ is literally ‘He is All Mandenkan variants seem to have two different at the place of coming.’ levels of tone, allowing for up to four meaningful tonal Several Mande peoples in Liberia and Sierra Leone patterns, e.g. Mauka in the Ivory Coast sá: ‘fontanelle’, developed writing systems where each letter stands for sá:` ‘kind of groundnut’, sà: ‘sheep’ and sà:´ ‘salary, a syllable. Only the writing system of the Vai in pay’. Odiennekan shows inverted tones compared to Liberia attracted broad public acceptance. The N'ko the other variants of Mandenkan: á mà dèn yè ‘He has script developed in 1949 by the Guinean Souleymane not seen the child’. (Odiennekan), à má dén yé Kanté has become quite popular in Upper Guinea. (Bambara). The tones of all known Mandekan variants are based on words rather than syllables. The sound inventory of language variants depends Bambara and Mandenkan (Manding) mainly on the language region. Malinke of Kankan in The so-called Mandenkan (literally: the language of Guinea shows a doubly articulated stop gb, unlike the Mali) or Manding is a denomination for some 40 closely related Malinke of Kita further to the North. closely related language variants within Central- ‘Heavy’ is thus gbíli~ in the Malinke of Kankan, but gíli ~ Southwestern Mande. Formerly regarded as different in Kita Malinke. Doubly articulated stops and implo- languages, these variants are now generally considered sives (, ) only appear in the southern regions of to be part of a language continuum. There are no clear Mandenkan (e.g. in Mauka). Western Mandenkan has language boundaries, although communication five different oral vowels, while Eastern Mandenkan between the extremes is rather difficult. Paradoxically, usually has seven. Western Mandenkan quite often such ‘dialects’ often show more differences than other shows older elements, such as intervocalic t, which has clearly distinct ‘languages’. A Bambara speaker in nearly disappeared in Eastern Mandenkan. Compare Bamako has a much better understanding of Ivory ‘lion’ in Mandinka jàta and in Bambara jàra. Coast Jula than of the Bambara of Beledougou 30 km Mandenkan variants are very much alike in word away from Bamako. Compare ‘first name’ in Bambara structure and syntax. of Bamako tɔ´ gɔ, in Dioula of Ivory Coast tɔ´ gɔ, in Bambara, which is mainly spoken in Mali and adja- Mauka (variant of Malinke) of the Ivory Coast tɔ´ ɔ and cent regions, is the most popular Mande language. in Bambara of the Beledougou cwáa. Used at different levels in Mali by approx. 90% of the Mandekan is subdivided into a western and an east- population, it offers excellent communication opportu- ern branch. Eastern Mandenkan comprises all impor- nities in the adjacent Ivory Coast and Burkina Faso, tant lingua francae within the Mande language group: both marked by the lingua franca Jula. Jula is actually Bambara, Malinke, and Jula. Speakers of one of these a Mandekan word for ‘dealer’. Jula therefore denotes variants can communicate with 30–35 millions mother not the member of an ethnic group, but rather people tongue and lingua-franca speakers in West Africa. The who speak Mandenkan traders’ languages as mother majority of the population in Mali, Guinea, Ivory tongues. As the language of (e.g. cola nut) traders was Coast, and Gambia speaks Mandenkan as first or sec- very often Bambara or Malinke, Bambara and Jula are ond language; important parts of the population of closely related. The Bambara language is used in all Senegal, Burkina Faso, Guinea-Bissau use Mandenkan parts of society: in tradition and modernity, in movies as means of communication. The reasons for the ascent like those of one of the leading African movie makers, 123 BAMBARA, MANDENKAN AND THE MANDE LANGUAGES Souleymane Cissé. Salif Keita and other worldwide Ebermann, Erwin. 1986. Kleines Wörterbuch der Bambara- acclaimed musicians prefer to sing in Bambara. The Sprache. Wien: Afro-Pub. Greenberg, Joseph. 1963. The languages of Africa. International dominant use of Bambara on television, including all Journal of Linguistics, supplement 29 (1). Part II. aspects of publicity, and in the radio of Mali also deci- Kastenholz, Raimund. 1996. Sprachgeschichte im West-Mande. sively contributes to its dissemination. Methoden und Rekonstruktionen. Mande languages and lin- Today, Bambara (Mandenkan) is one of the most guistics. Köln: Köppe. dynamic African languages with rapid expansion ten- Mukarovsky, Hans. 1966 Zur Stellung der Mandesprachen. Anthropos (Vienna) 61. 679–88. dencies to cover the next few decades. Nicolaï, Robert. 1989. Songhai et mandé. Mandenkan (Paris) 18. 69–80. References Pozdnyakow, K.I. 1980 Yazyki mande. Africana —Afrikansky Etnograifchesky Sbornik 12, ed. by D.A. Olderogge, Creissels, Denis. 1983. Éléments de Grammaire de la Langue 173–80. Leningrad: Nauka. Mandinka. Grenoble: Université des langues et lettres. Welmers, William E. 1958 The Mande languages. Linguistic Delafosse, Maurice. 1929. La langue mandingue et ses dialectes language studies 9th round table meeting, ed. by W. Austin, (malinké, bambara, dioula), Tome 1. Paris: P. Geuthner. 9–24. Georgetown University Monograph Series, 11. Dwyer, David. 1989. Mande.. The Niger-Congo languages. A Washington, DC. classification and description of Africa’s largest language ERWIN EBERMANN family, ed. by J. Bendor-Samuel, 46–65. New York and London: Lanham. See also Greenberg, Joseph Bar-Hillel, Yehoshua Yehoshua Bar-Hillel (1915–1975) was an Israeli logi- (including artificial languages)—in contrast to a lin- cian and philosopher of language who made significant guistics-based ‘general semantics’. contributions in a number of linguistic fields: formal Much of his work in the early 1950s was influenced and algebraic linguistics, logical aspects of natural lan- by Norbert Wiener’s cybernetics and by the possibili- guage, and computational linguistics, in particular, ties of the newly invented digital computers to test machine translation (MT) and information retrieval. speculations and theories about logic and language. In His principal essays are included in two collections May 1951, he was appointed to a position in the MIT Language and Information and Aspects of Language. Research Laboratory of Electronics, with the task of In most of his writings, Bar-Hillel’s aim was to investigating the application of computers to linguistic bridge the ‘disastrous’ gap between logic and linguis- work, and specifically MT and information retrieval. tics, believing that linguists (particularly semanticists) He visited all the US groups, wrote the first survey had ignored logic to their detriment, and that logicians (1951), and convened the first conference (June 1952). had ignored linguistics by creating a formal system The conference was a major factor in launching sub- devoid of any relevance to natural language in actual stantial MT research in the following years. There, use. He sought to extend the logical syntax of his men- Bar-Hillel often expressed ‘naïve optimism’, which he tor Rudolf Carnap, which he believed could serve as ‘a later regretted; however, he was never a believer in the methodological and terminological basis for structural full automation of translation. For practical reasons, he linguistics’, by describing a formalism for relating lex- advocated the collaboration of man and machine, the ical items to the specific entities (objects, persons, etc.) use of editors to revise MT output, the use of restrict- they refer to in particular contexts, both in natural and ed (unambiguous) forms of natural language, and the in artificial languages; in this, he anticipated the logical use of subject-specific dictionaries. and natural language semantics of Richard Montague. His major contribution to algebraic linguistics was With Carnap he explored the possibilities of a theory of categorial grammar, a ‘decision procedure’ for identi- ‘semantic information’, an extension of inductive and fying constituents in grammatically well-formed probabilistic logic to the semantics of communication. sentences, based on the logic of Kazimierz He also saw logic as a kind of ‘universal semantics’, Ajdukiewicz—a further example of bringing logic and providing the framework for expressing relationships linguistics closer. His 1953 essay is recognized as a among linguistic entities whatever the language pioneer article in the field, and he made a number of 124  BAR-HILLEL, YEHOSHUA important contributions to the theory of formal gram- data (linguistic, encyclopedic, inferential, referential) mar, much under the influence of Noam Chomsky—a could ever be sufficient to resolve all ambiguities in close friend from the 1950s until his death. His theory texts and to choose the best translations in context. of categorial grammar was first presented at the MT This article was perhaps the single most influential conference in 1952 as a way of dealing with the syn- publication in the early history of MT, convincing tax of natural language—a topic neglected by most many outside the field that MT was a misguided activ- linguists at the time. Later, Bar-Hillel demonstrated ity, and it continues today to be cited as ‘evidence’ of that categorial grammar was formally equivalent to the impossibility of MT. context-free grammars, and as such (following Later (1962), he lost his belief that even the Chomsky) inadequate for the description of natural man–machine partnership in MT could be cost effec- language; he had long recognized its weakness in deal- tive, and the last vestiges of his earlier enthusiasm for ing with discontinuous elements. Consequently cybernetics waned with the realization that the idea of (1962), he argued forcefully that MT researchers and a ‘learning machine’ was a delusion, since ‘all computational linguists should base parsers and gram- attempts at formalising… inference have completely mars on Chomsky’s type of ‘transformational gram- failed’. On the other hand, later still (1971), he with- mar’ (i.e. with rules for transforming active sentences drew his harsh judgment about the future of practical into passive, or for showing that look and up in, e.g. He MT, conceding that quality in MT could not be an looked the word up in the dictionary, form a single unit absolute and that, in practice, it varied according to ‘look up’). However, after his death, renewed interest recipients and uses. in nontransformational grammars has shown that cate- Bar-Hillel accepted that he often ‘trod on … toes’, gorial grammars can be extended to overcome the and he did not, in fact, like the role of devil’s advocate, deficiencies (cf. Wood 1993), and Bar-Hillel’s pioneer which he often found himself taking. In truth, his crit- work remains influential. icisms were often expressed forcefully (even bluntly) Although a firm supporter of Chomsky’s formal but they were always pertinent, well-argued, and grammar theory, he was not uncritical. He disliked the enlightening. confusing and overworked use of ‘theory’ for different conceptions and philosophies in linguistics, and he Biography argued for the development of grammars of linguistic performance (language in use) — i.e. parallel to his Yehoshua Bar-Hillel was born in Vienna on September extension of logic to deal with actual communication. 8, 1915. He moved to Palestine in 1933, and attended And he was a trenchant critic of the semantic theories Hebrew University from 1935. He received his M.A. of Jerrold Katz and Jerry Fodor, dismissing as simply in ‘The antinomies of logic’ in 1938, and a Ph.D. on false their thesis that meaning rules could be exhaus- ‘Theory of syntactic categories’ in 1939–1945 (inter- tively presented in the form of dictionary entries plus rupted by four years in Jewish Brigade Group, British rules of combination, since they had ignored the need Army). He worked as a teacher in high school in for inference rules and for mechanisms to relate words 1945–1947, and fought in the Israeli War of and their external referents. Independence in 1947–1949. He was a research fellow By the late 1950s, Bar-Hillel’s confidence in com- at Hebrew University in 1949–1953, and received a putational linguistics was waning. He wrote a report visiting fellowship to USA in 1950–1951. Bar-Hillel for the sponsors of MT research (published 1960), was appointed research associate in the Research which was highly critical of nearly all current projects. Laboratory of Electronics (RLE), Massachusetts He was particularly critical of those groups adopting Institute of Technology (MIT) in 1951–1952, and con- statistics-based analyses to ‘discover’ grammars for vened the first MT conference (at MIT) in June 1952. computer programs, of those investigating interlinguas He was a senior lecturer in Philosophy at Hebrew (i.e. intermediary language-neutral representations), University in 1953–1958, and held further appoint- indeed of any that did not keep to modest and realistic ment at RLE, MIT, 1955–1956. He returned to objectives. He was opposed to interlinguas not just in Hebrew University in 1956 and conducted a study tour practice (they did not lead to economies of program- of the United States (funded by the US Office of Naval ming) but even more in theory: only a logical seman- Research) in 1957–1960. He was Associate Professor tics could form a sound basis in his view. of Philosophy at Hebrew University in 1958–1961, His main criticism, however, was directed to the and visiting professor at the University of California at assumption that the goal of MT should be fully auto- Berkeley in 1960–1961, and then again at RLE, MIT matic high-quality translation, and he included a in 1961. He was Professor of Logic and Philosophy of ‘proof’ that such a goal was nonfeasible, not just in Science, Hebrew University, in 1961–1975; a member practice but in principle — arguing that no amount of of the Israel Academy of Sciences and Humanities in 125 BAR-HILLEL, YEHOSHUA 1963–1975; secretary and organizer of the third guistics. Jerusalem: Magnes Press, and Amsterdam: North- International Congress for Logic, Methodology and Holland. ———. 1971. Some reflections on the present outlook for Philosophy of Science, Jerusalem, in 1964; visiting high-quality machine translation. Feasibility study on fully professor, University of Michigan, in 1965; visiting automatic high quality translation, ed. by Winfred P. professor, University of Southern California, La Jolla, Lehmann and Rolf Stachowitz, 4pp. Griffiss Air Force Base, in 1966–1967; president, Division of Logic, NY: Rome Air Development Center, Air Force Systems Methodology and Philosophy of Science of the Command (RADC-TR-71-295). Bar-Hillel, Yehoshua, J. Malino, and A. Margalit. 1974. On International Union of History and Philosophy of logic and theoretical linguistics. Current trends in linguis- Science, in 1966–1968; president, International Union tics, ed. by T.A.Sebeok, Vol.12: Linguistics and adjacent arts of History and Philosophy of Science, in 1967; visit- and sciences, part.1: 37–101. The Hague: Mouton. ing professor, University of Konstanz, in 1971; and Carnap, Rudolf. 1934. Logische Syntax der Sprache. Wien: visiting professor, University of Berlin, in 1972. He Springer. (The logische syntax of language. London: Routledge, 1937.) died in Jerusalem on September 25, 1975. Hutchins, John. 2000. Yehoshua Bar-Hillel: a philosopher’s contribution to machine translation. Early years in machine translation: memoirs and biographies of pioneers, ed. by W. References John Hutchins, 299–312. Amsterdam: John Benjamins. Bar-Hillel, Yehoshua. 1950. On syntactic categories. Journal of Kasher, Asa (ed.) 1976. Language in focus: foundations, meth- Symbolic Logic 15. 1–16. ods and systems: essays in memory of Yehoshua Bar-Hillel. ———. 1951. The present state of research on mechanical Dordrecht: Reidel (Boston Studies in the Philosophy of translation. American Documentation 2 (4). 229–37. Science, 43). [Includes substantial bibliography of Bar- ———. 1953. A quasi-arithmetical notation for syntactic Hillel’s publications.] description. Language 29. 47–58. Koppel, Moshe, and Eli Shamir (eds.) 1996. Proceedings [of] ———. 1960. The present status of automatic translation of the Fourth Bar-Ilan Symposium on Foundations of Artificial languages. Advances in Computers 1. 91–163. Intelligence, June 20–22, 1995, Ramat Gan and Jerusalem, ———. 1966. Four lectures on algebraic linguistics and Israel. Menlo Park: American Association for Artificial machine translation. Automatic translation of languages, Intelligence. [Conference dedicated to Bar-Hillel and his papers presented at NATO Summer School held in Venice, ideas.] July 1962, 1–26. Oxford: Pergamon Press. Wood, Mary McGee. 1993. Categorial grammars. London: ———. 1964. Language and information: selected essays on Routledge. their theory and application. Reading, MA: Addison- W. JOHN HUTCHINS Wesley. ———. 1970. Aspects of language: essays in philosophy of See also Artificial Intelligence; Chomsky, Noam; language, linguistic philosophy, and methodology of lin- Computational Linguistics; Machine Translation Basque Basque is the oldest continuously spoken language in primary language. North central Spain is the home of western Europe. It was already an ancient language, the overwhelming majority of these speakers. Along a possibly a remnant from the Stone Age, when the bordering area in southwest France is a smaller group Romans first identified it over two millennia ago. of about a hundred thousand speakers. Crucial to its survival has been the historically remote The term Basque describes not only a language but habitat of its speakers in the rocky fastness of the west- also a distinct people and cultural region. This region ern Pyrenees Mountains between Spain and France. is called ‘Basque Country’ (Euskal Herria); however, The Basque language is related to no language family it is not a sovereign country. The effort of some to now extant. Its ancient, remote, and singular nature establish such an independent state has dominated has intrigued generations of scholars, producing as both recent and past Basque history. The struggle has much speculation as fact. been as futile as it has been violent. In Basque itself, the word for the language is The Basque Country has traditionally lain along the euskara. Speakers of euskara, according to the most French and Spanish sides of the farthest western edge recent census data, number over a million people. of the Pyrenees, descending to the Bay of Biscay. The However, only about two thirds use it as their daily size of the area occupied by the Basques has grown 126  BASQUE smaller over the millennia. Basque Country today Surviving from before the invasion of Europe by extends east from around the area of the Spanish city the Indo-European family of languages that dominate of Bilbao, past San Sebastian, and then veers north at the continent today, Basque falls on the ear as a some- the terminus of the Pyrenees. In France, it extends up what hard, guttural language. In looking at a Basque past Biarritz, ending around Bayonne. In length, there- text, one is struck by how often the letter ‘k’ appears. fore, Basque Country is no more than a hundred miles, Its morphology, in the marking of nouns and verbs with its extent inland being only several dozen miles. for person, number, or time, occurs predominantly by Its total area is a narrow, curvilinear band of less adding a suffix, rarely a prefix. Gender is not normal- than 4,000 square miles, smaller than the state of ly distinguished. Verbs are primarily transitive, Connecticut. In antiquity Basque speakers extended although there are also intransitive verbs and some that much farther north into France and farther east into the are both. The language includes adjectives and pro- Pyrenees. The Romans identified them as ‘vascones’. nouns. Moreover, in terms of number there is a dis- The Spanish termed them ‘vascos’, and the French, tinction between singular and plural, and there is hearing the ‘v’ as a ‘b’, as ‘basques’. With the influ- agreement between subject and verb. ence of Norman French in England, they were termed Basque syntax uses a standard sentence pattern of Basques in English. subject in the first position, an object in succession, Spain has the largest number of Basque speakers. and a verb following: S–O–V. The sentence ‘Ann hit The urban industrial complex of Bilbao is the largest John’ becomes ‘Ana Jonek jo zuen.’ Clauses also end Basque city, and is the fourth largest city in Spain. in verbs. There are, however, no relative pronouns. Spanish Basque Country comprises three north cen- The Basque vocabulary preserves the most ancient tral provinces, Bizkaiko, Gipuzkoako, and Araba. pre-Indo-European origin of the language. There are While not independent, this Spanish Basque Country an exceptional number of words for or related to stone. holds the status of an Autonomous Region, under the Nevertheless, it has successively accumulated words constitution of Spain. It is allowed limited local gov- over the millenia from, for example, Celtic, Latin, and erning authority and elects its own ‘president’, still Spanish. The word for ‘hello’ in Basque is ‘augur’, recognizing nonetheless the central authority of the derived from the Latin, ‘ave’. The initial nucleus of the Spanish monarchy. Euskadi is the name for the language was probably in southwestern France, in the Basque political state. There are also some Basque province of Acquataine. With the Roman invasions of speakers in the mountainous upper parts of the eastern Hispania and Gaul, Basques retreated south, moving province of Navarra, and in the wine-growing area higher and more deeply into the safety of Pyrenees, the southward of Riojo. In France, Basque speakers are strategy whereby the language survived. only in the far southwest province of Pyrénées- There are two major centers for the study and Atlantiques and have no autonomous status. research on Basque culture, history, and language. The Essentially, therefore, Basque Country consists of first is the University of the Basque Country (in four provinces in Spain and one in France. The total Spanish: Universidad del Pais Vasco; in Basque: population is three million people, but much less than EuskalHerriko Unibertsitatea), with branches in the a third of these use Basque for daily communication. three Spanish Basque provinces. The other is the Center Euskal Herria describes not only the cultural but also for Basque Studies at the University of Nevada in Reno. the geographic extent of the Basques. Basques emigrated to the New World in the last References century. Leaving their mountain pastures as noted Hualde, José Ignacio, Joseba A. Lakarra, and R. L. Trask (eds.) shepherds, a considerable number settled in Nevada. 1995. Towards [sic]a history of the Basque language. This state holds one of the major communities of Amsterdam: John Benjamins Publishing. Basques outside Spain. King, Alan. 1994. The Basque language. Reno: University of In traveling through Basque Country, one is more Nevada. likely to hear the language in the rural interior rather Kind, Alan. and Begotxu Olaisola Elordi. 1996. Colloquial Basque. London: Routledge. (with audio cassettes). than the urban, coastal area. The latter has given Michelena, Luis [Koldo Mitxelena]. See any of the many Basques a modern prosperity through the steel indus- dozens of works by him, the most noted authority of all time try of Bilbao and tourist beach resorts. However, it is on Basque. in the farms and pasturelands of the interior that Saltarelli, Mario. 1989. Basque. Croom helm descriptive gram- Basque has maintained its long-term economic and mar series. London: Routledge. Tovar, Antonio. 1957. The Basque language, translated by social survival. Centuries-old stone residences stand as Herbert Pierrepont Houghton. Philadelphia: University of proud homesteads (baserri and borda) that have been Pennsylvania Press. inhabited by the same families for many generations. Trask, R.L. 1997. The history of Basque. London: Routledge. Here lie the vigor and longevity of Basque. EDWARD A. RIEDINGER 127 BAUDOUIN DE COURTENAY, JAN IGNACY NIECISAAW Baudouin de Courtenay, Jan Ignacy NiecisBaw Jan Ignacy Niecis A aw Baudouin de Courtenay was a aspects of audition (as opposed to ‘anthropophonic pioneer of modern linguistics. He anticipated most of phonetics’, studying the production and the audition of the basic concepts of structuralist theories. In particu- sounds). In 1895, Baudouin de Courtenay published lar, his formative influence on modern phonology is his chief work, Versuch einer Theorie phonetischer generally acknowledged. However, his contribution to Alternationen (1895; Essay on a Theory of Phonetic linguistic science is sometimes undervalued since he Alternation). Despite its occasional ambiguity, the was not able to fully exploit his own discoveries and views expressed in this book have become part of never published a major synthetic work. phonology as we know it today. At the beginning of Baudouin de Courtenay is usually credited with the twentieth century, the idea of the phoneme started having introduced the linguistic term ‘phoneme’ to to infiltrate into international linguistics. It was later denote a speech sound that distinguishes meaning. He advanced by such linguists as Ferdinand de Saussure, initiated the discussion of what makes sounds distinc- Nikolai SergeeviJ Trubetzkoy, Leonard Bloomfield, tive in 1870, in his inaugural lecture at the University and Edward Sapir. Taking the lead from Baudouin de of St. Petersburg. In his hands, phonetics became more Courtenay, they cleared his theory of its psychological abstract. He pointed out that the sounds of language overtones. The distinction between the phoneme and are structural entities, rather than mere phonatory and the sound proposed by the Polish linguist may be said auditory phenomena. Baudouin de Courtenay attempt- to have laid the foundations of structural linguistics. ed to relate the study of phonemes to the linguistic It is worth noting that Baudouin de Courtenay’s intuition of speakers. He noted that the way we per- interest in what we might call psycholinguistics result- ceive different sounds is not always a matter of their ed in many analyses of his own children’s speech (over physical and physiological properties. The motor- 12,000 pages of manuscripts—most of them still acoustic properties are important elements of linguis- unpublished). tic structure as long as they contribute to the function Baudouin de Courtenay advocated a bold and indi- of sounds. vidual theory of the nature of language mixing and Around the turn of the century, international lin- change. It was put forward in his article ‘O smešan- guistic thought was strongly influenced by the nom xaraktere vsex jazykov’ (1901; On the Mixed neogrammarian school, centered on Leipzig. Many of Character of All Languages) and might be said to be a Baudouin de Courtenay’s ideas were inspired by oppo- reaction to Schleicher’s tree image of Indo-European sition to the nineteenth-century philology, concerned languages. Baudouin de Courtenay rejected the idea of exclusively with the history of languages. In particu- a pure language and attempted to analyze the mecha- lar, he stressed the importance of synchronic analysis nism of linguistic convergence. He claimed that lan- and suggested a distinction between language and guages mix both in space and in time (all languages speech (similar to Ferdinand de Saussure’s mix with their earlier stages). When languages of dif- langue/parole dichotomy). Still, Baudouin de ferent types mix, they can influence each other. Under Courtenay’s views of psychology remained within the such influence, a language may undergo many gram- dominant framework. He considered association psy- matical changes—it may lose inflection, level para- chology a fundamental achievement of modern sci- digms, regularize stress, etc. Therefore, in Baudouin ence. Therefore, Baudouin de Courtenay tried to build de Courtenay’s opinion, it is language contact that a bridge between his functional definition of the causes language change. phoneme and psychologism. It resulted in a concep- Although he occupied himself mainly with general tion of the phoneme as the psychological equivalent of problems—questions of theoretical phonetics, language a speech sound. This mentalistic approach to phonolo- mixture, relations between linguistics and psychology, gy might be considered parallel to the basic ideas put etc.—Baudouin de Courtenay was also a specialist in forward by generativism many years later. comparative and historical linguistics. He pioneered the Baudouin de Courtenay proposed a new linguistic issues of Old Polish in his Ph.D. dissertation. Since discipline, ‘psychophonetic phonetics’ (corresponding then, the history of the Polish language was among his to modern phonology). It would have to occupy itself permanent interests. He was also an outstanding with the relation between sound and the psychological Slavist. One of the three major palatalizations of velar 128  BAUDOUIN DE COURTENAY, JAN IGNACY NIECISAAW consonants that took place in Proto-Slavic is named 1870, and then worked in St. Petersburg, Russia, on after Baudouin de Courtenay since it was first described Old Polish and Indo-European, in 1870–1874. He was by him. Professor of Comparative Grammar of the Indo- Baudouin de Courtenay was a very charismatic European languages, University of Kazan, Russia, in teacher. He exerted a huge influence on several gener- 1875-1883 (there tutored MikoBaj Kruszewski). He ations of students. His lectures were usually very moved to Dorpat (now Tartu, Estonia), where he was informal and thought-provoking. They represented the Professor of Comparative Grammar of the Slavic lan- most advanced linguistics to be found in eastern guages, in 1883–1893; he was also Professor of Europe at that time. Baudouin de Courtenay’s long Comparative Linguistics and Sanskrit, Jagiellonian teaching career included professorships at several uni- University, Cracow (then the capital of the Austrian versities of eastern Europe. It began in 1874 when he partition of Poland), in 1893–1899. He lost his teach- was appointed to the faculty of the University of ing appointment because of his radical articles con- Kazan and lasted until his death. Many of de cerning social policy in the Austrian partition of Courtenay’s students became prominent linguists. A Poland. He moved again to St. Petersburg, where he number of them were active in OPOJAZ (the most worked as Professor, University of St. Petersburg, in important Russian circle devoted to literary theory and 1900–1918. He was imprisoned for three months for modern linguistics), e.g. Evgenij DmitreviJ Polivanov publishing a pamphlet on autonomy of nations in 1914. (1891–1938), one of the pioneers in historical phonol- He returned to Poland after World WarI in 1918. and ogy. Also, Kazimierz Nitsch (1874–1958), an out- worked as Professor of Comparative Linguistics, standing Polish dialectologist, was among the pupils Warsaw University, in 1920–1929. He stood as a can- of Baudouin de Courtenay. Most importantly, didate in a presidential election in 1922. He died in Baudouin de Courtenay’s theories were strongly influ- Warsaw on November 3, 1929. enced by his favorite student, MikoBaj Kruszewski (1851–1887). It is very difficult to separate the ideas of References the two linguists. Therefore, they are often referred to as the theories of ‘the Kazan School’. The time spent Adamska-SaBaciak, Arleta. 1996. Language change in the works in Kazan was the most prolific period in Baudouin de of Kruszewski, Baudouin de Courtenay and Rozwadowski. Poznan: Motivex. Courtenay’s life. However, partly due to Kruszewski’s Adamska-SaBaciak, Arleta. 1998. Jan Baudouin de Courtenay’s untimely death, Baudouin de Courtenay did not have contribution to general linguistics. Historiographia Lingui- any direct successors. stica XXV. Baudouin de Courtenay’s absorbing interest in lin- Baudouin de Courtenay, Jan Ignacy Niecislaw. 1868. Einige guistics did not deter him from becoming a political Fälle der Wirkung der Analogie in der polnischen Deklination. Beiträge zur vergleichenden sprachforschung activist. He supported emancipatory and democratic auf dem gebiete der arischen, celtischen und slavischen movements. After 1918, when Poland regained inde- sprachen VI. pendence, he emphasized the importance of securing ———.1870. O drevnepol’skom jazyke do XIV stoletija. the civil rights of national minorities (the minorities ———.1877. OtJëty o zanjatijax po jazykovedeniju v teJenie nominated him as their candidate for the presidency of 1872 i 1873 gg. ———.1895. Versuch einer Theorie phonetischer Alternationen. Poland). It was also reflected in his linguistic Ein Kapitel aus der Psychophonetik. research—he collected data on many minority lan- ———.1901. O smešannom xaraktere vsex jazykov. Curnal guages (e.g. Kashubian dialects in Poland, Slovene Ministerstva narodnego prosvešJenija 337. dialects in Italy, local dialects in Slovakia, and Yiddish ———.1910. Les lois phonétiques. Rocznik slawistyczny 3. in eastern Europe). ———.1922. Zarys historii jazyka polskiego. ———.1963. Izbrannye trudy po obšJemu jazykoznaniju, 2 vols. ———.1972. A Baudouin de Courtenay anthology. The begin- Biography nings of structural linguistics. ———.1974–1983. DzieBa wybrane, 6 vols. Warsaw: PWN. Jan Ignacy Niecislaw Baudouin de Courtenay was born Jakobson, Roman. 1960. Kazan´ska szkoBa polskiej lingwistyki in Radzymin, near Warsaw, Poland (then a province of i jej miejsce ws´wiatowym rozwoju fonologii. Biuletyn Polskiego Towarzystwa Je zykoznawczego 19. the Russian Empire) on March 13, 1845. He obtained Jakobson, Roman. 1929. Jan Baudouin de Courtenay. Slavische his M.A. (1866), Warsaw (at the Department of Rundschau 1. 809-12, reprinted in Portraits of linguists: a Philology and History). He received a scholarship of biographical source book for the history of western linguis- the Ministry of Education for outstanding achieve- tics, 1746–963, Vol.1: from Sir William Jones to Karl ments and continued his studies in Prague (Bohemia), Brugmann, ed. by Thomas A. Seboek. Bloomington: Indiana University Press, 1966. Jena, and Berlin (Germany), tutored by August Jakobson, Roman. On language, ed. by Linda R. Waugh and Schleicher; he obtained his Ph.D. for work on analogy Monique Monville-Burston. London and Cambridge, MA: in Polish declination, University of Leipzig, Germany, Harvard University Press. 129 BAUDOUIN DE COURTENAY, JAN IGNACY NIECISAAW Rothstein, Robert A. 1975. The linguist as a dissenter: Jan Stankiewicz, Edward. 1976. Baudouin de Courtenay and the Baudouin de Courtenay. For Wiktor Weintraub on the occa- foundations of structural linguistics. Lisse: Peter de Ridder sion of his 65th birthday, ed. by Viktor Erlich et al. The Press. Hague: Mouton. Toman, JindrK ich. 1991. Nationality as choice—Baudouin de Stankiewicz, Edward (ed.) 1972. Baudouin de Courtenay: his Courtenay’s individualistic approach. Cross Currents 10. life and work, introduction in a Baudouin de Courtenay PAWEA RUTKOWSKI anthology. The beginnings of structural linguistics. ed. by Bloomington: Indiana University Press. See also Bloomfiled, Leonard; Phoneme; Phonology Belgium Belgium is a small parliamentary and constitutional William of Orange, the Dutch king, to re-annex monarchy in northwest Europe; it shares borders with Belgium. However, this exemplary Belgian unity did the Netherlands (north and northeast), France (south not last very long: Flemish intellectuals started to and southwest), Germany (east), and Luxembourg voice claims for cultural freedom in the late 1830s. (southeast), with the North Sea forming a natural They were motivated mainly by a desire to protect the boundary on the northwest. It stretches over 32,547 Flemish cultural heritage. The movement took on a square kilometers and is divided into ten provinces: more social slant in later years. In the 1870s and five French-speaking provinces in the south (Walloon 1880s, three important linguistic laws were promul- Brabant, Hainaut, Namur, Liège, and Luxembourg), gated. First, in 1873, Flemish citizens were allowed to and five Flemish-speaking provinces in the north stand trial in their mother tongue if they wished. This (Antwerpen, Flemish Brabant, Limburg, Oost- came as a result of unfair trials in the 1860s where the Vlaanderen, and West-Vlaanderen). The Belgian pop- accused were condemned in French, a language they ulation numbers 10,213,752 (1999). From a linguistic did not understand. Then the administration in perspective, Belgium has a dual identity, straddling the Flanders became bilingual (1878), which disqualified divide between Romance and Germanic languages. Francophone civil servants from taking positions in Three national languages are officially recognized: the north of the country. Finally, part of the teaching in Dutch, French, and German. secondary schools of Flanders (1883) was to be under- Belgium became independent from the Netherlands taken in Flemish. With the growing influence of in 1830 and it was accepted internationally as a neutral Flemings in the government of the country, a Walloon state on the 20th of January the following year. French movement started to emerge, born of the worries of the became the language of the government, although Romance population who feared the loss of their priv- Germanic vernaculars (Flemish dialects) were spoken ileged position within the nation. in the north and east of the fledgling state. (German In 1898, Flemish claims were crowned by the dialectal varieties were used in what is known as ‘Alt- so-called loi d’égalité (the Equality Law) that estab- Belgien’, an area north of Liège around Vielsalm and lished the equivalence of the Flemish and French ver- Arlon, along the Luxembourg border.) At the time, the sions of laws. The weight of Flanders in the national higher bourgeoisie and the nobility throughout Europe political life was further increased when universal used French as the language of culture. Moreover, part suffrage was granted to all men after World War I. of the population spoke Romance vernaculars Given that the Flemish population was more numer- (Walloon, Picardy dialect, and Gaumais) and no other ous, Flemings started to occupy strategic functions major language appeared to be a reasonable alterna- in the government. Also, the linguistic composition tive: the Flemish dialects showed no unity, and the of the country was further complicated by annexation closest language, Dutch, was rejected as the Belgian of new German-speaking territories as a result of the Revolution broke out against Holland. Treaty of Versailles (June 28, 1919). This territory, In the 1830s, the prominent position of the French ‘Neubelgien’, comprises a southern and a northern language in Belgium was hardly challenged as lin- part separated by the Walloon cantons of Eupen, guistic and political factions were all united in a desire Malmédy, and St. Vith. The Flemish- and French- to firmly establish the new country as an international- speaking populations steadily grew apart during the ly recognized power, despite attempts made by twentieth century. In 1912, the ‘Lettre au roi’ (Letter to 130  BELGIUM the King) by Jules Destrée, a Walloon socialist, illus- between the two parts of the country, and in May 1998, trates the growing estrangement between the two parts the European Commission had to send a Swiss repre- of the Belgian population: it states that there are no sentative, Domino Columberg, to evaluate the situation Belgians, but only Flemings and Walloons. In 1932, a in Brussels and write a report on the whole linguistic new law drew a linguistic border between the two conflict in Belgium and its possible solution. main linguistic regions. This boundary could still be By the decree of December 10, 1973, Dutch is the modified according to the results of the census. official language of the Flemish Community However, the linguistic questions of the census were (5,912,000 inhabitants in Flanders and between 13 suppressed after World War II following their boycott [Francophone sources] and 25% [Neerlandophone by Flemish mayors. Flemings were concerned about sources] of Brussels’ 953,000 inhabitants). Although a losing Brussels—historically a Flemish city but with Language Union Treaty was signed between Flanders, an increasing Francophone population—and seeing Belgium, and the Netherlands on September 9, 1980, the Francophone ‘oil slick’ further expanding in sur- Flemish Dutch and Netherlandic Dutch are not identi- rounding areas. The Gilson laws painfully established cal. The main differences are found in pronunciation, the definitive linguistic border in 1962–1963: they but this does not hamper communication. Lexical dif- included facilities (the possibility of getting state serv- ferences also appear between the two variants, and ices in the other language in regions where there was Flemish Dutch possesses a more Frenchified vocabu- a linguistic minority of at least 30%) for Francophones lary. There are also minor differences in sentence in Brussels (Drogenbos, Kraainem, Linkebeek, structure. Flemish dialects are still very vivid, Rhode-St.-Genèse, Wemmel, Wezembeek-Oppem), although Dutch is the cultural vector. Standard French for Francophones in Flanders (Messines, Espierres, in Belgium is nowadays the mother tongue of Helchin, Renaix, Biévène, Herstappe, and the well- Wallonia (3,327,000 inhabitants) and of most Belgians known Fourons), Flemish in Wallonia (Comines, and immigrants in Brussels. Romance vernaculars are Dottignies, Enghien, Flobecq, Herseaux, Marcq, Petit- still used in informal situations among older people Enghien, and Warneton), and German in Wallonia (the and working classes; they can also be used by educat- six ‘Malmedian communes’: Belleveaux-Ligneville, ed people for sentimental, cultural, or amusement pur- Bevercé, Faymonville, Malmédy, Robertville, and poses. In any case, they are not the exclusive Waimes). The cohesion of the country was thus weak- communication means of any Belgian of the twenty- ening and it was necessary to reorganize the state. In first century. 1993, Belgium became a federal state after a series of French in Belgium presents a series of features that reforms: creation of communities (cultural attribu- distinguish it from Standard French (which is identi- tions) in 1970, reinforcement of communities and cre- fied as the Parisian variety). Belgian French—as well ation of regions (territorial responsibilities) in 1980, as regional varieties in France—possesses specific and creation of a third bilingual region, the Brussels words influenced by the dialects spoken in the area. Metropolitan Region (Bruxelles Capitale), in However, French in Belgium presents more specific 1988–1989. features that are linked with the institutions of the Present-day Belgium is therefore a federal state that country. This is the case, for example, in bourgmestre consists of three regions— Flanders, Wallonia, and ‘mayor’ (maire in France) or candidatures ‘first two Brussels— and three communities—Flemish, French, years of a degree’ (DEUG in France). In addition to and German. Regions and communities do not corre- these, Belgian varieties of French contain more words spond, which greatly complicates politics in Belgium. with foreign origins: Germanic in words such as ker- From a linguistic point of view, there are four areas: messe ‘kind of seasonal fairground’ (fête foraine in Flanders speaks Flemish only, Wallonia is France) or couque ‘bread roll’ (petit pain in France), Francophone—except for the German region—and Spanish as, for example, in bodega (bar found in a cel- Brussels is bilingual, although now mainly lar) or escavèche (a fish dish). Because of the numer- Francophone and very cosmopolitan, given the presence ous powers that have held sway over the territory of of bodies such as NATO and the European Union. contemporary Belgium, French in Belgium also seems Lately, the Communes à facilités have been a bone of more open to English than the Parisian variety. contention. The Peeters’ Circular (1997) tried to sup- Popular varieties of French in Belgium have fea- press facilities for French speakers around Brussels: tures of sentence structure that seem to be influenced Flemings do see facilities as a temporary solution to by Flemish: the position of the adjective before the allow integration in the other community, while noun (une propre chemise ‘one’s own shirt’ instead of Francophones understand them as a permanent state une chemise propre); interrogation in qu’est -ce que that could possibly be extended to other regions (espe- c’est pour (based on the Flemish wat is het voor ‘what cially around Brussels). This has led to ongoing tensions is that for’), or replacement of pouvoir ‘to be able to’ 131 BELGIUM by savoir ‘to know’. Finally, some words exist both in Languages have thus played a key role in the short France and Belgium but have different meanings in the history of Belgium. Linguistic tensions have shaped two countries. For example, torchon refers in France to the institutional evolution of the country in making a a tea towel but in Belgium to a cloth used for cleaning national state into a federal state. Although the coun- the floor; un rideau is a curtain in France and a net in try is nowadays divided between the two main linguis- Belgium. tic groups with increasing autonomy, linguistic In spite of some differences, Belgian and French questions (mainly concerning the future of Brussels) citizens understand each other without problems. still constitute a major driving force in Belgian politics Nowadays, specificities tend to fade away because of and are likely to remain so in the twenty-first century. the development of mass media, but there has always been a great interest in language correction in References Francophone Belgium. A feeling of linguistic insecu- rity (the conviction of not speaking the valued variety Baetens-Baerdsmore, Hugo, and E. Witte (eds.) 1986. The interdisciplinary study of urban bilingualism in Brussels. of a language) has resulted in the development of a Clevedon/Philadelphia: Multilingual Matters. strong grammatical tradition in Belgium, of which Bal, Willy, et al. 1994. Belgicismes. Inventaire des particular- Maurice Grevisse’s grammar, Le bon Usage, is the ités lexicales du français de Belgique. Louvain-la-Neuve: best-known example. There was a general tendency to Duculot. imitate French usage until recently; but now, Blampain, Daniel, et al. 1997. Le français en Belgique. Louvain: Duculot. Francophone Belgium appears more autonomous, Labeau, Emmanuelle (ed.) 2000. France–Belgique: des frères as shown in the law on the feminization of job ennemis de la langue de chez nous?, Québec: Centre interna- descriptions passed in 1993, without waiting for tional de Recherche en Aménagement linguistique (CIRAL). French legislation. Lemaire, Jacques (ed.) 1989. Le français et les Belges. German is spoken only by about 1% of the popula- Brussels: Éditions de l’Université de Bruxelles. Moreau, Marie-Louise, et al. 1999. Les Belges et la norme. tion. A separate German variety has not developed in Analyse d’un complexe linguistique. Brussels: Ministère de Belgium, although there are some specificities. These la culture, Service de la Langue française. come first from national institutions; for example, Pohl, Jacques. 1979. Les variétés régionales du français: études Permanent-deputation is used instead of Länder- belges (1945–1977). Bruxelles: Éditions de l’Université de regierung. The differences from Standard German also Bruxelles. Verdoodt, Albert. 1977. Les problèmes des groupes linguis- come from interferences with other vernaculars: tiques en Belgique. Louvain: Peeters. Germanic dialect, Walloon, and especially French. Wright, Sue (ed.) 1994. Languages in contact and conflict: These ‘gallicisms’ appear in borrowed words (e.g. ren- experiences in the Netherlands and Belgium. Clevedon: ovieren, based on French rénover instead of German Multilingual Matters, Current Issues in Language and erneuern) and in literal transpositions of phrases (e.g. Society, Vol.1, No.2. http://cfwb.be – site of the French Community of Belgium. the French téléphoner à has changed the Standard EMMANUELLE LABEAU German telephonieren mit ‘telephone with’ into tele- phonieren an ‘telephone to’). See also French Language Benveniste, Emile Benveniste is best remembered today for his writings mar of Old Persian (1931). In his thesis Origines de la on general linguistics and on semiology (following formation des noms en indo-européen (1935), which Saussure’s usage, this term has been used in Europe deals with Indo-European root morphemes, Benveniste more than ‘semiotics’). However, the bulk of his pub- broadens out from Iranian languages to the wider Indo- lications, of which we can only mention the most European field. Notes for the sequel to this research, important, focus on comparative Indo-European dealing with suffixal modals, were lost when philology. After the death at the front of fellow-student Benveniste’s apartment was ransacked during World Gauthiot, Benveniste completed his grammar of War II, but were painstakingly reconstructed and pub- Sogdian (published in 1929) and then Meillet’s gram- lished in 1948. Probably the best known of his works 132  BENVENISTE, EMILE in this area, both because of its less technical nature Biography and because of the fascinating link between language Benveniste was born in 1902 in Aleppo, but when he and culture, is his Le vocabulaire des institutions indo- was a baby the family moved to Paris. He attended européennes (1969; English translation 1973), in French and rabbinical schools and then went on to which six sections deal with topics such as the vocab- study Indo-European philology, and classical, Indian, ulary of kinship or of religion. and Celtic languages at the Sorbonne. Like his master, In his work on general linguistics, Benveniste elu- Meillet, and Saussure before that, he taught at the cidated and further developed the foundations laid by Ecole Pratique des Hautes Etudes, taking up a post in Saussure. He is a proponent of the structuralist notion Comparative Grammar and Iranian in 1927. Ten years of a system founded on a series of differences; he pro- later, he succeeded Meillet in the chair of Comparative motes the new discipline of semiology, and he extends Grammar at the Collège de France, having defended the remit of French linguistics to cover discours (dis- his thesis on the origin of Indo-European noun forma- course) and parole (speech). Benveniste participated tion in 1935. During World War II, he put his scholar- in the interwar debate on the arbitrary nature of the lin- ly knowledge to practical use, fleeing to Switzerland guistic sign, maintaining that the link between the sign with the aid of a priest who communicated with him in and external reality is arbitrary, although the link Sogdian. He became secretary of the Société de between signified and signifier is a necessary one. Linguistique de Paris, and editor of the society’s bul- Many of Benveniste’s most important articles are col- letin. Most of his time was devoted to writing and lected in his Problèmes de linguistique générale I & II teaching, although he went on a number of field trips, (1966, 1974; English translation of Vol. 1, 1971). The including visits to Iran, Afghanistan, and the American articles of the first volume were selected by the author North West, the results of which fed into his seminars himself, but the selection for the second volume, com- and into articles. Although a reclusive figure, piled with Benveniste’s help when he was in very poor Benveniste was in touch with many of the well-known health, has sometimes been questioned. It is particu- linguists of his time. His correspondence with larly for the pioneering work on énonciation that the Hjelmslev shows a common approach to a structural- Problems in General Linguistics is read today. The ist methodology; in 1947, Benveniste became one of French tradition in énonciation, aspects of which are the few foreign members of the Copenhagen dealt with under pragmatics and speech-act theory in Linguistic Circle. During the intellectual ferment of English-speaking circles, treats language not as a stat- the late 1960s in France, Benveniste became a slightly ic product, but as an interlocutionary and intersubjec- more public figure, giving interviews on linguistic tive phenomenon. Thus, Benveniste, in analyzing the structuralism, and being championed by such thinkers personal pronoun system, points to the difference as Kristéva and Barthes. On the founding of the between the I–you pronouns, which are discoursal International Association for Semiotic Studies in Paris units, and the third-person pronouns. The first- and in January 1969, he became its first president. second-person pronouns, moreover, are linked to a Benveniste had already suffered a debilitating coro- deictic system of demonstratives, and spatial and tem- nary thrombosis in 1956, and he had an incapacitating poral adverbs, as well as to the verbal system. French stroke in 1969, having seen his book on the vocabulary tenses are analyzed by Benveniste according to a dis- of Indo-European institutions through to completion tinction between histoire (story/history) and discours only months before. In his last difficult years of failing (discourse), adverbs such as ‘here’ and ‘today’, for health, he was nursed with dedication by his sister, and example, relating to a first-person pronoun and being his cause was promoted by contemporaries such as anchored in the discoursal tense system (whereas in a Barthes and Jakobsen. He died in October 1976. narrative we would hear that X was in such a place on such a date). This work on pronouns and on tense has been influential in both linguistics and literary criti- References cism. In the former, it feeds into various French tradi- tions, from the linguistics of Antoine Culioli to French Arrivé, Michel, and Claudine Normand (eds.) 1992. Emile Benveniste Vingt Ans Après, Linx no. 26, 1. discourse analysis. The histoire/discours distinction Benveniste, Emile. 1935. Origines de la formation des mots en was used to great effect (as récit/discours) by literary indo-européen, Paris: Adrien-Maisonneuve. critic Genette among others. It has been pointed out ———.1948. Noms d’agent et noms d’action en indo- that Benveniste’s work should not be seen as frag- européen, Paris: Adrien-Maisonneuve. mented, because even in his work on historical lin- ———.1962. Hittite et indo-européen, Paris: Adrien- Maisonneuve. guistics he shows the same concern to link language ———.1971. Problèmes de linguistique générale I, 1966, as and society and to account for the subjective nature of Problems in General Linguistics, transl. by M.E. Meek, language, as in his writings on general linguistics. Coral Gables, FL: University of Miami Press. 133 BENVENISTE, EMILE ———. 1973. Le vocabulaire des institutions européennes, Malkiel, Yakov. 1980. Lexis and grammar — necrological essay 1969, as Indo-European Language and Society, transl. by on Émile Benveniste (1902–76). Romance Philology 34(2). E.Palmer, London: Faber & Faber. 160–94. ———. 1974. Problèmes de linguistique générale II, Paris: Montaut, Annie, and Claudine Normand (eds.) 1997. Lectures Gallimard. d’Emile Benveniste, Linx no. spécial. Lotringer, Sylvère, and Thomas Gora (eds.) 1981. Polyphonic CAROL SANDERS linguistics: the many voices of Emile Benveniste, Semiotica, special supplement. Kristéva, Julia, et al. (eds.) 1975. Langue Discours Société. See also Indo-European 1: Overview; Semiotics; Pour Emile Benveniste, Paris: Seuil. Structuralism Bever, Thomas Within linguistics, Thomas Bever can be considered a do indeed have access to abstract grammatical infor- renaissance scholar. His work impacts the fields of mation. The experiments involved placing a click in a theoretical linguistics, psycholinguistics, cognitive neutral place in the stream of sounds. People tend to neuroscience, reading, esthetics, and language devel- perceive these clicks not in the place where they actu- opment. He is probably best known for his work that ally occur, but at some phrasal boundary. Bever distinguishes independent systems of language pro- extended these tests to even more abstract structures. cessing (production and perception) from the system The following two sentences, although apparently that represents grammatical knowledge. He argues that similar on the surface, differ in the deep semantic rela- the universal properties of language can come from a tions underlying them. In (2a), the noun phrase the variety of sources, including what can be said, under- troops is the patient of the main clause. In (2b), the stood, remembered, and learned, and are not necessar- troops is the agent of the embedded clause: ily the result of grammatical knowledge. (2) (a) The general defied the t/roopsi [ to fight] Originally interested in music and theater, Bever # reports that he decided to take a major in linguistics pri- marily because it was the course of study with the fewest (b) The general desired [the t/roops to fight] requirements that involved language courses. While an # undergraduate, he worked in the psycholinguistics lab of Margaret Bulowa (assisting in an early study of lan- (from Bever 1998:128) guage acquisition using a camera system he designed). His undergraduate thesis was on the stages of the acqui- Syntactic constituency is marked with square sition of phoneme perception in infants. He was one of brackets; the placement of the click is marked with a the first students in the linguistics doctoral program at slash. People perceive the click in different places MIT, where he studied both linguistics (mostly phono- (marked with a #) in the two sentences, corresponding logical theory under Morris Halle) and psychology. to the structure of the sentence. The click experiment In 1964, Bever was elected to the Harvard Society showed that speakers were aware of the deep semantic of Fellows, where he worked in the laboratory of difference in the structure of these clauses: the click George A. Miller. He continued working with Miller was perceived at a clausal constituent boundary when they both moved to Rockefeller University in reflecting the deep positioning of the noun phrase. 1967. It was in Miller’s lab that he first explored the These click experiments led Bever to another major question of the relationship between formal grammat- area of research, that of cerebral dominance (the idea ical models and psychological reality. Here, he collab- that one half of the brain is more powerful) in lan- orated extensively with Jerry Fodor, Merrill Garrett, guage and other cognitive skills, both in humans and and Jacques Mehler. Bever’s earliest work was partly in animals. a response to behaviorist claims that speakers and lan- Perhaps Bever’s most important contribution to guage learners do not have access to ‘abstract’ or linguistics showed that there is no single system that ‘deep’ levels of language. Using perceptual experi- contributes to our production and processing of lan- ments, Bever and his colleagues showed that speakers guage. In Generative Grammar, there are a set of rules 134  BEVER, THOMAS for generating the structure of sentences. These rules tions about the intersections among language acquisi- constitute our knowledge or ‘competence’ in language. tion, sentence processing, formal grammar, and neuro- Linguists tap this knowledge using acceptability judg- science (among other areas) and has shown that one ments. In early work, George Miller hypothesized that cannot thoroughly theorize about any one of these with- the set of generative rules were essentially identical to out considering the interactions with other domains. the set of processes we use to produce and decode sen- tences. One consequence of this idea is that the more Biography rules involved in generating a sentence, the harder it should be to produce or understand; this hypothesis is Thomas Bever was born in Boston, Massachusetts, on often known as the ‘Derivational Theory of December 9, 1939. He completed his B.A. (1961) Complexity’. While initial experimental results testing magna cum laude with highest honors in psychology this were promising, several studies by Bever and his and linguistics from Harvard, tutored by Roman colleagues showed that it was based on some faulty Jakobson, and his undergraduate thesis was on the assumptions about how the rules were organized. Bever stages of the acquisition of phoneme perception in (in work that continues up to the present time) suggests infants. His Ph.D. (1967), Massachusetts Institute of that perhaps more than one system is at work when pro- Technology, dissertation on Menomini phonology was cessing or producing a sentence. For example, when supervised by Morris Halle. He was Lecturer, MIT hearing a sentence, one uses a set of perceptual strategies Psychology department, 1964–1966; Assistant to provide an initial structure to the sentence. One exam- Professor, 1967–1969, and Associate Professor, ple of such a strategy is the idea that noun–verb–noun 1969–1970, Rockefeller University; Professor of (NVN) sequences map onto agent–action–patient Psychology, Columbia University, 1970–1986; semantics; another is that the first noun–verb sequence Chairman, Columbia University Interdisciplinary Ph.D. heard constitutes the subject and predicate of the main Program in Psychology and Linguistics, 1973–1986; clause. These perceptual strategies are not the same as Pulse Professor of Psychology and Professor of the generative rules that give us acceptability judgments. Linguistics, University of Rochester, 1985–1995; Head, Instead, these strategies form a separate cognitive sys- Language and Cognition Program 1986–1989 and tem. Evidence for this proposal comes from the fact that 1992–1994, Director, Cognitive Science Program at the stage when children seem to rigidly use perceptu- 1991–1992, Directory, Center for the Sciences of al strategies in other domains (such as in estimating Language 1988–1995 at the University of Rochester; numerosity)—at approximately 3 years of age—they Research Professor of Cognitive Science, Linguistics rigidly apply linguistic perceptual strategies to sentences and Psychology, University of Arizona, 1995–present; they hear, often doing much worse than younger chil- Head, Department of Linguistics, University of Arizona dren. For example, in tests where they are asked to iden- 1998–2003; Phi Beta Kappa 1961, National Institutes of tify which picture is best associated with a particular Health Predoctoral Fellowship 1962–1964, Harvard sentence (3), 2-and 4-year-olds both correctly identify Society of Fellows 1964–1967; National Science Big Bird as the agent and Cookie Monster as the theme. Foundation Faculty Fellowship, summers 1974–1977, However, 3-year-olds reverse the roles, consistent with Guggenheim Fellowship 1976–1977, Fellow, Center for the NVN strategy, which treats the first noun as an agent. Advanced Study in the Behavioral Sciences 1984–1985; Vice-President, the Rockefeller University (3) Cookie Monster was hit by Big Bird Chapter of the American Association of University Since there is a stage at which children seem to exclu- Professors, 1969–1970; and Co-founder and Associate sively use these perceptual strategies, it seems reason- Editor, Cognition, 1973–present. able to conclude that they exist. Obviously, these perceptual strategies must relate in References some way to grammatical rules. Bever claims that we use both systems in a kind of ‘analysis by synthesis’: we Berwick, Robert C., and Amy S. Weinberg. 1984. The gram- use perceptual strategies to provide an initial hypothesis matical basis of linguistic performance: language use and acquisition. Cambridge: MIT Press. about the structure of the sentence and the semantic Bever, Thomas. 1970. The cognitive basis for linguistic struc- relations that underlie it, then we go back with our gen- tures. Cognition and language development, ed. by R. erative rules and check the structure. One piece of Hayes, 279–362. New York: Wiley & Sons, Inc. experimental evidence that he presents is that tone and Bever, Thomas. 1980. Broca and Lashley were right: cerebral click perception at the end of clauses is poor, suggesting dominance is an accident of growth. Biological studies of mental processes, ed. by D. Caplan, 186–230. Cambridge, that this is the stage at which synthesis is occurring. MA: MIT Press. Bever’s wide-ranging interests reflect the complexity Bever, Thomas. 1988. The psychological reality of grammar: of the phenomena he studies. His work has raised ques- a student’s eye view of cognitive science. The making 135 BEVER, THOMAS of cognitive science: a festschrift for George A. Miller, ed. by Carroll, J.M., Bever, Thomas, and C.R. Pollock. 1981. The W. Hirst, 112–42. Cambridge: Cambridge University Press. non-uniqueness of linguistic intuitions, Language 57. Bever, Thomas, J.M. Carroll, and L.A. Miller (eds.) 1984. 368–82. Talking minds. Cambridge, MA: MIT Press. Carroll, David W. 1986. Psychology of language. Pacific Bever, Thomas, and J. Fodor. 1965. The psycholinguistic reali- Grove, CA: Brooks/Cole Publishing Company. ty of linguistic segments. Journal of Verbal Learning and Fodor, Jerry, Thomas G. Bever, and Merrill Garrett. 1974. Verbal Behavior 4. 414–20. The psychology of language: an introduction to psy- Bever, Thomas, and B. McElree. 1988. Empty categories cholinguistics and generative grammar. New York: access their antecedents during comprehension. Linguistic McGraw-Hill. Inquiry 19. Garrett, Merrill. 1990. Sentence processing. An invitation to Bever, Thomas, H.L. Roitblat, and H.S. Terrace (eds.) 1984. cognitive science. Vol 1: Language, ed. by Daniel N. Animal cognition. Hillsdale, NJ: Erlbaum. Osherson and Howard Lasnik, 133–75. Cambridge, MA: Bever, Thomas, M. Sanz, and D. Townsend. 1998. The emper- MIT Press. or’s psycholinguistics. Journal of Psycholinguistic Research Schütze, Carson. 1996. The empirical base of linguistics gram- 27. 261–84. maticality judgments and linguistic methodology. Chicago: Bever, Thomas, and D. Townsend. 2001. Sentence processing. The University of Chicago Press. Cambridge, MA: MIT Press. ANDREW CARNIE Bilingual Acquisition Children growing up exposed to two or more lan- occasionally been suggested. The controversial issue guages acquire each of them in much the same way as here is the age range during which changes occur, monolinguals (speakers of just one language). which result in qualitative differences between first Bilingual acquisition during childhood can thus be and second language acquisition. Although sufficient regarded as an instance of simultaneous acquisition of evidence is not yet available, the period around three two ‘first’ languages. In fact, if children are exposed to years of age appears to be of particular significance. more than two languages simultaneously, they are able The most widely accepted view is that one may qual- to acquire full competence of each that does not sub- ify as ‘simultaneous’ those instances of bilingual stantially differ from the speaking ability of monolin- acquisition where the child is exposed to two lan- guals; bilingualism is therefore a special case of guages before the age of three. In this case, each lan- multilingualism. guage of the bilingual is predicted to share crucial Comparison with monolinguals is one of the main similarities with that of the respective monolingual issues in research on bilingual acquisition. It can be speakers but to exhibit differences in comparison with extended to all types of acquisition where more than languages learned at a later age. one language is learned, independently of age of onset The insight that bilingual acquisition proceeds in and proficiency attained in each language. The pres- the same way and leads to the same type of linguistic ent discussion, however, is limited to child bilingual- knowledge as monolingual acquisition is largely due ism. Whether types of bilingual acquisition should, in to a number of case studies carried out in the late fact, be distinguished according to the age of onset of 1970s and 1980s, although the first monograph docu- learning (‘bilingualism’ vs. ‘foreign language learn- menting the linguistic development of a bilingual child ing’) depends ultimately on the results obtained in (Ronjat 1913) arrived at a similar conclusion. Until the these comparisons. Given our current state of knowl- 1970s, it was frequently assumed that bilinguals face edge, it is plausible to assume that age of onset is a serious difficulties in separating language systems and crucial factor causing fundamental differences that they run the risk of ending up with mixed lan- between child and adult language acquisition. guages. Such views were often the result of inadequate Consequently, it is necessary to distinguish between research methodologies. Large-scale studies, frequent- simultaneous and successive acquisition of bilingual- ly focusing on disfavored minority groups, failed to ism. ‘Simultaneous’ acquisition does not necessarily perform in-depth analyses of bilingual language use. imply that the child has been exposed to both lan- Early case studies, carried out by amateurs without the guages from birth, although this interpretation has appropriate training, analyzed the language use of 136  BILINGUAL ACQUISITION their own children. Some more satisfactory scholarly other sociolinguistic requirements. No later than dur- studies, however, arrived at similar conclusions, main- ing the first half of the third year, mixes within a sin- ly because they adopted a monolingual perspective in gle clause conform to grammatical constraints. Since their investigation of bilingualism. The resulting short- earlier examples consist almost exclusively of single- comings were amended by later research when, begin- word mixes, the role of structural constraints is diffi- ning in the 1980s, interest in bilingual acquisition cult to assess. The frequency of mixing does not increased dramatically. generally decrease with grammatical development, Other issues that have attracted the attention of both contrary to attempts to explain early mixing as result- researchers and the interested public involve alleged ing from deficient ability. Although not knowing a problems in separating the two languages. Bilinguals word in one language may initially lead to mixing, it is tend to mix languages within a conversation or even an also triggered by social context. Most importantly, the utterance; from a monolingual perspective, this language children are exposed to appears to influence appears to indicate an inability to keep the languages their mixing behavior. Not only do children mix more apart. It has also been suggested that bilingual children frequently if the persons they interact with mix freely, initially go through a phase during which they develop but also the nature of parent’s reaction to child mixing one ‘hybrid’ system before they are able to differenti- has been argued to be of crucial importance, i.e. par- ate the lexical and grammatical systems they are ents signal overtly or implicitly whether they encour- acquiring. As early as the 1970s, researchers agreed age or discourage language mixing. almost unanimously that these children eventually Quantitative and qualitative characteristics of lan- succeed in separating their languages, without much guage exposure are also relevant for the question if effort or specific pedagogical measures, but parents bilinguals can attain balanced, i.e. equally developed, and educators tended to regard the alleged single lin- knowledge of both languages. One language might guistic system phase as a potentially serious problem, tend to be dominant (preferred choice, faster recall of fearing the possibility of a lasting state of incomplete words, etc.), but dominance can shift over the lifespan, competence in both languages. More sophisticated and it does not seem to affect competence—no con- analyses of children’s language mixing eventually vincing evidence supports the idea of ‘semicompe- revealed no initial single-system stage. Children have tence’. Similarly, although it has been suggested that been shown to differentiate linguistic systems from bilingual acquisition is delayed in comparison to very early on. Grammatical differentiation, for exam- monolingual acquisition, onset as well as development ple, is evidenced as soon as multiword utterances are are clearly within the limits of monolingual norms, used. Thus, the available empirical evidence does not faster bilinguals outperforming slower monolinguals. support the idea of a temporary fusion of grammatical In both cases, differences concern details of speaking, systems, and mixing can for the most part be explained not qualitative differences in underlying knowledge. as a particular type of bilingual language use. Early This confirms the hypothesis that the human language grammatical differentiation, however, does not faculty predisposes the individual to become bilingual. exclude the possibility of crosslinguistic influence during later developmental phases, resulting either in vocabulary or grammatical transfer or in acceleration References or delay of development. The question of whether De Houwer, Annick. 1990. The acquisition of two languages such influence actually plays a major role in bilingual from birth: a case study. Cambridge: Cambridge University acquisition continues to be controversial. Press. Genesee, Fred. 1989. Early bilingual development: one lan- Language mixing by adult bilinguals is normally guage or two? Journal of Child Language 16. 161–79. the result of code-switching, a form of language use Grosjean, François. 1982. Life with two languages: an intro- determined by a complex network of sociolinguistic duction to bilingualism. Cambridge, MA and London: factors. Code-switching within a single clause is sub- Harvard University Press. ject to grammatical constraints depending on the struc- Köppe, Regina, and Jürgen M. Meisel. 1995. Code-switching in bilingual first language acquisition. One speaker, two lan- tural properties of both languages. Consequently, in guages: cross-disciplinary perspectives on code-switching, order to be able to code-switch, children have to ed. by Lesley Milroy and Pieter Muysken, 276–301. acquire the required social skills and grammatical Cambridge: Cambridge University Press. knowledge. It has therefore been suggested that mix- Lanza, Elizabeth. 1997. Language mixing in infant bilingual- ing during initial phases of linguistic development ism: a sociolinguistic perspective. Oxford: Clarendon Press. Leopold, Werner F. 1939–1949. Speech development of a bilin- may not yet be adult-like code-switching. However, gual child: a linguist’s record, vols. 1–4. Evanston, IL: before the age of 2, children choose the language Northwestern University Press; New York: AMS Press, according to the addressee and begin to comply with 1970. 137 BILINGUAL ACQUISITION Meisel, Jürgen M. 1989. Early differentiation of languages in Ronjat, Jules. 1913. Le développement du langage observé chez bilingual children. Bilingualism across the lifespan: aspects un enfant bilingue. Paris: Librairie Ancienne H. Champion. of acquisition, maturity, and loss, ed. by Kenneth Volterra, Virginia, and Traute Taeschner. 1978. The acquisition Hyltenstam and Loraine K. Obler, 13–40. Cambridge: and development of language by bilingual children. Journal Cambridge University Press. of Child Language 5. 311–26. Paradis, Johanne, and Fred Genesee. 1996. Syntactic acquisi- JÜRGEN M. MEISEL tion in bilingual children: autonomous or interdependent? Studies in Second Language Acquisition 18. 1–25. See also Acquisition; Acquisition Theories Bilingual Mixed Languages It is generally believed that a language’s basic vocabu- undetected. Any type of language can be involved in lary and grammatical inflections are relatively impervi- bilingual mixed languages, and there are no structural ous to foreign influence, and that they are therefore safe constraints that prevent the ‘intertwining’ of any two diagnostic tools for determining whether given lan- languages, although the result may differ according to guages are related. It is also said that ‘no language is so the typological properties of the languages involved. mixed that it cannot fit unambiguously into a family Media Lengua, spoken in Ecuador, is an example of tree: it will always be possible to show that the bulk of a bilingual mixed language. The vocabulary is of a language’s lexicon and grammatical structures come Spanish origin and the grammar stems from Quechua, from the same source.’According to these claims, there- a Native American language, as the following diagram fore, structurally mixed languages should not exist. for the sentence ‘I come to ask a favor’ shows: Nevertheless, mixed languages—languages that can- not be classified unequivocally—do in fact exist and are Media Lengua (Ecuador): Spanish lexicon—Quechua relatively common. The best-known examples are lan- grammar; Spanish items are in italics. Source: guages that arise as a direct result of language contact Muysken (1997). and that comprise linguistic material that cannot be traced back to a single source language. Such languages ML Unu fabur - ta pidi - nga - bu therefore do not belong to any language family and their ONE FAVOR - ACC ASK - NOM - BEN genesis is not a matter of descent with modification bini - xu - ni from a single parent language. It is generally accepted COME- PROG- 1P that mixed languages are of three types: pidgins, cre- oles, and bilingual mixtures (also known as ‘intertwined Q Shuk fabur - da mãa - nga - bu languages’). In pidgins and creoles, the vocabulary is ONE FAVOR-ACC ASK- NOM - BEN usually traceable to a single language, always the lan- guage of the dominant group involved in the contact, shamu - xu - ni while the grammatical structure (and often the sound COME - PROG - 1P system) generally cannot be traced to any single lan- guage. In bilingual mixed languages, the basic vocabu- S Vengo para pedir un favor lary can always be traced to a single source language COME 1SG FOR ASK-INF A FAVOR and the grammatical structure (including the sound sys- ‘I come to ask a favor’ tem) to another single source language. Furthermore, the linguistic material from each source language is Thus, Media Lengua uses the Quechua grammar as adopted wholesale without the kind of distortion that a structural grid into which the Spanish vocabulary occurs in pidgins or creoles. Also bilingual mixtures items are inserted. share a characteristic social feature: all of them are in- Michif, spoken in Western Canada and in North group languages used within the community as a sign of Dakota, represents a similar case. Here Cree, a Native community solidarity and are not understood by out- American language, provides the structural basis for siders, which explains why they have often remained the insertion of French vocabulary. 138  BILINGUAL MIXED LANGUAGES Other examples of bilingual mixed languages are: Second, a group of nomadic people settle and even- Source of Mixed Source of tually feel the need for a secret trade language. They Vocabulary Language Grammar will tend to use the grammatical system of the host Puquina Callahuya (Bolivia) Quechua society and the lexicon of the inherited language. In Cushitic Maⴕa / Inner each case, the language known best provides the Mbugu (Tanzania) Bantu grammar. The group creating the mixed language Japanese Senkyoshigo English must be highly bilingual when the language mixing (Japan) Dutch Krõjo / Javindo Low Javanese starts. Also, the resulting language is always intend- (Java) ed as an in-group language and therefore is not Dutch Pecuⴕ (Surabaja, Low Javanese intended to bridge a communication gap between Indonesia) speakers of different languages, in contrast to pidgins Malay Chindo (Java) Low Javanese and creoles. Cushitic Ilwana (Kenya) Bantu Angloromani (Great Britain, USA, Australia): Romani References lexicon—English grammar; Romani items are in ital- Arends, Jacques, Pieter Muysken, and Norval Smith (eds.) ics. Source: Bakker and Muysken (1994). 1994. Pidgins and Creoles: an introduction. Amsterdam and AR Palla bish besh - es apopli the Beng Philadelphia, PA: Benjamins. Bakker, Peter. 1996. Language intertwining and convergence: AFTER 20 YEAR – PLUR AGAIN DEVIL Typological aspects of the genesis of mixed languages. wel - d and pen-d : Av with man- di Sprachkontakt und Grammatikalisierung, ed. by Nicole Nau and Martin Haase, Special Issue of Sprachtypologie und COME-PAST SAY-PAST COME ME – DAT Universalienforschung 49(1). Bakker, Peter. 1997. A language of our own : the genesis of R Palla bish besh- aw apopli o Beng vi-as. Yov Michif, the mixed Cree–French language of the Canadian pen-das : Av man-tsa Métis. New York and Oxford: Oxford University Press. E After twenty years the Devil came back and Bakker, Peter. 2000. Rapid language change: creolization, inter- twining, convergence. Time depth in historical linguistics, said : Come with me ed. by Colin Renfrew, April McMahon, and Larry Trask. Over a dozen Romani mixed languages are spoken Cambridge: McDonald Institute for Archeological by Gypsies in Europe and the Middle East. In all cases, Research. the vocabulary is Romani and the grammar comes from Bakker, Peter, and Hein van der Voort. 1991. Para-Romani lan- guages: an overview and some speculations on their genesis. the language of the host country (Basque Romani, In the margin of romani: gypsy languages in contact, ed. by Spanish Romani, Swedish Romani, Armenian Romani, Peter Bakker and Marcel Cortiade. Amsterdam: Turkish Romani, etc.). Publications of the Institute for General Linguistics, The process responsible for such bilingual mix- University of Amsterdam. tures has been called ‘relexification’, where the entire Bakker, Peter, and Maarten Mous (eds.) 1994. Mixed lan- guages: Fifteen case studies in language intertwining. vocabulary or ‘lexicon’ is borrowed into another lan- Amsterdam: Institute for Functional Research into guage. Conversely, ‘regrammaticalization’, the intru- Language and Language Use, University of Amsterdam. sion of an entire grammar onto a native lexicon, is Bakker, Peter, and Pieter Muysken. 1994. Mixed languages equally possible. This seems to be true for the Romani and language intertwining In Arends, Muysken, and mixed languages, since it is most likely that the lan- Smith. Bakker, Peter, and Robert A. Papen. 1997. Michif: a mixed lan- guage of the host country was learned while main- guage based on Cree and French. In: Thomason. taining the original Romani vocabulary. This is why it Muysken, Pieter. 1981. Halfway between Quechua and Spanis: is important to distinguish the linguistic process from the case for relexification. Historicity and variation in the historical or social process of bilingual language Creole studies, ed. by Arnold Highfield and Albert Valdman. mixtures. Linguistically speaking, it is generally the Ann Arbor, MI: Karoma. Muysken, Pieter. 1997. Media Lengua. In Thomason. intrusion of a lexicon into a grammatical system Smith, Norval. 1994. An annotated list of Creoles, Pidgins and (relexification), whereas the sociohistorical process mixed languages. In Arends, Muysken, and Smith. may go either way. Thomason, Sarah G. (ed.) 1997. Contact languages: a wider Two specific circumstances for the creation of perspective. Amsterdam and Philadelphia, PA: Benjamins. bilingual mixtures have been proposed. First, the Thomason, Sarah G., and Terrence Kaufman. 1988. Language contact, creolization and genetic linguistics. Berkeley, CA: birth of a new ethnic group comes about after a first University of California Press. generation with mothers who speak one language and ROBERT A. PAPEN fathers who speak another. The mixed language will have the grammatical system of the mothers’ lan- See also Chinese Pidgin English; Chinese Pidgin guage and the vocabulary of the fathers’ language. Russian; Pidgins and Creoles 139 BILINGUALISM Bilingualism Broadly defined, bilingualism denotes the ability to use in which they originally learned them. In this way, lan- two languages. This can involve the ability to read, guage use may be context-specific despite a high level write, comprehend, or speak to varying levels of fluen- of fluency in more than one language. In summary, cy in each language. There is a wide degree of variation different sociolinguistic situations may determine the across bilingual speakers in terms of their language language that is primarily used and the manner in skills. This variation stems from the manner in which which it is used. the languages were acquired, the means by which they Code-switching, or the alternating use of more were learned, and the frequency of their use. The age of than one language, is common among bilingual acquisition and degree of exposure to languages will speakers. Individuals may either mix languages with- also determine the degree of bilingualism. in a statement or switch at defined points such as In most cases, bilingual speakers have one domi- breaks between topics. While the specific reasons for nant language, that is, they are more proficient in one switching are not yet clearly understood, it is known or more linguistic levels of language processing in one that bilingual speakers often switch languages in language than in the other. In some cases, the domi- order to facilitate communication or to express con- nant language is also the native or first language spo- cepts that can be language-specific. In this way, ken. The second language is often termed code-switching serves a pragmatic or strategic func- ‘nondominant’ or ‘subordinate’. Individuals can be tion. Code-switching is a complex, rule-governed equally proficient in two languages and are then called phenomenon that is guided by the particular topic, ‘balanced bilinguals’. When individuals learn both context, or circumstance regarding the communica- languages simultaneously, typically from birth, they tion at hand. The term ‘Spanglish’ has been accorded are said to be ‘compound bilinguals’. In other cases, to the intermixing of the Spanish and English lan- individuals learn one language first, and a second lan- guages in certain areas within the United States guage in a different context, such as in a different where both languages are often spoken (e.g. country or with a different parent. The latter bilinguals Southwestern states, New York, New Jersey, Florida). are often referred to as ‘coordinate bilinguals’. When Researchers view code-switching as advantageous in individuals learn a language in this way and the first some situations as it helps to preserve and maintain language remains relatively intact, the term ‘additive both languages simultaneously. Others argue that the bilingualism’ is used. Interestingly, an individual’s constant mixing of languages may make it difficult second language can sometimes develop into the more for individuals to know any single language with a dominant language and actually replace the first lan- high degree of proficiency. Nevertheless, the use of guage learned. For example, an individual born in languages in an interrelated way is common among France who is fluent in French may travel to the individuals who share a diverse but common lan- United States at a very young age and remain there for guage background. many years. This individual may come to know and It is agreed that bilingualism is much more the use English with greater fluency than their native lan- norm around the world than monolingualism. In guage, French. This type of bilingualism has often Indonesia, for example, several hundred languages been referred to as ‘subtractive’ and may actually are in constant use. In Central Europe, the ability to involve the attrition of skills in the earlier language. speak more than one language is often typical. Language ability can be tied to specific topics or Moreover, it has been documented that bilingual domains of knowledge that a bilingual possesses. For speakers possess a certain cognitive or mental flexi- example, an individual might use Spanish at home and bility that serves as an asset in cognitive processing in in conversation with friends, but English at work or for general. For example, bilingual children outperform technical discussion of a specific topic. Some individ- their monolingual counterparts in tasks that involve uals may be native speakers of a language such as role-playing, classifying objects, creativity, concept Chinese and have a strong or native-like proficiency in formation, memory, metalinguistic awareness, per- another language, such as English, yet perform all ceptual disembedding, problem-solving, role-taking, mathematical functions in their native language only. social sensitivity, and understanding complex instruc- These speakers claim that they have more automatic tions. High-performing five- and six-year-old bilin- processing of mathematical concepts in the language gual speakers exhibit a greater degree of divergent 140  BIOSEMIOTICS thinking, imagination, grammatical awareness, and Other terms that are often used to refer to bilingual- perceptual organization. Clearly, bilingual children ism include ‘polyglottism’ (although ‘polyglot’ often have more than one way of labeling their environment refers to people who allegedly command an improba- and have an early awareness of the multiple ways in bly large number of languages) and ‘multilingualism’. which concepts can be described and named across languages. Bilingual children simply have more References examples for words and grammar to choose from than Altarriba, J. 2000. Language processing and memory retrieval do monolinguals. Therefore, language competence in Spanish–English Bilinguals. Spanish Applied Linguistics often influences the range of interactive skills and 4. 215–45. capabilities that a bilingual can use. Baker, C. 1988. Key issues in bilingualism and bilingual edu- A distinction can also be drawn between ‘individual cation. Clevedon: Multilingual Matters. bilingualism’ (the bilingual competence of a single Bialystok, E. (ed.) 1991. Language processing in bilingual chil- dren. Cambridge: Cambridge University Press. individual) and ‘societal bilingualism’ (the degree to de Groot, A. M. B., and J. F. Kroll (eds.) 1997. Tutorials in which two languages are used within a society). In bilingualism. Mahwah, N J: Lawrence Erlbaum Associates. some cultures and language groups, bilingualism is Grosjean, F. 1982. Life with two languages. Cambridge, MA: widely accepted and languages are almost inter- Cambridge University Press. changeable across contexts. In other societies, multi- Hakuta, K. 1986. Mirror of languages: the debate on bilingual- ism. New York: Basic Books. ple languages may be used, but they may be directly Harris, R. J. (ed.) 1992. Cognitive processing in bilinguals. linked to specific contexts or uses. Diglossia is the Amsterdam: Elsevier. term used to describe two functionally different vari- Schreduer, R., and B. Weltens (eds.) 1993. The bilingual lexi- eties of the same language that are used for specific con. Philadelphia: John Benjamins. purposes. For example, some documents may be writ- Weinreich, U. 1953. Languages in contact. New York: Linguistic Circle of New York. ten in a more formal or Classical Arabic that is distinct JEANETTE ALTARRIBA from a colloquial Arabic that is more commonly used in conversation. See also Bilingual Acquisition Biosemiotics Biosemiotics deals with all kinds of sign-based (1) Cellular sign systems. This is the sign system processes in living systems. In other words, it is a the- of any bacterial (and eukaryotic) cell. Its char- ory of semiosis in living systems, or biology that inter- acteristic processes are enzymatic recognition prets living systems as sign systems; or the study of and transmembrane signaling. Its inheritance biological codes; or the study of prelinguistic sign system combines the epigenetic and genetic, processes and signs. These definitions can be taken as the latter being dominant. It is characterized (almost) synonymous. The appearance of human lan- by microsemiosis. According to Yates guage does not mean the primary appearance of signs, (1997:458), microsemiosis ‘does not address because there are signs preceding human language. communication between cells or among cell Accordingly, human language comprises a special complexes’. In a cell, true codes appear. The type among the sign systems. There exists the semiotic codes, as different from other relationships, threshold between living and nonliving, a natural bor- are the relationships that are not deducible der between the semiotic and nonsemiotic world, from universal physical laws because they which means that there are no sign processes per se in connect independent domains and are fixed the nonliving world, whereas life is perfused with due to a unique historical process (Barbieri signs, if it is not composed exclusively of signs. 2001). Codes as part of any single semiosis According to a biosemiotic view, life science and sign turn the web-like set of cellular processes science are coextensive. into a meaning-generating interpretation- Altogether, there exist five distinctively different like process—this is biointerpretation, or levels of sign systems. biosemiosis. 141 BIOSEMIOTICS (2) Vegetative sign systems. This includes the com- emphasis of the study in biosemiotics appears to be very munication between the tissues in a multicellu- different from the one in linguistic sciences. However, lar organism. Because the cells of an organism biosemiotic knowledge is important for linguistics due are genetically identical, the differences to its results in the analysis of more fundamental and between the cell types of an organism are based general communication phenomena. For instance, the on another inheritance system—the epigenetic general mechanism of categorization, as known in one. Vegetative sign systems are responsible for phonemics, has been shown to be analogous to the the genesis of a multicellular biological form, mechanisms of speciation in biparental organisms. the whole morphology of the body as a result of A pioneering approach to biosemiotics has been communication between cells (Kull 2000). established by the studies of a biologist and philoso- Morphogenesis and cell differentiation are its pher Jakov von Uexküll (1928, 1982, 1992). In the typical representatives. In this sense, vegetative framework of semiotics, biosemiotics has been devel- sign systems are not confined to plants—they oped basically together with a turn from a Saussurean occur in all multicellular organisms. Due to dyadic to Peircean triadic concept of sign (Deely them, organic forms are communicative struc- 2001) and with a development of concept of semios- tures. Vegetative sign systems provide catego- phere as a web of all kind of (including interspecific) rization without representation. communication processes. A key concept in it is an (3) Animal sign systems. This is the senso-neuro- analog-digital duality of all organic meaning-generat- muscular system, the one that is responsible for ing processes (Hoffmeyer, Emmeche 1991). Since the behavior of a motile animal organism. The 1990s, biosemiotics has established as a university dis- basic inheritance system here is neuro-humoral cipline (Sebeok, Umiker-Sebeok 1992, Kull 2000, (or behavioral). Its characteristic feature is the Sebeok 2001, Schult 2004). existence of representations, and thus— Umwelt, the model of the organism’s world. References Animal sign systems represent the primary modeling systems (according to Sebeok 1994) Barbieri, Marcello. 2001.The organic codes: the birth of seman- tic biology. Ancona: Pequod. and are studied by zoosemiotics. Bickerton, Derek. 1990. Language and species. Chicago: (4) Linguistic sign systems. This corresponds to the University of Chicago Press. secondary modeling system (natural language), Deacon, Terrence. 1997.The symbolic species: the co-evolution according to Sebeok (1994) (or primary model- of language and the brain. NY: Norton. ing systems, by Lotman). Different from the Deely, John. 2001. Umwelt. Semiotica 134(1/4). 125–35. Hoffmeyer, Jesper, and Emmeche Claus. 1991. Code-duality and animal sign system, it has syntactic signs the semiotics of nature. On semiotic modeling, ed. by M. (Bickerton 1990), i.e. the signs that are used to Anderson and F. Merrell, 117–66. Berlin: Mouton de Gruyter. refer to a type of relation between signs (e.g., a Kull, Kalevi. 1999. Biosemiotics in the twentieth century: a human ‘and’—a sign which other animals view from biology. Semiotica 127(1/4). 385–414. lack), and is principally symbolic (Deacon Kull, Kalevi. 2000. An introduction to phytosemiotics: semiotic botany and vegetative sign systems. Sign Systems Studies 1997). Its characteristic feature is the existence 28. 326–50. of sentences and thus symbolic language. Schult, Joachim (ed.) 2004. Biosemiotik: praktische Anwendung (5) Cultural sign systems. This corresponds to the und Konsequenzen für die Einzelwissenschaften. Berlin: tertiary modeling systems, according to Sebeok Verlag für Wissenchaft und Bildung. (or secondary modeling systems, according to Sebeok, Thomas A. 1994. Signs: an introduction to semiotics. Toronto: University of Toronto Press. Lotman). Their characteristic features include Sebeok, Thomas A. 2001. Biosemiotics: its roots, proliferation, the existence of artistic, ideological, ethical, and prospects. Semiotica 134(1/4). 61–78. and similar structures. Sebeok, Thomas A., and Jean Umiker-Sebeok (eds.) 1992. Biosemiotics. Berlin: Mouton de Gruyter. When characterizing the semiotic aspects of behavior Uexküll, Jakob von. 1928. Theoretische Biologie. Berlin: Springer. and communication of different groups of animals (e.g., Uexküll, Jakob von. 1982. The theory of meaning. Semiotica insects, birds, mammals), it is important to pay attention 42(1). 25–82. to (1) the general features of the body plan and sensory Uexküll, Jakob von. 1992. A stroll through the worlds of ani- apparatus of the animals, and the functional place of the mals and men: a picture book of invisible worlds. Semiotica animal group in ecosystems, (2) the main ways and 89(4). 319–91. Yates, F. Eugene. 1997. Microsemiosis. Semiotics: a handbook means of active communication, and (3) the main trends on the sign-theoretic foundations of nature and culture, Vol. 1, in the evolution of communication in the group. ed. by Roland Posner, Klaus Robering and Thomas A. Because nonhuman communication does not use Sebeok, 457–64. Berlin: Walter de Gruyter. sentences, that is, it is nonpropositional, the entire KALEVI KULL AND ALEKSEI TUROVSKI 142  BLOOMFIELD, LEONARD Bloomfield, Leonard Leonard Bloomfield may be considered the father of the neogrammarians, a group of nineteenth-century modern American linguistics. His masterwork on gen- German scholars who established the genetic relation- eral linguistics, Language (1933), determined the ship among Indo-European languages. Cognates direction the scientific study of language in the United (shared vocabulary) are compared to establish the States would take for the next three decades. A prolif- sound changes marking the transition from a common ic scholar, Leonard Bloomfield’s publications cover an hypothetical ancestor to one or more daughter lan- impressive spectrum: general linguistics, Indo- guages. A comparison of the median consonants in Old European and Germanic philology, dialect geography, English /brothor/, Old Saxon /brothar/, and Old High linguistic descriptions of the Algonquian American- German /bruoder/, for example, would have led schol- Indian languages Menomini, Cree, Fox, and Ojibwa, ars to the conclusion that /th/ was present in proto- and the Austronesian languages Tagalog and Illocano. Germanic, the shared ancestral language, but changed Bloomfield’s work spawned a generation of American to /d/ in Old High German. The comparative method scholars, the neo-Bloomfieldians, who dominated the rests on the assumption that sound change affects a lan- field of linguistics before Noam Chomsky’s theory of guage without exception and is therefore regular, never transformational grammar became widely accepted in occurring sporadically in just a few words. Using the the early 1960s: Bernard Bloch, Robert A. Hall Jr., neo-grammarian principle as a working hypothesis, Zellig Harris, Einar Haugen, Charles F. Hockett, Bloomfield applied the comparative method to Cree, Martin Joos, William Moulton, Eugene Nida, and W. Fox, Menomini, and Ojibwa, languages some Freeman Twadell, to name but a few. European scholars of the time might have considered An adherent of positivism and behaviorist psychol- primitive compared to Sanskrit, Gothic, Latin, or ogy, Bloomfield rejected mentalism, which takes into Greek, and proved that they were descended from a account nonphysical and hence nonobservable thought common ancestor, proto-Algonquian. In spite of criti- processes, insisting that only observable manifesta- cism, especially from abroad, Bloomfield remained a tions of language could be the subject of linguistic confirmed neo-grammarian throughout his career. investigation. He defined the meaning of a linguistic Bloomfield’s masterwork Language introduced form as the situation in which the speaker utters it and innovative methods for descriptive synchronic linguis- the response that it calls forth in the hearer. The strong tics and a new terminology soon to be adopted by his rejection of the mentalist approach to the study of American colleagues. The first half of the book is a meaning became a defining characteristic of state-of-the-art presentation of the methods and princi- Bloomfieldian linguistics and its weakest point, as ples of American structuralism; the second presents the Noam Chomsky would point out several decades later. findings of European historical linguists and dialectol- A strong believer in the scientific method, Bloomfield ogists supplemented with Bloomfield’s own research argued that language should be studied like a natural materials. He introduces the American reader to the science. In A Set of Postulates for the Science of phonemic principle established by the Prague School Language (1925), he outlined the assumptions under- of Linguistics. Phonemics is the study of the distinctive lying structural linguistics in a concise and com- sounds of a language, those that are recognized by the pressed form. native speaker as signaling differences in meaning. The As a historian of language, Bloomfield studied his initial consonants of pin, kin, sin, tin are phonemes of subject from the diachronic perspective; as a descrip- English since their substitution results in a change of tive structural linguist, he concentrated on the syn- meaning. The different articulations of the phoneme /p/ chronic perspective. A diachronic approach traces the in pit (aspirated), spit (unaspirated), and tip (unre- changes a language undergoes from a given point in leased), on the other hand, are phonetic variants of the time to a later date; a synchronic approach describes a same phoneme, which can never serve to distinguish language at one particular time. This distinction had meaning since their phonetic shape depends on their not been made by earlier generations of scholars, and position in the word. The phonemic principle was the Bloomfield, like the French structuralist Ferdinand the basis for all work in descriptive synchronic linguistics Saussure, insisted that both perspectives be kept strict- by Bloomfield and his disciples. In Language, the syn- ly separate. The classic diachronic approach to linguis- chronic phonemic principle is, for the first time, equat- tic analysis, the comparative method, was pioneered by ed with the diachronic neo-grammarian principle of the 143 BLOOMFIELD, LEONARD regularity of sound change. American reviewers of Bloomfieldian structuralism with its focus on Language were generally enthusiastic, whereas phonemics and morphology remained the dominant European scholars tended to be more critical citing the approach to linguistic analysis until the publication of lack of abstract theory, rejection of mentalism, new ter- Noam Chomsky’s treatises on transformational gener- minology (sememe, taxeme, glosseme, etc.), and ative grammar in the late 1950s and early 1960s. After Bloomfield’s defense of the neo-grammarian principle. several decades of relative obscurity, Bloomfield’s One reviewer considered the work to be typically contribution to the field of linguistics is once again American: data-oriented, dryly factual, but carefully appreciated. Leonard Bloomfield suffered a debilitat- and systematically executed. In the United States, ing stroke in 1946, which abruptly terminated his bril- Language became the preferred textbook in introducto- liant career and led to his death three years later. ry linguistics courses for the next quarter of a century following publication. It should be recommended read- Biography ing for graduate students in linguistics interested in the history of their discipline. Leonard Bloomfield was born on April 1, 1887 in Bloomfield began working with non-European lan- Chicago, IL. The son of Austrian immigrants and the guages early in his career. Tagalog Texts with nephew of Maurice Bloomfield, a noted scholar of Grammatic Analysis (1917) filled a void in a much neg- Indo-European philology and Indic studies, Leonard lected field. Based on the data provided by a single Bloomfield grew up in Chicago and rural Wisconsin. informant who dictated texts, it is still considered to be Fluent in both English and German, he quite natu- one of the best sources of information on any rally gravitated toward the study of Germanic and Austronesian language. While the Algonquian lan- Indo-European philology. After completing his under- guages had been studied systematically for three cen- graduate degree at Harvard in 1906, he began graduate turies, Bloomfield’s contribution was the systematic studies in German at the University of Wisconsin synchronic description of four languages, Cree, Fox, (1906–1908) and then transferred to the University of Menomini, and Ojibwa, and the application of the Chicago where he received his Ph.D. degree in 1909 diachronic comparative method to the reconstruction of with a thesis in historical Germanic linguistics (‘A the hypothetical ancestor, proto-Algonquian. semasiological differentiation in Germanic secondary Bloomfield provided a grammatical outline of each of Ablaut’). After several years of working as an instruc- the four living languages as well as of proto- tor of German at the Universities of Cincinnati (1909) Algonquian, and traced the historic development of and Illinois Champaign-Urbana (1910–1913), he spent each language, thus demonstrating the validity of the a year, between 1913 and 1914, of advanced study in comparative method in the process. He also produced Indo-European and Germanic philology at the German the first systematic description of the Algonquian mor- Universities of Leipzig and Göttingen, where he phology (the study of inflection and word formation), worked with Karl Brugner, August Leskien, and Jakob and thereby established the framework for all later stud- Wackernagel, some of the most highly regarded his- ies on the subject. Bloomfield’s field methods and torical linguists of the time. By 1914, he was Assistant methods of linguistic analysis have stood the test of Professor of comparative philology and German at the time, and are still used by modern scholars recording University of Illinois. For the rest of his career, unwritten languages or dialects. The bulk of his Bloomfield remained an adherent of their particular Algonquian material was published posthumously by view of language change. One of the founding mem- Charles F. Hockett, his literary executor. bers of the Linguistic Society of America (1924) with During World WarII, Bloomfield, like many of his George M. Bolling and Edgar H. Sturtevant, colleagues and students, devoted much of his time to Bloomfield held the chair of Germanic Philology at the development of a new foreign language teaching the University of Chicago from 1927 until 1940 when methodology for the War Department. The audiolin- he accepted a Sterling professorship in Germanic gual method focused on the spoken, colloquial form of Languages at Yale University. He became president of the language rather than the literary standard and the Linguistic Society of America in 1936. Bloomfield introduced practice sessions where the material was suffered a debilitating stroke in May 1946, which transmitted to the learner on the basis of inductive, ended his career. He died on April 18, 1949. behaviorist learning principles (oral repetition and memorization). By the 1960s, it had become the References almost universally accepted foreign language teaching Bloch, Bernard. 1949. Leonard Bloomfield. Language 25. methodology in the United States and to some extent 87–98. in Europe. Bloomfield authored and coauthored text Bloomfield, Leonard. 1917. Tagalog texts with grammatical books for elementary Dutch and Russian (1944/1945). analysis, 3 vols. Urbana: University of Illinois. 144  BOAS, FRANZ ———.1925. Why a linguistic society? Language 1. 1–5. Goddard, Ives. 1987. Leonard Bloomfield’s descriptive and ———.1926. A set of postulates for the science of language. comparative studies of Algonquian. In Hall. 179–217. Language 2. 153–64. Hall, Robert A. Jr. 1987. Leonard Bloomfield: essays on his life ———.1927. The word stems of Central Algonquian. and work. Philadelphia: John Benjamins. Festschrift Meinhoff. Hamburg: L. Friedrichsen. 303–403. Hall, Robert A. Jr. 1990. A life for language. A bibliographical ———.1928. Menomini texts. Publication of the American memoir of Leonard Bloomfield. Amsterdam Studies in the Ethnological Society, Vol. 12. New York: G.E. Stechert. Theory and History of Linguistic Science, Vol. 55. ———.1930. Linguistics as a science. Studies in Philology 27. Amsterdam/Philadelphia: Benjamins. 553–7. Hall, Robert A. Jr. 1987. Bloomfield and semantics. In Hall. ———.1933. Language. N Y: Holt, Rinehart and Winston. 155–60. ———.1942. Outline guide for the practical study of foreign Hockett, Charles F., and Robert A. Hall Jr. 1987. A new languages. Baltimore: Linguistic Society of America, Leonard Bloomfield bibliography. In Hall. 221–3. Special Publication. Hoenigswald, Henry M. 1987. Bloomfield and historical lin- ———.1942. Outline of Ilocano syntax. Language 18. guistics. In Hall. 73–88. 193–200. Lehmann, Winfried P. 1987. Bloomfield as an Indo- ———.1945. About foreign language teaching. Yale Review Europeanist. In Hall. 163–72. 34. 625–41. Moulton, William G. 1987. Bloomfield as a dialectologist. In ———.1962. The Menomini language. New Haven: Yale Hall. 139–54. University Press. Wolff, John U. 1987. Bloomfield as an Austronesianist. In Hall. ———.1970. A Leonard Bloomfield anthology, ed. by Charles 173–8. F. Hockett. Bloomington: Indiana University Press. RENATE BORN Boas, Franz Central to Franz Boas’ anthropology, with its subfields languages of this region resulted in grammars of of prehistoric archeology, physical anthropology, eth- Tsimshian, Chinook (1911a), Tlingit (1917a), and nology, and linguistics, was the idea of culture as a Kwakiutl (1911a, 1947), and collections of native lan- symbolic system or construct of the mind. guage texts. Boas further noted that some linguistic Accordingly, Boas focused on ethnology and linguis- characteristics were distributed geographically among tics while he paid less attention to prehistoric archeol- these languages, an observation that would eventually ogy and physical anthropology. Language had a lead to the establishment of areal-typological studies. special place in Boas’ work because it represented the Boas’ ideas on linguistic categorization were rooted prime example of a human symbolic system and was in nineteenth century German acoustics, psy- the best means for classifying the indigenous cultures chophysics, and psychology. His exposure to so-called of North America. Boas was primarily interested in the ‘alternating sounds’, speech sounds that were seeming- Native languages of the American Northwest. While ly vague, during his fieldwork in British Columbia was he was concerned with the genetic classification of the reason for a seminal paper, arguing that the vague- native languages, the main thrust of his work was ness of these sounds was a result of different catego- directed toward investigating categorization in lan- rizations of speech sounds in distinct languages and of guages and establishing a typology of the inner forms their varying apperception by speakers of other lan- or systems of unconscious grammatical categories of guages (1889). Eventually, Boas was to argue that lin- Native American languages. guistic categorization, in general, was the outcome of His early fieldwork for the British Association for grouping experiences into classes according to their the Advancement of Science under the supervision of similarity; however, the criteria underlying the similar- Horatio Hale (1817–1896) led to grammatical sketch- ity and class membership of experiences were not the es, vocabularies, and linguistic maps of the languages same in all languages. In English, for example, the of British Columbia. Boas’ lexical and grammatical existence of the words to grip, to kick, to pound, and to data of Nootka and Kwakiutl allowed him to unite the bite indicated that the different activities they designat- two languages as the Wakashan family (1890, 1891a, ed were classified as separate concepts. By contrast, in b). Based on grammatical evidence only, he also made Dakota the form xtaka ‘to grip’, which was present in the controversial proposal that Tlingit and Haida were the terms naxta’ka ‘to kick’, boxta’ka ‘to pound’, and related (1894). Boas’ lifelong preoccupation with the yaxta’ka ‘to bite’ suggested that the same activities 145 BOAS, FRANZ were classed under one concept only (1911a). Because gorization responsible for producing distinct phonetic, linguistic categories did not rise to the level of con- grammatical, and lexical systems characterizing dif- sciousness, they provided a unique view on the mental ferent languages. The Handbook grammars were to processes of abstraction and association. illustrate the ‘psychological principles’ underlying Influenced by the work of Heyman Steinthal Native American languages and to form the basis for (1823–1899), Boas suggested that the inner forms of establishing a typology of inner forms (1911:v). Boas languages molded their speakers’ thoughts. The uncon- provided an overview of the grammatical characteris- scious nature of these obligatory categories resulted in tics of Native American Languages and the principles their projection onto reality and their objectification. of language classification and reduced John W. Late in life Boas conceded, however, that although Powell’s (1834–1902) 58 Native American language thought was ‘directed in various channels’ by grammat- families to 55. ical categories, their influence was not to be overrated Early in his career, Boas (1894) had given more (1942:183). His linguistic relativism was further coun- weight to grammatical similarities in determining rela- terbalanced by the observation that all languages pos- tionships between languages. When he wrote his sessed the pronouns I, thou, and he and by his introduction to the Handbook (1911a), he had moved recognition of universal functional relations, such as to a more balanced position that equally stressed pho- subject–predicate, noun–attribute, and verb–adverb netic, lexical, and grammatical similarities in classify- (1901, 1838). Boas’ views suggest that he considered ing languages into families. Simultaneously, Boas the universal occurrence of these pronouns and func- increasingly emphasized an alternative to the common tional relations as examples of what Adolf Bastian origin of seemingly related languages: multiple origins (1826–1905) had termed Elementargedanken, elemen- resulting from borrowing of linguistic elements and tary ideas that necessarily occurred everywhere because language mixture (1911a, 1917, 1920, 1929). Boas’ of the constitution of the human mind. Nevertheless, late papers on the influence of Spanish on Nahuatl these ideas were not identical in different contexts but (1930) and of English on Kwakiutl (1932) document- were shaped by different psychological factors and ed instances of lexical and morphological intermixture environmental circumstances. One important task of between languages. anthropology was the identification of these ideas. Like Steinthal, Boas therefore objected to the use of Biography grammatical categories derived from Indo-European languages in describing Native American languages. Franz Boas was born in Minden, Westfalia, Germany Such a practice distorted the nature of these languages on July 9, 1858. He received a doctorate (1881) from and covered up their ‘characteristic psychological cat- the University of Kiel, Germany for his dissertation on egories’ (1917b:5–6). Only analytical grammars, the perception of the color of water. He undertook an which treated languages on their own terms, were expedition to Baffin Island in 1883–1884; habilitation acceptable; only they would provide access to the (1886): Berlin University, Germany with a thesis on the inner form and ‘psychological groupings’ of a lan- geographical results of his Baffin Island trip. He moved guage (1911a:81). These grammars formed the basis to the United States in 1886, and became a US citizen for tracing the ‘the history of the mental development in 1891. He was geographical editor of Science in of various branches of mankind’ and for identifying 1887–1888. He carried out intermittent fieldwork for grammatical concepts shared by the languages of the the British Association for the Advancement of Science world, albeit in modified form (1911a:71). (BAAS) in British Columbia in 1888–1894. He held a The Handbook of American Indian Languages docentship in physical anthropology, Clark University (1911–1941) was a monument to Boas’ ideas and in 1889–1892. He was assistant to F. W. Putnam, Head scholarship. He not only edited the four volumes com- of the Department of Ethnology and Archaeology at the prising the Handbook but also contributed extensively World Columbian Expedition in 1893–1894. He was to the first volume (1911a). Boas wrote the introduc- assistant curator, American Museum of Natural History tion and the grammars of Tsimshian, Kwakiutl, and in 1895–1905. In the field, he was a member of the Chinook and coauthored the grammar of Siouan (Boas Jesup Expedition in 1897, and 1900. He began lectur- and Swanton 1911). His introduction outlined his ing at Columbia University in 1896, and received full ideas on language and discussed its place within his Professorship in 1899–1936. He was a member of the anthropological framework. He showed that there was editorial board and publication committee of the no necessary correlation between culture, language, American Anthropologist, in 1899–1920; editor of the and race, but that the three were independent of each Publications of the Jesup North Pacific Expedition, in other. Boas highlighted at length the process of cate- 1900–1930; editor of Journal of American Folklore in 146  BOAS, FRANZ 1908–1925; founder and editor of the International ———.1917a. Grammatical notes on the language of the Journal of American Linguistics in 1917–1939; editor Tlingit Indians. The University Museum Anthropological Publications 8. of the Columbia University Contributions to ———.1917b. Introductory. International Journal of American Anthropology in 1913–1936; and editor of the Linguistics 1. Handbook of American Indian Languages in ———.1920. The classification of American languages. 1911–1941. He was Vice-President of Section H. of the American Anthropologist n.s. 22. American Association for the Advancement of Science ———.1929. Classification of American Indian languages. Language 5. (AAAS) in 1894; President of the American Folklore ———.1930. Spanish elements in Nahuatl, Todd Memorial Society in 1900–1901; Honorary Philologist, Bureau of Volumes Philological Studies, Vol. 1, ed. by John D. Fitz- American Ethnology in 1901–1920; President of the Gerald and Pauline Taylor. New York: Columbia University American Anthropological Association in 1907 and Press. 1908; Vice-President and Chairman of Section H. of ———.1932. Notes on some recent changes in the Kwakiutl language. International Journal of American Linguistics 7. the AAAS in 1909; President of the 23rd International ———.(ed.) 1938. Language. General Anthropology. Boston: Congress of Americanists in 1928; President of the D.C. Heath. Linguistic Society of America (LSA) in 1928–1929; ———.1942. Language and culture. Studies in the history of Chairman of the Committee of the American Council culture: the disciplines of the humanities, ed. by P. Long. of Learned Societies on Research in American Indian Freeport, NY: Books for Libraries Press. ———.1947. Kwakiutl grammar. Transactions of the Languages; Honorary President of the 24th American Philosophical Society 37, ed. by H. Boas International Congress of Americanists in 1930; and Yampolsky, Z.S. Harris and John Swanton; 1911a. Siouan. President of the American Association for the In Boaz. Advancement of Science in 1931. Franz Boas died in Cole, Douglas. 1999. Franz Boas: the early years 1858–1906. New York, NY, USA on December 21, 1942. Vancouver/Toronto: Douglas and Mcintyre; Seattle and London: University of Washington Press. Darnell, Regna. 1998. And along came Boas: continuity and References revolution in Americanist anthropology. Amsterdam and Philadelphia: John Benjamins. Boaz, Franz. 1911a. Handbook of American Indian languages, Goldschmidt, W. (ed.) 1959. The anthropology of Franz Boas. Part I, Bureau of American Ethnology, Bulletin 40, Essays on the centennial of his birth. Memoirs of the Washington, DC: Government Printing Office. American Anthropological Association 89. Boaz, Franz. 1889. On alternating sounds. American Gruber, J. W. 1967. Horatio Hale and the development of Anthropologist 2. American anthropology. Proceedings of the American ———.1890. First general report on the Indians of British Philosophical Society 111. Columbia. Report of the 59th Meeting of the British Mackert, Michael. 1993. The roots of Franz Boas view of lin- Association for the Advancement of Science, 1889. London: guistic categories as a window to the human mind. Murray. Historiographia Linguistica 20. ———.1891a. Second general report on the Indians of British Mackert, Michael. 1994. Franz Boas’ theory of phonetics. Columbia. Report of the 60th Meeting of the British Historiographia Linguistica 21. Association for the Advancement of Science, 1890. London: Mackert, Michael. 1995. The Hale–Boas debate on the origin Murray. and classification of languages. History of Linguistics 1993, ———.1891b. Vocabularies of the Tlingit, Haida, and ed. by Kurt R. Jankowsky. Amsterdam and Philadelphia: Tsimshian languages. Proceedings of the American John Benjamins. Philosophical Society 29. Mackert, Michael. 1999. Franz Boas’ early Northwest coast ———.1894. Classification of the languages of the North alphabet. Historiographia Linguistica 26. Pacific Coast. Memoirs of the International Congress of Rohner, Ronald. 1966. Franz Boas ethnographer on the Anthropology, Chicago 1893, ed. by C.S.Wake. Chicago: Northwest Coast. Pioneers of American Anthropology: The Schulte. Uses of Biography, ed. by June Helm. Seattle and London: ———.1901. The mind of primitive man. Journal of American University of Washington Press. Folklore 14. Stocking, George W. 1974. The Boas plan for the study of ———.1909. Notes on the Iroquois language. Putnam American Indian languages. Studies in the History of Anniversary Volume. New York: G. E. Stechert. Linguistics, ed. by Dell Hymes. Bloomington: Indiana ———.(ed.) 1911a, 1922, 1941. Handbook of American Indian University Press. languages, Parts 1–2, Bureau of American Ethnology, Stocking, George W. (ed.) 1996. History of anthropology. Vol. Bulletin 40, Washington, DC: Government Printing Office; 8: Volksgeist as method and ethic: essays on Boasian Part 3, Glückstadt–Hamburg–New York: J.J. Augustin, ethnography and the German anthropological tradition. 1933–1938; Part 4, New York: J.J. Augustin. Madison: University of Wisconsin Press. ———.(ed.) 1911a. Introduction. In Boaz. Williams, Vernon J. 1996. Rethinking race: Franz Boas and ———.(ed.) 1911a. Tsimshian. In Boaz. his contemporaries. Lexington: University of Kentucky ———.(ed.) 1911a. Kwakiutl. In Boaz. Press. ———.(ed.) 1911a. Chinook. In Boaz. ———.1911b. The mind of primitive man. New York: MICHAEL MACKERT MacMillan. See also Chinook Jargon 147 BOPP, FRANZ Body Language See Paralanguage Bopp, Franz Franz Bopp was one of the founders of nineteenth- tives was to investigate the oldest accessible form of a century comparative philology. He is often given cred- given language. it for establishing historical linguistics as a serious Bopp wrote many monographs on individual lan- area of scientific study. He was in a group of pioneer- guages (e.g. Celtic, Old Prussian, and Albanian) and ing German scholars (such as the brothers August on languages outside the Indo-European family (e.g. a Wilhelm and Friedrich von Schlegel and Wilhelm von paper on the relationship between Indo-European and Humboldt) that started their study of Sanskrit and Malayo-Polynesian). His investigation and systematic Indo-European at the University of Paris and, later on, comparison of a number of languages was a tremen- made it a German monopoly. dous refinement of descriptive techniques used by lin- Bopp’s chief activity centered on the study of the guists at that time. languages of the East. His knowledge of the literature In his most important work, Vergleichende written in Sanskrit and the language itself was truly Grammatik des Sanskrit, Zend, Griechischen, expert. Bopp was keenly interested in the great Indian Lateinischen, Litauischen, Altslavischen, Gothischen epics. He published several episodes from the und Deutschen (1833–1852 ; A Comparative Grammar MahâBhârata, both in Sanskrit and in translated form. of the Sanskrit, Zend, Greek, Latin, Lithuanian, Gothic, He also compiled a Sanskrit and Latin glossary. Many German, and Slavonic Languages), Bopp discussed and European researchers (e.g. William Carey or Henry established relationships between the Indo-European Thomas Colebrooke) had already proposed their own languages. In principle, a resemblance between related grammars of Sanskrit, based on the Indian tradition languages could be explained in two ways: either by that stretched back to Panini. Such grammars were claiming that one of them is derived from the other or obscure and unhelpful from the point of view of schol- by assuming that both are derived from another lan- ars who were used to the Western model of paradigms. guage. Contrary to many of his contemporaries, Bopp In his very influential Sanskrit grammar, Bopp aban- did not regard languages such as Greek or Latin as doned much of the original apparatus of the Indian derived from Sanskrit. He describes Indo-European lan- system of grammatical description. Instead, he adopt- guages as variations of one original language. However, ed the European framework that was grounded in the Sanskrit has preserved more features of the common traditional analyses of Greek and Latin. source than other cognate languages. Bopp’s Sanskrit studies were especially important Since the publication of his first major work Über as the requisite stimulus for great developments in the das Conjugationssystem der Sanskritsprache in field of comparative grammar. Bopp traced many sim- Vergleichung mit jenem der griechischen, lateinischen, ilarities between Sanskrit and its Western cognate lan- persischen und germanischen Sprache (1816; On the guages, both in terms of lexical items (roots) and on System of Conjugation in Sanskrit, compared with the the syntactic level. He denied the importance of sur- Greek, Latin, Persian, and German Languages), Bopp’s face differences between grammars, claiming that main interest was focused on detecting the common ori- they do not matter as long as it is possible to recon- gin of Indo-European inflections. Primarily, he applied struct the way languages have developed from the his analysis only to the verb. However, while in London, common source. Therefore, one of his prime objec- he published a very influential essay in the Annals of 148  BOPP, FRANZ Oriental Literature (entitled Analytical comparison of (many of them were reissued with important improve- the Sanskrit, Greek, Latin and Teutonic languages, ments). He wanted his conclusions to be as objective shewing the original identity of their grammatical struc- and nonspeculative as possible. Therefore, he was (and ture), in which he included analysis of nonverbal inflec- is) sometimes called ‘the factual Bopp’. tion. This essay received considerable public attention, which resulted in Bopp’s appointment to the newly Biography established chair of Oriental Literature and General Philology at the University of Berlin in 1821 (an addi- Franz Bopp was born in Mainz, Germany on tional factor was Bopp’s friendship with Wilhelm von September 14, 1791. He attended lectures in law, logic, Humboldt, who was at that time the Prussian Minister esthetics, history, and philosophy at the Aschaffenburg of Worship and Public Instruction). Ecole du Droit, Bavaria, Germany. He moved to Paris, Rejecting Friedrich Schlegel’s ideas, Bopp advocat- France, to study Sanskrit (together with August ed an agglutinative theory of the origin of morpholog- Wilhelm von Schlegel and Wilhelm von Humboldt, ical structure. He assumed that grammatical tutored by Antoine Léonard de Chézy) in 1812–1817, modifications can be expressed only by external addi- resided in London, England in 1817, and continued his tions to the root (i.e. by means of agglutination of sep- studies in Göttingen, Germany in 1818–1821. He was arate lexical items which become suffixes). According Extraordinary Professor of Oriental Literature and to Bopp, what Schlegel understood as ‘internal’ inflex- General Philology, University of Berlin, Germany in ion (a change of root vowels or reduplication) is an 1821 and Ordinary Professor of Oriental Literature and example of redundancy in Indo-European—it does not General Philology, University of Berlin, Germany in indicate any modification of meaning. Phenomena 1825 (he held this post till death). He was also a mem- such as apophony cannot be semantically relevant ber of the Royal Prussian Academy in 1822 and hon- since they are not separate elements. Bopp analyzed orary member of the Philological Society. Bopp died in roots as the primitive (monosyllabic) elements of Berlin on October 23, 1867. words that can be identified through examination of forms that contain them as a common base. There are References two types of roots: verbal (predicative) and pronomi- Bopp, Franz. 1816. Über das Conjugationssystem der nal (indicative). A root must be considered to be an Sanskritsprache in Vergleichung mit jenem der griechis- abstract entity, which means it does not have to be chen, lateinischen, persischen und germanischen Sprache. itself equivalent to any existing word. ———. 1820. Analytical comparison of the Sanskrit, Greek, Bopp distinguished those aspects of languages that Latin and Teutonic languages, showing the original identity can be formally described from those that are arbitrary. of their grammatical structure. Annals of Oriental Literature 1; reprinted as Analytical comparison of the Sanskrit, Greek, The former comprise what he called ‘mechanical’ (i.e. Latin and Teutonic languages, showing the original identity phonetic) laws and the origin of the forms indicating of their grammatical structure. Newly edited, together with grammatical relationships (morphology). On the other a bio-bibliographical account of Bopp by J. D. Guigniaut, an hand, in his Vergleichende Grammatik, Bopp refers to introduction to Analytical comparison by F. Techmer and a the relationship between form and meaning as the letter by W. von Humboldt, 1974. ———. 1827. Ausführliches Lehrgebäude der Sanskrit-Sprache. ‘secret of roots’—i.e. he notices that it is impossible to ———. 1829–1832. Glossarium sanscritum; new edition say why a given root means what it means, e.g. why Glossarium sanscritum, in quo omnes radices et vocabula the Indo-European root STHA/STA means ‘stand’ and usitatissima explicantur et cum vocabulis graeis, 1840–1847. not ‘go’. ———. 1833–1852. Vergleichende Grammatik des Sanskrit, Nineteenth century philologists tended to use bio- Zend, Griechischen, Lateinischen, Litauischen, Altslavischen, Gothischen und Deutschen; as A comparative logical terminology in describing the way that lan- grammar of the Sanskrit, Zend, Greek, Latin, Lithuanian, guages work. Being influenced by contemporary Gothic, German, and Slavonic languages, transl. by Edward developments of comparative anatomy, they viewed Backhouse Eastwick, 1856; revised edition Vergleichende languages as changing and organic. It could be said Grammatik des Sanskrit, Zend, Armenischen, Griechischen, that Bopp was no exception. He also tried to dissect Lateinischen, Litauischen, Altslavischen, Gothischen und Deutschen, 1857–1861. and classify languages like living organisms. ———. 1834. Kritische Grammatik der Sanskrita-Sprache in However, he examined them as static entities; the his- kürzerer Fassung. torical process of change did not interest him as much ———. 1841. Über die Verwandtschaft der malayisch-polyne- as the structure that resulted from it. sischen Sprachen mit den indisch-europäischen. Bopp devoted his life to laborious, quiet study in Lefmann, Salomon. 1891–1895. Franz Bopp, sein Leben und seine Wissenschaft, 2 vols. Berlin: G. Reimer. libraries, remaining aloof from any political or social Lehmann, Winfred P. 1991. Franz Bopp’s use of typology. events. He was a zealous researcher, very cautious and Zeitschrift für Phonetik, Sprachwissenschaft und careful, constantly maturing and correcting his works Kommunikationsforschung 44(3). 149 BOPP, FRANZ Lehmann, Winfred P. 1996. Theoretical bases of Indo-European tungen zum Boppschen Sprachvergleich aus Anlass linguistics. London and New York: Routledge. irriger Interpretationen in der linguistischen Literatur. Leonhard, Tafel. 1861. A review of some points in Bopp’s com- Innsbruck: Institut für Sprachwissenschaft der Universität parative grammar. Andover, MA: W.F. Draper. Innsbruck. Leskien, August. 1876. Bopp. Allgemeine deutsche Biographie Sternemann, Reinhard (ed.) 1994. Bopp-Symposium 1992 der 3. In Seboek. Humboldt-Universität zu Berlin: Akten der Konferenz vom Martineau, Russell. 1867. Obituary of Franz Bopp. Transactions 24.3-26.3.1992 aus Anlass von Franz Bopps zweihun- of the Philological Society 12, reprinted as Franz Bopp. In dertjährigem Geburtstag am 14.9.1991. Heidelberg: Seboek. Universitätverlag C. Winter. Seboek, Thomas A. (ed.) 1966. Portriats of linguists: a biog- Verburg, Pieter A. 1950. The background to the linguistic con- raphical source book for the history of Western linguists, ceptions of Franz Bopp. Lingua 2. 1746–1963, Vol. 1. From Sir William Jones to Karl PAWEA RUTKOWSKI Bruymann. Bloomington: Indiana University Press. Sternemann, Reinhard. 1984. Franz Bopp und die vergle- See also Indo-European 1: Overview; Panini; ichende indoeuropäische Sprachwissenschaft: Beobach- Philology; Sanskrit Brain Organization and Auditory Pathway Three major functions of the brain are the integration of adaptive behavior, and cognition. To reach its full sensory input, behavior in general, and coordination of functional state, the newborn’s brain has to ‘mature’ body functions. The development of the brain depends for some additional months, during which period the upon both the multicellular structure of creatures and neural networks for higher sensory and basic cognitive the necessity to control complex behavior of highly processes have to establish themselves. Beginning developed organisms. Before multicellular organisms already in the uterus, where the unborn child learns evolved, a unicellular organism (Protozoa) already rep- parts of the mother’s phoneme set, the child’s brain resented an omnipotent life form, which performed e.g. needs about two additional years to produce language movement, ingestion, and orientation with merely one after birth. This shows the spectrum of the different cell. After the evolutionary emergence of multicellular brain capabilities from controlling body functions to organisms (Metazoa), certain cells of these organisms cognitive processes. From an evolutionary point of adapted functionally and became e.g. secretory, sup- view, the brain consists of three overlaying brain parts: porting, sensory, or muscle cells. As a result, during this (1) the approximately 250-million-year-old ‘reptile developmental stage an increasing number of special- brain’, performing coordination of body functions and ized body cells had to coordinate with each other; simple sensory integration, (2) the ‘older mammalian hence, a communication network became necessary. brain’ performing e.g. complex integration of sensory Communication took place by an additional spe- input and emotions, and (3) the ‘only’ 50-million-year- cialized cell type—the nerve cell. Starting from the old ‘young mammalian brain’ (neocortex), the neu- stage of simple nerve nets (e.g. in jellyfish), the evolu- ronal substrate for higher cognitive processes. Even tion of nervous systems in animals led to small groups though these brain parts form a functional unit and act of nerve cells (ganglia), peripheral nervous systems, as a single organ, these three phylogenetic stages can and central nervous systems (brains). The human cen- be assigned to different levels of cognitive perform- tral nervous system represents the most complex inte- ance. Language corresponds to the most developed gration/coordination system: it comprises about 100 cognitive process and is only developed in humans. billion (1011) nerve cells, each connected, on average, Due to its importance, the human central nervous to about 1,000 others. During the nine months of preg- system, which consists of the brain and the spinal cord, nancy, the brain of a human fetus has to grow on aver- is protected by three membranous coverings (meninges) age by 250,000 new nerve cells per minute in order to and by bony capsules (cranium, backbone). Three main reach its final dimension (on average 1,400 gram). parts of the brain are shown in a side view (see Figure However, after birth the brain mainly controls the 1): the brain stem (B), the cerebellum (C), and the left basic body functions. The adult brain is not only cerebral hemisphere (H). The two cerebral hemispheres involved in the coordination of organs, body functions, represent 85% of the brain’s weight and are surrounded or parts but also in integrating sensory input, learning, by the 1.3 to 4.5 mm thick cerebral cortex. The cortex 150  BRAIN ORGANIZATION AND AUDITORY PATHWAY Telencephalon Corpus callosum Diencephalon Mesencephalon Cerebellum H Basal pons Cranial nerve VIII Medulla oblongata Primary Medulla spinalis auditory area B C Figure 1. Brain Organization and Auditory Pathway. consists of specially organized nerve and glial cells and the medulla, near the inferior border of the pons, into is compactly folded, which causes its surface enlarge- which the nerve inserts (see Figure 1). ment, although the volume remains unchanged. The At first, the sound signal is gathered and filtered by cortex shows an abundant structure of ridges (gyri) and the external ear. After mechanical amplification in the valleys (sulci) and represents the main neuronal struc- middle ear, the signal is transduced into an electrical ture for higher cognitive processes. Each hemisphere is signal by sensory cells (organ of Corti) in the inner ear. divided into four lobes: the frontal, the temporal, the The organ of Corti of each ear contains about 3,500 parietal, and the occipital lobe. Anatomically, the brain inner hair cells and about 12,000 outer hair cells, is divided into five parts, which are shown in Figure 1: which are innervated by 30–40,000 fibers of the the myelencephalon (medulla oblongata), the meten- cochlear branch of the VIII cranial nerve (auditory cephalon (pons, cerebellum), the mesencephalon, the nerve). About 90% of the fibers are connected to inner diencephalon, and the telencephalon (cerebral hemi- hair cells, and these mainly contribute to the informa- spheres). A major structure for transferring information tion on sound signals for subsequent analysis in the between both hemispheres is the corpus callosum, a brain. The analysis of the sound signal starts already in broad nerve tract that also belongs to the telencephalon. the cochlea, e.g. by separating the signal into the fre- One of the brain’s major function lies in the integra- quency domain (frequency dispersion), which will not tion of sensory information from the auditory channel. be discussed here in detail. The major part of the For primates, the auditory input can be seen as the most analysis (feature detection) is carried out during the important input from the environment. Like every other next stations of the projection to the auditory cortices. audible sound, language is perceived by the ears and The detailed function of the auditory cortex is still not transferred to certain cortical regions across the so- understood—however, it is the final place of sound called auditory pathway. The auditory pathway lead recognition, especially for speech analysis (e.g. from the nerve fibers of the cochlear branch of the VIII phoneme detection). Starting at the cochlea, at each cranial nerve (vestibulocochlear nerve) via the primary level of the auditory pathway, the neurons are arranged auditory nuclei of the medulla, to several neural centers in a well-ordered map that reflects the frequencies of of the brain stem, up to the primary and the secondary the stimuli (tonotopic organization). Consequently, auditory cortex. Most of the structures analyzing the neurons responsive to low and high frequencies are auditory signal are located in the brain stem, which located at a maximum distance from each other. comprises the medulla, the pons, and the mesen- While each hemisphere receives sensory input and cephalon. Since the internal ear lies close to the brain controls motor movement of the opposite (contralater- stem, the auditory nerve is only about 5 mm away from al) side of the body, the information of each ear goes 151 BRAIN ORGANIZATION AND AUDITORY PATHWAY to both hemispheres, whereby the majority crosses to tory system exhibits increased stability against dam- the contralateral side. First, all afferent fibers of the age to lower neural structures. The next station of the auditory nerve terminate in the ipsilateral cochlear pathway, the superior olivary complex, already nucleus, whereas during projection a major part of the receives bilateral input from both cochlear nuclei. The afferences change to the contralateral side of the brain main processing stations in the superior olivary com- (indicated by line thickness in Figure 2). Due to this plex are the medial superior olive, the lateral superior crossing of fibers at several stations and the presence olive, and the medial nucleus of the trapezoid body. of decreasing fibers and parallel processing, the audi- The next major station of the auditory pathway is the Region Function Time Left side Right side (examples) (approx.) 0 ms Gathering sound waves External ear Passive amplification Sensory cells Middle ear and cochlea Middle ear and cochlea Active amplification Sound transduction Frequency dispersion Cranial nerve VIII Pars cochlearis Pars cochlearis 2 ms Basic feature Nucleus cochlearis Nucleus cochlearis detection Medulla oblongata Vocalization Oliva superior Oliva superior 4 ms Sound localization Basal pons Lat. lemniscus Lat. lemniscus Mesencephalon Tectum (inf. colliculus) Tectum (inf. colliculus) Complex feature detection Language Medical Medical 8 ms geniculate body geniculate body Diencephalon Thalamus, Limbic system Basic emotional analysis Left primary Right primary Basic phoneme auditory cortex 10 ms auditory cortex analysis Corpus callosum Areas of the left secondary Areas of the right secondary auditory cortex auditory cortex Telencephalon Left temporo-parietal cortex Right temporo-parietal cortex Phonematic analysis Limited semantic analysis of Lexicon of categorial word concrete nomen Hemispheric meaning Prosodic analysis specific Semantic analysis analyses Syntactic analysis Right frontal cortex (see boxes) Metaphoric analysis Left frontal cortex Episodic memory Lexicon of word form 10 to several Working memory hundred ms Semantic memory Working memory Syntactic analysis Simplified illustration of the auditory pathway of the left ear and some examples of known processing stages for spoken language comprehension. The identical, mirrored pathway of the right ear is not shown. Figure 2. Brain Organization and Auditory Pathway. 152  BRESNAN, JOAN lateral lemniscus, which is divided into a ventral and a C. Landis, James L. Roberts and Larry R. Squire. San dorsal nucleus. The following two major stations are Diego: Academic Press. Dearmond, Stephen J., Madeline M. Fusco, and Maynard M. the inferior colliculus and the medial geniculate body. Dewey. 1989. Structure of the human brain: a photographic The primary auditory cortex, which is identical to atlas, 3rd edition. Oxford: Oxford University Press. Brodmans area 41, comprises only a small area of the Durrant, John D., and Jean H. Lovrinic. 1995. Bases of hearing superior surface of the superior temporal gyrus science, 3rd edition. Baltimore: Williams & Wilkins. (Heschl’s gyrus) and is the principal target of all Gazzaniga, Michael S. (ed.) 2000. The new cognitive neuro- sciences. Cambridge, MA: MIT Press. ascending auditory pathways (see Figure 2). Gummer, Anthony W., and Hans-Peter Zenner. 1996. Central In most people, the left hemisphere plays a more processing of auditory information. Comprehensive human important role than the right hemisphere with regard to physiology, Vol. 1, ed. by Rainer Greger and Uwe language processing. Due to the multiple crossing of Windhorst. Berlin: Springer. the signals of each ear, damage in the lower auditory Kandel, Eric R., James H. Schwartz, and Thomas M. Jessel. 1991. Principles of neural science, 3rd edition. New York: nucleus of one hemisphere does not lead to a unilater- Elsevier. al, complete loss of acoustic processing, but results in Moore, Brain C.J. (ed.) 1995. Hearing 2nd edition. San Diego: a bilateral decreased hearing ability. However, substan- Academic Press. tial left hemispheric brain lesions of language-related Nolte, John. 1999. The human brain: an introduction to its func- cortical areas do not reduce the hearing ability, but only tional anatomy, 4th edition. St. Louis: Mosby. Purves, Dale, George J. Augustine, David Fitzpatrick, lead to language-specific impairments (see chapters Lawrence C. Katz, Anthony-Samuel LaMantia, and James ‘Lateralization and Handedness’ and ‘Aphasia’). O. McNamara (eds.) 1997. Neuroscience. Sunderland: Sinauer Associates. Webster, William R., and Laurence J. Garey. 1990. Auditory References system. The human nervous system, ed. by George Paxinos. Aitkin, Lindsay. 1990. The auditory cortex: structural and func- San Diego: Academic Press. tional bases of auditory perception. London: Chapman & Hall. Zemlin, Willard R. 1998. Speech and hearing science: anatomy Brown, M. Christian. 1999. Audition. Fundamental neuro- and physiology, 4th edition. Boston: Allyn &Bacon. science, ed. by Michael J. Zigmond, Floyd E. Bloom, Story HORST M. MÜLLER Bresnan, Joan Few students enter Joan Bresnan’s classes in syntactic approaches to the analysis of syntax were necessary to theory and leave without being fascinated about the address the problems of the transformational approach. subject matter of syntax. She is one of the most influ- It was at this point that Joan Bresnan drew attention ential teachers and researchers in the field of syntax in to the role of the lexicon in syntactic analysis. The lex- the late twentieth and early twenty-first centuries. Joan icon in early days of transformational grammar played Bresnan entered graduate school at the time when only a derivative role in syntactic analysis as it was transformational approaches to syntax, as developed regarded as having too many lexical idiosyncracies in mainly by her teacher at MIT, Noam Chomsky, consti- a framework that stressed the role of transformational tuted the dominant paradigm. Transformational rules. With evidence from various languages, such as approaches to the analysis of syntactic representation Chichewa and other Bantu languages, native often take one level of representation as the basic point Australian languages, and native American languages, of departure and attempts are then made to derive other Joan Bresnan demonstrated the important role that lex- levels of representation from this basic level through a ical information ought to play in syntactic theory. This series of transformational operations, summarized as led to her first major revolution in syntatic theory: the move alpha. The most classical example is the passive development of the idea of Lexical-Functional transformation whereby a passive sentence in a lan- Grammar (LFG), and then making it one of the most guage is derived by applying a passive rule on an active influential linguistic theories/frameworks of our time. sentence. However, confronted with a lot more natural Joan Bresnan developed LFG with Ronald Kaplan in language data from languages with structures very dif- the late 1970s and early 1980s. The term ‘Lexical- ferent from those of the known European languages Functional Grammar’ first appeared in Bresnan’s edit- like English and French, it became clear that other ed volume in 1982, The mental representation of 153 BRESNAN, JOAN grammatical relations. It is a nontransformational and theory of complementation in English under the super- nonderivational approach to the representation of vision of Noam Chomsky at MIT in 1972. From 1972 grammatical information. It is lexically oriented and to 1973, she was Assistant Professor of Linguistics and assumes the basic idea that different dimensions of Philosophy at Stanford University. In 1973, she moved grammatical information can be represented as sepa- to the University of Massachusetts at Amherst and rate but parallel structures. Grammatical information served as Assistant Professor from that time until is represented in LFG in three main structures: the 1975. Between 1979 and 1983, she was full Professor Constituent Structure, where precedence and domi- at MIT. Joan Bresnan moved back to Stanford nance relations are indicated, the Functional Structure, University in 1983 as Professor of Linguistics and has where grammatical functions such as SUBJECT and remained there to date. Professor Bresnan held a OBJECT are indicated, and the Argument Structure, Howard H. and Jessie T. Watkins University where grammatical roles such as AGENT and THEME Professorship between 1992 and 1997, and in 2000 are expressed. These levels interact through a system she was named Sadie Dernham Patek Professor in of mapping or correspondence rules. Humanities. Joan Bresnan has received many awards Joan Bresnan’s second major revolutionary contri- and fellowships, and is a member of many societies bution to syntactic theory came with the rise of and boards. She was a Guggenheim Fellow from 1975 Optimality Theory (OT) in the early 1990s. In this the- to 1976, a Fellow at the Center for Advanced Study in ory of grammar, Universal Grammar is characterized the Behavioral Sciences, Stanford University from as a set of constraint interactions, with the basic 1982 to 1983, and a principal investigator of many hypothesis that there is a set of universal constraints. projects financed by National Science Foundation Languages satisfy these constraints differently and dif- Grants. She is a member of the editorial boards of ferences in languages will result from this fact. This many journals and publishing houses such as approach to grammar is potentially compatible with Cambridge University Press, Linguistic Inquiry, and many theories of grammar that assume an input from the Journal of Japanese Linguistics. She is a past mem- which a set of candidate outputs are generated. It was ber of the Executive Committee of the International at this point that Joan Bresnan came up with the idea Lexical-Functional Grammar Association, and Past of marrying the basic tenets of OT with those of LFG President of the Linguistic Society of America. to form the notion of OT-LFG or what she termed ‘Optimal Syntax’. In Optimal Syntax, the universal References input is assumed to be modeled by an abstract set of functional structures representing morphosyntactic Aissen, Judith. 1999. Markedness and subject choice in opti- content in a language-independent format. The univer- mality theory. Natural language & linguistic theory, Vol. 17, also in Optimality-theoretic syntax, ed. by Geraldine sal candidate set in Optimal Syntax should consist of Legendre, Sten Vikner, and Jane Grimshaw. Cambridge, pairs of constituent structures and their corresponding MA: MIT Press. functional structures, which are then matched to the Bresnan, Joan. 1979. The theory of complementation in English input functional structure through a set of correspon- syntax. New York, NY: Garland Publishing. dence principles and constraints that decide the opti- ———.(ed.) 1982. The mental representation of grammatical relations. Cambridge, MA: MIT Press. mal candidate from the set of candidates. Optimal ———.1997. Complex predicates, co-edited with Alex Alsina Syntax or OT-LFG is one of the most exciting and Peter Sells. Stanford, CA: CSLI Publications. approaches to grammatical studies in the early part of ———.2000. Optimal syntax. Optimality theory: phonology, the twenty-first century and this is being led by Joan syntax, and acquisition, ed. by Joost Dekkers, Frank van der Bresnan. Leeuw and Jeroen van de Weijer. Oxford: Oxford University Press. In addition to these major contributions in syntactic ———.2001. Explaining morphosyntactic competition. The theory, Joan Bresnan, with her students and col- handbook of contemporary syntactic theory, ed. by Mark leagues, has also made profound contributions to the Baltin and Chris Collins. Malden, MA: Blackwell description and analysis of particular languages and Publishers. language groups, including Bantu languages such as ———.2001. Lexical-functional syntax. Malden, MA: Blackwell Publishers. Chichewa, and Australian aboriginal languages such Kaplan, Ronald M. 1989. The formal architecture of lexical- as Warlpiri. functional grammar. Proceedings of ROCLING II, ed. by Chu-Ren Huang and Keh-Jiann Chen, also in Journal of Information Science and Engineering 5(4). 305–22. Biography Kaplan, Ronald M. 1995. The formal architecture of lexical- functional grammar. ed. by Dalrymple, et al., 7–27. Joan Bresnan was born in Chicago, Illinois, on August Stanford, CA: CSLI Publications. 22, 1945. She received a B.A. degree in Philosophy at Kaplan, Ronald M., and Joan Bresnan. 1982. Lexical-functional Reed College in 1966. She completed a Ph.D. on the grammar: a formal system for grammatical representation. 154  BRITISH SIGN LANGUAGE The mental representation of grammatical relations, ed. by ed. by Flip G. Droste and John E. Joseph, Current issues in Joan Bresnan. Cambridge, MA: MIT Press. linguistic theory, Vol. 75. Amsterdam: Benjamins. Sadler, Louisa. 1996. New developments in LFG. Concise encyclopedia of syntactic theories, ed. by Keith Brown and ADAMS BODOMO Jim Miller. Oxford: Elsevier Science. Wescoat, Michael, and Annie Zaenen. 1991. Lexical functional See also Chomsky, Noam; Grammar, Theories; grammar. Linguistic theory and grammatical description, Lexicon: Overview; Universal Grammar British Sign Language British Sign Language (BSL) is the language of the BSL was exported by signers to other countries, and British Deaf community, with its associated social and was the predecessor of Auslan (Australian Sign cultural organizations. There are no accurate figures of Language) and New Zealand Sign Language. It has the number of users of BSL, but estimates range been suggested that these three languages should be between 30,000 and 60,000 deaf signers. Hearing chil- considered as dialects of a single language. There is also dren of signing deaf parents may also be included. The evidence of the early influence of BSL on signing in number of signers who are not members of the com- South Africa, Hong Kong, and Newfoundland. Perhaps munity has increased enormously over the past 15 surprisingly, there are few known links between BSL years, through the widespread uptake of further educa- and American Sign Language (ASL) and the languages tion classes in BSL. In addition to signers in Britain, are not mutually intelligible. BSL is also not genetical- BSL is also used by some signers in Ireland, and it is ly related to Irish Sign Language (ISL) despite their taught as a foreign language to deaf school children in close geographical proximity. Norway and Russia. Regional Dialects Historical Records of BSL There is considerable regional lexical variation in BSL has no written form. Consequently, historical BSL, including core semantic areas such as color records of the language are rare. Of the few early terms, days of the week, and numerals. Despite this recorded references to sign language use in Britain, variation, most signers are familiar with different most are in the form of descriptions in English by dialects of BSL, and it appears likely that the use of hearing writers. We do know that deaf people have signing on television over the past 20 years is gradual- been using some form of signing in Britain since at ly leading to the development of a national standard. least the sixteenth century. A description of the signs used by a deaf man to make his marriage vows in Grammar of BSL 1575 is recorded in the parish records of St. Martin’s in Leicester. From the mid-seventeeth century BSL and other sign languages exhibit structural differ- onwards, books such as those by Bulwer (1644, ences from spoken languages, as use of the visual–spa- 1648) were published, which included drawings of tial modality provides a different range of structural signs and lexical and grammatical descriptions possibilities, including visual imagery, movement in of BSL. space, and multiple articulators. Also, in the mid-seventeenth century, the first man- In common with all sign languages, BSL signs ual alphabets appeared. These provide a means of often reflect visual properties of referents, such as reproducing the orthography of the written language their shape, movement, or how they are handled. using various configurations of the hands and fingers. Whether or not a sign is visually motivated, all Fingerspelling serves as a resource for lexical borrow- signs exhibit a conventionalized relationship between ing and new sign formation. Unlike most deaf com- the form and the referent. munities, the BSL community uses a two-handed The placement and movement of signs within a lin- alphabet, the earliest version of which was published guistically determined area of space around the signer anonymously in 1698. are a central feature of BSL, having considerable 155 BRITISH SIGN LANGUAGE grammatical functions. As in other sign languages, the Social Situation of BSL movement of verbs between points in space can reflect The education of deaf children has had a great impact on grammatical relations between referents at the concep- BSL. The first schools for deaf children in Britain, using tual level, differentiating semantic roles and grammat- BSL as the language of instruction, were established in ical classes (subjects and agents, patients and objects the mid-eighteenth century. Although BSL enjoyed of actions), without any implication of a spatially con- widespread use in British deaf schools throughout the strained relationship. nineteenth century, there were always educationalists Signed language can also convey spatial relations who believed that deaf children should use English more directly: sentences can be constructed topograph- rather than BSL. This philosophy gradually gained ically, with the spatial relationships between signs cor- strength, and by the early twentieth century signing was responding to actual relationships among the referents. banned in schools and children were punished for its use. The availability of two hands, a head, and a face Signing was reintroduced into deaf education from the provides BSL and other sign languages with the pos- 1970s onward, with the acceptance of bilingual (BSL sibility of using simultaneously articulated structures and English) education. However, the closure of many to place referents in space and represent their relative deaf schools has reduced opportunities for deaf children locations and movements. BSL can also directly rep- to use sign language at an early age, and this has been resent the timing of two events relative to each other, viewed as a threat to the viability of BSL in the future, by showing the two events on two different ‘chan- with the removal of children’s access to the community nels’. Sign languages differ in the extent to which of signers and the linguistic and cultural role model that they exploit such devices. In a comparison of sen- the community provides to young deaf children. tences generated from the same picture materials in Despite the changes in education policy, the past 20 ISL and BSL, simultaneous signs appeared in 20% of years have witnessed a substantial improvement in the ISL utterances and 80% of BSL utterances. status of BSL. A weekly program in BSL has been Simultaneity can also be extended for poetic effect. In broadcast by the BBC since 1981; a comprehensive a BSL poem by Dorothy Miles, ‘Trio’, the poet repre- dictionary of BSL was published in 1992; and the sents herself, a dog, and a bird all dozing together British government appears poised to offer some for- after a good lunch. This is signed by using three artic- mal recognition to BSL as a minority language, and ulators simultaneously, using one hand to refer to the public interest in and acceptability of BSL has proba- dog, one hand to refer to the bird, and the head to refer bly never been higher. to the poet. References BSL as a Minority Language Boyes-Braem, Penny, and Rachel Sutton-Spence (eds.) 2001. Although BSL is fully independent of English, both The hands are the head of the mouth. Hamburg: Signum lexically and grammatically, English has influenced Press. BSL, as might be expected when any powerful major- Brennan, Mary, Martin Colville, and Lillian Lawson. 1984. ity language surrounds a minority language. English Words in hand: a structural analysis of the signs of BSL. Edinburgh: Moray House College of Education. provides loan signs to BSL through fingerspelling and Brien, David (ed.) 1992. Dictionary of BSL/English. London: loan translation. Faber & Faber. Family terms, calendar vocabulary items, and signs Bulwer, John B. 1644. Chirologia: or the natural language of for units of measurement of time and space in BSL are the hand. London: R. Whitaker. frequently of English origin. BSL also reflects the Bulwer, John B. 1648. Philocophus: or the deafe and dumbe man’s friend. London: Humphrey Moseley. influence of English in its use of mouth patterns Johnston, Trevor. 1989. AUSLAN dictionary. A dictionary of derived from spoken English (‘mouthings’). These are the sign language of the Australian deaf community. used in a wide variety of ways and in conjunction with Victroria: Australian Print Group. other mouth patterns unrelated to English (‘mouth ges- Kennedy, Graeme, Robert Arnold, Pat Dugdale, and David tures’). The use of mouthings varies with age, and Moskowitz. 1998. A dictionary of NZSL. Auckland: Auckland University Press. social and linguistic background, as well as situation- Kyle, Jim, and Bencie Woll. 1985. Sign language: the study of al variety. Although mouthings feature in all languages deaf people and their language. Cambridge: Cambridge and function in similar ways, anecdotal evidence sug- University Press. gests that BSL uses many more mouthings than, for Rachel Sutton-Spence, and Bencie Woll. 1999. The linguistics example, ISL or ASL. of British sign language: an introduction. Cambridge: Cambridge University Press. While BSL is a minority language with respect to BENCIE WOLL English and borrows from English, it also acts as a donor to other sign languages, for example, ISL. See also American Sign Language 156  BRUGMANN, KARL Brugmann, Karl Best known as the leading member of the so-called caused the most profound break with the scholars of the ‘Junggrammatiker’ (‘neogrammarians’), Karl Brugmann preceding generation. Brugmann’s personal vicissitudes played a crucial role in laying the foundations of histor- are an example of this break: at the inception of his sci- ical linguistics in the modern sense. In one of his most entific career, he produced two articles, Nasalis sonans famous publications, the preface to the Morphologische in der indogermanischen Grundsprache and Zur Untersuchungen auf dem Gebiete der indogermanischen Geschichte der stammabstufenden Deklinationen, which Sprachen, Brugmann and his co-editor, H. Osthoff, provoked a harsh reaction from his former teacher Georg reacted to current ideas about language change, outlin- Curtius, who, feeling criticized as methodologically ing a theory based on the regularity of sound laws, inadequate, never wanted to acknowledge his pupil’s whose apparent irregularities could only be explained as achievements and, still years later, published a thorough, either the outcome of other sound laws, or as the effect but by then outdated critique of the Neogrammarians. of analogy. Although his major interest was in the fields of The importance of the concept of sound laws for the phonology and morphology, Brugmann also published development of historical linguistics can hardly be some pioneering work on syntax. He was deeply con- overstated. Before the neogrammarians, phonological vinced of the importance of a global vision of utter- change was conceived as applicable in most cases, but ances, prior to the study of forms in isolation, as shown no theoretical statement was made about its regularity, by his statement, ‘In fact, all speech is brought about in and the (apparent) exceptions were left unexplained. sentences’ (‘In der Tat geschiet alles Sprechen in By giving regularity the status of a basic requisite, the Sätzen’), in the introduction to his posthumously pub- neogrammarians opened the way to much more lished volume Die Syntax des einfachen Satzes im refined research in the field of sound change and lan- Indogermanischen. Besides the history of ancient lan- guage change in general, based on the idea that a phe- guages, some more recent changes also attracted his nomenon was not exhaustively described, until all attention, most notably the creation of dummy subjects possible exceptions or counterexamples had been sat- in the modern Germanic languages and in French, to isfyingly explained. which he devoted an essay (Der Ursprung des Although the achievements of the neogrammarians Scheinsubjekts ‘es’ in den germanischen und romanis- are best described as the result of in-depth collabora- chen Sprachen). tion among a group of scholars who shared the same Brugmann’s work on syntax also shows that, in ideas, Brugmann certainly stood out of the group, with spite of his central concern with language change, he his 400 odd publications, covering various areas of lin- could also give insightful synchronic descriptions of guistics and philology. The most famous remains his linguistic phenomena. This is mostly shown in his Grundriß der vergleichenden Grammatik der indoger- analysis of word order in the ancient languages, based manischen Sprachen, whose first volume appeared in on what could be termed a functional view of sentence 1886. The work, completed in 1901 with the addition structure. Here, as well as in his assessment of the of the second volume by Karl Brugmann and three vol- motivation of language change, Brugmann stressed the umes about syntax by Berthold Delbrück, is still the importance of the psychological dimension of lan- most authoritative reference book in the field. guage use, thus highlighting the role of the speaker in By shifting emphasis from language classification the process of language structuring. and reconstruction to language change, the Neogrammarians broke with the tradition of Indo- Biography European linguistics in two ways. In the first place, they brought forth a conception of linguistics as a historical Karl Brugman was born in Wiesbaden (Germany), on science, thus closer to human sciences, as sociology and March 16, 1849 as Karl Friedrich Christian Brugman psychology, than to natural sciences, to which it was (his last name was changed to Brugmann in 1882). He closer before. Furthermore, they provided linguistics studied in Halle and Leipzig, and received his Ph.D. with a methodology of its own, no longer borrowed from Leipzig (1871) for his thesis on suppletion in Greek. natural sciences, and not directly dependent on the meth- From 1872 to 1877, he worked as a teacher in ods of classical philology. This last point is of crucial Wiesbaden and Leipzig, habilitated in Leipzig in 1877, importance for the emancipation of linguistics, and and was immediately appointed extraordinary professor. 157 BRUGMANN, KARL From 1874 to 1877, he was ordinary professor in ———.1925. Die Syntax des einfachen Satzes im Freiburg i. Br. and in 1887 he worked as ordinary pro- Indogermanischen. Berlin – Leipzig: De Gruyter. Brugmann, Karl, and Berthold Delbrück. 1897–1916. fessor in Leipzig, where a new chair of ‘indogermanis- Grundriss der vergleichenden Grammatik der indogermanis- che Sprachwissenschaft’ (Indo-European Linguistics) chen Sprachen, 2nd edition. Straßburg: Trübner, reprinted. had been created specifically for him. He held this posi- Berlin: De Gruyter, 1930 and 1967. tion until his death, in Leipzig, on June 29, 1919. He was Brugmann, Karl, and Hermann Osthoff. 1878 Morphologische the founder and editor of ‘Indogermanische Untersuchungen auf dem Gebiete der indogermanischen Sprachen, Vol. 1. Leipzig: Hirzel. Forschungen’ (founded in 1891) and the first chairman Collinge, N. E. 1985. The laws of Indo-European. Amsterdam, of the Indogermanische Gesellschaft in 1912. Philadelphia: Benjamins. Curtius, Georg. 1885. Zur Kritik der neuesten Sprachforschung. Leipzig: Hirzel. References Morpurgo Davies, Anna. 1986. Karl Brugmann and late nine- Brugmann, Karl. 1917. Der Ursprung des Scheinsubjekts es teenth-century linguistics. Studies in the history of in den germanischen und romanischen Sprachen. Western linguistics. In honour of R.H. Robins, ed. by Th. Sitzungsberichte der sächsischen Gesellschaft der Bynon and F. R. Palmer. Cambridge: Cambridge Wissenschaften, Vol. 5. University Press. ———.1876. Nasalis sonans in der indogermanischen Morpurgo Davies, Anna. 1996. Nineteenth century linguistics. Grundsprache. Studien zur griechischen und lateinischen London: Longman (History of linguistics, Vol. 4, ed. by, Grammatik 9. Giulio Lepschy). ———.1876. Zur Geschichte der stammabstufenden Sebeok, Thomas A. (ed.) 1966. Portraits of linguists. Deklinationen. Studien zur griechischen und lateinischen Bloomington, London: Indiana University Press. Grammatik 9. SILVIA LURAGHI Bühler, Karl The psychologist and philosopher Karl Bühler started aspects of language. This research perspective devel- as a member of the Würzburg school of oped into Bühler’s ‘Sprachtheorie’ (1934), which is ‘Denkpsychologie’, where Oswald Külpe was his one of the most important forerunners of semiotics and teacher. Bühler later followed Külpe to Bonn and contemporary cognitive linguistics. Munich. In the 1930s, he built up the Psychological Bühler’s most remembered contribution to linguis- Institute of Vienna University, which became one of tics is his organon model. See the figure, adapted from the most flourishing psychological institutes in ‘Sprachtheorie’ (1934). Europe. Together with his wife Charlotte, who also Communication takes place between the speaker worked as a psychologist, he emigrated to Norway and (sender) and the listener (receiver), and it is about the later to the United States when the Nazis had occupied objects around us. In Bühler’s view, language is an Austria. He spent his last years as a professor of psy- ‘organum’ or tool for one persons communicating with chiatry in Los Angeles. another about the world. The three main functions of Bühler contributed in many ways to the social and language that Bühler distinguishes in his organon cognitive sciences. At the beginning of his career, he model are Darstellung (representation, of states of worked on the psychology of thinking. Next, he stud- affairs), Ausdruck (expression, of the sender’s feel- ied human perception and developed a new notion of ings), and Appell (appeal, to the receiver). All functions Gestalt psychology, which he understood as a com- exist in every single utterance. However, usually one petitor to the Berlin schools view of Gestalt, as exem- prevails. When the focus is on the feelings of the plified by Kurt Koffka and Max Wertheimer. Another sender, the expressive function of communication important part of Bühler’s work in collaboration with dominates. An object-oriented communication is very Charlotte Bühler was developmental psychology. neutral or representative. If the focus is on the receiver, Bühler wrote the most read German textbook on the we deal with an appeal. The circle in the middle of the issue, titled ‘Die geistige Entwicklung des Kindes’ illustration above symbolizes the concrete, sensibly (1918). Bühler’s treatment of the mental development given sound. The overlapping triangle symbolizes the of the child shows a strong concern for the cognitive meaning of the sound and its Gestalt-like features. In 158  BÜHLER, KARL Objects and States of Affairs Representation Expression S Sender Receiver Appeal those places where the circle is larger than the triangle, the representational function of language. A symbol the sound (circle) contains information that lacks bears an invariable content, a content that is independ- meaning (triangle). Where the triangle is larger than the ent of the actual situation. The denotation of a deictic circle, there is a meaning (triangle) lacking an expres- (or indexical) expression like ‘I’ varies from speaker to sion in sound (circle). Both phenomena occur all the speaker and from situation to situation. That is the rea- time in everyday communication. The linguistic sign son why Roman Jakobson calls these expressions itself is a symptom, a signal, or a symbol. If it is a ‘shifters’. The denotation of a symbolic expression does symptom, it reveals the interiority or consciousness of not vary or shift. Instead, it rigidly denotates always the the sender. If the sign is a signal, it is directed to the same thing, like for instance ‘Downing Street 10’. behavior of the receiver. Signs that are mere bearers of Karl Bühler’s Sprachtheorie is an important work information about the states of affairs are symbols, just of twentieth-century linguistics. A historical reevalua- representing the objects themselves. Roman Jakobson tion of his work as a whole began in the 1970s. His expanded Bühler’s model and assumes six functions of unpublished works (lectures, manuscripts) are also language, adding the poetic, phatic, and metalingual collected now and edited. It is striking how Bühler’s function to Bühler’s three functions. Karl Popper, who work still stimulates contemporary linguistic research. was a student of Bühler’s in Vienna, proposes an addi- tional argumentative function. Biography According to Bühler, the lexicon of a language can be split into the field of symbol words (Symbolwörter) and Karl Bühler was born in Meckesheim, Germany on the field of deictic words (Zeigwörter). As for deixis, May 27, 1879. He studied human medicine and Bühler points out that the listener starts an orientation philosophy at the Universities of Freiburg and procedure when the speaker uses a deictic expression. Strasbourg. He received a Ph.D. in human medicine Bühler shows that the reference is not only based on (1903, Freiburg) for work related to color perception, a a sensual perception (when seeing the pointing finger, Ph.D. in philosophy (1904, Strasbourg) for work relat- e.g.). He argues that the pointing finger is the natural ed to Henry Home, and Habilitation (1907, Würzburg) instrument of visual demonstration, but it may be for work on cognitive science. He was assistant profes- replaced by other pointing aids. However, some point- sor at the Universities of Würzburg (1906–1909) and ing needs to be done. Bühler’s analysis covers the Bonn (1909–1913), associate professor at the deixis of pronouns (‘I’, ‘you’, ‘he’) and the deixis of University of Munich (1913–1918), full professor for time and place expressions (‘now’, ‘here’, ‘there’). philosophy and pedagogy at the Technical University The other field of language, the symbol field, differs of Dresden (1918–1922), and full professor for psy- radically from deixis. The symbol field is the place for chology at the University of Vienna (1922–1938); he 159 BÜHLER, KARL emigrated to Norway and later to the United States ———.1933. Ausdruckstheorie. Das System an der Geschichte when the Nazis occupied Austria. He was professor at aufgezeigt. Reprinted with an introduction by Albert Wellek, Stuttgart: Gustav Fischer, Verlag, 1968. St. Scholastica Duluth College and St. Thomas College ———.1934. Sprachtheorie. Jena: Fischer. at St. Paul, Minnesota (1940–1945), and professor of ———.1960. Das Gestaltprinzip im Leben der Menschen und psychiatry at the University of Southern California, der Tiere. Bern: Huber. Los Angeles (1945–1955). He died in Los Angeles on Eschbach, Achim (ed.) 1988. Karl Bühler’s theory of language. October 24, 1963. Amsterdam: Benjamins. Fiske, John. 1987. Introduction to communication studies. London: Methün. References Hülzer, Heike. 1989. Karl Bühler und Wilhelm Stählin: Psychologische Fundamente der Metapherntheorie im Brunswick, Egon (ed.) 1929. Beiträge zur Problemgeschichte ersten Drittel des 20. Jahrhunderts. Münster: Nodus. der Psychologie. Festschrift zu Karl Bühler’s 50. Innis, Robert E. 1982. Karl Bühler. Semiotic foundations of Geburtstag. Jena: Fischer. language theory. New York: Plenum Press. Bühler, Karl. 1918. Die geistige Entwicklung des Kindes. Jena: MONIKA RATHERT Fischer. ———.1927. Die Krise der Psychologie. Jena: Fischer. See also Jakobson, Roman; Semiotics Burmese Burmese is the national language of the country of plain and delta regions. Other varieties of Burmese, Burma (also known as Myanmar), located to the namely Northern Burmese, Arakanese (Yanbye, southwest of China and to the east of Bangladesh and Chuangtha), Intha, Danu, Taungyo, Yaw, Tavoyan, and India. It is estimated that 68% of the population, or Magui, are found in geographically isolated areas between 22 and 31 million people speak Burmese, blocked by either mountains or larger bodies of water, ranking it 43rd in the world by population. Burmese is such as along the long peninsula toward Malaysia or directly related to Tibetan and indirectly related to across the Pegu mountains of Central Burma proper, Chinese. It belongs to the Tibeto-Burman language or the more effectively isolating Arakan mountain branch of the larger phylum of Sino-Tibetan. Within range in southwestern Burma. the Tibeto-Burman grouping of languages, Burmese Ancestors of the modern Burmese-speaking people belongs to the Loloish (Yi) Branch and shares typo- migrated during the second-to-ninth century CE from logical characteristics with lesser-known languages of the Tibetan plateau southward into areas of southwestern western China, Thailand, Laos, and Vietnam. Some of China. Following riverine routes along the eastern end of the better-known minority languages related to the Himalayan mountains, they eventually descended Burmese are Lisu, Akha, and Lahu that are found into the hot, dry regions of the Irrawaddy River flood spread across four Southeast Asian countries. plain between 800 and 1000 CE. An earlier migration of As a result of internal political changes since 1990, a Tibeto-Burman speaking population, the Pyu, preced- an alternate name of Myanmar has been used for both ed the Burmese by 400–600 years. When the Burmese the Burmese language and the country of Burma, but migrated to the plains, the Pyu civilization was in most linguists continue to refer to the language as decline. Other language contact with Mon-Khmer- Burmese. Motivation for the name change is unclear, speaking populations, principally the Mon, resulted in since the name of the language and country has not significant cultural and linguistic changes for the changed in Burmese itself. There are two speech lev- Burmese. The Mon had already successfully developed els in Burmese. In the more colloquial, common, spo- an advanced Buddhist civilization, which the Burmese ken form of the language, the name for the country and eventually emulated, adopting similarly Indic-influ- speech is a form of the word ba ma a, from which the enced systems of government, kingship, religion, art, English form ‘Burma’ is derived. In the literary, more architecture, and a South Indian variety of the Bra¯hm¯ formal Burmese, the same name is myanma. alphabet. The first record of writing Burmese was found There are six to ten recognized regional dialects of on a stone inscription dated 1113 CE. The influence of Burmese, some of which are not mutually intelligible Mon (and Pali) is found lexically in borrowed words for with Standard Burmese. Standard Burmese is spoken cultural and technological items, as well as common, in the highly populated, central Irrawaddy River flood everyday idioms. Phonologically, Mon’s pervasive 160  BURMESE influence is marked in the Burmese tonal system, which ment form is used while maintaining the grammatical has adopted voice phonation or register as additional function. Lexical forms have been recruited over time tonal features in addition to simple pitch. to serve standard grammatical functions, demonstrat- ing the flexibility of Burmese semantics and the sta- bility of the grammar. Significance of Burmese Burmese is one of the two oldest written languages of Dictionaries and Grammars the Tibeto-Burman language family. The other is Tibetan. These two languages provide an invaluable The first English–Burmese dictionary was compiled by window on the language and culture of an ancient lin- Adoniram Judson, a missionary linguist, during the guistic world. Without some written forms of ancient middle of the nineteenth century and is still used today. languages, the reconstructed genetic relationships pro- In 1993, the Myanmar Language Commission produced posed between living languages would be far more a highly popular Myanmar–English Dictionary, which speculative. Burmese texts also provide an insight into includes a history of the Myanmar alphabet. During the the political and cultural relationships of a Southeast 1970s, the Burmese Language Commission released a Asian world shortly after the first millenium. Stone five-volume, Burmese– Burmese dictionary set. inscriptions are the principal media to have survived. Numerous grammars of Burmese have been written. The most famous is the Myazedi inscription dated The first in English was Adoniram Judson’s grammar of 1113 CE, which records the same text in four lan- 1866. During the British colonial period of 1885–1948, guages: Pyu, Burmese, Mon, and Pali. various grammatical studies produced by local and for- The Burmese alphabet is adapted from the Mon eign scholars attempted to grapple with Burmese gram- script and was one of the war trophies resulting from matical constructions using Indo-European, or Latinate the conquest of the Mon empire in 1044 CE. The dis- models. During the 1960s, various revisionist Burmese tinctive beauty and harmony of the alphabet is dis- linguists proposed alternate grammatical categories played in each character shape, variations of a circle. using only Colloquial Burmese, rather than the higher No word space is used; hence, each word is connected status, Literary Burmese. This attempt to legitimate to the previous in a continuous right-to-left flow until a modern Colloquial Burmese met with little official semantic juncture is encountered. A brief space is then acceptance. The School of Oriental and African Studies, used, followed by another long string of elegant circles. University of London, has made a most significant con- tribution to the study and analysis of colloquial Burmese for English speakers. Okell’s Colloquial Literary Form and Colloquial Language Burmese grammar volumes and language lessons have The term diglossia describes a situation where two made accessible what was once distant and intangible extremely different forms of the same language are for the language learner. Unpublished doctoral disserta- used for different social functions within a society. tions since the 1980s have further examined the beauty This situation arose because the ancient written form of the Burmese language. exerted a conservative force restricting natural lan- guage change. Over the centuries, the written form Phonology changed little, while the spoken language more freely followed regular sound changes. The extent to which Burmese has four lexical tones, which are character- the written form has constrained changes in the collo- ized by both pitch and voice phonation (plain, breathy, quial form is unknown. and creaky). The syllable consists of two types: a minor Literary Burmese, also called Written Burmese, dif- and major syllable. Minor syllables occur only with a fers from Colloquial Burmese (Spoken) in style, gram- major syllable, have only a schwa vowel, neutral tone, mar, lexicon, and grammatical particles, which are pattern as only consonant and vowel (CV), and have a similar to English prepositions. The most obvious dif- reduced inventory of initial consonants that may occur. ference between Literary Burmese and Colloquial The major syllable need not occur with a minor sylla- Burmese is seen in a unique set of postpositional parti- ble, bears one of the four lexical tones (T), patterns cles, which serve as spatial, temporal, and logical oper- with the potential canonical structure of initial conso- ators orienting nominal units within the sentence. They nant (Ci), optional medial consonants (Cm), vowel (V), also serve as case markers indicating sentence relations and optional final glottal or nasalization (Cf). The such as semantic subject, object, and indirect object. major syllable canonical shape is Ci(Cm)V(Cf)T. Each style has its own unique set of postpositionals. There are 32 possible initial consonants, two medial In some cases, a historical relationship exists consonants, ten vowels, and two types of final conso- between the particle sets, and in other cases a replace- nants. The syllable rhyme (VCf) is historically important 161 BURMESE in determining vowel quality changes such as off-glid- mranma tá-khyin- bríthísh- lak-Бawk kyà- ing, nasalization, and one of the four tones (Table 1). nong- ká tó-Б í -só rawk- Some of the more unusual features of Burmese are ngam-sany khái-bù- the voiceless nasals, hnim [n
e˜ ] ‘to press down’, but sany Burma- one-time- British- hand- fell- nim [ne˜] ‘settle’. The final nasals, while written, are country- past plural- under- arrived pronounced only as nasalization on the preceding topic genitive toward past- vowel. Final stops, while written, are pronounced as a previou glottal stop with an abrupt closure and low tone. This sly- has traditionally been called the stopped or checked realis tone. The voiceless lateral and the aspirated voiceless ‘At one time Burma fell under subjugation of the sibilant are somewhat universally rare sounds. British.’ Surprisingly, Burmese shares with English the moder- Noun phrases are normally structured with the ately unusual voiceless and voiced dental fricative. modifier preceding the head noun. Relative clauses function as a modifier of the head noun but with a Grammar postpositional particle marking the relationship between the preceding clause and the following Burmese is a verb-final language. Typical sentence noun. Possessive phrases also function in the same word order is subject, object, verb (SOV). way. Subject Object Verb Possessor Possessed (Head) su tha ma àng cà-sany bríthísh-tó-Б í lak-Бawk s/he rice eat-realis British-plural-genitive hand-under ‘S/he eats rice.’ ‘subjugation of the British’ Sentential nominal elements, such as the subject, Burmese is one of the languages in the world that object, location, or manner, typically occur with a requires the use of semantic classifiers when nouns are postpositional marker indicating the sentential role of enumerated. In enumerated noun phrases modifiers the nominal. Overt marking of subject and object roles follow the nominal head, with the resulting order: head is not obligatory in Burmese as in European lan- noun, numeral, classifier. This contrary ordering is a guages. Often a postpositional is used to disambiguate kind of appositional phrase in juxtaposition with the or contrastively emphasize the role of the nominal. nominal phrase. TABLE 1 Burmese Sounds Consonants (Ci) Labial Dental Alveolar Alveo-palatal Palatal Velar Glottal Plosive pH p b t t d k k   Fricative   s s z  h Affricate t  t d Nasal m6 m n
n 
 
 Glide w
Lateral l
l Vowels (V) Consonants (Cf) i u Historically as writtern -m-n- - -p-t-c-k e o Phonetic realization Nasalization   a  a a Tone (T) Consonants (Cm) ´ (creaky) (low, plain) ` (high, -p, -t, -k w  breathy) (stopped) 162  BURUSHASKI Nominal Numeral Classifier almost an endless creative potential for creating beau- nghak tac kawng ty, humor, and powerful description of visual sense bird one body (use of animals) and emotional feeling. ‘one bird’ O Verb Phrase In the same way, pluralization of nouns also follows Бaphàm kham-swà-rá-hra-sany the noun. capture undergo-go-must-pity-realis ‘(He) was caught, poor fellow.’ Nominal (Head) Quantifier Compounding as a process is extremely productive Бalup-samà myà work-person many in Burmese, occurring with words, phrases, and claus- ‘workers’ es. Compounding occurs with both nominal and ver- bal constructions. The semantics of the compound Burmese treats quantification, plurals, and numeral construct are not necessarily predictable from the lex- classifiers, uniquely from other types of specification. ical parts but form a semantic blending of various Plurals are not as common, nor as grammatically selected components of meaning. Burmese has a required as in English. Rather, plurals and enumerative noted preference for doublets, such that one com- phrases make a statement about the topic, and give a pound can be elegantly doubled into a compound of sense of inclusion, or definiteness, as articles do in two compounds. Then, these double compounds can English. be further doubled in a more formal, expressive The verb phrase as the last element of the sentence speech style; veritable doubled pleasure of semantic carries the bulk of the action and attitude of the speak- juxtaposition. er. A series of postpositions refer to the attitude of the speaker to the content of the sentence, the attitude to the hearer, the truth value of the sentence content, and References intentionality. Other functions marked by postposi- Becker, Alton L. 1993. The elusive figures of Burmese gram- tions are verbal aspect (punctiliar, process, complet- mar: an essay. The role of theory in language description, ed), mood (declarative, two types of interrogative, ed. by William A. Foley. Berlin: Mouton de Gruyter. Okell, John. 1969. A reference grammar of colloquial Burmese, imperative), and polarity (positive or negative). 2 vols. London: Oxford University Press. O Verb Phrase ———. 1994. Burmese: an introduction to the spoken lan- guage, 2 vols. Dekalb, II: Northern Illinois University, bà lup-yá-sè-sá-lè Center for Southeast Asian Studies. what do-must-yet-realis-question ———. 1994. Burmese: an introduction to the script. Dekalb, ‘What else must (you) do?’ II: Northern Illinois University, Center for Southeast Asian Studies. Preceding the head verb a limited set of modifiers ———. 1994. Burmese: an introduction to the literary style. may occur, which serve as intensifiers, comparatives, Dekalb, II: Northern Illinois University, Center for manner, or directionals that indicate both actual and Southeast Asian Studies. Roop, D.H. 1972. An introduction to the Burmese writing sys- metaphorical attunement of the action or state of the tem. New Haven: Yale University Press. main verbal proposition. Following the verb head a PAULETTE HOPPLE more open class of verbal auxiliaries may occur. The juxtaposition of such verbs in Burmese seems to have See also Chinese; Diglossia Burushaski Burushaski (also ‘Khajuna, Kanjut’) is spoken mostly other, while the Yasin dialect differs from HB and NB in the Northern Areas of Pakistan. There are two main in lexicon, phonology, and morphology. Many HB dialects of Burushaski: Hunza (HB)/Nager (NB) speakers also speak Shina; YB speakers are bilingual (approximately 45,000 speakers) and Yasin (YB) (also in Khowar. Educated Burushaski speakers also speak ‘Werchikwar’) (approximately 15,000 speakers). The and read Urdu, the national and link language of Hunza and Nager variants are quite similar to each Pakistan. With increasing literacy and bilingualism in 163 BURUSHASKI Urdu, many features of Burushaski lexicon and syntax Vowels are being influenced by Urdu. Both HB/NB and YB have five basic short vowels, i, e, Burushaski is still regarded as a linguistic isolate, a, o, u, and two sets of corresponding long vowels although its possible relationship with other languages (consisting of two time units, or moras), one stressed has long been the subject of intense research. Early on the first mora, represented as áa, and another attempts to link it with other languages include com- stressed on the second mora, represented as aá. First- parisons with Caucasian and with the Yeneseian lan- mora stress results in a high-falling pitch, while sec- guages. Most recently, Hašule (1988) argues that ond-mora stress results in a low-rising pitch contour. Burushaski is Indo-European, most closely related to Vowel length and moraic stress are contrastive, e.g. bat the pre-Balkanic languages, especially Phrygian. The ‘stone’, baát ‘cooked rice’, and báat ‘after’ (Berger most comprehensive grammars of HB are Lorimer 1998). This places Burushaski in the areal grouping of (1935a, b, 1938) and Berger (1998); Tiffou (1999) north Pakistani languages, including at least Shina, deals with both HB and YB. Anderson (1997, to Khowar, and Burushaski, with pitch accent systems. appear) are English-language analytical summaries of Burushaski phonology and morphology. Morphology Burushaski has recently begun to move toward becoming a literary language. Some writers have cho- Nouns sen to sometimes write in a Roman-based script—in Burushaski nouns fall into a four-valued gender class 1980, a Roman-based Burushaski primer was pub- system. The four classes are human female (hf), human lished—while others write in a modified Perso-Arabic male (hm), x, and y. The x class includes nonhuman script, especially when dealing with religious themes. animates and most count nouns, while the y class The Burushaski Research Academy, a local organiza- includes inanimates and mass nouns. The class mem- tion, is encouraging the development of writing in bership of a noun determines verb and sometimes Burushaski. Collections of riddles and proverbs have adjective agreement; in some cases, the stem form of a been published in Roman Burushaski, while poetry on verb also varies according to the class of the subject or religious themes has been published in Perso-Arabic the direct object, e.g. yáaltas ‘to wash (h or x object) vs. script. báaltas ‘to wash (y object), and wálas ‘to fall (x subject) vs. balás ‘to fall (y subject). There are numerous plural suffixes in Bur., which are correlated Phonology with the class of the noun. For example, -tiŋ (hm, hf) in Consonants daraγá-tiŋ ‘canal overseers’; -ŋ (y) in hála-ŋ ‘goals HB has the consonants shown in Table 1 (Tiffou 1999: (polo)’ -muts (x) in bas.á-muts ‘turbans’. Bur. has a suf- 118–9). fixal singular article -an ‘a, one’ (from han ‘one’), and In HB, f and x appear in loan words; y is a retroflex a plural article –ik; thus, haγ úr-an ‘a horse’; and harált- glide, unique to HB and NB. Except for y, YB has the iŋ-ik ‘(some) rain-showers’ (Berger 1998, I:39, 43). same consonants as HB; it also has x, corresponding to Nouns have an elaborate case-marking system, con- HB qh. sisting of several levels. The basic case endings are TABLE 1 Manner of Point of articulation articulation Bilabial Dental Palatal Retroflex Velar Postvelar Glottal Stops Voiceless p t t· k q Voiceless, aspirated ph th t·h kh qh Voiced b d d· g Affricates Voiceless (pf) ts  c· Voiceless aspirated tsh h c·h Voiced j j Fricatives · Voiceless (f) s š s· h Voiced z (D) γ Nasals m n Continuants w l y y Vibrant r 164  BURUSHASKI TABLE 2 TABLE 3 2 3 4 5 Type I, Type I, Type Type Case ‘man’ ‘woman’ ‘horse’ ‘sword’ unstressed stressed II III (hm) (hf) (x) (y) 1 a- á- á- áa- Nominative/ hir gus haγ ur γ aténc. 2 gu- gú-/-kú- gó-/-kó- góo-/ absolute kóo- hm i- í- é- ée- Oblique hir gus-mu haγ ur γ aténc. hf mu- mú- mó- móo- Ergative hir-e gus-e haγ ur-e γ aténc. -e x i- í- é- ée- Genitive hir-e gus-mu haγ ur-e γ aténc. -e y (sg/pl) i- í- é- ée- Dative hir-ar gus-mo-r haγ ur-ar γ aténc. -ar 1 (pl) mi- mí- mé- mée- Ablative hir-tsum gus-mu- haγ ur γ aténc. 2 (pl) ma- má- má- máa- tsum -tsum -tsum hm, hf, x (pl) u- u´ - ó- óo- illustrated in Table 2 for one noun of each class causative verbs. There can be more than one pronom- (Anderson, to appear). inal prefix in a sentence. For example, in the following The ergative case is used for the subjects of transi- sentence, the first-person singular pronominal prefix tive verbs, such as ‘hit’, while the nominative/absolu- áa indexes the affected object ‘me’, and the second- tive case is used for the subject of intransitive verbs, person singular prefix kóo indexes the subject ‘you’. such as ‘go’. For some speakers, though, the nomina- áa-lji du-kóo -s.qalc -um -a tive/absolutive can be used in the future and the present me-up.to d -you -overtake(DURATIVE) -PPL-you of transitive verbs. The genitive indicates possessors; ‘You will overtake me.’ the dative marks indirect objects, some direct objects, (Bashir 1985:15, from Lorimer 1938:119) and direction toward something. The ablative indicates direction away from something and appears in compar- There are four series of these prefixes (Berger ative constructions. Two examples follow. 1998:91). See Table 3. jáa-r madát a -c h -í Independent pronouns also exist. They are: I(OBLIQUE)-DATIVE help I-give Sg. Pl. ‘help me’ 1. je, já ‘I’ mi ‘we’ (Berger 1998,I:69) 2. un ‘you’ ma ‘you’ uŋ -tsum je kam a- p -a 3. iné (h) ‘he, she’ ué (h) ‘they’ you-ABLATIVE I small not-be-I isé (x) ‘it’ itsé (x) ‘they’ ‘I am not smaller/less than you.’ (Anderson, to appear) ité (y) ‘it’ iké (y) ‘they’ In addition, there are several locative cases, and a Adjectives fairly large set of ‘relational nouns’. These words are Some plural adjectives agree in class with nouns, e.g. nominal in origin, but are undergoing the process of burúm ‘white (sg.)’, burúm-išo ‘white-x pl’, burúm-iŋ grammaticization into postpositions. ‘white-y pl’ (Berger 1998:47). Comparative and superlative constructions are formed analytically with Pronouns the ablative case. Pronominal prefixes are a central element of Burushaski morphology. In general, these prefixes are Verbs used to indicate an affected, animate entity. They are Burushaski has a relatively small number of simple obligatory with certain nouns, e.g. names for body verbs; the majority of verbs consist of a noun or adjec- parts and kinship terms, indicating inalienable posses- tive combined with one of a small set of basic verbs sion. For example, -me ‘tooth’ cannot stand alone; it (sometimes called “light verbs”), which function to must include a prefix indicating ‘whose tooth’, e.g. gu- change the noun or adjective into a verb. Of these, the me ‘your tooth’. Some adjectives and postpositions, most frequently used are mánas ‘to be’, which forms and a few numbers require pronominal prefixes, as do intransitive verbs, and étas ‘to do, make’, which forms many verbs. For example, -apat ‘side of body’ is a transitives. For example, nat. étas ‘to dance’, kárpat. body part term requiring a pronominal prefix, from étas ‘to quarrel’, kárpat. mánas ‘to quarrel’. which the adjective -pac im ‘being alongside of __’, The Burushaski verb is agglutinative; that is, it is go-pac im ‘by your side’ (Bashir 1985:3). With verbs, formed of several separate meaningful units (mor- pronominal prefixes index the subject of some intran- phemes), each of which corresponds to a specific sitive verbs, the object of some transitive verbs, or the meaning element. The general structure of a verb form indirect object or causee of some ditransitive or can be schematized as follows (Anderson, to appear). 165 BURUSHASKI These positions are ordered from left to right in the a -tú -ku -man -um -a verb form (See Table 4). NEG -d -you(sg). -be.born -NON- you(sg) One example of a fairly complex verb form, in itself DURATIVE- a complete sentence, is given here. Position numbers [–4 –3 –2 0 +4 +5] and morpheme names are indicated; note that not all ‘You weren’t born’ the positions are filled in any one verb. (Berger 1998,I:91) Tense forms fall into two categories: those formed TABLE 4 on the nondurative (sometimes called ‘past’) stem and those formed on the durative (sometimes called Position Morpheme Function ‘present’) stem. –4 a-, oó- Forms a negative verb –3 d-, n- Prefixes/preverbs. d- indicates (abstract) Sentence Structure motion in the direction of speaker or terminus Basic Word Order (Bashir 1985) but is no Burushaski has postpositions, and adjectives precede longer productive. n- nouns. The basic word order of a simple sentence is marks the conjunctive SUBJECT – OBJECT – VERB. For example: participial (also ‘con- verb’, ‘absolutive’) form híre gus mu-yeétsimi of the verb. –2 Pronominal prefixes Index the subject, indirect man woman saw object, or direct object of ‘The man saw the woman.’ a sentence, depending on (Berger 1998,I:177) the transitivity of the verb. Observe that the direct object ‘woman’ is expressed –1 Causative morpheme Indicates either a on the verb by the prefix mu-, for human females. -s- causative verb, or a bene- factive meaning. 0 verb root Contains the basic mean- ing of the verb. Complex Sentences +1 plurality marker -ya- Marks an action as dis- In complex sentences, the subordinate clause usually tributed in some way, precedes the main clause. The indigenous strategy for either by having a plural forming subordinate clauses involves using one of the subject or by being per- nonfinite verb forms. For example, one type of relative formed on multiple occa- sions or in multiple clause is formed with the infinitive: places. +2 durative, Markers of durative [but. bulá dél-as] iné hír-e nonpast participle (cf. ‘present’) base. They much polo hit-INFINITIVE that man marker: c, š, .c, j, y; indicate that an action is ‘the man who plays much polo…’ durative, i.e. has temporal (Berger 1998,I:186) extent. +3 1st person singular Marks only first-person Another construction uses the -m participle subject -(y)a- singular subjects; thus, position 3 and position 5 (position +4). cannot both be filled. +4 (a) m- participle marker; These all mark nonfinite ja baríŋ etum ne hir šua bái (b) infinitive -as; (c) verb forms, or optatives. I speech done that man good is conjunctive participle ‘The man with whom I spoke is good.’ suffix -n; (d) optative -?; (Tiffou 1999:2000) (e) optative -áa. +5 (a) subject suffixes The auxiliary ba- ‘be’ In addition, some relative clauses are formed with other than 1st sg.; (b) forms complex verb imperative endings; tenses and itself has finite verbs and a relative (-like) element. For example, (c) auxiliary verb ba-. personal endings. the following YB sentence: +6 (a) interrogative; (a) Marks the sentence (b) case ending as a question. (b) Enables ámet caγ á-ule gambúrimu bién (te) s.iéli duá a sentence to be embed- that place-in flowers are (that) beautiful is ded in a more complex ‘The place where there are flowers is beautiful.’ sentence. (Tiffou 1999:201) 166  BURUSHASKI The conjunctive participial form, marked by n/d- South Asia and adjacent areas, ed. by Arlene R.K. Zide, VERB-n (Positions: –3, 0, 2) forms many adverbial David Magier, and Eric Schiller, 1–32. Bloomington, IN: Indiana University Linguistics Club. clauses. For example: _____. 2000. A thematic survey of Burushaski research. íne hír-e d-ítal-in s.apík s.ími History of Language 6(1). 1–14. Berger, Hermann. 1974. Das Yasin-Burushaski (Werchikwar): that man d-get up(CONJUNCTIVE ate Grammatik, Texte, Wörterbuch. Neuindische Studien,. Bd. PARTICIPLE) bread 3. Wiesbaden: O. Harrassowitz. ‘That man got up and ate bread.’ Berger, Hermann, with collaboration of Nasiruddin Nasir (Berger 1998, I:189) Hunzai. 1998. Die Burushaski-Sprache von Hunza und Nager. Neuindische Studien, Bd. 13. Wiesbaden: In many types of adverbial clauses, a case marker Harrasowitz. follows a nonfinite verb form. For example: Haule, Ilija. 1998. Basic Burushaski etymologies: the Indo- European and Paleo-Balkanic affinities of Burushaski. LIN- sén-as-ar COM Etymological Studies, Vol. 01. München: LINCOM say-INFINITIVE-DATIVE Europa. Lorimer, D.L.R. 1935–1938. The Burushaski language. Vol. I, ‘When he said’ Introduction and Grammar. Vol. II, Texts and translations. (Berger 1998,I:190) Vol. III. Vocabularies and index. Publikationer Serie B: Skrifter - Instituttet for sammenlignende kulturforskning. 29, Subordinating constructions augmented by the 1–3. Oslo: Instituttet for Sammenlignende Kulturforskning, complementizer ke ‘that’ (probably an influence from H. Aschehoug. Khowar and/or Urdu) are increasingly frequent. For ———. 1962. Werchikwar English vocabulary; with a few example, Werchikwar texts, Vol. 51. Instituttet for sammenlignende kulturforskning. Serie B: Skrifter. Oslo: Norwegian je girámar ke in ními Universities Press. I village-to when came he-went Tiffou, Étienne. 1999. Parlons Bourouchaski. Paris: L’harmattan. ‘When I came to the village, he left.’ Tiffou, Étienne, in collaboration with, Yves-Charles Morin, Hermann Berger, D.L.R. Lorimer, et al. 1993. Hunza (Berger 1998, I:191) proverbs. Calgary: University of Calgary Press. Tikkanen, Bertil. 1995. Burushaski converbs in their South and Central Asian areal context. Converbs in cross-linguistic References perspective: structure and meaning of adverbial verb forms- Anderson, Gregory D.S. 1997. Burushaski phonology. adverbial participles, gerunds. Vol. 13, Empirical approach- Phonologies of Asia and Africa (including the es to language typology, ed. by Martin Haspelmath and Caucasus),Vol. 2, ed. by Alan S. Kaye, 1021–41. Winona Ekkehard König, 487–528. Berlin: Mouton de Gruyter. Lake, IN: Eisenbrauns. Willson, Stephen R. 1999. Basic Burushaski vocabulary. _____. to appear. Burushaski morphology. Morphologies of Studies in Languages of Northern Pakistan, Vol. 6. Asia and Africa. Winona Lake, IN: Eisenbrauns. Islamabad: National Institute of Pakistan Studies, Quaid-i- Bashir, Elena. 1985. Toward a semantics of the Burushaski Azam University and Summer Institute of Linguistics. verb. Proceedings of the conference on participant roles: ELENA BASHIR 167  C Canada Canada, whose name derives from the Iroquois word Home language Mother tongue [ganá:da] ‘settlement’, is a confederation of ten English 20,011,535 67.5% 17,521,880 59.1% provinces and three territories in northern North America. Formerly a British colony, Canada was cre- French 6,531,375 22.0% 6,782,320 22.9% ated by the British North America Act (1867), which Nonofficial 3,096,110 10.5% 5,334,770 18.0% language joined the provinces of Upper Canada (Ontario), Lower Canada (Québec), New Brunswick, and Nova Scotia into a single dominion. Other provinces joined The British North America Act provided for the use the confederation at later dates: Manitoba (1870), of both English and French in Parliament and the British Columbia (1871), Prince Edward Island Québec legislature, provisions that were later extended (1873), Alberta (1905), Saskatchewan (1905), and to other provinces and territories, but in practice most Newfoundland and Labrador (1949). In 1875, the of the provincial legislatures and much of the federal Yukon and Northwest Territories were incorporated government were English dominant until the 1970s. into the confederation, and the territory of Nunavut Since the adoption of the Official Languages Act was created in 1999. Although Canada is the second- (1969), Canada has maintained a policy of official largest country in the world, encompassing an area of bilingualism in English and French at the federal level, 9,922,385 km2, it has a population of only 30,007,094 although all provinces except New Brunswick are still (according to the 2001 census), most of which is con- officially monolingual in either French (Québec) or centrated along its southern border with the United English. Unlike the ‘melting pot’ policy of the United States. States, the Canadian policy since 1971 has been one of In light of its vast size, Canada’s present multilin- ‘bilingualism within a multicultural framework’. This gual situation should not be surprising. A distinction is framework attempts to promote conditions for stable normally made between the ‘founding languages’ bilingualism, develop the prestige of French in Canada, (English and French), the aboriginal (or First Nations) and protect individual and minority language rights. languages, and the ‘heritage’ (or ‘international’) lan- These rights were further enshrined in the Constitution guages of immigrant groups, which are neither and the Charter of Rights and Freedoms (1982). The Anglophone (English speaking) nor Francophone effect of federal policy has been to make competence in (French speaking). As the following figures from the both official languages a valuable asset in politics, law, 2001 census demonstrate, English is claimed by the culture, and business. French immersion courses have majority of Canadians both as their mother tongue and increased official-language bilingualism among as the language they use at home on a daily basis, with Anglophones, especially in Québec, although French as the second-largest language. Francophones are still the most bilingual nationally. 169 CANADA Canada’s first inhabitants arrived from Siberia The varieties of French spoken in Canada can be across the Bering land bridge between 7,000 and traced to two colonies founded in the early seventeenth 12,000 years ago, bringing languages belonging to lin- century: Acadia (1604; now Nova Scotia, New guistic families that are now spread throughout the Brunswick, and Prince Edward Island) and New hemisphere. Although there is disagreement about the France (1609, now Québec). Although Acadia was exact historical affiliation of these families, Canadian ceded to the British in 1716 and many of the Acadians aboriginal languages are usually grouped into 11 fam- were later deported, Acadian French survived and ilies and isolates, most of which are also spoken in the spread to Newfoundland, the Îles de la Madeleine, and United States, Greenland, or Siberia. Varieties of the Gaspé Peninsula (Québec). By 1762, the British Inuktitut cover a wide area across the north, from the had also conquered New France, but in this case they Yukon to Labrador. British Columbia is home to the guaranteed the colonists’ right to the French language. widest array of languages, including Kutenai and Québec French is nowadays the mother tongue of most Tlingit, the Haida dialects, and the Tsimshian, inhabitants of the province and, because of westward Wakashan, and Salish families. The Athabaskan fami- migration in the nineteenth and twentieth centuries, ly stretches across the west from British Columbia to served as the basis of the French spoken in Ontario, Manitoba and north into the Yukon and Northwest Manitoba, and the western provinces. Because of their Territories. The Algonquian family, which includes different patterns of settlement and subsequent devel- Blackfoot, Cree, Ojibwa, Micmac, and possibly opment, Acadian and Québec French show a number Beothuk (an extinct language of Newfoundland), is the of distinctions from each other in grammar and pro- most widespread, with languages spoken from Alberta nunciation, especially in their vowel systems and ver- to Labrador, as well as in New Brunswick and Nova bal inflection. Both differ from European French, most Scotia. Michif, a mixed language that arose from inter- noticeably in vocabulary and pronunciation, especial- marriage among speakers of Cree and French, is spo- ly Acadian French and colloquial Québec French, ken in Manitoba. The Dakotan languages (a branch of which is also known as joual (from cheval ‘horse’). the Siouan family) are spoken in Manitoba, Although there is a perception of massive and perva- Saskatchewan, and Alberta, and a number of Iroquoian sive influence from English, anglicisms and English languages are spoken in Ontario and Québec. borrowings in fact account for a very small percentage According to the 2001 census, the aboriginal lan- of the total vocabulary. guages with the largest number of speakers are Cree, Although French was originally the majority lan- Inuktitut, and Ojibwa. guage of Canada, concentrated in Lower Canada and When the Europeans arrived in North America, Acadia, the influx of Anglophone settlers to Upper there were up to 300 languages spoken in what is now Canada and the Atlantic provinces resulted in a Canada, although nowadays there are fewer than 100. national decline in the Francophone percentage: from This attrition, the result of infectious diseases brought 50% to approximately 25% in the nineteenth century. by Europeans, genocide, and government policies, In addition, in all provinces except Québec and New has only recently begun to be addressed. After Brunswick, the Francophone population has gradually Confederation, little attention was given to the lin- been shifting to English. These patterns have guistic rights of aboriginal peoples, most of whom did decreased the presence of French in Canada while not have full citizenship. Until the 1960s, the federal simultaneously increasing its territorialization. By the government pursued an assimilationist policy, confin- middle of the twentieth century, these factors, plus the ing aboriginal peoples to reserves and giving educa- declining birthrate among Québec Francophones, led tional authority to religious groups, who enforced to the perception of a ‘doomsday scenario’ in which instruction in the majority language (English or French would eventually disappear from Canada. In French). More recently, aboriginal groups such as the 1976, this perception prompted the newly elected Assembly of First Nations have called for a greater nationalist Parti Québécois government of Québec to degree of self-determination, closing down the resi- pass Bill 101, the Charte de la Langue Française. Bill dential schools and transferring educational responsi- 101 ensures the rights of the Francophone majority by bility to local authorities. Some aboriginal languages making French the sole official language and requiring now have official status in the Northwest Territories all non-Anglophone immigrants to attend French-lan- and Nunavut. Although federal and provincial fund- guage schools, all public signs to be in French, and all ing for aboriginal-language instruction has helped businesses to conduct their affairs in French. This pol- slow language loss to some degree, these programs icy is implemented by three boards responsible for are inconsistent across the country and do not always codifying and developing French, monitoring its sta- promote fluency. tus, and enforcing the language laws. Despite studies 170  CANADA indicating that Québec Francophones still do not feel After 1860, immigration was widened to European that the French language in the province is secure, the countries outside of the British Isles, drawing settlers language laws have increased the status of French in from Germany, Italy, the Scandinavian nations, and Québec to the point that it is possible to live and work Ukraine. Such settlers, who increased the population entirely in French, even in the bilingual city of of the eastern cities and were instrumental in settling Montréal, and the percentage of Québécois reporting the western provinces, tended to give up their ethnic French as their mother tongue has remained stable languages in favor of English, further increasing the (81%) from 1986 to 2001. However, the prestige of Anglophone majority. After 1910, immigration English as the language of North American business widened to settlers from Russia, Iceland, Finland, and culture poses a continuing threat to the mainte- Poland, Greece, and the Netherlands. Since World War nance of French in Canada. II, immigrants have also arrived from southeastern The English presence in Canada dates from 1497, Europe, eastern Africa, India, Pakistan, Korea, when Newfoundland was claimed for the British Vietnam, China, Hong Kong, and the Philippines. Crown. Newfoundland was settled between the six- With the adoption of multiculturalism as a federal pol- teenth and eighteenth centuries almost entirely from icy in 1971, immigration was seen as a response to Ireland and southwestern England and remained rela- falling birthrates and a higher average age in the pop- tively isolated until World War II, making ulation. The annual rate of immigration has varied Newfoundland English distinctive in pronunciation, considerably, but currently Canada accepts approxi- grammar, and vocabulary from varieties of Canadian mately 200,000 immigrants every year. English spoken elsewhere. Present-day Standard The short-term effect of immigration has been to Canadian English is usually traced to the two waves of increase the number of nonofficial languages spoken: immigration of United Empire Loyalists (1776–1793), in the 2001 census, almost 100 nonofficial languages Americans loyal to the British monarch who fled the were claimed as a mother tongue by 18% of the popu- American Revolution: one wave came from coastal lation. The ten largest groups are (in descending order) New England (Connecticut, Massachusetts, Maine, Chinese, Italian, German, Punjabi, Spanish, and Rhode Island) and settled in Nova Scotia and New Portuguese, Polish, Ukrainian, Arabic, and Tagalog Brunswick, whereas another came from the midland (Pilipino). Immigrants tend to settle in southern (Pennsylvania, New Jersey, New York, and Vermont) Ontario (especially Toronto), Vancouver, and and settled in Ontario. The Loyalists thus served as Montréal, making Canada’s urban population increas- linguistic models for subsequent settlers recruited by ingly multiethnic. However, because most immigrants the British government in the nineteenth century from can speak at least one official language within a short England, Scotland, and Ireland to augment the time of arrival, the long-term effect of immigration is Anglophone population and supply agricultural and more likely to be social than linguistic. Indeed, construction workers. Because the western provinces although immigration has contributed to social con- were settled largely by migration from the east, flict, the ongoing educational and political accommo- Standard Canadian English is nowadays considered dation of so many different languages within one state extremely uniform across the country, without the reflects favorably on the Canadian linguistic marked regional standards characteristic of the United experiment. States. Despite this apparent uniformity, several non- standard regional varieties persist, such as Newfoundland English, the African-Canadian English References spoken by the descendants of Black Loyalists in Nova Chambers, J.K. 1991. Canada. English around the world: soci- Scotia, Ottawa Valley English, and the noticeably dis- olinguistic perspectives, ed. by Jenny Cheshire. Cambridge: tinctive pronunciation of certain vowels, especially Cambridge University Press. Churchill, Stacey. 1998. Official languages in Canada: chang- before [r], in the Atlantic provinces. Although the ing the language landscape. Ottawa: Canadian Heritage. shared origins of Canadian English and U.S. English Clarke, Sandra (ed.) 1993. Focus on Canada. Amsterdam and make them grammatically very similar, there are dif- Philadelphia: John Benjamins. ferences in pronunciation, vocabulary, and spelling Edwards, John (ed.) 1998. Language in Canada. Cambridge: practices. Most noticeable in pronunciation is the rais- Cambridge University Press. Marmen, Louise, and Jean-Pierre Corbeil. 1999. Languages in ing of the first part of the vowel in words such as Canada: 1996 Census/Langues au Canada: recensement de ‘house’, which Americans often hear as ‘hoose’. There 1996. Ottawa: Canadian Heritage. is considerable variation among Canadians between JAMES A. WALKER British and American spelling standards in words such as colour/color and cheque/check. See also English; French Language 171 CAPE VERDEAN CREOLE Cape Verdean Creole The Cape Verde islands belong to the northwestern In the phonological domain, all dialects of CVC union of archipelagoes in the Atlantic ocean (including contain the vowels {u, ε, a, e, ó, o, i}. The Barlavento the Canary Islands, the Azores, and Madeira) and are group also preserves the Portuguese diphthongs ei and situated approximately 450 kilometers from Senegal oi, which have merged with e and o in Sotavento. and Mauritania. The archipelago is composed of two (Compare Port. Manteiga ‘butter’ : Sot. mantega, Port. main clusters of islands: the windward islands and the açoitar ‘to whip’ : Sot. sota.) Sotavento varieties also leeward islands, known locally as Barlavento and typically denasalize word-final vowels; compare Port. Sotavento, respectively. The Barlavento islands também ‘too’ : Sot. tanbe. The Sotavento varieties include Boavista, Sal, São Nicolau, Santa Luzia, São change v into b, as in Port. vestir ‘to dress’ : CVC bisti. Vicente, and Santo Antão. The Sotavento islands con- The Portuguese palatal liquid lh typically becomes the sist of Brava, Fogo, Santiago, and Maio. Some affricate dj in CVC: Port. velho ‘old’ : CVC bedju. 450,000 Cape Verdeans currently reside in the islands In the morphological domain, Sotavento differs and one million reside abroad (primarily Portugal, the from Barlavento in making productive use of redupli- Netherlands, New England, and California). cation, a morphological process also found in African Although the Portuguese and the Italians officially languages, whereby a reduplicated adjective or adverb discovered the islands in approximately 1460, there is expresses emphasis, as in moku moku ‘very drunk’ or cartographic evidence that the Greeks and the Arabs faxi faxi ‘very quickly’. Noun reduplication may yield were already aware of their presence as early as in a distributive interpretation, as in dia dia ‘every day’, 1413 (Andrade 1996). Furthermore, the Jalofos tribe is or may simply lead to a change in meaning, as in boka reported to have lived on the island of Santiago prior ‘mouth’ → boka boka ‘in secret’. Lexical categories to the arrival of the Portuguese (Brásio 1962). may shift grammatical category when reduplicated, as Strategically located at the crossroads of Europe, in mansu ‘quiet’ (adjective) → mansu mansu ‘secrecy’ Africa, and America, the archipelago played a critical (noun). Reduplicative processes of this type are rela- role in the slave trade from the fifteenth to nineteenth tively scarce in Barlavento. centuries. The Portuguese initially brought the slaves to The lexicon of Cape Verdean Creole comes prima- Cape Verde from Guinea, including Senegal and Sierra rily from Portuguese, with a smattering of African Leone. However, competition from other western pow- forms, such as banbu ‘to carry a child on one’s back’ ers reduced the West African source of slaves for Cape from Manding and Manjaku banbu (same meaning), Verde to the limits of present-day Guinea-Bissau by the fepu ‘entirely’ from Wolof fepp ‘everywhere’, lokoti middle of the sixteenth century (Carreira 1982). ‘to get out of a hole’ from Wolof loqoti (same mean- Historical sources state that most of the slaves in Cape ing) (Santos 1979). Many of the words of African ori- Verde came from the tribes of the Mandingos, Balantes, gin refer to flora and fauna. Bijagos, Feloupes, Beafadas, Pepels, Quissis, Brames, The verbal system is fairly intricate. The bare verb Banhuns, Peuls, Jalofos, Bambaras, Bololas, and stem assumes various functions, depending on whether Manjakus (Brasio 1962). The white settlers came from the verb is stative or nonstative. The bare stem of the Algarve and Alentejo in Portugal, and also included verb ten ‘have, own, possess’ has a present tense reading, Jews, Spaniards, Italians, and French (Martinus 1996). as in N ten dos fidju ‘I have two sons’. In contrast, the A number of morphophonological and syntactic bare stem of a nonstative verb yields a past tense inter- features distinguish the Barlavento varieties from their pretation, as in N papia ku bu pai ‘I spoke with your Sotavento counterparts, which can, to a certain extent, father’. Creole languages typically have preverbal be related to the greater influence of African languages unbound markers expressing Tense, Mood, or Aspect in the Sotavento varieties. This disparity is in part due (TMA). Cape Verdean Creole has two TMA markers, ta to the fact that the Barlavento islands were settled a and sta, which occur preverbally, as illustrated in algen hundred years later than the Sotavento. As a result, the ku si boka ka ta era kaminhu ‘one who asks does not get sociolinguistic situation in Cape Verde is fairly com- lost’ and pamodi bu sta txora? ‘why are you crying?’ plex; it also involves an ideological dimension, as The marker dja in contrast precedes subject clitics but Portuguese (which provided the language with most of must follow full noun phrases; compare Dja’N atxa bon- its lexicon) is the only official language in the archi- beru ‘I found the exterminator’ vs. N atxa bonberu dja pelago to date. poi nunbru na porta ‘I found that the exterminator had 172  CARIB AND CARIBAN LANGUAGES already put the number on the door’. In contrast to other References creoles, Cape Verdean Creole possesses the inflected Andrade Silva, Elisa. 1996. Les Iles du Cap-Vert: De la anterior marker –ba, which expresses simple past with Découverte à l’Indépendance Nationale (1460–1975). Paris: stative verbs and pluperfect with nonstative verbs, as in L’Harmattan. Mininu tenba febri ‘The child had fever’ and João kume- Baptista, Marlyse. 1997. The morpho-syntax of nominal and ba tudu galinha ‘John had eaten the whole chicken’. verbal categories in Cape Verdean Creole. Ph.D. disserta- tion. Harvard University, Ann Arbor, MI: UMI Dissertation While the importance of inflectional morphology in Services. creoles has been minimized by a number of creolists, Baptista, Marlyse. 2002. The syntax of cape verdean creole: the Cape Verdean Creole and a few others display inflec- sotavento varieties. Philadelphia, Amsterdam: John tional morphology not only on verb stems, as seen with Benjamins. the –ba verbal inflection above, but also on noun stems. Brásio, António. 1962. Descobrimento, Povoamento, Evangelização do Arquipélago de Cabo Verde. Lisbon: CVC shares this feature with a number of other creoles Centro de Estudos Históricos Ultramarinos. and pidgins, such as Guinea-Bissau Creole, Casamance Brito, Ana Paulo. 1967. Dialectos Crioulos-Portugueses: Creole, Ghanaian Pidgin English, and Nigerian Pidgin Apontamentos para a Gramática do Crioulo que se Fala na English. Plural suffixation in Cape Verdean is a pro- Ilha de Santiago de Cabo Verde. Estudos Linguísticos ductive process subject to definiteness and the animacy Crioulos, revised by Adolpho Coelho, ed. by Jorge Morais- Barbosa [originally published in 1887]. Lisbon: Academia hierarchy. (The animacy hierarchy entails that animate International da Cultura Portuguesa. entities are more likely than inanimate objects to carry Carreira, Antonio. 1982. O Crioulo de Cabo Verde: Surto e agreement markers such as the plural suffix.) From a Expansão. Lisbon: Carreira edition. cognitive perspective, animate and/or definite entities Duarte, Dulce. 1994. Crioulo Caboverdiano e Diglossia. Papia 3. are more salient in the speaker’s consciousness and 32–45. Ferraz, Luiz, and Marius Valkhoff. 1975. A comparative study hence are more likely to be pluralized (Lyons 1999). of São-Tomense and Cabo-Verdiano Creole. Miscelanea The following examples illustrate this feature: Bu tra- Luso-Africana: Colectanea de Estudos Coligidos, ed. by badja ku nha povus ‘you worked with my people’, Ami Marius Valkhoff. Lisbon: Junta de Investigações Científicas nha mininus, es fika tudu la pa fora ‘As for me, my do Ultramar. children, they are all over the place’. Although the Lyons, Christopher. 1999. Definiteness, Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press. study of large corpora reveals that plural suffixation in Martinus, Frank. 1996. The kiss of a slave: Papiamentu’s West- monolingual speech is sensitive to variables such as African connections. Ph.D. Dissertation. University of [+/-animate], [+/-human], and [+/-definite], it is impor- Amsterdam. tant to keep in mind that monolingual speech is not a Petit Futé. 1999. Coutry guide: Cap-Vert. Paris: Nouvelles monolithic whole and that speakers may have multiple Editions de l’Université. Santos, Rosine. 1979. Le Créole des Iles du Cap-Vert: grammars. Variation in general has been a central Comparaison avec les Langues Africaines. Réalités theme in the study of Cape Verdean Creole (Baptista Africaines et Langue Française 11. Dakar: Centre de 2002). In addition, social class and education, higher Linguistique Appliquée de Dakar. degrees of which tend to be bound up with diglossia Veiga, Manuel. 1998. Le Créole du Cap-Vert: Etude and bilingualism (cf. Duarte 1994), combine with other Grammaticale Descriptive et Contrastive. Thèse Doctorale. Aix-en-Provence: Université Aix-Marseille. social factors to make the Cape Verdean language quite heterogeneous, as is expected of any natural language. BERT VAUX AND MARLYSE BAPTISTA Carib and Cariban Languages Languages belonging to the Cariban language family lists, brief collections of utterances) exists for maybe are spoken in northern South America, the bulk in half. There is no question that many of the languages northern Brazil, Venezuela, Guyana, Surinam, and spoken at the time of first contact with Europeans have French Guiana, with outliers to the west in Colombia become extinct, but we will never know how many. and to the south in Central Brazil. The historical liter- Currently, some 25 Cariban languages remain, with a ature on the Cariban family names over 100 lan- cumulative total of between 60,000 and 100,000 guages; however, linguistic information (e.g. word speakers; well over half of these speak Carib proper or 173 CARIB AND CARIBAN LANGUAGES one of three closely related languages belonging to a TABLE 1 single subgroup (the Pemóng Proper Subgroup), with Venezuelan Branch (A–B–C–D–E–F) most Cariban languages having only between 100 and 3,000 speakers. Carib proper (its English name, called Pemóng–Panare–Mapoyo–Tamanaku Sub–Branch (A–B–C–D) Galibi, Kaliña, Kari’na, and Cariña in other countries) Pemóng–Panare Macro-Group (A–B) is spoken by between 10,000 and 25,000 people along A. Pemóng Group a roughly 1,000-mile arc of the coast and inland from A1. Pemóng Proper Subgroup: Makushi, Pemóng the northeasternmost point of Brazil through French (Taurepang, Kamarakóto, Arekuna), Kapóng (Akawaio, Patamuna, Ingarikó) Guiana, Surinam, and Guyana, to the easternmost 300 miles of the Venezuelan coast. Several Caribbean lan- A2. (†)Purukotó guages historically claimed to be Cariban (Black B. Panare Carib, Island Carib, Garifuna) are linguistically Mapoyo-Tamanaku Macro-Group (C–D) Arawakan, with some Carib features (mostly vocabu- C. Mapoyo (Mapoyo, Wanai, Yawarana, Pémono) lary) due to intensive contact with Carib invaders. D. †Tamanaku Most languages of the Cariban family are still poorly E. †Kumaná (Chaima, Cumanagota) documented, with unreliably transcribed word lists and F. Makiritare (De’kwana, Ye’kwana, Maiongong) poor or no grammatical description. The two best docu- mented Cariban languages are Carib (due mostly to Southern Branch (G–H–I) work by Berend Hoff) and Hixkaryana (due exclusively G. Nahukwa Group: Kuikúru, Kalapalo, Amonap, Matipú to work by Desmond Derbyshire); work by Sergio Meira H. Bakarí on Tiriyó is rapidly bringing a third language into this I. Arara Group: Arara-Pirirí, Ikpéng (Txikão), †Apiaká- select group. The Carib language boasts of two gram- Apingi, †Juma, †Yarumá mars, two dictionaries, a collection of texts, several arti- cles on selected linguistic topics and work in progress in North Amazonian Branch (J–K) both French Guiana and Venezuela; Hixkaryana has two grammars, a collection of texts and several articles, all J. Jawaperi Group: Waimirí-Atroarí, †Jawaperi, †Bonari by Derbyshire. Other Cariban languages with grammars K. Paravilyana Group (of uneven quality and comprehensiveness) include K1. †Sapara Apalaí, De’kwana, Ikpéng (Txikao), Makushi, Pemón, K2. Paravilyana Subgroup: †Pawishiana, †Paravilyana and Waiwai. In the domain of the lexicon, Cariban lan- guages have been represented almost exclusively by Residue (groups and Languages still in search of branches, in bilingual word lists (some labeled dictionaries), with alphabetical order) true dictionaries existing for only Carib, Panare, and Groups Pemón. Scholars and students are actively working on L. Parukotoan Group projects in every country where Cariban languages are L1. Katxúyana (Chikena, Cachuena, Ingarüna, spoken. Ongoing research projects that ought to result in Shikuyana, Warikyana) solid primary descriptive materials (text collections, L2. Waiwai SubGroup: Waiwai (Wabui), Hixkaryana grammars, dictionaries) include: in Brazil, Arara, Ikpéng M. Taranoan Group (Txikão), Ingarikó (Kapóng), Katxúyana, Kuikúru, M1. Tiriyo Subgroup: Akuriyo, Tiriyo (Trio, Saluma, Tiriyó, Waimiri-Atroari, and Wayana; in Colombia, Pianakoto) Yukpa; in French Guiana, Wayana and Kali’na; in M2. Karihona (Carijona, Hianácoto) Guyana, Akawaio (Kapóng); in Surinam, Tiriyó; and in N. Wayana Group: Wayana, †Arakuajú Venezuela, Japrería, Kari’ña, Mapoyo/Yawarana, O. †Yao Group: †Tiverikoto, †Yao Panare, Pemón, and Yukpa. P. Yukpa Group: Yukpa, Japreria, †Koyama Given the poor documentation of most Cariban lan- Languages guages, current classifications of the Cariban family Q. Apalaí are all unconvincing, as are reconstructions of ances- tral homelands and migration patterns. The most R. Carib (Kari’nya, Kalinya, Cariña, Galibi) recent classificatory hypothesis (which will certainly S. †Opon-Karare evolve as more data become available) is reproduced T. †Palmella here in Table 1; it is derived by modifying Terrence U. †Pimenteira Kaufman’s (1994) classification (i) adding suggestions from recent work in Venezuela and the Guiana Plateau, local names for each language, and (iv) conservatively (ii) marking as extinct (†) every language that has not isolating several languages/groups that Kaufman had been attested in the last 25 years, (iii) using current linked into tentative branches. 174  CARIB AND CARIBAN LANGUAGES In the absence of a solid linguistic classification, has suffixes that change nouns to verbs, nouns or verbs discussions of prehistory — especially ancestral to adverbs, and adverbs, verbs, or postpositions to nouns. homelands and migration patterns — are perforce Most Cariban languages match this profile, with changes somewhat speculative. One hypothesis suggests that usually in the direction of simplification. There are two the ancestral homeland of Proto-Carib was in the word classes in Cariban languages, particles and sound Guiana plateau, from which the outlying groups symbolic words, to which no affixes may be attached; moved southward and westward. Another posits a cen- both are rare in elicited speech but ubiquitous in texts. tral Brazilian origin with one or more subsequent northward migrations and rapid expansion upon Syntax arrival in the Guiana Plateau. In either scenario, In Carib and most languages across the family, certain Cariban peoples appear to have occupied the Guiana two-word combinations—possessor–possessed, object– Plateau for at least three millennia. postposition, and object–verb—are clearly grammatical units, the tight bond between them demonstrated by invariant order, the inability to interpose words between Characteristics of Carib and them, and the fact that they seem to be pronounced as a Cariban languages single word. Beyond these two-word sequences, the Phonology word order appears to be quite variable, in Carib or in The Carib language has a relatively straightforward other Cariban languages. sound system, with three voiceless stops /p, t, k/, a voice- The Carib main clause consists of an inflected verb, less fricative /s/, two nasals /m, n/, two glides /w, j/, and with affixes indicating the type of both subject and a sound that—depending on the neighboring sounds— object. Noun phrases can optionally occur as subject and surfaces strictly in syllable-final position as either a velar object. When both subject and object of a transitive verb fricative /x/ or a glottal stop /ʔ/. All consonants are are third person, the object immediately precedes the palatalized (with tongue movement toward the palate) verb, forming a tight verb phrase in a subject– when adjacent to [i]; /p, t, k/ are pronounced as /b, d, g/ object–verb construction. The subject usually precedes between vowels or after nasals and glides; and the pro- the verb phrase, but this is not always required. When nounciation of nasals depends on the following conso- either subject or object is not third person, the subject– nant. Some Cariban languages (and even some Carib object–verb ordering (SOV) remains more frequent, yet dialects) add further consonants to the above inventory. all other orders of subject, object, and verb are permitted. The six-vowel system of Carib (and most other Carib distinguishes three degrees of past tense, and each Cariban languages) includes three high vowels /i, , u/, tense has both a perfective and an imperfective form. A two mid-vowels /e, o/ (usually phonetically [ε, ɔ]), and similar syntax is found in 19 other Cariban languages, one low vowel /a/. Across the family, the most common belonging to virtually every proposed branch of the fam- addition to this inventory is /ə/, with a few languages ily, which indicates that the common ancestral language adding a second set of mid-vowels [ε, ɔ]. (Proto-Carib) also had this structure. Several Cariban Independently, nasalized vowels are attested only in languages present the rare basic order OVS, document- Apalaí. The difference between long and short vowels ed first — and most extensively — in Hixkaryana. The length is a differentiating factor for words in Carib and OVS order is commonly observed also in languages several other Cariban languages, but vowels may also where it is not clearly ‘basic’. While SVO is attested in be lengthened due to speech rhythm: in an iambic foot several Cariban languages, the predominance of SVO in structure, the vowel of every second light (short, open) a few languages is most likely attributable to contact syllable is lengthened (excepting the final syllable, with languages of colonization. which is never lengthened). Across the family, primary Ergativity is pervasive in the Cariban family, and intonational stress has been difficult to identify reliably, indeed throughout much of South America. In ergative and sometimes it appears not to fall on the head of a constructions, the subject of an intransitive verb and foot (e.g. in Carib, Hixkaryana, Pemón, and Tiriyó). the direct object of a transitive verb (called the absolu- tive) receive the same grammatical treatment, that is, Morphology they receive the same case-marking, verb agreement, Carib makes heavy use of affixes, i.e. word particles are word order, etc. In contrast, the subject of a transitive added to other words to either add to or change the verb (called the ergative) takes different case-marking, meaning. In some cases, words together with their affix- verb agreement, word order, etc. Languages across the es may express meanings equivalent to phrases or family do not use relative clauses, i.e. subordinate sentences in other languages. Personal prefixes (‘I’, clauses that are introduced by words like that, which, ‘you’, ‘he’/’she’/’it’) and grammatical prefixes or suf- who, etc. Instead, Cariban languages use affixes to fixes appear on verbs, nouns, and postpositions. Carib turn the subordinate verb into a noun or an adverb. 175 CARIB AND CARIBAN LANGUAGES Early hypotheses about historical evolution of References Cariban grammar held that the ergativity seen in subor- Derbyshire, Desmond C. 1985. Hixkaryana and linguistic dinate clauses indicates that Proto-Carib had ergative typology. Arlington: Summer Institute of Linguistics and the syntax in both main and subordinate clauses, with the University of Texas at Arlington. nonergative main clause in today’s languages being an Derbyshire, Desmond C. 1999. The Cariban language family. innovation. Most recent work argues that the few cases Languages of South America, ed. by R.M.W. Dixon and Alexandra Aikhenvald. Cambridge: University of of main clause ergativity in the Cariban family are Cambridge Press. readily explained as innovations, in which ergatively Gildea, Spike. 1998. On reconstructing grammar: comparative organized subordinate clause morphosyntax is ‘elevat- Cariban morphosyntax. Oxford: Oxford University Press. ed’ to main clause status. The origins of all main clause Hoff, Berend J. 1968. The Carib language. The Hague: types observed so far in the northern Cariban languages Martinus Nijhoff. Kaufman, Terrence K. 1994. The Native languages of South fit this latter hypothesis; however, each new descrip- America. Atlas of the world’s languages, ed. by tion, especially from the less-known southern Cariban Christopher Mosely and R.E. Asher. New York: languages of the Xingu valley, will provide new chal- Routledge. lenges to the historical reconstructions in place so far. SPIKE GILDEA Caribbean The Caribbean islands (collectively called the West Antilles (i.e. Anguilla, Antigua and Barbados, the Indies) extend over a thousand miles from the tip of Bahamas, Barbados, British Virgin Islands, Cayman the Florida Peninsula to the northern coast of South Islands, Dominica, Grenada, Jamaica, Montserrat, St. America. They vary in size and population, ranging Kitts-Nevis, St. Lucia, St. Vincent and the Grenadines, from Cuba, with an area of 42,804 square miles and Trinidad and Tobago, Turks and Caicos islands, and more than ten million inhabitants, to the small, unin- the United States Virgin Islands), as well as two main- habited, rocky islets of the Grenadines and the Virgin land nations: Belize in Central America and Guyana in Islands. Most islands are smaller than Barbados (only South America. The Hispanophone Caribbean 166 square miles), with a quarter of a million inhabi- includes Puerto Rico, Cuba, and the Dominican tants. Although the islands share in common a long Republic. The Francophone Caribbean includes Haiti history of colonialism beginning in the fifteenth cen- and the French Antilles (Guadeloupe, Dominica, tury, slavery, and economies once dominated exclu- Martinique, and St. Lucia) in the Lesser Antilles island sively by sugar plantations, each island has its group. The Netherland Antilles (Aruba, Bonaire, distinctive history, culture, and language or languages. Curaçao, St. Eustatius, Saba, and St. Maarten) are all Politically, they comprise 23 entities, ranging from that remain of the former Dutch empire in the those that are still quasicolonies, such as the French Caribbean. Antilles (a French département) and Puerto Rico (offi- The native American population was exterminat- cially a commonwealth of the United States), to those ed first by the Spanish in the Greater Antilles (Cuba, that are independent, such as Haiti (formerly a French Puerto Rico, and the Dominican Republic) and by colony), the Dominican Republic (originally a Spanish the British and French in the Lesser Antilles. The colony), and Jamaica (formerly a British colony). indigenous Arawakan languages became extinct as Some territories, such as Guyana, have passed through their speakers died as a result of conquest, disease, a series of colonial regimes before becoming inde- or forced labor, and their remnants were absorbed in pendent. the Hispanicized population. The islands were Divided linguistically and culturally, the West repopulated through the introduction of African Indies can be categorized into four distinct groups slaves to cultivate sugar and coffee plantations according to their officially recognized languages, established by European settlers during the seven- usually the language of the former or current coloniz- teenth and eighteenth centuries. After the abolition er. The largest group is the Anglophone Caribbean, of slavery in the British, French, and Dutch territo- including the majority of islands outside the Greater ries in the nineteenth century, indentured labor was 176  CARIBBEAN introduced for plantation work, particularly from In their everyday lives, the majority of people in the northern India. Caribbean speak creole languages, which have devel- The historical partition of the West Indies among oped from contact between the languages of the colo- the various European colonizers (Spanish, French, nizers and those of the slave laborers. Nevertheless, British, Dutch, and Danish) has had a number of the majority of children throughout the Caribbean con- important consequences. Eric Williams, first and tinue to be educated in metropolitan languages rather long-time Prime Minister of newly independent than in their own native creole languages. The dispar- Trinidad and Tobago (1962–1981) and founder of the ity between home and school language has contributed People’s National Movement, which led the country to a high rate of educational failure and low rates of to independence, believed that the absence of an literacy. Compare Haitian Creole French Mwen indigenous Caribbean language was a major obstacle grangou with standard French J’ai faim ‘I’m hungry’. to decolonization. This meant that there was no rally- Public attitudes to the creoles still tend to be largely ing point for the nationalist movement. Moreover, the negative, and the comments made by some speakers of absence of a common language was a hindrance to French Creole from the Eastern Caribbean are typical: political, economic, and cultural contacts between ‘It’s broken French, you can’t write it down. No, it’s the islands. The communication gap existing between not a language’. Speakers often have no special names the elite, who speak metropolitan European lan- for their languages other than patois (patwa) or broken guages, and the masses, who speak creoles, also English/French. poses serious problems of social and political organ- Despite the school’s rejection of the children’s ization. Interisland connections have remained weak, native language, creoles today are the most powerful because each colonizing power tended to dominate vehicles by means of which local Caribbean identities relationships with its own colonies. Where interis- are formed and expressed. The absence of a high cul- land linkage has occurred, it has been confined large- ture expressed through an indigenous language has ly to islands belonging to the same European meant that the voice of the masses and popular culture language group (each of these usually being of has generally been expressed through creoles, in both the same colonial affiliation). The Hispanophone spoken and written forms. islands also have a link to Latin America, which is In Jamaica, in particular, a strong, black national lacking in the English-, French-, and Dutch-speaking identity has gone hand in hand with nationalist politics territories. Although plantations continue to occupy to foster a powerfully vibrant creative arts movement, large portions of land, their social significance whose influence has extended far beyond Jamaica and has declined as people have left the countryside, the Caribbean. Rejecting the term dialect because it either moving to the towns or emigrating to other suggested inferiority, Edward Kamau Brathwaite countries. (1984) argued for the use of what he called ‘nation lan- Major metropolitan languages of European origin guage’ (Jamaican Creole English) in poetry as a way that play a role in the Caribbean today are the lan- of capturing the sounds and rhythm of oral traditions guages of the former and present colonial administra- of performance. Inspired by hearing a recording of tions: English, Dutch, French, and Spanish. In addition T.S. Eliot reading from The Waste Land, Brathwaite to the standard varieties of these languages, there are a urged poets to model their poetry on the African- number of colonial varieties of French in St. derived rhythms of calypso to break the pentameter, Barthélemy (St. Barts) and St. Thomas. Standard which other New World poets before him, such as Walt Dutch is an official language in the Netherlands Whitman, had also sought to undermine. Brathwaite Antilles, the only remnant of the Dutch empire. The was one of the founders of the Caribbean Artists Caribbean basin constitutes a clearly definable dialect Movement in 1966 in London, who fostered the devel- area of American Spanish, including the varieties spo- opment of a West Indian literature rooted in the lan- ken in Cuba, Puerto Rico, the Dominican Republic, a guages and experiences of the islands. Through the few places in Trinidad, Guyana, the Caribbean coast of commercial success of performers such as Bob Venezuela and Colombia, and some coastal areas of Marley, Mikey Smith, Linton Kwesi Johnson, and Central America and Mexico. Benjamin Zepphaniah, or Mutabaruka, whose music Languages of Southeast Asian and Asian origin and sound poems were on the British reggae music include varieties of Hindi, Urdu, and Chinese brought charts in the late 1970s and early 1980s, the once his- by indentured laborers. These languages are primarily torically devalued Caribbean popular culture has nonstandard varieties used as home languages with no become part of multicultural Britain. wider currency or status, and the younger generation The Anglophone Caribbean has a population are often no longer fluent in their use. of approximately six million. The three largest 177 CARIBBEAN English-creole-speaking communities are Jamaica, tradition of written literature, particularly poetry, in Trinidad and Tobago, and Guyana. In most of these the French Antilles. Haiti, occupying the western communities, standard (usually British) English is the third of the island of Hispaniola, which it shares with official language and is used in most public sectors of the Dominican Republic, has the largest number of communication. Despite the fact that the creoles of the creole speakers (more than six million) in the eastern Caribbean are older, they are closer to standard Caribbean. Independent since 1984, 90% of Haitians English than the creoles of the western Caribbean. speak creole as their only language, and are cultural- This is because of the presence of fewer Africans in ly and linguistically isolated from France. A bilingual the early settler population, allowing regional forms of elite has continued to maintain French, and French English to be established. continues to dominate in the public and formal sec- In a small number of places, the creoles coexist tors, such as government and education. Since 1983, with a French creole (e.g. in St. Lucia and Dominica) Haitian Creole has been a co-official language with or with other ethnic languages (e.g. Belize, Guyana, French and has been used in some primary schools as United States Virgin Islands, and Trinidad). In most well as in adult literacy campaigns. Haitian Creole cases, there is a broad spectrum of varieties, which can has one of the richest literatures of any creole lan- be arranged along a continuum from those that are guages, dating back to the nineteenth century. A cre- ‘deepest’, i.e. furthest removed from standard English ole variety of Spanish called Palenquero is spoken by and spoken by the rural uneducated, to those that are older people in an isolated village called El Palenque closest to standard English. Most people know several de San Basilio, south of the city of Cartagena on the varieties and are able to switch between them depend- Caribbean coast of Colombia. Papiamentu is a creole ing on the context. Compare some of the variants from based on Portuguese and Spanish spoken on the Guyanese Creole, mi gii am/mi bin gii am/mi bin gii ii Leeward Islands of the Netherland Antilles (Aruba, with a geev ii/a geev him ‘I gave him’. The deepest Bonaire, and Curaçao) just north of Venezuela. form of creole uses the object form me instead of stan- Papiamentu has the highest status of any of the dard English I; past tense marking is optional by Caribbean creoles; it is spoken by virtually all social means of bin. classes across a wide range of contexts and is used in The Afro-Caribbean region has been one of the the media. vibrant areas in the world producing literature in Only three creole forms of Dutch are known to English, as evidenced by the fact that Nobel prizes and have resulted from the presence of Dutch traders and other prestigious awards have gone to some of its writ- settlers in the Caribbean during the seventeenth and ers, such as Wole Soyinka from Nigeria and Derek eighteenth centuries. A variety called Negerhollands Walcott (born 1930) from St. Lucia, some of whose (‘Black Dutch’) was present until just recently in the works draw on pidgin and creole English. Virgin Islands, ruled first by Denmark and then by French creoles are spoken primarily on the French the United States. There it was replaced by local Antilles islands (Guadeloupe, Dominica, Martinique, forms of English. The other two nearly extinct and St. Lucia) of the Lesser Antilles and Haiti, as well varieties, Berbice Creole Dutch and Skepi Creole as in former French Guaiana on the South American Dutch, arose in what were once separate colonies iso- mainland between Suriname and Brazil, today a lated by great rivers and wildernesses, but became département of France. Although each creole variety subsequently British Guiana (later independent spoken in the French Antilles is slightly different, Guyana) in northern South America. They are now they are sufficiently similar to one another to consti- spoken only by older people and are not mutually tute a group distinct from Haitian and from the vari- intelligible. eties spoken in the Commonwealth Antilles (Trinidad and the Windward Islands of Dominica, St. Lucia, and References Grenada), which were British from the eighteenth century until their recent independence. In these Allsopp, Richard. 1996. Dictionary of Caribbean usage. countries, English is now the official language. Oxford: Oxford University Press. Brathwaite, Edward Kamau. 1984. The history of the voice: Hence, the creoles coexist with English creoles and The development of nation language in Anglophone standard English and have been influenced by Caribbean poetry. London: New Beacon Books. English, whereas Haitian and the French Antilles vari- Dalphinis, Morgan. 1985. Caribbean & African languages. eties have continued to be influenced by French. London: Karia Press. Creole French is still the main vernacular on Devonish, Hubert. 1986. Language and liberation: Creole lan- guage politics in the Caribbean. London: Karia Press. Dominica and St. Lucia, although English and creole Holm, John. 1989. Pidgins & Creoles. Vol. II. Reference sur- English are gaining ground. There is a fairly strong vey. Cambridge: Cambridge University Press. 178  CASE Holm, John. 1994. English in the Caribbean. Cambridge histo- Ramchand, Kenneth. 1970. The West Indian novel and its back- ry of the English language, Vol. V, English in Britain and ground. London: Faber. overseas: origins and development, ed. by Robert Williams, Eric. 1970. From Columbus to Castro: the history of Burchfield. Cambridge: Cambridge University Press. the Caribbean 1492–1969. New York: Harper & Row. Lawton, David L. 1982. English in the Caribbean. English Winford, Donald. 1991. The Caribbean. English around the as a world language, ed. by Richard W. Bailey and world. Sociolinguistic perspectives, ed. by Jenny Cheshire. Manfred Görlach. Ann Arbor: University of Michigan Cambridge: Cambridge University Press. Press. SUZANNE ROMAINE Morris, Carol and Hubisi Nwenmely. 1993. The Kwéyôl lan- guage and literacy project. Language and Education 7(4). See also English; French Language; Pidgins and 259–70. Creoles Case The notion of case was inherited from traditional syntactic relations only. In other words, case marking grammars describing languages such as Latin, shows only whether or not a given noun is e.g. the sub- Sanskrit, Russian, and German. In these languages, ject or an object, independent of the meaning of the nouns (or noun phrases) differ according to their par- words involved. Noam Chomsky’s Lectures on ticular function in the sentence; i.e. a subject noun Government and Binding (1981) gives a first explicit would be slightly different when used as an object. formulation of case theory as a separate module of Languages differ very much with respect to how case Universal Grammar. is manifested. There are languages, such as Chinese, The Chomskyan theory of case relies crucially on that do not mark nouns for case at all. There are also the assumption that case is present in all languages (in languages with a partial manifestation of case. In Chinese as well as in English or Russian) and that English, for example, regular noun phrases do not overt case marking is just a manifestation of an show case marking, but pronouns do: he (nominative abstract case, which is assigned to noun phrases in a case, subject) vs. him (accusative case, object) vs. his particular syntactic environment. Although traditional (genitive, or possessive, case). Then, there are lan- grammars use a fairly large number of cases, Chomsky guages that have a rich case system, such as German, eliminated all but three: nominative (assigned to the in which regular noun phrases are marked for case: subject position of a finite clause, e.g. John /he speaks Der Hund bellt ‘The dog barks’ (nominative, subject) well), accusative (assigned to an object position of a vs. der Kopf des Hundes ‘the dog’s head’ (genitive) vs. transitive construction, e.g. We see John/him), and Ich gebe dem Hund einen Ball ‘I give a ball to the dog’ oblique (assigned by prepositions, e.g. of John/him, to (dative, indirect object) vs. Ich sehe den Hund ‘I see John/him, etc.). the dog’ (accusative, direct object). For case to be functional, the syntactic environ- Given such a variation, there was a long debate in ment has to fulfill certain conditions, among them the linguistics on whether case is universally active in all requirement that a ‘case assigner’ has to be present. In languages or whether this notion is relevant only in the English, verbs, prepositions, and tense can function as description of languages that actually mark their nouns case assigners. A sentence such as I am happy John is accordingly. In the first half of the twentieth century, ungrammatical, because there is no word adjacent to the most influential theories on case ignored its syn- John that could possibly assign oblique case to it. I tactic nature and aimed to specify the meaning of dif- am happy for John, on the other hand, does contain ferent case markers (prepositions, inflections, etc.). the case-assigning preposition for and is thus gram- Charles Fillmore (1968) introduced the notion of case matical. into the framework of generative grammar, arguing Furthermore, a noun phrase must stand in a partic- that case reflects meaning relationships between e.g. ular relation with its case assigner. A sentence such as the verb and the subject or object in a sentence. In the John believes that him to be happy is ungrammatical, late 1970s, however, Jean-Roger Vergnaud shifted per- because the noun him is structurally too far from the spective and claimed that case is a manifestation of case-assigning verb believes to receive the proper 179 CASE accusative case. However, the verb is able to exert the The notion of inherent case implies that nouns, as proper influence on the noun phrase if the comple- well as adjectives, can be regarded as case assigners. mentizer that is absent: John believes him to be Thus, in a gift to my friend and a gift of my friend, the happy. case relationship is one between the two noun phrases Thus, the case assigner and the case-receiving a gift and my friend. Thus, a gift is said to be the case noun phrase have to be directly adjacent. This is also assigner, whereas my friend receives case. The prepo- shown by the contrast between the following sen- sitions to and of are interpreted as realizations of the tences: I see rarely him and I see him rarely. The first dative and genitive cases, respectively. one is ungrammatical, because the adverb rarely ‘Abstract case’ remains a very important concept in intervenes between the noun phrase him and the case- linguistics. The question about the nature of case and assigning verb see. The second one shows rarely at its role in syntax is still open. the sentence-final position, see is adjacent to him, and accusative case can be assigned. The sentence is thus grammatical. References According to early syntactic theories, case was dis- Blake, Barry J. 1994. Case. Cambridge: Cambridge University cussed solely with respect to the form of nouns. Press, 2nd edition, 2001. However, Chomsky argued in Knowledge of Chomsky, Noam. 1981. Lectures on government and binding. Language (1986) that case seems to be crucial for the Dordrecht: Foris. proper interpretation of the sentence. Thus, case is not Chomsky, Noam. 1986. Knowledge of language: its nature, origin, and use. New York: Praeger. only a formal requirement but also plays a role in Fillmore, Charles. 1968. The case for case. Universals in lin- meaning. guistic theory, ed. by Emmon Bach and Robert T. Harms. Case is an abstract notion and can be divided into New York: Holt, Rinehart, and Winston. two kinds: ‘structural case’ and ‘inherent case’. Haegeman, Liliane. 1991. Introduction to government and Nominative and accusative are considered structural binding theory. Oxford and Cambridge, MA: Blackwell; 2nd edition, 1994. cases, because they refer only to the formal proper- Lasnik, Howard. 1999. Minimalist analysis. Oxford and ties of the sentence: they mark whether a given noun Malden, MA: Blackwell. is a subject or an object. Dative, genitive, etc., i.e. the EGOR TSEDRYK oblique cases, are inherent cases because they depend on a meaning relationship between the case See also Chinese (Mandarin); Chomsky, Noam; assigner and the noun phrase that is assigned the English; German; Grammar, Traditional; Univer- case. sal Grammar Causation The notion of CAUSATION, whereby an agent causes a productive process by which a language encodes an event to occur has long been considered to be a causation (PRODUCTIVE CAUSATIVES). Lexical fundamental semantic primitive. This can be seen in causatives include predicates that inherently denote a work in a variety of traditions: Generative causative relation (1) or can form part of an inchoat- Semantics, Functional Linguistics, Cognitive ive/causative alternation (2). Productive causatives Linguistics, Conceptual/Event Structure, Role and involve a separate causative predicate, either a sepa- Reference Grammar, and Proto-Roles, to name a rate word (ANALYTIC CAUSATIVES—3a) or a causative few. Many of these approaches assume that there is morpheme (SYNTHETIC CAUSATIVES—3b): a fundamental relation between an agent of causa- Lexical Causatives: tion (CAUSER) and the CAUSED EVENT. In some cases, the causation is effect through an intermediary (1) The farmer killed the duckling. Inherent (CAUSEE). causative The expression of causation can be intrinsic to cer- (2) a. The window broke. Inchoative tain predicates (LEXICAL CAUSATIVES) or can be part of b. The child broke the glass. Causative 180  CAUSATION Productive Causatives: causee argument is encoded, there is a good deal of cross-linguistic consistency; this is true regardless of (3) a. The farmer made the duckling die. Analytic whether a productive causative is expressed analytical- causative ˆ ly or synthetically (see Gibson and Raposo 1986). Two b. Adam çocuga kay aç-tür-d Synthetic basic patterns are widespread—one having to do with causative (Turkish) the valence of the base predicate, and the other with ‘The man made the child open the door.’ the semantics of the causative predicate: A good deal of interest with respect to the linguistics (6) Causee encoding—valence based: of causation comes from the manners in which lan- guages tend to productively encode the notion of causa- (a) If the base predicate is INTRANSITIVE, the tion. This article surveys two dimensions of the encoding causee is encoded as a DIRECT OBJECT. of causation: CLAUSALITY and CAUSEE ENCODING. (b) If the base predicate is TRANSITIVE, the A striking aspect of productive causative construc- causee is encoded as an INDIRECT OBJECT or tions is the manner in which they simultaneously exhibit as an OBLIQUE. both biclausal and monoclausal characteristics. Because (7) Causee encoding—semantically based: productive analytic causatives have two distinct predi- (a) If the causee undergoes DIRECT CAUSATION, cates, a base predicate (e.g. die in 3a) and a causative it is encoded as a DIRECT OBJECT. predicate (e.g. make), it is not surprising that they behave (b) If the causee undergoes INDIRECT CAUSA- as two clauses. More surprising are the monoclausal TION, it is encoded as an INDIRECT OBJECT or aspects of these constructions. For example, analytic as an OBLIQUE. causatives in Spanish appear to be standard biclausal Spanish causatives exhibit both of these encoding constructions with embedded infinitival clauses: patterns: (4) El maestro les hizo leer ese libro a los estudi- (8) Causee encoding based on valence of base antes. predicate: ‘The teacher made the students read that book.’ a. José los hizo trabajar. Intransitive base pred- However, on closer inspection, various local icate dependencies, which are normally restricted to single ‘Jose made themDO work.’ clauses, can span the two predicates. For example, the b José les hizo leer el libro. placement of object clitics is normally clause-bound- Transitive base predicate ed; the clitic occurs with the verb of which it is an ‘Jose made themIO read the book.’ object. However, in the case of causative construc- (9) Causee encoding based on directness of causa- tions, a clitic may (and sometimes must) occur with tion: the causative predicate (known as CLITIC CLIMBING): a. La hice probarlo a la fuerza. (5) El maestro se lo hizo leer a los estudiantes. Direct causation ‘The teacher made the students read it.’ ‘I made herDO try/taste it by force.’ This is one of several clause-bounded phenomena b. Le hice probarlo diciéndole que era that are mitigated across causative and base predicates, riquísimo. and a way in which causative constructions, even Indirect causation when expressed analytically, share characteristics with ‘I had herIO try/taste it by telling her it monoclauses (cf. Aissen 1979; Aissen and Perlmutter was delicious.’ 1983; Rizzi 1978; Burzio 1986, among others). This Analyses that deal with the two types of encoding has led to analyses in terms of COMPLEX PREDICATE for- patterns include those of Comrie (1981), Cole (1983), mation. While the mechanics of such analyses have and Ackerman and Moore (1999). differed, the basic insight is the same: two predicates Causation is a pervasive semantic notion that define a single clausal domain (see Moore (1996) for exhibits strikingly consistent patterns of grammati- a survey of approaches). cization cross-linguistically. Because the causative and base predicates form something like a single, complex predicate, the argu- References ments of the two predicates become coarguments. In this way, the argument that would have been the Ackerman Farrell, and John Moore. 1999. Syntagmatic and paradigmatic dimensions of causee encodings. Linguistics subject of the base predicate (the CAUSEE argument) and Philosophy 22. 1–44. can no longer be a subject, as that role is taken by the Aissen, Judith L. 1979. The syntax of causative constructions. CAUSER argument. While languages vary in how the New York: Garland. 181 CAUSATION Aissen, Judith L., and David M. Perlmutter. 1983. Clause Jackendoff, Ray. 1991. Semantic structures. Cambridge, MA: reduction in Spanish. Studies in relational grammar, Vol. 1, MIT Press. ed. by David M. Perlmutter. Chicago: The University of Moore, John. 1996. Reduced constructions in Spanish. New Chicago Press. York: Garland. Burzio, Luigi. 1986. Italian syntax: a government and binding Rizzi, Lugi. 1978. A restructuring rule in Italian syntax. Recent approach. Dordrecht: Reidel. transformational studies in European languages, ed. by Cole, Peter. 1983. The grammatical role of the causee in uni- Samuel J. Keyser. New York: Academic Press. versal grammar. International Journal of American Strozer, Judith Reina. 1976. Clitics in Spanish. Doctoral Linguistics 49. 115–33. Dissertation. Los Angeles: University of California. Comrie, Bernard. 1981. Language universals and linguistic Talmy, Leonard. 1985. Force dynamics in language and typology: syntax and morphology. Chicago: The University thought. Papers from the parasession on causatives of Chicago Press. and sgentivity, ed. by William H. Eilfort, Paul D. Kroeber, Gibson, Jeanne, and Eduardo Raposo. 1986. Clause union, the and Karen L. Peterson. Chicago: Chicago Linguistic stratal uniqueness law, and the Chômeur relation. Natural Society. Language and Linguistic Theory 4. 295–331. JOHN MOORE Celtic Languages ‘Gallia est omnis divisa in partes tres, quarum unam Scotland and the now extinct Manx, whose alleged incolunt Belgae, aliam Aquitani, tertiam qui ipsorum last native speaker died in 1974. The Brittonic Celtic linguae Celtae, nostra Galli appellantur. Hi omnes languages are Welsh in Wales, Breton in Britanny lingua, institutis, legibus se differunt.’ and Cornish, which died out in the eighteenth or ‘All Gaul is divided into three parts, one of which nineteenth century. the Belgae inhabit, the Aquitani another, those who in In pre-Roman times, Continental Celtic languages their own language are called Celts, in our Gauls, the were spoken all over Europe. Lepontic is attested in third. All these differ from each other in language, cus- northern Italy before the arrival of the Gauls. Gaulish toms and laws.’ (Caesar, De bello gallico I,1) (in today’s France) is known through personal names Celtic is classified as a separate branch of the Indo- in Latin texts and some inscriptions in Greek script European (IE) language family. Like Greek and Latin, and a considerable amount in Latin; Celtiberian (or Celtic is a Centum language. It can be further subdi- Hispano-Celtic) on the Iberian peninsula in Iberian vided into Continental and Insular Celtic. The division script (Iberian being a non-Indo-European language), into Continental and Insular Celtic languages is not Galatian in Asia Minor, today’s Turkey. A contested merely geographically based, but also genetically issue is the ‘celticity’ of Lusitanian, attested in three based. While the Gaulish and Brittonic languages pre-Roman inscriptions in the west of the Iberian share some features, excluding Celtiberian and peninsula. Archeologically, the original native place of Goidelic, the Insular and Continental division is based the Celts must have been in Central Europe, some- on commonalties in sound shifts from IE. where between Bohemia and Bavaria. The earliest Although scholars agree on the relationship clearly datable sources go back to the sixth century between ancient and modern Celtic languages, it BCE. must be stressed that speakers of different Today, the surviving Celtic languages are confined Continental Celtic languages did not refer to them- to the westernmost parts of Europe, the Celtic fringe: selves as mutually related ‘Celts’. These languages the west, northwest and southwest of Ireland, Scottish died out before the end of the Roman Empire, while Highlands and Western Isles, large parts of Wales and evidence of the Celtic languages in the British Isles western Britanny. Remnants of Celtic-speaking people starts in the Early Middle Ages. This makes compar- are also to be found in the Americas: Welsh-speaking ison of ancient modern Celtic languages problemat- people in Patagonia (Argentina) and some Scottish ic. The living Celtic languages can be further Gaelic in Nova Scotia (Canada). Exactly when and in subdivided into Goidelic (Gaelic) and Brittonic how many migration waves the Celts came to Ireland (British). The descendants of Goidelic are Modern is hard to determine. The oldest Irish coherent manu- Irish and Scottish Gaelic spoken in Ireland and script text Amra Choluim Cille goes back to 600 CE, 182  CELTIC LANGUAGES making Irish the oldest attested continental European Scottish Gaelic language after Greek and Latin. Within the written Before the Middle Ages (approximately tenth century Irish tradition, three main periods can be distin- CE), there was no difference between the Gaelic spo- guished: Old Irish, Middle Irish, and Modern Irish. ken in Scotland and the one in Ireland, since they both The Great Famine (1845–1948) nearly eradicated the developed from the same parent language, Old Irish. Modern Irish language. More than one million people The Reformation played a key role in linguistic self- starved, while another million fled to America. Today, definition. After the Jacobite Rising in 1745, a huge after efforts to revive the endangered language, only deportation of Scottish Gaelic-speaking population to three main dialects are distinguished according to the Overseas initiated the blackest chapter in the history of provinces in which they are spoken: Connacht, Ulster, Scotland: the Clearances. In the first half of the nine- and Munster Irish. A standard language, mainly based teenth century, Gaelic-speaking peasants were system- on Munster Irish, also exists. Irish has been the official atically driven from their land in order to give way to language of Ireland since the establishment of the Free extensive sheep breeding. Today, Gaelic is a commu- State in 1922. nity language only in the Western Isles and on parts of In the study of Celtic languages, phonology plays Europe’s western seaboard. Many linguistic phenome- a large role as Old Irish preserves few case or per- na encountered in Modern Irish, especially the Ulster sonal endings. As early as Common Celtic times, the Dialect, are found in Scottish Gaelic as well, including sound p had disappeared. Therefore, the IE word for sound changes, mutations, and spelling. father appears in Old Irish as athair. According to some scholars (see Untermann and Wodtko 1997), the existence of p in Lusitan makes it probable that Welsh this loss was not a common Celtic phenomenon. Another peculiarity of Celtic languages are the so- Written Old Welsh texts are attested from the ninth called initial mutations. Simply speaking, mutation century. Until 1288, Welsh had status in administra- means that the sound dropped at the end of one word tion and government. The Bible translations in 1567 affects the first sound of the following word. One of (New Testament by William Salesbury and Richard the most interesting characteristics of Celtic lan- Davies) and 1588 (the whole Bible by William guages is their syntax: the basic word order is Morgan) played a crucial role in Welsh history. Today, Verb–Subject–Object, where the verb stands at the linguists distinguish two main dialects based on lexi- beginning of the sentence. For example, in Irish, cal variation: Northwalian Welsh and Southwalian D’oscail mé an doras (lit. ‘Opened I the door’), the Welsh, each comprising a number of more localized verb not the subject is first. dialects, and the Literary or Formal Welsh language, itself comprising a number of written varieties or ‘dialects’. Irish Welsh has a few distinctive consonant sounds, the The sound system of Irish is fairly complicated com- most unusual being the voiceless lateral fricative ll pared to English. For each consonant, there is a and the voiceless alveolar trill rh. The unusual palatal equivalent, making it similar to Russian. spellings w for [u] and u for [i] makes it seem as if Lenition, a sound assimilation, and eclipsis, the omis- there are too few vowels in a word. Three initial muta- sion of symbols, make the orthography look some- tions, soft mutation, aspirate mutation, and nasal what strange to the native speaker of English. mutation, are present, and Welsh shares the VSO syn- Examples include phrases like i nGaillimh and ó tactic pattern: Agorodd y dyn y drws (lit. ‘Opened the Bhaile Átha Cliath for ‘in Galway’ and ‘from Dublin’ man the door’). and surnames like Ní Dhomhnaill for female ‘Mac Donald’. Two verbs for ‘to be’ exist in Irish, one being Breton the copula is and the other the substantive verb tá. Their distribution is very similar to Spanish ser and Brittany was settled by people fleeing from estar. There is no word for ‘to have’ in Irish, as for Anglo-Saxons between the fourth and sixth century many—in other languages clearly verbal concepts CE. Breton is therefore a Britannic language, even like ‘to love’, ‘to owe’, ‘to know’, etc., are all phrasal though Britanny is geographically European main- verbs containing the verb ‘to be’ (e.g. Ir. Tá grá agam land. There is no unique standard written language. duit lit, ‘there is love at me to you’). This gives us a The main dialects are Cornouaille, Leon and Tregor, picture of a language that on the one hand has a clear- and Vannetais. Due to historical circumstances, ly verb-centered syntax, but on the other very few Britanny has always been bilingual. As a consequence strictly verbal concepts. of permanent language contact, Breton has a very 183 CELTIC LANGUAGES French-sounding accentuation and sound system and References also the strongest tendencies of all Celtic languages Ball, Martin J. (ed.) 1993. The Celtic languages. London and toward dropping its verb initial sentence structure. New York: Routledge. Celtic Scholarship began in 1707 with the Welsh Black, Ronald, William Gillies, and Roibeard Maolalaigh (eds.) historian Edward Lhuyd (1660–1709) and his unfin- 1999. Celtic connections. Proceedings of the Tenth ished work Archaeologia Britannica. In 1853, Johann International Congress of Celtic Studies (Vol. 1, Language, literature, history, culture). East Linton: Tuckwell. Kaspar Zeuss (1806–1856) published his Grammatica Durkacz, Victor Edward. 1983. The decline of the Celtic lan- Celtica, which must be regarded as a further milestone guages. Edinburgh: John Donald. in the establishment of Celtic Studies. A few more Hindley, Reg. 1990. The death of the Irish language. London prominent Celticists were German native speakers: and New York: Routledge. Ernst Windisch (1844–1918), Rudolf Thurneysen Macaulay, Donald. 1992. The Celtic languages. Cambridge: Cambridge University Press. (1857–1940), Kuno Meyer (1858–1919), Julius Maier, Bernhard. 2000. Die Kelten. München: Beck. Pokorny (1887–1970), and last but not the least Dane McCone, Kim. 1992. Relative Chronologie: Keltisch. Holger Pedersen (1867–1953). Old Irish is most Rekonstruktion und relative Chronologie, ed. by Robert prominently studied, followed by Middle Welsh. Beekes, Alexander Lubotsky and Jos Weitenberg, Innsbruck: Recent developments in European archeology have led Innsbrucker Beiträge zur Sprachwissenschaft. McCone, Kim. 1996. Prehistoric, old and middle Irish. Progress to an increase in often-controversial works on in medieval Irish studies, ed. by Kim McCone and Katherine Continental Celtic. Simms. Maynooth: Department of Old Irish. What all Celtic languages spoken nowadays have in Schrijver, Peter. 1995. Studies in British Celtic historical common is that they are endangered minority lan- phonology. Amsterdam, Atlanta: Rodopi. guages, due to the predominance of English or, in Sims-Williams, Patrick. 1998. The Celtic languages. The Indo- European languages, ed. by Anna Giacalone Ramat and Britanny, French. Celtic linguistic history must be Paolo Ramat. London and New York: Routledge. examined within the context of industrialization, and Uhlich, Jürgen. 1999. Zur sprachlichen Einordnung des its negative influence on indigenous languages. Oral Lepontischen. Akten des zweiten deutschen Keltologen- tradition is dying out in the wake of modern media, Symposiums, ed. by Stefan Zimmer, Rolf Ködderitzsch and and today nearly all adult Celtic native speakers are Arndt Wigger. Tübingen: Niemeyer Untermann, Jürgen, and Dagmar Wodtko. 1997. Monumenta bilingual. The British Labour Party has been regarded Linguarum Hispanicarum, Vol. IV (Die tartessischen, by some as the ‘rescuer’ of the Celts in Great Britain keltiberischen und lusitanischen Inschriften). Wiesbaden: since it has engaged in decentralization and the found- Reichert. ing of regional parliaments and assemblies. However, ELVIRA M. VESELINOVIC the future of the Celts as well as the languages remains to be seen. See also Indo-European 1: Overview Chafe, Wallace From the very beginning of his linguistic career language with the agent–patient system of role mark- Wallace Chafe has been involved in studies of ing on the verb. Seneca examples cited in Chafe’s American Indian languages, in particular, of the work often demonstrate vast differences in conceptu- Iroquoian and Caddoan language families. In 1962, alization possible across human languages, especially he published a general survey of the sociolinguistic as compared to European languages. Chafe also stud- state of the native languages of North America, indi- ied languages of the Caddoan language family, dis- cating the number of speakers and the age of the tantly related to Iroquoian. youngest speakers, which remains relevant even now, In 1970, Chafe published a monograph entitled 40 years later. A semantically based sketch of Onondaga (Iroquoian) Among American Indian languages, Chafe has that was revolutionary for its time, being markedly dif- been involved most of all with Seneca, an Iroquoian ferent from both traditional American descriptivism language spoken in New York state, a polysynthetic with its emphasis on formal patterns and the growing 184  CHAFE, WALLACE generative approach based on axioms of universal syn- explored in isolation from each other. Chafe evalu- tax. Chafe chose meaning as his starting point and pre- ates both disciplines studying those two domains, sented the linguistic material as a system of mappings that is, modern linguistics and psychology, in a very ‘meaning => form’. In general, Chafe is unique among critical vein, in particular, for their reliance upon arti- the American linguists of his generation in never being ficial data and distrust toward introspective methods tempted to accept the generative Chomskyan of research. Chafe emphasizes the use of introspec- approach. tion as a source of insights into language and cogni- This orientation toward semantics was also the tive processes, as well as the priority of natural data. chief element of his well-known book Meaning and The latter concern results from Chafe’s inherent con- the structure of language (1970), later translated into nection with the American ethnographic tradition five languages. One of the main components of that that has always emphasized the importance of natural book is a semantic classification of verbs. Along empirical data. with Charles Fillmore, Chafe was among the first In this 1994 book, Chafe explores two main prob- who realized the importance of semantic roles. The lems. The first is the explanation of linguistic phe- book also clearly demonstrated a typical feature of nomena on the basis of processes occurring in the Chafe’s subsequent publications — a very clear and speaker’s consciousness. Consciousness can be imme- simple style of writing that is of great persuasive diate (a reflection of what is here and now) and dis- force. placed (memories, imagination). Immediate In the 1970s, the main focus of Chafe’s theoretical consciousness is more basic and elementary, while research interests shifted to natural discourse. At that displaced consciousness can only be understood as a time, the trend that later acquired the label ‘function- modification or complication of immediate conscious- alism’ and became a radical alternative to generativism ness. The second problem is the comparison of spoken started to take shape in the United States. Although and written language. Spoken language is more basic Chafe was never an ideolog, he became one of the and universal, despite the fact that for decades lin- leading functionalists. guists have directed their attention to written or quasi- From the very beginning of his discourse studies, written language only. Chafe gives priority to the Chafe relied on a cognitive approach and tried to spoken mode, but has also studied properties that dis- reconstruct the mental processes of a speaking per- tinguish written language. son. In that respect, he was far ahead of the other stu- The fundamental unit of discourse, according to dents of text. In a widely known 1976 article, Chafe Chafe, is the intonation unit, which is a quantum of approached such familiar phenomena as referential discourse corresponding to one focus of conscious- devices, word order, and choice of subject from a cog- ness. Each intonation unit typically contains one ele- nitive perspective. His studies of the 1970s culminat- ment of new information. The opposition ‘given vs. ed in the 1980 monograph The pear stories in which accessible vs. new’ is responsible for prosodic an unusual method of data collection was used: Chafe (stressed vs. unstressed) and lexical (pronoun vs. and his younger associates (D. Tannen, J. du Bois, P. noun) realizations of referents. The status of syntactic Clancy, among others) showed a short film of a boy subject, central for English grammar, is explained on collecting pears to speakers of different languages and the basis of the notion of the starting point for the ages who later retold what they saw. Thus, the input expression of a focus of consciousness. The important was controlled, and the different versions of the pear light subject constraint suggests that subjects are nor- stories were available for comparison. Many innova- mally given or accessible, and can be new only under tive conclusions concerning processes of the verbal- very special conditions. The intermediate status of ization of visual experience, the dynamics of human accessible information is the base for defining a dis- consciousness, cultural differences in emphasizing course topic, which, in its turn, underlies a definition different layers of information, and the cognitive of sentence. basis of grammatical choices were drawn in that Among various modifications related to written lan- study. guage and the displaced mode of consciousness, prose The following stage of Chafe’s work is reflected in written in the third person and quoted speech are dis- his recent (1994) book Discourse, consciousness, cussed. and time. This work looks into relationships between At present, Chafe is engaged in producing descrip- language and consciousness, the latter understood by tions of the Seneca and Caddo languages, and he is Chafe as the mental system crucially responsible for also investigating the uses of prosody in ordinary discourse production. According to Chafe, language speech, particularly for the expression of emotions and and consciousness cannot be understood and humor, and relations between language and music. 185 CHAFE, WALLACE Among the main characteristics of Chafe’s style of sional societies: Linguistic Society of America; scientific thinking are the ability to look afresh at American Anthropological Association; American ordinary phenomena and to integrate a broad range of Psychological Association; American Psychological facts into a coherent picture, as well as independence Society; Society for Linguistic Anthropology; from the dominant trends of thought. There is no Society for the Study of the Indigenous Languages doubt that Chafe is among the leading figures and a of the Americas; International Society for Historical classic of modern linguistics, but like Sapir he has Linguistics; and International Pragmatics not created a circle of immediate followers and Association. He was Skomp Distinguished Lecturer, his works are not in the mainstream of modern Indiana University, 1997, and received the medal of linguistics. the University of Helsinki in 1998. Biography References Wallace Chafe was born in Cambridge, Chafe, Wallace. 1962. Estimates regarding the present speakers Massachusetts on September 3, 1927. He served in of North American Indian languages. International Journal the US Navy, 1945–1946, and Central Intelligence of American Linguistics 28. Agency, 1950–1954. He received his B.A. in ———. 1967. Seneca morphology and dictionary. Washington: German from Yale University in 1950 and M.A. in Smithsonian Press. ———. 1970. A semantically based sketch of Onondaga. linguistics from Yale University in 1956. He International Journal of American Linguistics, Memoir 25 received his Ph.D. in linguistics from Yale (Supplement to 36 (2)). University in 1958 for a dissertation on the Seneca ———. 1970. Meaning and the structure of language. Chicago: language. He was Assistant Professor of Modern University of Chicago Press. Languages at the University of Buffalo, 1958–1959, ———. 1974. Language and consciousness. Language 50. ———. 1976. The Caddoan, Iroquoian, and Siouan languages. and linguist in the Bureau of American Ethnology, The Hague: Mouton. Smithsonian Institution, 1959–1962. In 1962, he ———. 1976. Givenness, contrastiveness, definiteness, sub- moved to the University of California at Berkeley jects, topics, and point of view. Subject and topic, ed. by and worked there in the Department of Linguistics C.N. Li. New York: Academic Press. until 1986. From 1986 onward, he worked in the ——— (ed.) 1980. The pear stories: cognitive, cultural, and linguistic aspects of narrative production. Norwood, NJ: Department of Linguistics at the University of Ablex. California at Santa Barbara. At present, he is Chafe, Wallace, and Johanna Nichols (eds.) 1986. Evidentiality: Professor Emeritus at UCSB. He was Fellow of the the linguistic encoding of epistemology. Norwood, NJ: Center for Advanced Study in the Behavioral Ablex. Sciences, Palo, 1976–1977. He held a number of Chafe, Wallace. 1994. Discourse, consciousness, and time: the flow and displacement of conscious experience in additional teaching appointments, including the speaking and writing. Chicago: University of Chicago Linguistic Institute of the Linguistic Society of Press. America, in the summer of 1995, Albuquerque, New ———. 1996. Sketch of Seneca, an Iroquoian language. Mexico. Chafe was editor of the Mouton Grammar Handbook of North American Indians. Vol. 17, Languages, Library, 1984–1991; Topic Editor, Oxford ed. by W.C. Sturtevant. Washington, DC: Smithsonian Institution. International Encyclopedia of Linguistics; President ———. 1998. How a historical linguist and a native speaker of the Society for the Study of the Indigenous understand a complex morphology. Historical linguistics Languages of the Americas, 1985–1986; Associate 1997, ed. by M.S. Schmid, J.R. Austin, and D. Stein. Director of the Linguistic Institute of the Linguistic Amsterdam/Philadelphia: John Benjamins. Society of America, Georgetown University, 1985; ———. 1998. Language and the flow of thought. The new psy- chology of language: cognitive and functional approaches to co-Director of the Center for the Study of Writing, language structure, ed. by M. Tomasello. Mahwah, NJ: Berkeley, 1985–1986; and member of the Linguistic Lawrence Erlbaum. Society of America Executive Committee, ———. 2000. Loci of diversity and convergence in thought and 1987–1989. He was also Member of the advisory language. Explorations in linguistic relativity, ed. by M. and editorial boards of: International Journal of Pütz, and M. Verspoor. Amsterdam/Philadelphia: John Benjamins. American Linguistics; Anthropological Linguistics; Chafe, Wallace. Autobiography. Historiographia Linguistica (in John Benjamins Typological Studies in Language press). Series; International Pragmatics Association; Parret, Herman (ed.) 1974. Interview with Wallace Chafe. Humor: International Journal of Humor Research; Discussing language.The Hague: Mouton. Text: An Interdisciplinary Journal for the Study of Kibrik, Andrej A. 1994. Cognitive studies in discourse. Voprosy jazykoznanija 5 [in Russian]. Discourse; and Consciousness and Cognition. In addition, he was a member of the following profes- ANDREJ KIBRIK 186  CHAO YUEN REN Chao Yuen Ren A Chinese–American linguist and one of the leading Europe to work and study with the phoneticians figures in American Structuralism, Chao Yuen Ren Daniel Jones and Stephen Jones, the linguists Joseph made important contributions in the general theory of Vendryes and Antoine Meillet, and the sinologists language and by introducing the Chinese language(s) as Bernhard Karlgren, Paul Pelliot, and Henri Maspero. an object of research to a broader linguistic community. After his return to China, primarily to teach Chinese His work covers a wide range of issues both in theoret- phonology, he spent more than ten years making and ical and in applied linguistics, including fieldwork on organizing surveys of the Chinese dialects, which Chinese dialects (usually conceived today as separate resulted in a number of pioneering studies: the first one languages), the grammar of Modern Chinese, Chinese was on the Wu dialects of Jiangsu and Zhejiang lexicography, practical material for Chinese language provinces (1926–1927), published as his ground-break- teaching, language standardization in China, the gener- ing Studies in the modern Wu dialects (1928), followed al theory of language, and studies in sociolinguistics. by surveys on the Yue dialects of Guangdong and Born into a family belonging to the old scholar and Guangxi provinces (1928–1929), Minnan dialects of mandarin class, Chao’s interest in linguistic variability Shantou and Chaozhou and the dialects of Anhui was aroused during early childhood, when he came into (1934), Jiangxi and Hunan (1935), and Hubei provinces contact with, and learned to speak, a number of differ- (1936). During the late 1920s, he was also active in the ent dialects. His family originally came from the town Committee on Unification of the National Language of Changzhou (between Nanjing and Shanghai), where and headed a group of Chinese linguists who developed the westernmost variety of the Wu dialect group is spo- a new system of Latinization for Chinese, National ken. During the first half of his childhood, however, the Language Romanization (Gwoyeu Romatzyh), which family lived in various places in Hebei province, was unique in using spelling variants instead of diacrit- Northern China, and frequently moved around between ics to express the tones of Chinese. the then provincial capital Baoding and smaller places. Chao had plans for a dialectal survey of the whole In those years, Chao was therefore exposed not only to country, but his fieldwork was eventually interrupted the Wu dialects spoken by the older generations of his by the political development in China and the outbreak family but also to different varieties of Northern of war with Japan. Following these events, in 1938 he Mandarin, which he heard from his surroundings. came to the United States for good and subsequently Although he majored in mathematics and physics at taught at the University of Hawaii; at Yale, where he Cornell University, he increasingly took an interest in met Leonard Bloomfield, Edward Sapir, and Bernard philosophy and was influenced in particular by the Bloch; at Harvard; and, finally, at the University of writings of Bertrand Russell. However, he kept up his California, from where he retired in 1960. strong interest in languages, further stimulated by a Chao’s pioneering fieldwork on the Chinese course in phonetics and other linguistic courses. dialects is clearly one of his main contributions to lin- During those years, he also learned German and guistics. Before him, the European sinologists Henri French, and picked up more Chinese dialects from Maspero (1883–1945) and Bernard Karlgren Chinese fellow students. During his graduate studies (1889–1978) had used Chinese dialect data in compar- in the Department of Philosophy at Harvard, he attend- ative studies for reconstructing earlier stages of ed an introductory course in linguistics and took quite Chinese (in particular, what Karlgren called ‘Ancient a few language courses on the side, among them a Chinese’), but they were interested in Chinese dialects course in Sanskrit. only to the degree that they had preserved the phonet- During the winter of 1920/1921, Chao served as an ic features of those earlier stages. Chao, however, was interpreter to Bertrand Russell during his lecture tour the first to look at the Chinese dialects from a truly in China and made full use of his mastery of the modern perspective, recognizing them as objects dialects. This experience provided a challenge to Chao worth studying for their own sake. He carried out in a way physics never had, and it was then that he wide-ranging dialect surveys using the most modern decided upon linguistics as a major field of study and sound recording devices of the times, and he included research. Therefore, while teaching philosophy at real dialect words found in nonlearned, colloquial Harvard during the next three years, he took linguistic speech. Although Chao’s fieldwork was disrupted by courses at the same time, and subsequently went to the historical events in China and thus remained 187 CHAO YUEN REN fragmentary, it provided a solid foundation for later family returned to Changzhou in 1901, and after both his work, much of which was done by his students. parents died in 1904, he spent a year in his aunt’s home In the field of Chinese grammar and language teach- in Suzhou. After attending a primary school in ing, Chao wrote two epoch-making books in which he Changzhou in 1906, he was sent to Nanjing for three applied the grammatical approach of American struc- years to continue his studies. In 1910, he passed a gov- turalism to the description of Modern Chinese: ernment examination, which earned him a scholarship Mandarin primer, published in 1948 as a textbook for for studying in the United States of America, and he American students learning Chinese, contains an intro- went to the Tsing Hua College in Beijing to prepare him- ductory grammatical sketch that was translated into self. Chao thus studied mathematics and physics at Chinese and had a tremendous impact in China, effec- Cornell University, Ithaca, B.A. (1914), followed by tively introducing the methodology of structuralism to physics and philosophy at Harvard, where he earned his a wide Chinese audience. The grammatical sketch of Ph.D. degree in 1918 with Continuity, a study in method- the Primer provided the best linguistic description of ology, which was supervised by the philosopher and Modern Chinese until 1968, when Chao published A logician Henry Maurice Sheffer. Chao received the post- grammar of spoken Chinese, a monumental work that doctorate Sheldon Fellowship, 1918–1919 and went on was the first comprehensive reference grammar of to teach physics at Cornell in 1919–1920. He then went Chinese written in English, and that is arguably the best back to China to teach mathematics at Tsing Hua description of Modern Chinese to this day. College in Beijing, but was soon asked to interpret for In Chinese lexicography, Chao and Yang’s Concise Bertrand Russell on his lecture tour of China, dictionary of spoken Chinese (1947) made a number 1920–1921. Afterwards, he returned to Harvard to teach of important lexicographic innovations, restricting philosophy and Chinese, at the same time taking lin- entries to those morphemes that occur in the modern guistic courses (1921–1924). He went to Europe to visit spoken or written language, marking the morphemes the sinologist Bernhard Karlgren in Gothenburg, took as ‘free’ or ‘bound’ from the point of view of the mod- courses in phonetics with Daniel Jones and Stephen ern language, and as ‘literary’ if they occur in the writ- Jones in London, and attended lectures by Joseph ten language only. Pronunciation of the entries is given Vendryes, Antoine Meillet, Henri Maspero, and Paul in National Language Romanization. Pelliot in Paris (1924). Returning to China, he taught Among Chao’s works on theoretical phonology, the Chinese phonology and music at Tsing Hua College’s most influential one was his classical paper ‘The non- newly established Institute of Sinology (1925), and then uniqueness of phonemic solutions of phonetic systems’ conducted and organized fieldwork on Chinese dialects (1934/1957), which deals with the definition of the between 1926 and 1936. He was an active member of the phoneme and tries to show that the principles for estab- Committee on Unification of the National Language and lishing a system of phonemes for a given language the chairman of the Linguistics Section of the Institute of depend on a number of possible objectives set up by the History and Philology, Academia Sinica (1929–1938). linguist, some of which are external to the language sys- He was visiting professor at the Sunyatsen University, tem itself. Chao also introduced a convenient device of Canton (1928—1929) and then temporarily returned to representing tones in phonetic or phonological transcrip- the United States of America as the director of the tions, which became known as ‘tone letters’ and was Chinese Educational Mission in 1932–1933. He later adopted for the International Phonetic Alphabet. obtained his American citizenship in 1938 and then Over the years, a number of important articles on the taught literary Chinese as a visiting professor at the problems of phonetics and transcription followed. University of Hawaii (1938–1939) and Yale More than anyone else, Chao laid the foundation (1939–1941) before returning to Harvard once again, as for studying Chinese within the international perspec- a researcher with the Harvard-Yenching Institute’s tive of modern linguistics, and played a key role in Chinese–English Dictionary Project. He also taught introducing the ideas of Western linguistics to linguis- Chinese and Cantonese, and served as head of Harvard’s tic scholars inside China. Outside the area of linguis- wartime Chinese language program (1941–1947). He tics, Chao is remembered as an outstanding translator became Professor of Oriental Languages and Linguistics (Alice’s adventures in wonderland, 1922) and as a at the University of California, Berkeley in 1947 and was composer of Chinese and Western music. made Agassiz Professor of Oriental Languages and Linguistics in 1952. A guest lecturer at National Taiwan University and Kyoto University, Japan in 1959, he Biography retired the following year, in 1960. Chao was born in Tianjin, China, on November 3, 1892. Chao was also President of the Linguistic Society From the age of seven, he was given a traditional of America, 1945; full member of the Academia Chinese education in the family school at home. The Sinica, 1947; Fellow of the American Academy of Arts 188  CHINA and Sciences, 1948; and President of the American ———.1947. Cantonese primer. Oriental Society, 1960. He earned honorary degrees ———.1948. Mandarin primer. ———.1968. A grammar of spoken Chinese. from Princeton University (1946), the University of ———.1968. Language and symbolic systems. California (1962), and Ohio State University (1970) ———.1968. Readings in sayable Chinese. and was named Honorary Professor at Beijing ———. 1975. Yuen Ren Chao’s autobiography: first 30 years, University in 1981. Chao died in Cambridge, from 1892–1921. Ithaca, New York: Spoken Language Massachusetts, on February 24, 1982. Services. ———. 1976. My linguistic autobiography. In Chao. 1–20. ———. 1976. My fieldwork on the Chinese dialects. In Chao. References 26–33. ———. 1976. Aspects of Chinese sociolinguistics, ed. by Bibliography of Yuen Ren Chao’s works, In Chao. 1976. Anwar S. Dil. Stanford: Stanford University Press. 402–15. Chao, Yuen Ren. and Yang Lien-Sheng. 1947. Concise diction- Chao Yuen Ren. 1928. Xiandai Wuyu de Yanjiu (Studies in the ary of spoken chinese. modern Wu dialects). Dil, Anwar S. 1976. Introduction. In Chao. xi–xiv. ———.1930. A system of tone letters. Le Maître Phonétique Egerod, Søren, 1996. Chao Yuen Ren. Lexicon 45. 24–7. Grammaticorum, ed. by Harro Stammerjohann et al., 178–9. ———.1934. The non-uniqueness of phonemic solutions of pho- Tübingen: Niemeyer. netic systems. Bulletin of the Institute of History and Wang, W. S.-Y. 1983. Chao Yuen-Ren. Language 59. 605–7. Philology (Academia Sinica) 4. 363–97; reprinted in Readings ROBIN SACKMANN in linguistics, ed. by Martin Joos, 38–54. Washington, DC: American Council of Learned Societies, 1957. See also Structuralism China The region of China discussed in this essay covers the south throughout much of the history of China. entire territory claimed by the People’s Republic of Southward invasion by Altaic speakers sometimes leads China, including Taiwan. It consists of 23 provinces, to the spreading of Chinese further north, as in the case five autonomous regions, four municipalities, and two of Manchu’s government. On the other hand, southward special administrative regions in an area of over expansions and escapes by Sinitic speakers are respon- 9,630,000 square kilometers. With a population of sible, in part, for the migration of Hmong-Mien and Tai- 1,295,330,000 (in the 2000 census), China is the most Kadai speakers to their present-day habitation. populous country and one of the most diversified Scholars in China generally consider the Sino- regions in terms of languages and cultures in the Tibetan family to consist of four branches: Sinitic, world. No accurate number for languages spoken in Tibeto-Burman, Tai-Kadai, and Hmong-Mien (Miao- China is available, but at least 235 languages (includ- Yao). While the typological profiles of the latter two ing signed languages and ten extinct languages) have are rather similar to that of Sinitic, their genetic rela- existed in this region, cf. Grimes’s Ethnologue (2000). tionship with the family remains to be proven. Table 1 Barring a few exceptions, they belong to one of the summarizes three aspects of the major language following groups (parenthetical figures indicate num- groups as to whether their members (confined to those ber of members and the group’s percentage of total liv- in China) are tone languages, whether morphemes (the ing languages): Sinitic (13, 5.78%), Tibeto-Burman smallest unit that carries a meaning) in the languages (72, 32%), Tai-Kadai (31, 13.78%), Hmong-Mien (32, are predominantly monosyllabic, and in what order the 14.22%), Altaic (25, 11.11%), Austro-Asiatic (24, Subject (S), Verb (V), and Object (O) appear in canon- 10.67%), Austronesian (18, 8%), and Indo-European ical sentences. (4, 1.78%). Figure 1 outlines their approximate distri- The boundary between language and dialect is bution in China. fuzzy in nature, and highly controversial when politi- The map shows that Sinitic sits in the center, but it cal factors are taken into consideration. The case of has spread out to every corner of China. There has been Chinese is a good example of this. The Chinese iden- a general migration flow of people from the north to the tify themselves ethnically as Han, and Hanyu is the 189 CHINA Altaic Altaic River Yellow itic Tibeto-Burman Sin River Yangtze Hmong-Mien Austro-Asiatic Tai-Kadai Austronesian Figure 1. The distribution of major language groups in China. TABLE 1 A Brief Typological Comparison of the the standard language of the new republic. Later in Major Language Groups in China the 1950s, the Government of the People’s Republic Language Tone Monosyllabicity Word Order of China promoted Putonghua ‘common language’ Group Language of Morphemes (also based on Beijing Mandarin) as the national lan- guage. Gwoyu (still in use in Taiwan) and Putonghua Sinitic All Yes S–V–O have little difference in phonology. Their divergence Tibeto- Most Yes S–O–V has largely resulted from further language standardi- Burman zation introduced on the mainland, notably the Tai-Kadai All Yes S–V–O replacement of phonetic characters by pinyin in the Hmong- All Yes S–V–O Roman alphabet and especially the use of simplified Mien characters. The latter has compromised the antedilu- Altaic None No S–O–V vian tradition of a unified written form for the Austro- Few Most S–V–O Chinese since the unification of the warring states in Asiatic China in 221 BCE. Austronesian None No V–S–O Encyclopedia Sinica (1988) estimates the number of speakers of seven major dialect groups of Hanyu in the early 1980s as follows: over 700 million for official language of China. From a strictly linguistic Mandarin, 70 million for Wu, 40 million for Yue, 40 point of view (based on mutual intelligibility), how- million for Min, 25 million for Xiang, 30 million for ever, at least 13 Chinese languages are spoken in Gan, and 37 million for Hakka. Figure 2 shows the modern China (see Figure 2). In the Chinese tradition, relation between 13 identified Sinitic languages. The these are construed as dialects of Hanyu or subdi- tentative placement of Wan Nan is geographically alects therein. Prior to the twentieth century, Classical motivated. Chinese had been used in writing for all purposes, The Yangtze River is the watershed between serving as the official language of the emperor for Northern Chinese and Southern Chinese. Mandarin many dynasties. At the May 4th movement in 1919, stretches from the vast region north of the Yangtze student protesters in Beijing held among their goals a River to southwestern China provinces of Sichuan, standard written language based on the modern Yunnan, Guizhou, western Hunan, and northwestern Chinese speech. The then Government of the Guangxi. Following the escape of the Nationalist gov- Republic of China responded by designating Gwoyu ernment from the mainland, a sizable number of ‘national language’, based on Beijing Mandarin, as Mandarin speakers settled down in Taiwan in the 190  CHINA Mandarin Northen Chinese Jinyu Min Nan (Southern Min) Wu Sinitic Pu-xian Yue Min Dong (Eastern Min) Min Southern Min Bei (Northern Min) Chinese Xiang Min Zhong (Central Min) Gan Hakka Figure 2. A linguistic tree of iden- Wan Nan (Huizhou) tified Sinitic languages in China. 1950s. More than 870 million people (in the late ern half contains Min Nan and Pu-xian (the latter 1990s) speak Mandarin as a first language, including occupies only a corner in the east), and the northern nearly all Manchu and the entire population of Hui, the half contains Min Dong (in the eastern half), Min Bei second and third largest minority nationalities of (about 60% of the remainder in the north), and Min China. The figure rises well beyond one billion if the Zhong (the western half of the leftover). A transition- number of fluent speakers of Mandarin as a second al language between Min Nan and Min Dong exists in language is taken into account. Mandarin has four Youxi county, the very center of the province. major dialects: Northern Mandarin, Northwestern Min Nan is the largest Min language, with 25.7 mil- Mandarin, Southwestern Mandarin, and Lower lion speakers (in 1984) in southern Fujian, eastern and Yangtze Mandarin. western parts of Guangdong, parts of Hainan and Found in coastal provinces, Wu is spoken on the Zhejiang, and an additional 15 million speakers (in east coast, Yue on the south coast, and Min on the 1993) in Taiwan. Major Min Nan dialects include southeast. The Yangtze Delta can be considered as the Hokkien, Chao-Shan, Leizhou, Hainan, and Zhenan homeland of Wu. Wu is spoken nowadays in Zhejian Min. Many can be subdivided, e.g. Hokkien has four province and much of the area south of the Yangtze subdialects: Amoy, Quanzhou, Zhangzhou, and River in Jiangsu province. It has five dialects: Taihu Longyan. Pu-xian and Min Zhong occupy the smallest (which includes Shanghai as one of its six subdi- areas among the Sinitic, but Pu-xian has around three alects), Taizhou, Wenzhou, Wuzhou, and Liqu. The million speakers, whereas speakers of Min Zhong are homeland of Yue (Cantonese) is in the Pearl Delta; less than one million. They each have two dialects, and from there, it spreads to a large part of Guangdong so do Min Dong and Min Bei. province and as far as southeastern Guangxi. Yue Xiang is mainly spoken in much of the eastern half dialects include Yuehai (which is spoken in of Hunan province. With Mandarin dominant in the Guangzhou, Hong Kong, and Macau), Siyi, Gaoyang, north and in the west, Xiang has been under great pres- and Guinan. sure of assimilation. The influence of Mandarin repre- Min is regarded as the fourth largest dialect of sents a primary factor for variation between its two Hanyu, and also the most diversified one with respect dialects: Northern Xiang is subject to more innovation, to dialectal difference. Indeed, the diversion is so sig- while Southern Xiang is more conservative. Gan is nificant that it warrants the treatment of Min as a lin- largely spoken in Jiangxi province, with five major guistic group comprising five consanguineous, but dialects: Chang-Jing, Yi-Liu, Ji-Cha, Fu-Guang, and distinct, languages. Li’s Dialects of Fujian (1997) Ying-Yi. It is the closest relative of Hakka. describes the distribution of these languages as fol- The distribution of Hakka is unique among lows: excluding a small Wu-spoken area in the north- Southern Chinese in that it covers the widest area in ern corner and about one third of the province in China, but often with a low concentration; it does not western Fujian (where Hakka and Gan are spoken), the have a home base in any province. While the border rest of Fujian can be divided into six areas: the south- area between northeastern Guangdong, southern 191 CHINA Jiangxi, and southwestern Fujian has a higher concen- standard characters based on Sichuan Yi was intro- tration of Hakka speakers, small Hakka-speaking duced in 1975. Tomba is the traditional writing devel- communities are scattered in parts of Taiwan, Hunan, oped by Naxi nationality. The hieroglyphic writing, Sichuan, Guangxi, and throughout Guangdong. Major however, is restricted to religious texts only. Hakka dialects include Min-ke, Yue-Tai, Yuezhong, Himalayish languages have about 4.1 million Yuebei, Yugui, Tonggu, etc. The majority of She speakers in China. Over 82% of them speak Tibetan, nationality also speak Hakka as their first language. Khams, or Amdo. These three consanguineous lan- Jinyu and Wan Nan are two of the less-known guages (regarded as ‘dialects’ of Tibetan in China) Chinese languages. Spoken in Shanxi province and share a common writing script adapted from the adjacent areas in Shaanxi and Henan provinces, Jinyu Indic syllabic writing system. Besides Tibetan nation- is generally assumed to be a subdialect of ality, the other official nationality for speakers of Northwestern Mandarin. Wan Nan is mainly spoken in Himalayish is Moinba (whose population was less an area called Huizhou in southern Anhui province. It than 8,000 in the 1990 census). Exclusive to China, is usually treated as a subdialect of Lower Yangtze Qiangic languages have about 490,000 speakers in Mandarin. total. Approximately one fifth of Qiangic speakers are The Chinese government recognizes only 55 minor- Qiang or Pumi nationalities; the rest identify them- ity nationalities. Their total population is about 106.8 selves officially as being of Tibetan nationality. million, or 8.25% of China’s population (in the 2000 Historically, their languages have no written form, census). Around 60% of them are native speakers of with the exception of Tangut (an extinct relative), over 94% of languages in China. Table 2 shows the top which had a logographic writing modeled after 12 minority languages, each with one million or more Chinese characters. Bai, Tujia, Jingpo, Drung, and speakers (in the early 1980s or 1990s). Lhoba are among the 17 official nationalities speaking Spreading across Xizang Tibetan Autonomous Tibeto-Burman languages in China. Region, southeastern Qinghai, western Sichuan, and Spoken in much of Guangxi Zhuang Autonomous much of the Yunnan provinces, Tibeto-Burman Region, Hainan, southern Guizhou, and southwestern accounts for almost one third of identified languages Yunnan, Tai-Kadai has the second greatest number of in China. The three largest groups within it are Yi- speakers in China. Chinese scholars regard Northern Burmese (28 languages), Himalayish (17 languages), Zhuang and Southern Zhuang as dialects of Zhuang, and Qiangic (15 languages). Speakers of Yi-Burmese which was once written in logographic characters akin in China exceed 6.4 million, including such nationali- to Chinese characters, but a new Roman orthography ties as Yi, Hani, Lisu, Lahu, Naxi, Nu, Jino, and was introduced in the late 1950s. Together with 20 Achang. Chinese scholars consider that Yi nationali- others, the Zhuang languages belong to the Kam-Tai ties speak six rather distinct “dialects” (just as the group. Seven nationalities are designated for the 20 Chinese themselves do). Yi nationalities use an ideo- million speakers of Kam-Tai, namely Zhuang, Buyi, graphic system for writing their languages, but charac- Dai, Dong, Sui, Mulam, and Maonan. The Dai have ters vary considerably from district to district over adapted Brahmi script (a syllabic Indic script) for writ- southeastern Sichuan, Guizhou, and Yunnan. A set of ing Lü and Tai Nüa for many centuries. Kadai (81,000 TABLE 2 Top Twelve Minority Languages of China Language Name Linguistic Affiliation Nationality No. of Speakers Northern Zhuang Tai-Kadai, Kam-Tai Zhuang 10,000,000 Uyghur Altaic, Turkic Uygur 7,200,000 Southern Zhuang Tai-Kadai, Kam-Tai Zhuang 4,000,000 Mongolian Altaic, Mongolic Mongol 3,381,000 Bouyei Tai-Kadai, Kam-Tai Buyi 2,000,000 Korean Altaic Korean 1,920,000 Sichuan Yi Tibeto-Burman, Yi-Burmese Yi 1,600,000 Khams Tibeto-Burman, Himalayish Tibetan 1,487,000 Southern Dong Tai-Kadai, Kam-Tai Dong 1,480,750 Kazakh Altaic, Turkic Kazak 1,100,000 Tibetan Tibeto-Burman, Himalayish Tibetan 1,066,200 Hmong Njua Hmong-Mien, Hmongic Miao 1,000,000 192  CHINA speakers in five languages) and Hlai (747,000 speak- Amis (with 130,000 speakers) of the Paiwanic group, ers in two languages) are two smaller groups within which lost seven members during the colonial rule of Tai-Kadai. Gelao and Li are the two official nationali- Japan (from 1895 to 1945). Grimes’s Ethnologue ties speaking Kadai and Hlai, respectively. reports that another four Paiwanic languages and two All 32 Hmong-Mien languages found in the world Tsouic languages are nearly extinct. Unlike mainland have speakers in China, scattered in various parts of China, Taiwan has witnessed language extinction sim- Guizhou, Hunan, Guangdong, Guangxi, and Yunnan. ilar to that found in Australia and North America, in The three official nationalities—Miao, Yao, and She— the violent way of supplanting an aboriginal language correspond to the three language groups within with a colonial one. Under the rule of the Nationalist Hmong-Mien. Hmongic is the largest, comprising 26 government in the 1950s, linguistic imperialism was languages spoken by 4.2 million people who are most- imposed on the Taiwanese people with Mandarin in ly Miao, but some are Yao. Over 1.1 million Yao speak lieu of Japanese. As the Taiwan government has advo- five Mienic languages, but less than 1,000 She speak cated multiculturalism in recent years, the environ- the singleton language in the Ho Nte group. Hmong- ment has began to change to one propitious for Mien languages have no traditional writing, but linguistic diversity. Yet, many educated Chinese par- Roman orthographies have been devised for several ents in Taipei (the capital city) prefer speaking larger languages since the 1950s. Mandarin to their young children, even though they With 11 Turkic languages, eight Mongolic lan- are native speakers of Min Nan. Mandarin has guages, and five Tungusic languages, all three major undoubtedly established its superior status in Taiwan branches of the Altaic family are present in China. over the half century. Spoken along the northern and northwestern border of The firmly established status of Mandarin is also China—mostly in Heilongjiang province, Inner unchallengeable in mainland China. While the Chinese Mongolian Autonomous Region, and Xinjiang Uygur government does not adopt any aggressive language Autonomous Region—by 18 nationalities, the Altaic policies to wipe out linguistic minorities, policies that languages amount to 25, if Korean is included. Khakas, would sustain linguistic diversity are equally lacking. It Nanai, and Manchu each have less than 50 speakers in is stated in the Chinese Constitution that minority China. According to estimates of the 1990s, Turkic has nationalities have the right to maintain their own lan- over 8.49 million speakers among Uygur, Kazak, Kirgiz, guages, but this by no means implies active promotion of Salar, Bonan, Ozbek, Tatar, and Yugur nationalities; minority languages. In spite of the availability of newly Mongolic has nearly 4.15 million speakers among devised orthographies for languages of a few larger Mongol, Tu, Daur, Dongxiang, and Yugur nationalities; nationalities, the priority for mastering Putonghua, and Tungusic has about 49,700 speakers among Xibe, which is often the sole language in education and mass Ewinki, Oroqen, Hezhen, and Manchu nationalities. media, cannot be mistaken. Ethnic minorities have been Arabic script is used for writing Uyghur, Kazakh, and in a plight parallel to that of new immigrants in countries Kirghiz, whereas traditional Mongolian script is used for such as the United States. In order to become part of the writing Mongolian and Xibe. Koreans write in Hangul, mainstream society, they have to give up their native lan- a phonetic-based spelling system created in 1444. guages, sometimes eagerly. Interethnic marriage (not Va, Blang, and De’ang nationalities in western necessarily with Han-Chinese) in cities is another factor Yunnan, plus Jing nationality on the south coast of for prompting children to become monolingual in Guangxi, are speakers of Austro-Asiatic languages. Mandarin. This kind of language shift is usually com- All 24 Austro-Asiatic languages of China belong to pleted within several generations, characterized by bilin- the Mon-Khmer branch, with around 381,000 speakers gual speakers during the peaceful transition. in total. Only one sixth of them have more than 10,000 Bilingual speakers of autochthonous languages are speakers. Parauk is the largest, with about 180,000 legion in China but the bias toward Mandarin is con- speakers (in 1990). spicuous. Most of the bilinguals, except for a small The mainland government designates an fraction of minority nationalities, are literate only in Austronesian-speaking minority in Taiwan as Gaoshan standard Chinese based on Mandarin. Linguistic nationality, but on the island they are further divided minorities surrounded by larger communities of differ- into nine ethnic groups, collectively referred to as ent languages are likely to become multilingual in sev- Aborigines. The total number of Austronesian speak- eral oral languages, e.g. some Zhuang nationalities ers in China is around 339,800 for 18 Austronesian speak Yue and Mandarin in addition to their native lan- languages. Those spoken on the Taiwan island belong guage and some Pumi nationalities are fluent in to the Formosan branch, which comprises three Prinmi, Mandarin, and Sichuan Yi. While considerable groups: Paiwanic (ten languages), Tsouic (four lan- ethnic minorities (especially those above the age of guages), and Atayalic (two languages). The largest is 50) are fluent in two or more languages, the ratio of 193 CHINA monolingual Mandarin speakers in minority nationali- References ties is steadily rising. Grimes, Barbara (ed., Joseph Grimes, consulting editor) 2000. Ethnic minorities are not alone under the pressure Ethnologue: languages of the world, 14th edition. Dallas, of linguistic absorption. All languages other than TX: Summer Institute of Linguistics. Mandarin are facing varied degrees of endangerment, Ji, Xianlin (editor-in-charge) 1988. Zhongguo Da Baikequanshu: as Putonghua is upheld as the national language at the Yuyan, Wenzi (Encyclopedia sinica: languages and scripts). Beijing: China Encyclopedia Publisher. expense of the others. For instance, vernacular pro- Li, Rulong. 1997. Fujian Fangyan (Dialects of Fujian). Fuzhou: nunciation of Chinese characters is no longer taught in Fujian People’s Publisher. school (except in Hong Kong and Macau). In the past, Norman, Jerry. 1998. Chinese. Cambridge: Cambridge Wu, Yue, Min Nan, and other Sinitic languages were University Press. featured in vernacular operas in different parts of the Ramsey, Robert. 1987. The languages of China. Princeton, N J: Princeton University Press. country, but nowadays popular songs are sung exclu- Wang, William S.-Y. (ed.) 1991. Languages and dialects of sively in Putonghua in mainland China. These suggest China. Berkeley, CA: Journal of Chinese Linguistics. an ever-shrinking room for the survival of other lan- Wurm, Stephen, et al. (general editors) 1987. Language atlas of guages in China. China. Hong Kong: Longman. The linguistic landscape in China is byzantine. Wurm, Stephen, Peter Muhlhausler, and Darrell Tryon (eds.) 1996. Atlas of languages of intercultural communication in Many languages are awaiting in-depth studies and the Pacific, Asia, and the Americas. Berlin and New York: proper classification, while some are yet to be discov- Mouton de Gruyter. ered. Hopefully, these would be accomplished before PICUS SIZHI DING language death accelerates under the momentous impact of the on-going economic development and See also Altaic; Austronesian; Chinese (Mandarin); modernization. Chinese and Japanese Traditional Grammar Chinese (Mandarin) The Chinese language is represented by groups of that the term Mandarin in its wide sense actually cov- dialects that are approximately as distant from each ers many varieties of speech. other as English from German, or French and Spanish. There are many ways to refer to this language. The traditional classification has been into seven Traditionally, when there is no need to differentiate it major groups: Mandarin, Wu, Xiang, Gan, Min, Yue, from the other major dialects, Mandarin is simply and Hakka. Collectively, these dialects are spoken by called ‘Chinese’. We will follow this tradition in the more speakers than any other language in the world. present article. The word ‘Mandarin’ actually means Of these, Mandarin has the special status of having ‘official’, corresponding to the Chinese word ‘guan’. been the official language of China for most of Thus, the Chinese counterpart to ‘Mandarin’ as a lan- Chinese history. As such, Mandarin is the first dialect guage is ‘Guanhua’, or ‘official speech’. To emphasize learned by the great majority of people in China, and its wide usage and enhance its populist intent, the term is acquired to various degrees of proficiency by speak- currently preferred is ‘Putonghua’, where ‘putong’ ers of all other dialects. Thus, there are well over a bil- means ‘common’. The term ‘Guoyu’, or ‘national lan- lion people who use Mandarin in some way or other in guage’, is used in Taiwan. The term ‘Huayu’ is used in their lives. Singapore, where ‘hua’ is an auspicious word that has Whereas the other dialects are concentrated mostly been used to designate ethnic Chinese since the along the southeastern coast, Mandarin is distributed ancient Xia dynasty, some 4,000 years ago. In fact, the widely all over China. The Mandarin spoken in Harbin two elements ‘hua’ and ‘xia’ are often used together as in the northeast is clearly different from the Mandarin a compound to refer to the long ethnic heritage. spoken in Chengdu in the southwest, a thousand miles Lastly, the term ‘Sinitic’ is used in more technical away. Because Beijing has been the capital of China discussions, such as ‘the Sinitic language’, meaning for most of the past millennium, its speech has become ‘the Chinese language’. The stem ‘Sin-’ probably an official linguistic standard, recognizing of course derives from the name of the Qin dynasty (221–206 194  CHINESE (MANDARIN) BCE), when China was first unified politically. ‘Sino- have accumulated over the centuries into an impressive Tibetan’ refers to the family of languages one branch body of philological scholarship in China. This scholar- of which is Sinitic; another branch is Tibeto-Burman, ship received a major impetus from abroad some 2000 which contains several hundred languages spoken in years ago, when Buddhism first came into China. With Western China and around the Himalaya mountains. the religious teachings came the Sanskrit language and The traditional classification of Sino-Tibetan is that its quasialphabetic spelling system, which in turn stimu- proposed by F.K. Li; it includes a third branch of lated the phonetic analysis of the Chinese language. Zhuang-Dong languages, and a fourth branch of Miao- Under this influence, Chinese scholars divided into Yao languages. This four-branch hypothesis has the syllable for the time in terms of the fanqie system. A recently become controversial as more data on minor- character X can be spelled by two other characters, A ity languages became available, although the relation and B, where A indicates the initial consonant of X and between Chinese and Tibeto-Burman has never been B indicates the remainder of the syllable of X. The fan- disputed. It has been estimated that the Chinese and qie system fails to help when the reader does not know Tibet-Burman retained their unity until about 6,000 either A or B. Around 1,700 years ago, Chinese schol- years ago, when extensive migrations radiated out- ars first began writing about the four tones that their lan- ward from its homeland, presumably in northern guage had, which were named ping, shang, qu, and ru. China, which ultimately gave rise to the hundreds of Rime dictionaries were compiled around 1,500 years languages we see today. ago, phonetic charts were constructed 1,000 years ago, There have been efforts to connect the Sino-Tibetan and by 400 years ago, scholars had achieved a reason- languages to much more distant origins. In the early part able understanding of the history of the language. of the twentieth century, the American linguist Edward Chinese and western traditions began to come Sapir remarked on the deep similarities between Tibetan together around the beginning of the twentieth centu- and some of the native languages of North America. ry. In 1898, the first western style grammar of Chinese More recently, research is being done on the Dene- was published by Ma Jianzhong, who studied several Caucasian hypothesis, which posits that groups as dis- European languages while in Shanghai, and who later tant as the Na-Dene languages of North America, the received training in Paris. Early in the twentieth cen- Yeniseian languages in Siberia, Basque in Western tury, the Swedish scholar Bernhard Karlgren traveled Europe, among others, are all related to Sino-Tibetan. to China to conduct fieldwork on Chinese dialects, A competing hypothesis links Chinese to the which he published as part of his dissertation, Austronesian languages, thus implying an ancient 1915–1926. Karlgren was the first to integrate the homeland more to the south. Such hypotheses on dis- philological achievements of the Chinese scholars tant prehistories are intrinsically fascinating, although with the real sounds of modern dialects to arrive at naturally they are more difficult to verify, often for systematic reconstructions. Although many aspects of lack of adequate data or precise methodology. An these reconstructions have been superseded, his work encouraging development in recent years is the has provided an invaluable stimulus for integrating increase in interdisciplinary collaboration, particularly western ideas into Chinese scholarship. among linguists, anthropologists, and geneticists, A very important source of data for reconstructions, which significantly broadens our perspectives on investigated by Chinese scholars over many centuries, human evolution, and brings new promise that the con- is the Shijing, a body of some 300 poems and songs troversies surrounding the ancestry of the Chinese lan- that date back to around 3,000 years ago. This work is guage can eventually be resolved. variously known in English as the Book of Poetry, the The earliest specimens of Chinese date back some Book of Songs, or the Book of Odes. By investigating 3,400 years, in the form of inscriptions on animal bones which sets of characters rimed in the Shijing, and by and on bronze artifacts. These inscriptions were made integrating this knowledge with the phonetic informa- for the purpose of divination or for various historical or tion contained in the individual characters, we can go ritual purposes. Although both the inscribed characters a long way toward inferring how they were pro- and the language they represented have changed over nounced. The language of the Shijing is referred to as the centuries, their ancestral relations with modern char- Old Chinese. Thanks to the early rime dictionaries, the acters and contemporary Chinese can be clearly estab- best known of which is the Qiejun, we can also recon- lished. Thus of all the modern languages, Chinese can struct the language of the seventh century, which is boast of the longest period of continuous use. referred to as Middle Chinese. Since the characters often contain a phonetic compo- Mandarin can be traced back to another famous nent, they also gave valuable indirect clues for the recon- rime dictionary, the Zhongyuan Yinyun, compiled in struction of the spoken language. Efforts to understand the early fourteenth century. Another important source the shape, sound, and semantics of the ancient characters of information on the sounds of this period is the 195 CHINESE (MANDARIN) ‘Phags-pa alphabet, invented by a Tibetan lama in Morphology 1260, by order of Kubli Khan. The language revealed In Chinese, by far the great majority of morphemes are by these sources is called Old Mandarin. represented by single syllables. Since single syllables are written by single characters, there is good corre- Phonology spondence between the minimal units of the sound system, the writing system, and the grammatical sys- Some of the distinct aspects of Mandarin can be dis- tem, i.e. syllable = morpheme = character. cussed by reference to the structure of the syllable, as In contrast to European languages, word formation shown in the following chart: in Chinese does not involve any inflectional morphol- ogy to speak of, such as one finds in ‘books’, Tone ‘walked’, etc. This poses no problem, of course, since Initial Final information regarding gender, number, and case in Medial Rime nouns and tense in verbs is usually provided by the Nucleus Ending context. The speaker is not forced by the grammar, for instance, to choose ‘he’ or ‘she’ when referring to a All the dialects of Chinese can be described accord- friend, since both of these are ta in Chinese. On the ing to the above chart, since they all share the same syl- other hand, Chinese does make use of derivational lable structure. The three basic building blocks are the morphology, such as: tone, the initial, and the final. In European languages, a syllable is distinguished from other syllables by its Prefixes— di wu ‘the fifth’, tou wu ‘the first five’, vowels, such as ‘me’ vs. ‘moo’, or by its consonants, Suffixes— zhuantou ‘brick’, zhitou ‘finger’, such as ‘me’ vs. ‘bee’. In a tone language, a syllable is haizi ‘child’, maozi ‘hat’. distinguished along an additional dimension, as illus- trated in the illustration below. A Chinese syllable is There is an interesting compounding process, where- also distinguished by its vowels and consonants. by opposites are conjoined to form a word. For exam- These are illustrated for Mandarin as spoken in ple, da ‘big’ and xiao ‘small’ are conjoined to mean Beijing in Figure 1 by ‘ma’ vs. ‘mi’ for vowel distinc- ‘size; lai ‘come’ and wang ‘go’ combine to mean ‘social tions, and by ‘ma’ vs. ‘pa’ for consonant distinctions. visitation’. A particularly interesting word formed by Additionally, ‘ma’ can carry any of four distinct tones, this process is from dong ‘east’ and xi ‘west’; the word yielding four completely unrelated words with the dongxi now has the very general meaning of ‘thing’ and meanings of ‘to scold’, ‘hemp’, ‘mother’, and ‘horse’. is used very productively. While English has similar Tone is produced primarily by shaping the pitch con- words like ‘flip-flop’ and ‘zig-zag’, the two processes tour of the voice by varying the rate of vibration of the are different in that the two members of the Chinese vocal folds. The pitch contours of these four words, as compounding are not phonetically related. pronounced by the present author, are shown in the Another interesting process of derivational mor- computer traces in Figure 1. We can see, for example, phology is reduplication. The process works in sever- that the pitch contour for ‘to scold’ is primarily falling al different ways. One area of reduplication has to do while for ‘hemp’ it is rising. For other varieties of with the formation of kinship terms. Whereas most Mandarin, both the number of tones and the shapes of languages have ‘papa’ and ‘mama’, such terms are their pitch contours will be different. much more abundant in Chinese, e.g. didi ‘younger Going back to the chart of the syllable, the initial rep- brother’, nainai ‘grandmother’, jiujiu ‘maternal resents at most one consonant. Thus, there are no con- uncle’, etc. Another area of reduplication has to do sonant clusters in Chinese, such as in English ‘spy’ or with certain nouns to convey the meaning of ‘every’, ‘pry’. The medial may be any one of the three onglides: e.g. tiantian ‘every day’, renren ‘every person’, jiajia ‘i’, ‘u’, or ‘ü’. The nucleus in Mandarin is a simple ‘every family’, etc. vowel, although it may be a diphthong in other dialects. Verbs reduplicate, to convey a range of meanings, The ending may be any one of the three consonants ‘r’, such as transitory action, e.g. kankan ‘to take a look’, ‘n’, or ‘ng’. Thus, a Chinese syllable ‘liàng’, which changchang ‘to have a taste’, tantan ‘to have a chat’, means ‘light’, contains each of the following elements: etc. When the verb is disyllabic, the reduplication Tone: high falling, applies to the two syllables together, e.g shangliang- Initial: l, shangliang ‘to discuss a little’, or yanjiuyanjiu ‘to Medial: i, investigate a little’. Adjectives also reduplicate to form Nucleus: a, adverbs by taking the particle de. Thus, kuai means Ending: ng. ‘quick’ and kuaikuaide pao means ‘run quickly’; qing 196  CHINESE (MANDARIN) means ‘light’ and qingqingde fangxia means ‘put In forming yes–no questions, the typical process for down lightly’. However, when the adjective is disyl- European languages is the inversion of word order, labic, the pattern of reduplication is different from that such as ‘can he sell books?’ formed from ‘he can sell of the verb. The reduplication here is syllable by sylla- books’. English is unusual in the respect that a peri- ble, e.g. an’anjingjingde from anjing means ‘quietly’ phrastic ‘do’ is used to carry the tense, such as ‘did he and gaogaoxingxingde from gaoxing means ‘happily’. sell books?’ formed from ‘he sold books’. Rather than In addition to the compounding and reduplication inversion, the typical process for Chinese is to conjoin discussed above, another distinctive aspect of Chinese an affirmative verb phrase with its negative counter- morphology is the use of classifiers. In English, one part. The negative counterpart to [1] is [4]: uses measure words to quantify mass nouns, in expres- sions such as ‘a grain of sand’, ‘a head of cattle’, or ‘a [4] Ta mei-you mai shu. ‘He neg-aspect sell book’ piece of cloth’. Many classifiers in Chinese, however, Conjoining [1] and [4], we obtain either the ques- have very little semantic content to speak of, such as tion forms [5] or [6] below, depending on whether we in yige ren ‘one-classifier person’, yipi ma ‘one-classi- delete repeated materials from the affirmative or nega- fier horse’, yijian dayi ‘one-classifier coat’, yiben shu tive verb phrase: ‘one-classifier book’, etc. In learning how to use a noun in Chinese, one needs to learn the appropriate [5] Ta mai-le shu mei-you? ‘He sell-aspect book classifier for it. neg-aspect?’ [6] Ta you mei-you mai shu? ‘He aspect neg-aspect sell book?’ Syntax Corresponding to the WH question words in Since the language has very little inflectional mor- English, i.e. ‘who, what, when’, Chinese has shei, phology, grammatical relations among the words are shemme, and shemme shihou, respectively. These have primarily indicated by word order. The basic syntax in been called SH words. In English, WH words need to Chinese is subject–verb–object, much as in English. A be moved to the front of the sentence; in Chinese there typical declarative sentence with the perfective aspect is no such movement. Going back to [1], we may form marker le would be: the questions [7] and [8] as follows: [1] Ta mai-le shu. ‘He sell-aspect book.’ [7] Shei mai-le shu? ‘Who sell-aspect book?’ [8] Ta mai-li shemme? ‘He sell-aspect what?’ A major feature of Chinese syntax is that modifica- tional structures precede the modified. This means that Because of the required movement, the English the head is always at the end of phrases. We have seen translation of [8] would be ‘What did he sell?’ above that adverbs precede the verb in the verb phrase Every language has a stock of prefabricated phras- kuaikuaide pao ‘run quickly’. In noun phrases, numer- es that have a richer cultural content, sometimes by als, possessive, adjectives all precede the noun, such as means of their historical allusions. Such phrases are tade sanben hong shu ‘his three red books’. This learned as complete units, rather than formed sponta- applies to relative clauses as well, with the help of the neously for the moment. Thus, in English, we might particle de, as in zuotian zai xianggang kande shu say ‘He is the early bird that caught the worm this ‘yesterday in Hong Kong read-particle book’. time’; or ‘Aren’t you the hostess with the mostess?’ There are two syntactic variants of a declarative Chinese is extremely rich in this regard, particularly in sentence: the disposal form with the particle ba shown phrases formed with four syllables. These range from in [2], and the passive form with the particle bei shown colloquial expressions such as you tou hua nao ‘oily in [3]: head slippery brain’, used to describe someone devi- ous, to forms that are more literary such as pao zhuan [2] Ta ba shu mai-le. yin yu ‘throw brick induce jade’, used as a gesture of [3] Shu bei ta mai-le. humility to say that what you have to offer is not wor- In these variants, the object is fronted to precede the thy of what you are about to receive. verb. The meaning of the passive sentence [3] is Because they can have a wide range of cultural asso- comparable to that of passive sentences in English. ciations, prefabricated phrases are sometimes difficult However, the disposal form has no ready counterpart to understand, especially when they are used as part of in the syntax of European languages. Although it is a language games. A striking case of misunderstanding form that is frequently used, its meaning and range of took place when the American writer Edgar Snow use are topics for which there is no general agreement interviewed Mao Zedong. Mao described himself with as yet among Chinese grammarians. the expression heshang da shan, which literally means 197 CHINESE (MANDARIN) ‘monk with umbrella’. This led Snow to evoke a Chao, Y.R. 1968. A grammar of spoken Chinese. Berkaley, CA: poignant image: ‘He was, he said, only a lone monk University of California Press. Huang, Churen, et al. (eds.) 1996. Readings in Chinese natural walking in the world with a leaky umbrella’ [Life language processing. Journal of Chinese Linguistics Magazine, April 4, 1971]. In actuality, the second part Monograph 9. of this four-syllable expression is wu fa wu tian, which Huang, James C.-T., and A. Y.-H. Li (eds.) 1996. New horizons could mean ‘no hair no sky’ or ‘no law no heaven.’ The in Chinese linguistics. Dordrecht: Kluwer. concept ‘heaven’ is especially important in this context Karlgren, Bernhard. 1954. Compendium of phonetics in ancient and archaic Chinese. Bulletin of the Museum of Far Eastern because in traditional Chinese culture, the ruler gov- Antiquities 26. 211–367. erns by tianming, i.e. ‘mandate of heaven’. The key to Li, F.K. 1973. Languages and dialects of China. Journal of this expression is the fact that fa is a homonym that Chinese Linguistics 1. 1–13. means either ‘hair’, which a monk lacks, or ‘law’. The Masini, Federico. 1993. The formation of modern Chinese lex- true message Mao was conveying is one of bravado— icon and its evolution toward a national language. Journal of Chinese Linguistics Monograph 6. defying heaven—which Snow completely missed Norman, Jerry. 1988. Chinese. Cambridge. because of the cultural and linguistic gap. Peyraube, Alain. 1996. Recent issues in Chinese historical syn- There are many other interesting aspects of the tax, ed. by Huang and Li, 161–214. Chinese language: Boltz (1994) gives a good survey of Sun, Chaofen (ed.) 1997. Studies in the history of Chinese syn- the Chinese writing system, for example, and Huang et tax. Journal of Chinese Linguistics Monograph 10. Tzeng, O.J.L., and William S.-Y.Wang. 1983. The first two R’s. al. (1996) discuss the special challenges that Chinese American Scientist 71. 238–43. poses to natural language processing—interfacing the Wang, William S.-Y. 1973. The Chinese language. Scientific language with the computer. American, February issue. Wang, William S.-Y. (ed.) 1991. Languages and dialects of China. Journal of Chinese Linguistics Monograph 3. References Wang, William S.-Y. (ed.) 1995. The ancestry of the Chinese Baxter, William H. 1992. A handbook of old Chinese phonolo- language. Journal of Chinese Linguistics Monograph 8. gy. The Hague: Mouton. WILLIAM S.-Y. WANG Boltz, William G. 1994. The origin and early development of the Chinese writing system. American Oriental Society. See also Austronesian; China Chinese and Japanese Traditional Grammar The Chinese linguistic tradition bequeathed an analy- The Chinese debate on ‘rectification of names’ sis of tone, foreshadowed twentieth-century views of originates in several passages from the Analects of syllable structure, and contributed a moral and politi- Confucius, wherein the belief that use of words should cal perspective on the question of discrepancy between reflect reality is asserted. When it does not, the people word meaning and word usage. In addition, explica- cannot be properly governed. Xun Zi (c. 313–238 tion of the formation of Chinese graphs made an BCE), regarded as China’s first major philosopher of important contribution to the understanding of writing language, contributed to this debate in a work titled systems. The Japanese tradition, on the other hand, Zheng Ming (Rectification of names), which addresses offers unique perspectives on defining parts of speech, the origin of names, in terms similar to those found in and analyzing the ways in which the parts combine to Plato’s Cratylus. Unlike Cratylus, however, the main form larger wholes. Both traditions bear a debt to the focus does not concern whether language is to be trust- linguistic tradition of India, some of which was trans- ed as epistemology. In Xun Zi’s view, names are estab- mitted through the powerful vehicle of Buddhism, in lished by convention, and have no inherent the intersection of phonetic analysis and sacred sound. correctness. They thus cannot represent absolute stan- Also common to both was the impetus of early poetry, dards for human behavior. in the desire to know the precise sounds and meanings Hsü Shen (58–147 CE) compiled the Shuo wen jie zi of the past. (first century CE, Explanations of simple graphs and 198  CHINESE AND JAPANESE TRADITIONAL GRAMMAR analyses of composite graphs), the first major lexicon, ways in which certain emotive particles trigger con- and one of the most significant works of the Chinese cord with inflectional endings in the verb, and coined tradition. The 9,353 graphs contained in the work are the term kakari-musubi, still in use today, to refer to it. organized under 540 radicals. One reason for its land- In Mikuni kotoba katsuyôshô (1782, On inflections in mark status derives from the fact that this is the first our language), Motoori laid the groundwork for surviving account of the six principles of formation of understanding the patterns of verbal conjugation. the Chinese graph, including that of semantic radical Suzuki Akira (1764–1837), student of Motoori plus phonetic element, which accounts for more than Norinaga, in his Gengo shishûron (1824, On the four 90% of modern graphs. categories in language), breaks down parts of speech A poet, Shen Yue (441–513), is traditionally credited into nouns, verbs (activities), and adjectives (states), on with the first analysis of Chinese tones, recognizing and the one hand, and particles and bound suffixes, on the naming four types. The tradition of rhyme books is other. Underlying this categorization is a conceptualiza- thought to have originated as an aid to composition of tion of content vs. function words, said to have been poetry, and the technique of fan qie (reverse cutting) drawn under the influence of such a classification for used therein prefigures twentieth-century notions of the Chinese. These grammatical studies from the early mod- syllable as composed of an internal structure, with onset ern period fall under the scholarly domain of kokugaku and rhyme. The purpose of the rhyme books was to (nativism), while the work of those continuing in the indicate the pronunciation of the monosyllabic graphs same vein in the twentieth century is termed kokugogaku of Chinese, and did so by first making four major tonal (national language studies), as separate from the west- divisions, and then within each tonal section, listing ern-influenced discipline of linguistics (gengogaku). graphs in charts arranged by initial sound, and by the The most influential twentieth-century figure to remainder, or rhyme. The so-called ‘cutting’ involved continue along nativist lines was Tokieda Motoki indicating pronunciation of a given graph by providing (1900–1967), who expanded the framework to encom- two other graphs having the same initial and rhyme por- pass all languages, without restriction to Japanese. He tions, respectively. The Qie yun (601 CE), compiled by added a subjective, performative dimension to the Lu Fayan, is the earliest major example of the genre. bifurcation of content words, or shi (in his view, objec- Rhyme tables, the earliest known example of which tively conceptualized) vs. function words, or ji, which is the Yunjing (twelfth century, Mirror of rhymes), pro- express the emotions, attitude, or judgments of the vided a more detailed classification of all of the sylla- speaker. Under this view, content words are nested bles of Chinese. In these works, initials are arranged within function words in an iconic display, which according to articulatory categories such as ‘tongue leaves the subjective elements rightmost at each level sounds’, or dentals, and characterized in terms of of grammatical construction. phonation by such terms as ‘clear’ (voiceless) or Major work on Japanese dialects was carried out by ‘muddy’ (voiced). the folklorist Yanagita Kunio (1875–1962) in his Kagyû- In Japan, early linguistic studies were concerned kô (1927, Thoughts on ‘Snail’), where he posited the with determining the sound correlates to different sets ‘circle theory’ of linguistic diffusion, and linguist Hattori of phonograms in Old Japanese, which had merged into Shirô (1908–), in his investigations of the geographical single sounds in later stages of Japanese. Notable in distribution of accent patterns. The National Language their efforts in this regard were the poet Fujiwara Teika Research Institute has published two major dialect (1162–1241) and Buddhist monk Keichû (1640–1701). atlases: the Linguistic atlas of Japan (1981–1985) and Analysis of morphology and syntax in traditional the Grammar atlas of Japanese dialects (1989–). Japanese grammar is largely a concern with tactics: the combinatorial properties and functions of various References verbal and adjectival suffixes, on the one hand, and postpositions or particles, on the other. Fujitani Baxter, William H. 1992. A handbook of old Chinese phonolo- Nariakira (1738–1779) created the first comprehensive gy. Berlin and New York: Mouton de Gruyter. Chaudhuri, Saroj Kumar. 1998. Siddham in China and Japan. grammatical framework, using the metaphor of cloth- Sino-Platonic Papers, No. 88. ing: he categorized nouns as ‘names’, verbs and adjec- Furuta, Tôsaku, and Hiroshi Tsukishima. 1972. Kokugogaku tives as ‘clothes’, adverbs as ‘hairpins’, and shi (History of Japanese language studies). Tokyo: Tokyo postpositions and suffixes as ‘binding cords’. His best- Daigaku Shuppan Kai. known works are the Kazashishô (1767, On hairpins) Hansen, Chad. 1983. Language and logic in ancient China. Ann Arbor: The University of Michigan Press. and the Ayuishô (1773, On binding cords). Malmqvist, Göran. 1994. Chinese linguistics. History of lin- In his Teniwoha himokagami (1771, Survey of par- guistics, Vol.1: The Eastern traditions of linguistics, ed. by ticles), Motoori Norinaga (1730–1801) explicated the Giulio Lepschy. London and New York: Longman. 199 CHINESE AND JAPANESE TRADITIONAL GRAMMAR Miller, Roy Andrew. 1975. The Far East. Current trends in lin- Ramsey, S. Robert. 1987. The languages of China. Princeton, guistics, Vol.13: Historiography of linguistics, Part 2, ed. by NJ: Princeton University Press. Thomas A. Sebeok. The Hague and Paris: Mouton. Sakai, Naoki. 1992. Voices of the past: the status of language in Motoori, Norinaga. 1997. Kojiki-den, Book 1, translated by eighteenth-century Japanese discourse. Ithaca, New York: Ann Wehmeyer. Ithaca, New York: Cornell East Asia Cornell University Press. Series. ANN WEHMEYER Ramsey, S. Robert. 1982. Language change in Japan and the odyssey of a Teisetsu. Journal of Japanese Studies 8. See also China; Chinese (Mandarin); Japanese Chinese Pidgin English Were Chinese Pidgin English not widely known by tance of Cantonese syntax and phonology, such that to that name, it would be preferable to use the alternative, most English speakers, Chinese Pidgin English was China Coast Pidgin. While having significant influ- hardly more intelligible than Chinese itself. There is ence, English alone did not provide the core of con- convincing documentary evidence that the right guess- ventions that came to constitute the language. Instead, es that set the parameters of Chinese Pidgin English there was convergence of English and Cantonese fea- took place within the trade factories at Canton tures in a process of mutual accommodation. Indeed, between 1699 and 1748, where its inchoate forms the hybrid nature of the structures of Chinese Pidgin were the instrument of limited exchange between the English makes it a crucial piece of evidence in debate ‘supercargoes’ or British East India Company proper- within creole linguistics, reflecting a shift away from ty agents, and the Cantonese-speaking domestics particularist notions of pidgins and creoles to one in assigned to them by their authorized ‘Hong merchant’ which they are considered merely relative, although interlocutors. related, forms of contact languages. One objection to such a scenario is the undeniable Chinese Pidgin English originated during the first impact of Portuguese on Chinese Pidgin English, one locally regulated phase of trade between Chinese and that might lead us to suspect that its origins go back to English in the late seventeenth century (c. the arrival of the Portuguese along the south China 1689–1748). It found a stable form in and around coast in 1557—English ‘Canton’ is derived from the Canton for about another century (1748–1842), and Portuguese pronunciation of Guangzhou. Some of the was thereafter disseminated further up the Chinese earliest fragments of putative Chinese Pidgin English coast during the remainder of the nineteenth century, appear to contain more Portuguese than English, and beyond by the Chinese diaspora to the Pacific and which would tend to point toward a Portuguese base, even the United States (1842–1890). Finally, Chinese e.g. this phrase from 1748: Pidgin English fell into obsolescence during the twen- (1) Carei grandi hola, pickenini hola? tieth century when it was heavily anglicized and lost Want large whore, small whore many of its original structures as numerous speakers Quer grande [puta], pequena [puta] (Ptg) came to target English as a lingua franca. ‘Do you prefer a big or a small prostitute?’ Unfortunately, with the exception of slender sec- ond-hand evidence from the first period of genesis, There is no doubt that Portuguese or some pidginized mostly in European travel accounts, almost all the data form of it was spoken along the China coast up to and we have fall between 1836 and 1901, and, heavily throughout the time Chinese Pidgin English was formed. anglicized, are not reliable. The early data, however In Macao, there eventually formed Macaista, a variety of tenuous, nonetheless reflect a process of mutual creolized Portuguese. Maritime and pidginized accommodation in which speakers had to make Portuguese were used from the Indian subcontinent guesses about what their interlocutors would under- across through Malaysia and Indonesia. As for Canton stand and ‘right’ guesses would be incorporated into itself, we know that the first requirement of a supercargo the grammar of the developing contact language. on English ships trading to China was a knowledge of Balanced against the predominance of English ety- Portuguese. Yet, in the final analysis it is unlikely that a mons in Chinese Pidgin English is the relative impor- prior Portuguese pidgin served as the base of incipient 200  CHINESE PIDGIN RUSSIAN Chinese Pidgin English, not so much because the traders taking their Cantonese servants with them as Portuguese contribution was limited to an ever-decreas- China was forced open to trade. After economic con- ing percentage of its lexicon but rather because the role ditions worsened throughout the nineteenth century, of Cantonese is predominant both in pronunciation and there was a large Cantonese diaspora to California, word structure, from which fact we can suppose that the Hawaii, Australia, and Singapore. At the same time, majority of early pidgin learners were native Cantonese the core of Chinese Pidgin English, learned and trans- speakers who adopted the vocabulary of the supercargo mitted as an auxiliary second language for several gen- masters to whom they were assigned. erations, began to dissolve. It was this late period of its The Canton-based pattern of trade that formed history, roughly from 1842 onward, that gave rise to Chinese Pidgin English continued until the end of the stereotypes and misunderstandings, which have tend- eighteenth century. In 1796, the Emperor banned the ed not only to mask the originality of this language but opium trade, at which point the English turned to also to confuse it with all forms of imperfectly learned smuggling. In 1840, a second attempt was made to ban English, which it was not. opium, which led to the Opium War and the Nanking Treaty of 1842, opening the doors to China. The focus References of contact shifted from Canton to the Treaty cities and Bauer, A. 1975. Das Kanton-Englisch: ein Pidginidiom als to Hong Kong. As the monopoly of the Hong Kong Beispiel für ein Kulturkontaktphänomen. Bern: Peter Lang. merchants broke down, contacts began to proliferate Holm, John. 1988–1989. Pidgins and Creoles, 2 vols. between aspiring Chinese entrepreneurs and the quick- Cambridge: Cambridge University Press. ly increasing number of foreigners in south China. The Lang, George. 2000. Hardly more intelligible than Chinese habit of resorting to Chinese Pidgin English was car- itself: a brief account of Chinese Pidgin english. Asian Englishes 3(1). 21–38. ried into new sites of trade where the status and prac- Morse, Hosea Ballou. 1966. The chronicles of the East India tice of the language changed. As speakers of standard Company trading to China, 1635–1834. Taipei: Ch’eng-Wen English began entering the Chinese community from Publishing Company. the late nineteenth century onward, Chinese Pidgin Shi, Deng-Xi. 1991. Chinese Pidgin English: its origins and lin- English became increasingly associated with the ser- guistic features. Journal of Chinese Linguistics 19. 1–40. Thomason, Sarah Grey, and Terrence Kaufman. 1988. vants of foreigners, and was deprecated. Eventually, Language contact, creolization, and genetic linguistics. Chinese Pidgin English was conveyed first as far north Berkeley: University of California Press. along the coast as Shanghai and Tientsin, British GEORGE LANG Chinese Pidgin Russian Not surprisingly, the Russian expansion across the the Strait of Bering, since, although a contact lan- Ural mountains into Siberia and America led to a num- guage, it is not a pidgin (but rather a variety of Aleut ber of contact languages to a greater or lesser extent with Russian verbal morphology). lexically based on Russian. We have documentation of MPR emerged in the very beginning of the twentieth such varieties from four locations—Kyakhta (on the century, when Russians managed the Manchurian rail- Mongolian border), Manchuria, the Taymyr Peninsula, way. It began to decline after World War II, when both and in the Ussuri area just north of Machuria. The parties took an increasing interest in learning the other’s respective pidgins are hereafter abbreviated as KPR, language, and definitively dropped out of use when the MPR, TPR, and UPR. new political climate put an end to Sino-Russian friend- It is also possible that further varieties exist or have ship. People who remember the pidgin are still alive, existed in Siberia, as well as in areas of (attempted) however; thus, research is still possible, but urgent. Russian colonization in Alaska, California, and KPR emerged as a result of trade between Russians Hawaii. In addition, mention will be made of and Chinese in the eighteenth century. The pidgin has Russenorsk (RN), a trade pidgin of the Arctic Ocean been considered extinct, but in 1990 some speakers based partly on Russian and partly on Norwegian. were found among Chinese merchants in the I will not, however, discuss Mednyj Aleut, spoken in Mongolian capital of Ulan Bator. 201 CHINESE PIDGIN RUSSIAN TPR, also called Govorka, is used between speak- Tense/Mode/Aspect (TMA) marking is usually ers of Samoyedic, Turkic, and Tungusic languages on rather restricted. TPR encodes tense, but not aspect. If the Taymyr Peninsula in northernmost Siberia. It has the little data available there are indeed reliable, UPR been declining in favor of Russian since World War II, is remarkable, and possibly unique among pidgin lan- and is now spoken only by elderly people. Fortunately, gauges, in having a grammaticalized evidentiality research on TPR is currently carried out by German marker. This, however, seems to have been the only linguist Dieter Stern. TMA category in the language. Russenorsk had no UPR is the least well documented of the Russian grammaticalized TMA marking at all, whereas the pidgins, at it is known virtually exclusively through other Far Eastern varieties seem to have made option- accounts of Russian expeditions to the area in the late al use of nonbound morphemes for this purpose. nineteenth and early twentieth centuries. Negation is Russian style in MPR, but postverbal in Russenorsk, finally, was used in the trade between the other Asian varieties. Russenorsk had a preverbal Russians and Norwegians in the border area until the negation, or rather two of them, which usually occu- beginning of the twentieth century. There is a relative pied the second position in the sentence. Nominal wealth of material on this pidgin, which has been ana- number is usually left unmarked, although TPR has lyzed primarily by Norwegian linguists. optional morphemes derived from Russian words Although Russenorsk and TPR are no doubt local meaning ‘much’ and ‘they all’. Grammatical gender developments, it should be emphasized that the rela- has not survived in any of the varieties. tionship between the Far Eastern varieties is unclear, One of the shibboleths of pidginhood is a limited and, indeed, many authors treat them all as ‘Chinese adpositional inventory. Russenorsk had po as its only Pidgin Russian’. preposition, corresponding for the most part to za in In most of the pidgins, there is a tendency toward the Far East. These items expressed more or less any syllabic simplification, favoring CV syllables. Other spatial relationship imaginable. The same job is car- phonological aspects are highly variable, as they dis- ried out by mesto in TPR, which, however, also has a play influences from surrounding languages, such as sociative adposition meste. It is noteworthy that adpo- the devoicing of Russian plosives in MPR, the replace- sitions in TPR are post- rather than prenominal. ment of /f/ with /p/ in KPR, and the merger of /s, z, ʃ, Reduplication is absent from Russenorsk and MPR, tʃ  / in TPR. but has been reported for KPR. Lexically, some of the varieties are remarkably As in many other pidgins, juxtaposition is frequent- mixed. This is particularly true for RN, in which only ly used to indicate possessive relationship, and in a about half of the vocabulary was of Russian origin, similar manner, clauses are to a great extent joined and MPR, where approximately one third of the lexi- paratactically without any overt conjunction or sub- con was non-Russian (mainly from Mandarin). Some junction (although such clauses do exist). lexical influence from local languages is reported for Again, as one would expect from pidgins, bound the other varieties as well. In addition to the lexical morphology is scarce, although there are a couple of items themselves, some calquing from substrate lan- potential candidates for the status of bound mor- guages has also been observed. phemes. As in all pidgins, the small lexicon entails an excep- The pronominal systems also merit mention. The tional degree of polysemy and multifunctionality, the forms themselves, invariable with regard to syntactic serious semantic underspecification leading to a strong function, are mostly derived from Russian genitives, reliance on context for disambiguation. possibly a carryover from Russian foreigner talk. As for the derivation of the Russian forms, it is Exceptions to this are the TPR forms, and the 2sg form noteworthy that the verbs are often developed from elsewhere, which derive from Russian accusatives. imperatives rather than from infinitives, as in pidgins They have been subject to some rather far-reaching lexically based on Western European languages. restructuring, however, in particular in TPR. There, The basic word order is of particular interest, for only the singular forms have been taken over, with only MPR was consistently SVO, like both Russian additional morphemes marking number. Even more and Mandarin. RN was basically SVO, but used SOV conspicuous is the presence of an inclusive/exclusive with adverbials. The other varieties are predominant- distinction for nonsingular pronouns. ly SOV, presumably due to substrate influences. TPR is also spectacular in having the order REL N. Another un-Indo-European feature is the optional lack of References fronting of interrogative pronouns. As would be JabBonska, Alina, and Anatole Lyovin. 1969. The Sino-Russian expected from SOV languages, postverbal auxiliaries mixed language in Manchuria. University of Hawaii are attested. Working Papers in Linguistics 1. 135–64. 202  CHINOOK JARGON Neumann, Günther. 1966. Zur chinesisch-russischen tion in the Pacific, Asia, and the Americas. Berlin and New Behelfssprache von Kjachta. Die Sprache 12. 237–51. York: Mouton de Gruyter. Nichols, Johanna. 1980. Pidginization and foreigner talk: Wurm, Stephen. 1992. Some contact languages and Pidgin and Chinese Pidgin Russian. Papers from the 4th International Creole languages in the Siberian region. Language Sciences Conference on Historical Linguistics, ed. by Elizabeth Closs 14(3). 249–85. Traugott, Rebecca Labrum, Susan Shepherd, and Paul Wurm, Stephen. 1996. The Taimyr Peninsula Russian-based Kiparsky, 397–407. Amsterdam: Benjamins. pidgin. Language contact in the Arctic. Northern Pidgins Nichols, Johanna. 1986. The bottom line: Chinese Pidgin and contact languages, ed. by Ernst Håkon Jahr and Ingvild Russian. Evidentiality: The linguistic coding of epistemolo- Broch, 79–90. Berlin and New York: Mouton de Gruyter. gy, ed. by Wallace Chafe and Johanna Nichols, 239–57. Xelimskij, Evgenij. 1996. ‘Govorka’ — the pidgin Russian of Norwood: Ablex. the Taymyr Peninsular Area, In Mühlhäusler Tryon and Peter Mühlhäusler, Darrell T. Tryon, and Stephen A. Wurm Wurm, pp. 1033–4. (eds.) 1997. Atlas of languages of intercultural communica- MIKAEL PARKVALL Chinook Jargon Chinook Jargon is a pidgin language of the Pacific formed its traditional function as a lingua franca even Northwest that is first attested reliably from the first toward the end of its continuous existence. decade of the nineteenth century, in the journals of Like other classic trade pidgins, Chinook Jargon has Lewis and Clark; the earliest extensive documentation a limited vocabulary. Over 600 Chinook Jargon words is by Horatio Hale (1846). Its lexifier language—the are reliably attested—that is, they occur in at least two language from which most of its vocabulary is independent sources, and usually in more than two. Of drawn—was Lower Chinook (Shoalwater), the lan- these, perhaps a third come from French or English, guage of a once-powerful tribe at the mouth of the reflecting the widespread use of Chinook Jargon as a lin- Columbia River. At its peak, Chinook Jargon was spo- gua franca between Whites and Natives in the late nine- ken as far north as southern Alaska, as far south as the teenth and early twentieth century. The earliest Chinook northern border of California, and as far east as the Jargon sources, from the first half of the nineteenth cen- Idaho panhandle in the United States and interior tury, also list English and French words, but not nearly British Columbia in Canada. It flourished especially as many as in the later documentation. A rough count of between c. 1850 and 1950, when it was a primary c. 650 Chinook Jargon words must certainly omit many medium of communication between Whites, on the that were in common use: even the most ambitious dic- one hand, and Native Americans and Native Canadians tionaries inevitably omit many of a language’s words, on the other. Its use declined sharply in the late twen- and the published Chinook Jargon dictionaries (e.g. tieth century as English replaced it everywhere in the Shaw 1909) are not especially ambitious. Northwest. As late as 1980, however, monthly ser- Chinook Jargon phonemes are typical for an indige- mons were delivered in Chinook Jargon in at least one nous language of the Pacific Northwest. More than 30 British Columbian church, and a few elderly fluent consonant phonemes are attested, in a few to numer- speakers on the Grand Ronde Reservation in Oregon ous words, in at least two independent sources each. contributed to its survival by teaching it to younger Many of the consonants are unfamiliar to speakers of tribal members. Efforts are currently under way to English and other European languages: glottalized revitalize Chinook Jargon, especially at Grand Ronde. (specifically, ejective) stops and affricates, uvular as By the late twentieth century, Chinook Jargon was well as velar obstruents, a glottalized lateral affricate, the main or only Native language spoken on the reser- and a lateral fricative. Moreover, Chinook Jargon vation, where members of six Native tribes descended words often contain consonant clusters that are com- from speakers of three completely different indige- pletely foreign to European languages, e.g. in /tk’up/ nous language families. As Henry Zenk has noted ‘white, light in color’, /ptSix/ ‘thin’, and /tL’m@n/ (1984), Chinook Jargon became ‘an important factor ‘soft, ground up’. Chinook Jargon thus presents a in the sense of identity and solidarity that many striking counterexample to the often-repeated claim Natives of the reservation period came to feel as that pidgin structures are maximally simple, and the “Grand Ronde Indians”.’ Chinook Jargon thus per- consistency of many of its phonological features 203 CHINOOK JARGON across Native speakers from numerous tribes offers a tion, quite a few words entered the pidgin from French counterexample to the common claim that pidgin and then English, especially after about 1850. structures show more internal variation than nonpidgin The hypothesized contribution of the Nootka languages. To judge by the surviving documentation Jargon to the initial formation of Chinook Jargon is and the few existing tape recordings from the mid- predicated on the assumption that such basic words twentieth century, most White speakers of Chinook would necessarily have been in Chinook Jargon from Jargon did not learn either the non-European sounds or the beginning, not added later after Chinook Jargon the non-European consonant clusters of the pidgin (but was fully formed. Against this assumption, however, is see Demers et al. (1871) for a notable exception). the undoubted fact that, while the Nootka words in Also, like other classic trade pidgins, Chinook Chinook Jargon show clear signs of transmission from Jargon has limited morphosyntactic resources. It Whites, the Chinookan and Salishan words in Chinook entirely lacks the complex morphological (word) Jargon were clearly transmitted from Natives to other structures that characterize Native languages in the Natives. Words of Chinookan and Salishan origin region, and its range of syntactic constructions is not show all the elaborate features of typical Northwest large. As with the sound system, its syntactic struc- phonological systems, including glottalized stops, tures closely resemble those of Native Northwest lan- velar vs. uvular dorsal obstruents, and lateral fricatives guages—with one possible major exception, the and affricates. This is not the case with Nootka-origin dominant S(ubject)–V(erb)–O(bject) word order. Most words: they are significantly distorted in comparison Northwest languages are verb-initial; Chinook Jargon to their Nootka source words, with virtually no sounds syntax is verb-initial only with adjectival predicates, that would be foreign to English and French speakers. e.g. _Hayas ulu tsuq nayka_ ‘I am very thirsty’ (lit. The transmission of some of the Nootka jargon/pidgin ‘much hungry water I’). Besides the SVO word order, words to Chinook Jargon is therefore much more like- the constructions are clearly Native, not based on ly to have occurred after, not before, the crystallization either English or French. They include, among others, of Chinook Jargon as a pidgin language. sentence-initial negation, yes/no questions formed Both origin theories are plausible. The postcontact with an optional question particle, and an imperative theory has been popular in part because Chinook construction (‘it would be good if you would do X’). Jargon, from the time it was first documented, already The question of the pidgin’s origin remains highly had the Nootka words and a fair number of French and controversial. One hypothesis is that Chinook Jargon English words. Also, of course, all Chinook Jargon predates extensive contact with Whites in the documentation is necessarily postcontact, since Northwest—that it was used as a means of intertribal Natives in the Northwest had no writing before Whites communication, perhaps at first between speakers of arrived. The precontact origin theory is preferable if Lower Chinook and their Native slaves. (‘Slave jar- one adopts the standard simplicity criterion of histori- gons’ are reported elsewhere in the Northwest, for cal linguistics: the Native phonology and syntax are instance among the Nez Perce.) Based on this theory, easily accounted for if Natives created Chinook Jargon Chinook Jargon achieved its later spread when Whites without significant participation by Whites, but if the adopted it for use as a lingua franca, shortly after pidgin arose postcontact, with some French- and/or 1806. The second origin hypothesis is that Chinook English-influenced structural features, those features Jargon arose as a lingua franca only after Whites must have been lost before the pidgin was document- arrived in large numbers in the Northwest. A common ed. Based on this criterion, the precontact origin is the feature of this theory is the proposal that a Nootka simpler hypothesis. But not all pidgin/creole special- trade jargon (or pidgin) arose first on Vancouver ists accept this criterion; hence, the controversy Island, at the end of the eighteenth century or early in continues. the nineteenth century, and then spread to the mouth of The future of Chinook Jargon is in some doubt, the Columbia River when Whites ventured there for because the Grand Ronde elders and others who spoke trade; the Nootka trade jargon/pidgin was then relexi- the language as part of their ordinary daily lives are fied through the replacement of most Nootka words by now gone. Its fate rests with the younger enthusiasts, Chinook words. especially younger tribal members, who are now Evidence adduced in support of the second theory working to prevent the pidgin from disappearing. is primarily lexical. Most of the lexicon of early Chinook Jargon, including the bulk of the basic vocab- ulary, comes from Lower Chinook, but two or three References dozen words (some of them quite basic) are from Demers, Modeste, F. N. Blanchet, and L. N. St. Onge. 1871. Nootka, and a smaller number of words come from Chinook dictionary, catechism, prayers, and hymns. Salishan and other languages of the region. In addi- Montreal: Quebec Mission. 204  CHOMSKY, NOAM Hale, Horatio. 1846. United States exploring expedition during Thomason, Sarah G. 1983. Chinook Jargon in areal and histor- the years 1838–1842; ethnography and philology. ical context. Language 59. 820–70. Philadelphia: Sherman. Zenk, Henry Benjamin. 1984. Chinook Jargon and Native cul- Samarin, William J. 1986. Chinook Jargon and pidgin histori- tural persistence in the Grand Ronde Indian community, ography. Canadian Journal of Anthropology 5. 23–34. 1856–1907: a special case of creolization. Eugene, OR: Shaw, George C. 1909. The Chinook Jargon and how to use it. University of Oregon dissertation. Seattle: Rainier Printing Co., Inc. SARAH G. THOMASON Splawn, A.J. 1944. Ka-Mi-Akin: last hero of the Yakimas, 2nd edition. Yakima, WA, Caldwell, Idaho: The Caxton Printers; See also Language: Contact—Overview; Pidgins 1st edition, 1917. and Creoles Chomsky, Noam Noam Chomsky is one of the most profound and influ- January 1947) records that ‘N. Chomsky has given ential thinkers of our time, ‘arguably the most impor- much-needed assistance with the manuscript.’ tant intellectual alive’ (New York Times Book Review, Since the appearance of his first published book, February 25, 1979), an evaluation in accord with the Syntactic Structures, in 1957, Chomsky has been rec- fact that he is the world’s most cited living author in ognized as ‘an eminent and revolutionary scholar in several citation indexes. the field of linguistics’, to quote from a representative In his early years, Chomsky read Hebrew literature blurb. Chomsky almost single-handedly assimilated with his father, a professor of Hebrew and Jewish edu- psychology to the natural sciences by formulating cation at Gratz College in Philadelphia (and president ‘transformational generative grammar’ (a model for of its faculty for 45 years), considered ‘one of the the ‘cognitive sciences’, of which it was the first). This world’s foremost Hebrew grammarians’ (New York theory takes for granted the now widely (although not Times obituary, July 22, 1977). His immersion in the unanimously) held view that all human languages Jewish cultural tradition was deep. Out of this experi- operate under the same general principles in spite of ence, his budding political interests converged toward their superficial diversity. Particularly important in this what was then mainstream Zionism (now widely con- context is that the universal foundation for language is sidered ‘anti-Zionism’). taken to be inherent in (and an essential part of) Politics brought Chomsky into linguistics. As a ‘human nature’, this being the central notion in both of teenager, he was deeply interested in radical politics with Chomsky’s main endeavors. Thus, Chomsky rejected an anarchist or left-wing Marxist flavor (strongly anti- behaviorism, unequivocally so in his renowned 1959 Leninist, and more generally anti-Bolshevik). Through review of B.F. Skinner’s 1957 Verbal behavior, a these political interests, he met Zellig Harris, whom review that demonstrated the inability of the behavior- Chomsky has described as ‘a person of unusual bril- ist approach to account for any but the most superficial liance and originality’ and very broad interests. Harris aspects of human language. was not only an acute left-libertarian thinker and analyst Chomsky redefined the nature and scope of linguis- but also a leading figure in linguistics, teaching at the tics. As the Encyclopedia Britannica (1992) says, ‘... University of Pennsylvania, and the most rigorous prac- there is no major theoretical issue in linguistics today titioner of the reigning structuralist methodology. Harris that is debated in terms other than those in which he was surely the linguist best prepared to initiate someone has chosen to define it’. A recent book on the future of with Chomsky’s mind and inclinations into the field of science describes him, accurately, as ‘the most impor- linguistics (as then understood). He also interested him tant linguist who has ever lived’. As John Lyons says in the study of philosophy, logic, and mathematics, fields in Chomsky (1970), he has spoken ‘with unrivaled that were to open for him the avenue to major discover- authority in all aspects of grammatical theory’ since ies. The first reading Chomsky did in linguistics, before Syntactic structures ‘revolutionized the scientific he had taken any courses, was the proofreading of study of language’, fundamentally changing the then Harris’s important Methods of structural linguistics current understanding of language and the mind/brain. (1951), the most exacting exposition of immediately pre- Chomsky’s most fundamental contribution has been Chomskyan linguistic theory. The Preface (dated to open the way for the cognitive natural sciences. His 205 CHOMSKY, NOAM specific model for human language (transformational Chomsky explained had never even been noted before. generative grammar) can be regarded as a confluence Further, many of the analyses he presented have yet to of traditional (and long-forgotten) concerns of the be surpassed. However, like the M.A. thesis, LSLT had study of language and the mind (as in the work of no direct impact on the field at the time. Within two Wilhelm von Humboldt or Otto Jespersen) and new years, though, LSLT had a revolutionary indirect influ- understanding provided by the formal sciences in the ence, since the material contained in it formed the late 1930s, particularly recursive function theory. The basis for the lecture notes that were published in 1957 basic idea is that a sentence is the result of a computa- as Syntactic structures, a book whose unprecedented tion producing a ‘derivation’, beginning with an influence was noted above. abstract structural representation sequentially altered The year 1965 witnessed the publication of by structure-dependent operations (‘transformations’). Chomsky’s next major book, Aspects of the theory of Chomsky’s unusual knowledge of philosophy, logic, syntax, which made fully explicit the role of linguistics and mathematics allowed him to make this model pre- in the investigation of the human mind. Its first chap- cise, and to go on to develop, in the late 1950s, alge- ter, originally written during the year Chomsky spent braic linguistics, a branch of abstract algebra, which is at the Institute for Advanced Studies at Princeton now part of computer science. (For Israeli logician and (1958–1959), remains one of the clearest and most mathematician Yehoshua Bar-Hillel, one of its most forceful expositions of the major goal of generative eminent practitioners, Chomsky was not just the theorizing: providing an account of how the child pre- founder of algebraic linguistics but also ‘by far the best sented with limited evidence develops a computation- man in this exciting new field’.) al system making possible the production and Chomsky’s written output is truly prodigious, comprehension of an unlimited number of brand new encompassing, in linguistics alone, dozens of books sentences (‘explanatory adequacy’). The book laid out and nearly 200 articles (and several times that in polit- what came to be called the ‘standard theory’ of syntax. ical analysis). His first substantial contribution to the In this theory, the syntactic derivation of a sentence development of generative grammar was his B.A. the- begins with deep structure (DS), a representation con- sis (1949), an examination of certain morphophonemic structed via phrase structure rules and lexical insertion alternations in Modern Hebrew, revised and expanded rules. Grammatical relations are fully determined at in his M.A. thesis (1951), and revised once more that this level. Transformations then successively, and year (that version was eventually published in 1979). cyclically (‘bottom up’), modify this representation, Although this work had no impact on the field of lin- eventually producing surface structure (SS), the input guistics at the time, it did introduce numerous technical to the phonological component. This contrasts with devices that would eventually be of major significance, the LSLT model, wherein the recursive component is including abstract underlying syntactic and phonologi- the transformational one: ever larger structures are cre- cal forms and crucially ordered rules deriving surface ated not by phrase structure rules but by generalized forms from these underlying forms. Significantly, transformations, which embed one structure inside already in this early work, Chomsky was emphasizing another. (Interestingly, in Chomsky’s ‘minimalist pro- that a grammar is a finite characterization of an infinite gram’ of the 1990s, there is a return to generalized set of sentences. This theme continues to be of pro- transformations.) Versions of the standard theory found importance in linguistics. (including the ‘extended standard theory’ and the Over the next few years, while a Junior Fellow of ‘revised extended standard theory’) dominated syntac- the Society of Fellows at Harvard University, tic theorizing for more than two decades. Chomsky wrote The logical structure of linguistic the- While Chomsky is most closely associated with ory (LSLT), a monumental work (completed in 1955, syntactic theory, much of his early work was con- but published, and only in part, 20 years later) that laid cerned with phonology, culminating in the ground- out the formal basis for a complete theory of linguistic breaking book in 1968, The sound pattern of English structure. The concepts and technical notions that were (co-authored by Morris Halle), which was a detailed to become central to theoretical linguistics for the next study of the phonology of English, but, just as impor- several decades were developed in this paper, includ- tantly, a full-blown theory of the phonological compo- ing the crucial idea of abstract underlying structure. nent of linguistic theory and its interaction with the Made precise are such concepts as level of representa- syntactic component. Also noteworthy is the elaborat- tion, phrase structure rule, phrase marker, grammatical ed outline of a theory of markedness, which sought to transformation, and derived constituent structure. explain why certain phonological processes are much LSLT is also rich in conceptual and empirical argu- more common than others. ments, and it provides an extraordinarily detailed Chomsky’s concern with explanatory adequacy led account of the syntax of English. Many of the patterns in due course to the ‘Government-Binding’ model, 206  CHOMSKY, NOAM articulated in Lectures on government and binding age of modern philosophy (a not always recognized (1981). In this model, the language (and dialect)-spe- antecedent of the cognitive revolution of the 1950s), cific and construction-specific rules of earlier frame- on which Chomsky has shed much light as an intellec- works are replaced by operations of great generality tual historian. It seems fair to describe him as the whose functioning is constrained by universal condi- scholar who is to the period initiated by the cognitive tions. The rules and conditions are ‘principles’, and revolution of the mid-1950s what Descartes was to the the limited range of variation available for them con- first phase of the age of modern philosophy. stitutes the ‘parameters’. (Chomsky came to prefer the name ‘Principles and Parameters’ for the model, rea- soning that government and binding are just two Biography among many technical devices in the theory, and not Noam Chomsky was born in Philadelphia, necessarily the most important ones.) The P&P Pennsylvania on December 7, 1928. His father, model—the first one ever to suggest a substantive William Chomsky, was a noted Hebrew scholar who solution to the fundamental problem of language wrote Hebrew, the eternal language. While studying at acquisition—represents a radical break from the rich the University of Pennsylvania under the supervision of tradition of thousands of years of linguistic inquiry. It Zellig Harris, he was also a Junior Fellow of the is true, though, that the entire (modern) generative Harvard Society of Fellows from 1951 to 1955. After grammar period is in many ways a new era. receiving a Ph.D. in linguistics from the University of The P&P model was in turn further refined in Pennsylvania in 1955, Chomsky joined the faculty at Chomsky’s ‘minimalist program’ of the 1990s and early MIT under the sponsorship of Morris Halle and was 2000s (see Chomsky (1995), the second chapter of promoted to full professor of Foreign Languages and which was written and circulated in 1988, based on lec- Linguistics in 1961, appointed Ferrari Ward Professor tures in 1986 and 1987). The P&P model had DS, SS, of Linguistics in 1966, and Institute Professor in 1976. Logical Form (LF), and Phonetic Form (PF) as signifi- Since 1967, when the University of London and the cant levels of representation. Given that a human lan- University of Chicago awarded him his first two hon- guage is a way of relating sound (or, more generally, orary doctorates, he has been the recipient of scores of gesture, as in sign languages) and meaning, minimalism honorary degrees throughout the world. In 1969, he seeks to establish that there are no levels except the delivered the John Locke Lectures at Oxford ‘interface’ levels PF and LF. In the most recent develop- University, and in 1970, the Bertrand Russell Memorial ments of the theory, Chomsky suggests, in another par- Lecture at Cambridge University. Among the almost tial return to an earlier formulation, that even PF and LF, countless other honors that he has received is the 1988 as specific levels of representation in the technical sense, Kyoto Prize in basic science, created in 1984 for the do not exist. Rather, throughout the derivation, the syn- purpose of recognizing outstanding achievements in tactic structure thus far created is encapsulated and sent categories not named by the Nobel Prizes. In 1984, he off to the interface components for phonetic and seman- received the Distinguished Scientific Contribution tic interpretation. The minimalist program further main- Award of the American Psychological Association ‘for tains that derivations and representations conform to an enlarging our definition of scientific psychology’. ‘economy’ criterion demanding that they be minimal in a sense determined by the language faculty: no extra steps in derivations and no extra symbols in representa- References tions. A major technical goal is to reduce all constraints on representation to ‘bare output conditions’, deter- Anthony, Louise M., and Norbert Hornstein (eds.) 2003. Chomsky and his critics (Philosophers and their critics). mined by the properties of the mental systems that the Oxford: Blackwell Publishers. linguistic computational system must interface with that Barsky, Robert F. 1997. Noam Chomsky: a life of dissent. are external to the computational system itself. Cambridge, MA: MIT Press. A direct consequence of Chomsky’s scientific work Chomsky, Noam. 1957. Syntactic structures. The Hague: is that it provides strong evidence in support of episte- Mouton. ———. 1959. A review of B.F. Skinner’s verbal behavior. mological rationalism, as he was quick to point out. Language 35. 26–58. Early on, he turned to the serious study of the ———. 1965. Aspects of the theory of syntax. Cambridge, Cartesian tradition, which he was to revive and update, MA: MIT Press. shortly after he made his initial discoveries. These dis- ———. 1966. Cartesian linguistics. Cambridge, MA: MIT coveries made it possible for him to go well beyond Press. ———. 1972. Problems of knowledge and freedom: the the programmatic insights of the Cartesians, and give Russell lectures. New York: Random House. substance to their central claims, in the process recon- ———. 1975. The logical structure of linguistic theory. New structing the enduring ideas of the first phase of the York: Plenum. 207 CHOMSKY, NOAM ———. 1979. The morphophonemics of modern Hebrew. New Harris, Zellig. 1951. Methods in structural linguistics. Chicago: York: Garland. University of Chicago Press. ———. 1981. Lectures on government and binding. Dordrecht: Jenkins, Lyle. 2000. Biolinguistics: exploring the biology of Foris. language. Cambridge: Cambridge University Press. ———. 1986. Knowledge of language. New York: Praeger. Lyons, John. 1970. Chomsky, 3rd edition. London: Fontana. ———. 1995. The minimalist program. Cambridge, MA: MIT McGilvray, James. 1999. Chomsky: language, mind, and poli- Press. tics. Cambridge: Polity Press. ———. 2000. New horizons in the study of language and mind. Otero, C. P. (ed.) 1994. Noam Chomsky: critical assessments (8 Cambridge: Cambridge University Press. tomes). London: Routledge. ———. 2000. Minimalist inquiries: the framework. Step by Smith, Neil. 1999. Chomsky: ideas and ideals. Cambridge: step: essays on minimalist syntax in honor of Howard Cambridge University Press. Lasnik, ed. by Roger Martin, David Michaels, and Juan HOWARD LASNIK AND CARLOS OTERO Uriagereka, 89–155. Cambridge, MA: MIT Press. Chomsky, Noam, and Morris Halle. 1968. The sound pattern of See also Bar-Hillel, Yehoshua; Computational Lin- English. New York: Harper and Row. guistics; Grammar, Theories; Structuralism Clark, Eve V. Educated in France and the United Kingdom, Eve V. or negative value. A positive value of Time shows that Clark started her linguistics career in the United the word refers to some aspect of time. Similarly, the States at Stanford University. She worked on the values of Prior indicate that one event precedes Language Universals Project led by Joseph Greenberg (positive) and follows (negative) the other. In light of and Charles Ferguson and became one of the first fac- the semantic analysis of these two words, one can see ulty members of the Linguistics Department. She has that children do not understand the component Time at worked extensively in various subfields of linguistics first. Once they understand that both before and after including pragmatics, lexical semantics, and psy- contain the component Time, they may still not have cholinguistics. Her greatest contribution to linguistics acquired the feature Prior. At the third stage, they and psychology is her groundbreaking research on understand the positive member of Prior (i.e. before) meaning acquisition—in particular, her earlier work but not the negative member (i.e. after)—in other on the semantic feature hypothesis, her comparative words, the positive value of Prior is learned before the studies of children’s word formation, and her current negative one. Finally, they understand both. Clark research on adults’ offers of and children’s uptake of showed that this type of analysis could explain the lexical and semantic information. findings of other studies on the acquisition of other In the 1970s, Eve Clark’s research focused mostly relational terms such as more–less, high–low, and on the semantic feature hypothesis. Simply put, the tall–short. semantic feature hypothesis states that in learning a In the 1980s, Clark’s research focused on word new word, children only learn some of its semantic formation in acquisition. With colleagues Ruth features and then add to this knowledge as they find Berman, Barbara Hecht, and Randa Mulford, she out more about what the word means. In one of her conducted extensive comparative studies of chil- earliest studies, Clark looked at the acquisition of the dren’s word formation in English, French, and meaning of before and after. The acquisition order of Hebrew. She proposed several principles—semantic these two words is as follows: (1) children did not transparency, productivity, and conventionality—that understand either word; (2) they understood before but influence children’s acquisition of word-formation not after; (3) they interpreted after as if it meant devices (e.g. compounding, affixation) in language. before; and finally, (4) they understood both words For instance, younger English-speaking children correctly. Notice that the meaning of these two words (e.g. aged three) often rely on simple compounds is made up of two components—namely, Time and (e.g. plant-man) to coin new nouns; their use of -er Prior, and these components can have either a positive for agents (e.g. wall-builder) is rather inconsistent. 208  CLARK, EVE V. This finding is in accordance with the principle of studies on meaning acquisition focus on specific con- semantic transparency: while the head noun -man is straints, the principles that Clark proposed for mean- semantically transparent to them (i.e. they know what ing acquisition apply to other facets of language the noun man means), younger children still need to acquisition and to adult language use as well. For analyze -er to obtain the agentive meaning of the suf- instance, the pragmatic principles of conventionality fix. As children get older, they replace words that and contrast (i.e., a difference in meaning is marked by they have coined through compounding with those a difference in form) are not only essential to chil- conventionally used (e.g. plant-man with gardener). dren’s acquisition of meaning but also guide adult This provides evidence for the principle of conven- speakers in their uses of language. The explanatory tionality: ‘For certain meanings, a conventional word power of these principles reflects the strength of or word formation device exists that should be used Clark’s research. in the language community’ (Clark and Berman 1984:549). Another principle is productivity: [T]hose word-formation devices used most often by adults in Biography word innovations are preferred in the language for Eve V. Clark was born in Camberley, UK, July 26, constructing new word forms’ (ibid.: 548). Empirical 1942. She received her M.A. Hons. (1965) in French evidence shows that children rely on this in acquiring Language and Literature, with minors in Spanish and word-formation devices. For instance, among the Phonetics, University of Edinburgh; Postgraduate agentive suffixes -er, -ist, and -ian in English, -er is Diploma in General Linguistics (1966), University of the most frequent. According to the principle of pro- Edinburgh; and Ph.D. in Linguistics (1969), ductivity, children should acquire -er before the oth- University of Edinburgh. She was Assistant Instructor ers, and this turns out to be the case (see Clark and and Instructor in the Department of French at the Cohen 1984). Cross-linguistic studies have shown University of Pittsburgh (1967–1969); Research that regardless of the language being acquired, these Associate, Language Universals Project, Stanford principles play a significant role in children’s acqui- University (1969–1970); Lecturer, Assistant Professor, sition and use of word formation. Associate Professor, and Full Professor, Stanford Since the 1990s, Clark’s research has centered on University (1970–present); Visiting Scientist, Max- the pragmatics of child-directed speech, especially its Planck-Institut für Psycholinguistik Nijmegen, The relation to children’s meaning acquisition. In acquir- Netherlands (1981, 1983–1984, 1990–1991, ing vocabulary, children need to learn to link specific 1997–1998); Fellow, Center for Advanced Study in the terms with their conventional meanings. In the last Behavioral Sciences, Stanford, California few decades, some researchers have proposed that (1979–1980); Guggenheim Fellow, John Simon children observe certain built-in constraints in map- Guggenheim Memorial Foundation (1983–1984); and ping meanings onto word forms, e.g. the mutual Foreign Member, Royal Netherlands Academy of Arts exclusivity constraint where each referent is picked and Sciences (KNAW) (elected 1991). out by just one word. However, studies have shown that the constraint approach underestimates the importance of adult contribution. Children as young References as two years old can and do make use of lexical and Clark, Eve. 1971. On the acquisition of the meaning of before semantic information offered to them by adults. For and after. Journal of Verbal Learning and Verbal Behavior example, two-year-olds readily accept and use multi- 10. 266–75. ———. 1973. What’s in a word? On the child’s acquisition of ple terms for the same referent (e.g. dog and spaniel) semantics in his first language. Cognitive Development and when adults provide them with this information. An the Acquisition of Language, ed. by Timothy E. Moore, alternative to the constraint approach is to examine 65–110. New York: Academic Press. the role of adult-directed speech in meaning acquisi- ———. 1983. Meaning and concepts. Handbook of child psy- tion and children’s uptake of adults’ offers of words chology, Vol. 3: Cognitive development, 4th edition, ed. by John H. Flavell and Ellen M. Markman, 787–840. New and relations among words. Clark’s current research York: Wiley. goals are to (1) establish the range of adults’ offers of ———. 1991. Acquisitional principles in lexical development. words and relations and (2) determine how effective Perspectives on thought and language: interrelations in different types of offers are for word learning, as development, ed. by Susan A. Gelman and James P. Byrnes, measured by children’s uptake. 31–71. Cambridge: Cambridge University Press. ———. 1993. The lexicon in acquisition, Cambridge and New The central themes in much of Clark’s research are York: Cambridge University Press. the emphases on general cognitive principles and on ———. 1999. Acquisition in the course of conversation. social interaction in language acquisition. While many Studies in the Linguistic Sciences 29. 1–18. 209 CLARK, EVE V. Clark, Eve, and Ruth Berman. 1984. Structure and use in the Markman, E. 1987. How children constrain the possible mean- acquisition of word-formation. Language 60. 547–90. ings of words. Concepts and conceptual development: econo- Clark, Eve, and Sophia Cohen. 1984. Productivity and memory logical and intellectual factors in categorization, ed. by U. for newly formed words. Journal of Child Language 11. Neisser, 255–287. Cambridge: Cambridge University Press. 611–25. Markman, E., and G. Wachtel. 1988. Children’s use of mutual Donaldson, M., and G. Balfour. 1968. Less is more: a study of exclusivity to constrain the meanings of words. Cognitive language comprehension in children. British Journal of Psychology 20. 121–157. Psychology, 59. 461–471. Merriman, W., and L. Bowman. 1989. The mutual exclusivity Donaldson, M., and R. Wales. 1970. On the acquisition of some bias in children’s word learning. Monographs of the Society relational terms. Cognition and the development of lan- for Research in Child Development. 54. guage, ed. by J.R. Hayes, 235–268. New York: Wiley. Gelman, S., J. Coley, K. Rosengren, E. Hartman, and A. Pappas. ANDREW WONG 1998. Beyond labeling: the role of maternal input in the acquisition of richly structured categories. Monographs of See also Acquisition; Pragmatics; Psycholinguis- the Society for Research in Child Development. 63. tics; Semantics Clause Sentences can be of arbitrary length. Every well- verb phrase VP). In X’-theory, the canonical subject formed sentence, however, can be adequately position is the specifier of IP. The sentence John likes described in terms of its internal structure: the words it Mary can adequately be described in terms of an IP (and contains, the phrases these build and the morphemes further internal structure). The CP-level is invoked for they are made up from, and other such relations. additional material, such as the question Who does John Beyond the regular phrase level—noun phrase (NP), like?, which is transformationally derived from the for- verb phrase (VP), etc.—sentences can be divided into mer sentence. Here, the subject John occupies the same clauses. Above all, a sentence is always a clause; but position in a structural description (i.e. the tree or the some sentences may consist of more than one clause. phrase-marker) of the sentence as the original sentence. In traditional grammar, the distinction is made But it is assumed that the object who has moved from its between main (subordinate) and subordinate (depend- canonical predicate- or VP-internal position to the ent) clauses. In a sentence like John likes Mary, the beginning of the sentence, the specifier of CP. whole structure is also one clause, but in the slightly We can thus see that IP and CP can be well-formed longer sentence John likes Mary when she plays foot- main clauses. Finiteness plays a role for, but is no ball, the boldfaced part is a separate clause all by itself. essential identifier of, main vs. subordinate clauses. In In generative grammar, the term embedded is used to principle, any clause can be finite or nonfinite. As a indicate that the subordinate clause is embedded under rule of thumb, unmarked, declarative sentences—what the main clause, but still part of the same sentence. This have sometimes been dubbed kernel sentences—are is more than mere notation: traditional grammar does IPs, while more complicated or derived structures not consider subordinate clauses to be clauses proper; (including interrogatives, imperatives, or exclamatives, in the generative approach, however, all well-formed such as Does John likes Mary?, Play better soccer sentences are clauses, and every (grammatical) embed- now!, or What a great game she played!) are usually ded clause is by definition a well-formed structure on CPs. Another rule of thumb is that a well-formed the sentential level, hence a clause. clause contains only one subject and one predicate (i.e. The sentence–clause distinction in generative terms one main verb and its potential complements, depend- (as in the Principles-and-Parameters theory) is often one ing on whether the verb is an intransitive, a transitive, of categorial projection, selection, and complementa- or a ditransitive verb, for example). But virtually every tion. Two main categories are decisive for the type of clause can be embedded under or subordinated to a structure: the Inflectional Phrase (or IP for short, for- higher clause. This depends on the type of clause merly S) and the Complementizer Phrase (CP or S’). IP and/or on the potential main clause predicate. denotes the traditional concept that a well-formed sen- Addressing the former first, subordinate or embed- tence consists of a subject and a predicate (often the ded clauses come in various types. We commonly 210  CLAUSE-TYPE INDICATORS distinguish between adverbial, complement, and rela- it, and the relative clause form one constituent. tive clauses. An adverbial clause (also oblique clause) Moreover, relative clauses are optional, not obligatory. relates to the main clause through the semantic mean- Thus, while in a noun-complement clause, the comple- ings expressed by adverbs, such as time, manner, place, ment clause must be used, relative clauses need not be. instrument, circumstance, concession, purpose, result, Consider, for example, the sentence John likes the girl cause, or condition. In English, adverbial clauses are who plays soccer . The boldfaced part is a relative typically introduced by a subordinating conjunction (or clause, adding further information about the NP, the girl. subordinator), an element that links the adverbial This type is a restrictive relative clause, which is charac- clause to the higher main clause. An example is John terized by being essential for identification (otherwise likes Mary because she plays soccer well. Other subor- we would not know which girl John likes). A nonre- dinators are after, when, whenever, while, as, although, strictive relative clause is not required for identification or if. In generative analysis, subordinators invariably of the NP, as in Mary, who plays soccer, is a nice girl. head a CP-projection; adverbial clauses are thus CPs. To sum up, the property of human language that sen- Complement clauses are characterized as clauses tences can be of infinite length does not mean that that serve as a complement to a lexical item, i.e. without longer sentences cannot be analyzed. Rather, a property which that lexical item would not be complete. This of language is that sentences can consist of more than addresses the role of the potential main clause predicate one clause, but each clause can be clearly identified. We mentioned above. Two major classifications of comple- distinguish between a (unique) main clause and (any ment clauses exist, namely noun-complement and verb- number of) subordinate clauses. The latter come in three complement clauses. Just as some physical object must types: adverbial, complement, and relative clauses. be expressed to form a complete structure of the verb throw (as in The goalkeeper threw the ball), as part of References the definition or subcategorization frame of that verb, some clause must be expressed to satisfy the subcatego- Haegeman, Liliane. 1991. Introduction to government and binding, 1st edition. Oxford: Blackwell; 2nd edition, 1994. rization frame of a verb like tell. The boldfaced part in Huddleston, Rodney. 1984. Introduction to the grammar of Mary told John (that) she played soccer is a comple- English. Cambridge: Cambridge University Press. ment clause, namely complement to the main verb tell. Huddleston, Rodney, and Geoffrey K. Pullum. 2002. The (Naturally, tell can also take an NP as a complement, as Cambridge grammar of the English language. Cambridge: in Mary told John a lie, but we concentrate on clauses Cambridge University Press. Keenan, Edward. 1985. Relative clauses Vol. 2: Complex con- here.) In this case, the complementizer that is optional; structions, ed. by Shopen, pp. 141–70. it can be sued or left out. In this instance, other elements Noonan, Michael. 1985. Complemenatation. Language typolo- could also be used instead, such as why or how. The gy and syntactic description, Vol. 2: Complex construc- important relation is that between the main verb tell and tions, ed. by Timothy Shopen, 42–140. Cambridge: the proposition expressed by the following clause. A Cambridge University Press. Radford, Andrew. 1988. Transformational grammar. Cambridge: similar relationship can be found with certain nouns. A Cambridge University Press. noun like fact takes a complement clause. For example, Shopen, Timothy (ed.) 1985. Language typology and syntactic we say John likes the fact that Mary plays soccer. It is description, Vol.2: Complex constructions. Cambridge: thus part of the subcategorization frame of a noun like Cambridge University Press. fact that it must be followed by a complement clause. Thompson, Sandra A., and Ronald E. Longacre. 1985. Adverbial clauses. ed. by Shopen, pp. 171–234. Relative clauses, finally, modify, describe, or further KLEANTHES K. GROHMANN specify a noun, but are structurally part of the entire noun phrase, i.e. the head noun, any other modifiers of See also Grammar, Traditional Clause-Type Indicators Communicating by language is a highly structured and expressing surprise, gathering information, giving goal-directed social activity that serves many purpos- orders, and so forth. For verbal interaction to be effec- es: making contact, delivering a speech, telling a joke, tive, the speaker has to formulate what he or she wants 211 CLAUSE-TYPE INDICATORS to say in such a way that its communicative intention a, neither of which can occur together with y(r). Yet, the is recognizable or identifiable to the other participants particle y(r) is absent in embedded and verb-initial in the speech situation. Making a statement will natu- declarative constructions. In Maale (an Omotic language rally take a different form than posing a question or of Southern Ethiopia), all clause types are marked by a issuing a command. In language theory, the systemat- special verbal affix. Thus, the main verb of affirmative ic relation between the grammatical form of sentences declarative clauses contains the affix –ne (e.g. ʔ–atsí zig- and their conventionalized conversational use is called inó mukk–é–ne [man yesterday come-PERFECTIVE-AFFIR- clause type. MATIVE. DECLARATIVE] ‘The man came yesterday’), The three major clause types that occur most fre- while the corresponding interrogatives are characterized quently in the world’s languages are declaratives, by the affix –iya when the main verb is marked for interrogatives, and imperatives, which correlate with Perfective aspect, which indicates the completion of the statements of facts, inquiries, and directives (a cover event that is described (e.g. ʔ–atsí ziginó mukk–é–iya? term for commands, requests, orders, and the like), [man yesterday come-PERFECTIVE-INTERROGATIVE] ‘Did respectively. Clause types represent a grammatical the man come yesterday?’). system in at least two respects. First, one can easily A more complex situation is obtained in languages construct triplets of corresponding declaratives, with so-called evidential systems. Evidential particles yes–no questions, and imperatives (e.g. John left in a (or affixes) indicate the source reliability of the hurry; did John leave in a hurry?; leave us alone!). speaker’s knowledge and the kind of evidence he or Second, clause types are mutually exclusive, no sen- she adduces for what is being communicated. In lan- tence belonging simultaneously to two different types. guages with elaborate evidential marking, such as To distinguish one clause type from another on a Tuyuca (a Tucanoan language spoken in Colombia formal basis, languages resort to a large variety of and Brazil), there are no unmarked declaratives, since grammatical devices, which include changes in the every sentence must contain an evidential marker shape of verbs, word-order alternations, different qualifying the information on which an assertion is intonation patterns, and special particles and affixes. based. A simple declarative like he played soccer cor- The concern here is with such dedicated affixes and responds to five different verbal constructions, each particles, which are put to use as clause-type indica- associated with a specific evidential marker. tors. Clause-type indicators constitute a subset of Compare: díiga apé–wi [soccer play-VISUAL] ‘he function words with a primarily classificatory func- played soccer’ (I say him play), díiga apé–ti [soccer tion that mark the sentence they modify as a declara- play-NON-VISUAL] ‘he played soccer’ (I heard the tive or an interrogative construction. Out of the game and him, but I didn’t see it or him), díiga apé–yi ordinary as sentence-type indicators may look at first [soccer play-APPARENT] ‘he played soccer’ (I have sight, they are also attested in English. Consider, for evidence that he played soccer, but I didn’t actually instance, the subordinating conjunctions that and see the game), díiga apé–yig! [soccer play-HEARSAY] whether, which introduce embedded declarative and ‘he played soccer’ (I obtained the information from interrogative sentences, respectively (e.g. John said someone else), díiga apé–hîyi [soccer play-DEDUC- that there would be enough food at the party vs. John TIVE] ‘he played soccer’ (it is reasonable to assume it). wondered whether there would be enough food at the Interrogative sentences come in two varieties: party). yes–no questions, where the truth of the questioned As many linguists have observed, declarative sen- statement is at issue (e.g. Was John invited to the tences tend to be the most unmarked of these clause party?) and the constituent question, where the ques- types: they typically occur without any clause type indi- tion word or phrase signals the missing piece of cator or specific ordering and impose the fewest restric- information (e.g. Who was invited to the Party?). tions on what verbal categories can be selected. There In Polish, the question marker. czy is used to distin- are, however, many languages with marked declaratives. guish main and embedded yes-or-no questions from For example, Welsh (a Celtic language of the British declarative sentences (e.g. Pan Kowalski by w Austrii Isles) has a clause-initial particle y(r), which is placed in ‘Mr. Kowalski has been to Austria’ vs. czy Pan declarative constructions with periphrastic verb forms, Kowalski by w Austrii ‘Has Mr. Kowalski been to which are composed of the present and the imperfect Austria?’ and nie pytaem czy Pan Kowalski by w forms of bod ‘to be’ (third-person masc. sing. mae ‘he Austrii ‘I did not ask whether Mr. Kowalski had been is’) and a verbal noun (e.g. Y mae Siôn yn gweld draig to Austria’). In Mandarin Chinese, question words like [PARTICLE is John PROGRESSIVE.PARTICLE see VERBAL-NOUN] shei and shemne convey not only a question interpre- ‘John sees (lit. is seeing) a dragon’). The declarative par- tation ‘who’ and ‘what’ but may also be interpreted as ticle y(r) contrasts with other sentence-type indicators indefinite pronouns ‘someone, anyone’ and ‘some- like the negative particle ní(d) and the question particle thing, anything’ (e.g. shei mai–le yi–ben-shu? [who 212  CLINICAL LINGUISTICS buy-ASPECT one-CLASSIFIER -book] ‘who bought a Barnes, Janet. 1984. Evidentials in the Tuyuca verb. book?’ vs. ta bu xiang jian shei [S/he not want see International Journal of American Linguistics 50. 255–71. Bielec, Dana. 1998. Polish—an essential grammar. London and who] ‘She (or he) does not want to see anyone’). If, on New York: Routledge. the other hand, the question particle ne occurs in sen- Cheng, Lisa Lai-Shen. 1991. On the typology of wh-ques- tence-final position, only the interrogative and not the tions. Ph.D. Dissertation, Massachusetts Institute of indefinite interpretation of the question word is avail- Technology. (reprinted in 1997, New York and London: able (e.g. Hufei mai–le shemne ne? [Hufei buy-ASPECT Garland). Cinque, Guglielmo. 1999. Adverbs and functional projections: what INTERROGATIVE.PARTICLE] ‘What did Hufei buy?’ a cross-linguistic perspective. Oxford and New York: [NOT ‘Hufei bought something’]). Languages with Oxford University Press. dedicated interrogative particles like Chinese leave the Palmer, Frank R. 1986. Mood and modality. Cambridge and question word in exactly the same syntactic position as New York, Cambridge University Press. the corresponding declarative sentence, while lan- Sadock, Jerrold M., and Arnold M. Zwicky. 1985. Speech act distinctions in syntax. Language typology and syntactic guages without such clause-typing devices, like description, Vol. I: Clause structure, (ed. by Timothy English, place the question word in front of the clause Shopen,) 155–96. Cambridge and New York: Cambridge and require the inversion of the auxiliary verb and sub- University Press. ject (and insertion of do if there is no auxiliary verb). Sproat, Richard. 1985. Welsh syntax and VSO structure. With evidential particles and affixes being a relative- Natural Language and Linguistic Theory 3. 173–216. Willett, Thomas. 1988. A cross-linguistic survey of the gram- ly recent discovery, more research needs to be done to maticalization of evidentiality. Studies in Language 12. clarify the functional role of such clause-type indicators. 51–97. Yule, George. 1996. Pragmatics, Oxford introductions to lan- guage study. Oxford and New York: Oxford University References Press. CHRIS H. REINTGES Amha, Azeb. 2001. The Maale language. Ph.D. Dissertation. Leiden University, CNWS Publications 99. See also Clause Clinical Linguistics Clinical linguistics, a branch of applied linguistics, is the area. Other milestones were the establishment of the the use of linguistics to describe, analyze, and treat journal Clinical linguistics and phonetics in 1987 and language disability. The study of linguistic aspects of the founding of the International Clinical Phonetics and communication disorders is of relevance to a broader Linguistics Association (ICPLA) in 1991. understanding of language and linguistic theory. Rather than diagnosing using medical criteria, clini- Practitioners of clinical linguistics range from aca- cal linguistics focuses on the linguistic manifestations demic linguists with research and teaching interests in of the disorder, aiming to provide a comprehensive language disability to practicing professionals such as typology of language disorders based on their linguistic speech and language pathologists/therapists, educa- characteristics. This approach aids in diagnosing and tional and clinical psychologists, and neurologists. treating impairments of unknown underlying causes. Research in the field tends to be multifaceted, drawing For example, there is as yet no agreed medical explana- on a wide range of disciplines in addition to linguis- tion for Specific Language Impairment (SLI), a condi- tics, such as psycholinguistics, cognitive science, neu- tion found in children who have problems with spoken roscience, and biomedical science. language but no other obvious cognitive or neurological Although systematic phonetic descriptions of speech deficit. In spite of this, it is still possible to describe the disorders have been routine since the 1950s, clinical lin- linguistic characteristics of SLI precisely enough for guistics did not emerge as a coherent discipline inclu- research purposes and for devising remedial programs. sive of phonology, grammar, semantics, and pragmatics There appears to be no level of language organiza- until the late 1970s. This was in part due to the pioneer- tion that is immune to impairment. Language disor- ing work of David Crystal whose book Clinical linguis- ders are equally likely to be found in adults and in tics (1981) has been particularly influential in defining children who are still acquiring language, and in both 213 CLINICAL LINGUISTICS the production and comprehension of spoken, written, intervention. Nevertheless, clinical linguistics is still a and signed language. Anatomical deficits such as cleft new field and a great deal of exploratory work remains palate may affect a speaker’s phonetics and phonolo- to be done, particularly in the form of case studies and gy, resulting in an inability to differentiate pairs of in areas such as semantics. Often, it is only when a words such as ‘bat’ and ‘mat’. Grammar, semantics, complex system goes wrong that we become aware of and pragmatics may be impaired as a result of devel- the contribution—and even the existence—of its com- opmental disorders in children or a stroke in adults. A ponent subsystems. One potential growth area for clin- child with SLI might repeat ‘Has the mouse been ical linguistics, therefore, is its role in informing and chased by the cat?’ as ‘A mouse chasing cat’. A stroke evaluating linguistic theory and in illuminating our patient might see nothing wrong with adding a past understanding of normal language structure and use. tense suffix to nouns, resulting in words like ‘towned’ and ‘faithed’, due to problems with word formation. People with autism find it difficult to make use of con- References text to infer nonliteral meaning. Ball, Martin J. 1989. Phonetics for speech pathology. London: Any linguistic theory that purports to throw light on Taylor & Francis, 2nd edition, London: Whurr, 1993. Ball, Martin J., and Raymond D. Kent (eds.) 1997. The new the nature of the human mind should describe both the phonologies: developments in clinical linguistics. San normal and the pathological. Some have argued that Diego: Singular. certain syntactic anomalies found in the language of Chiat, Shula. 2000. Understanding children with language prob- people with Broca’s aphasia, a condition resulting lems. Cambridge and New York: Cambridge University Press. from damage to Broca’s area of the brain, can be Crystal, David. 1980. Introduction to language pathology. London: Edward Arnold, 4th edition (co-author Rosemary explained using Chomsky’s ‘Principles and Varley), London: Whurr, and San Diego: Singular, 1998. Parameters’ theory of syntactic structure. These Crystal, David. 1981. Clinical linguistics. Vienna: Springer- researchers believe that such anomalies provide evi- Verlag, London: Whurr, 1989. dence that syntactic abilities are separate from other Grodzinsky, Yosef. 1990. Theoretical perspectives on language cognitive skills. In addition, data from children with deficits. Cambridge, MA: MIT Press. Grundy, Kim (ed.) 1989. Linguistics in clinical practice. SLI and Williams Syndrome, a genetic condition in London: Taylor & Francis; 2nd edition, London: Whurr and which linguistic proficiency develops despite poor San Diego: Singular, 1995. cognitive abilities, have been use d to support a simi- Hewlett, Nigel, Louise Kelly, and Fay Windsor (eds.) 2001. lar view of syntax as an autonomous mental ‘module’. Themes in clinical linguistics and phonetics. New York: Others, however, cite evidence from a range of lan- Lawrence Erlbaum. Howard, Sara J., and Barry C. Heselwood. 2005. Clinical pho- guage disorders in support of a nonmodularist, func- netics and phonology: analysing speech in a clinical context. tionalist account of language suggesting a New York: Academic Press. considerable degree of codependency between linguis- Perkins, Michael R. 2005. Pragmatics and communication impair- tic and cognitive processes. Clinical linguistic research ment. Cambridge and New York: Cambridge University Press. will continue to play a key role in debates of this kind. Perkins, Michael R., and Sara J. Howard (eds.) 1995. Case stud- ies in clinical linguistics. London: Whurr and San Diego: The range of analytical methods used in clinical lin- Singular. guistics is comparable to that found in other areas of MICHAEL PERKINS linguistics. Both theory-based and data-driven approaches are used, with a wide selection of materials See also Aphasia; Applied Linguistics: Overview; produced for purposes of diagnosis, assessment, and Psycholinguistics Code-Switching Speakers of more than one language (e.g. bilinguals) are guage with a phrase or word from another language. known for their ability to code-switch or mix their lan- While some linguists suggest that people code-switch as guages during communication. This phenomenon occurs a strategy in order to be better understood and to enhance when bilinguals substitute a word or phrase from one lan- the listeners’ comprehension, code-switching among 214  CODE-SWITCHING bilinguals has traditionally been viewed as a strategy to second language. In other words, these bilinguals compensate for diminished language proficiency. The code-switch more when they communicate in Spanish premise behind this theory is that bilinguals code-switch than when they use English. Psycholinguistic evidence because they do not know either language completely. also suggests that bilinguals retrieve English code- This argument is also known as semilingualism, meaning switched words faster when they listen to Spanish sen- bilinguals ‘almost’ speak both languages correctly. tences, whereas they are slower to retrieve Spanish However, language proficiency is not clearly defined, as code-switched words as they listen to English sen- it is unclear whether reading and writing language skills tences. This evidence suggests that the bilingual relies should take precedence over spoken skills in determining on the second language as opposed to the first. How language proficiency. This reliance on reading and writ- are these findings explained? The general premise ing is problematic because most bilinguals receive their behind this view is that after a certain level of fluency formal education in one language, while a majority of and frequent use of the second language, a language their social interactions take place in the other language. shift occurs. That is, during early stages of bilingual- Thus, when their reading and writing abilities are tested ism, Spanish–English bilinguals rely on their first lan- in both languages, the language in which bilinguals guage when they communicate in their second received more formal education will usually fare better. language. As a result, bilinguals are more likely to Recent psycholinguistic research has focused on code-switch to Spanish, when they communicate in how code-switching is a natural product of the inter- English. However, as the second language becomes action of the bilingual’s two languages. Early the dominant language, bilinguals rely on the second researchers viewed code-switching as evidence that language when they communicate in the first lan- the bilinguals’ two languages were organized in sepa- guage. In this case, bilinguals code-switch to English rate and distinct mental dictionaries. For example, a when they communicate in Spanish. The second lan- general finding throughout the literature is that guage becomes more readily accessible and bilinguals bilinguals take longer to read and comprehend sen- come to rely on it more. Regardless of which language tences containing code-switched words as compared the bilingual learned first, the more active (dominant) to monolingual sentences. Apparently, this time-con- language determines which mental dictionary is going suming process is due to a ‘mental switch mechanism’ to be accessed faster. This argument is reasonable that determines which of the bilingual’s two mental since most bilinguals in the United States, whose first dictionaries are ‘on’ or ‘off’ during the course of lan- language is Spanish, obtain their formal education in guage comprehension. This mental switch is responsi- English. Likewise, many of their everyday interactions ble for selecting the appropriate mental dictionary to involve the second language. As a result, words and be used during the comprehension of a sentence. Other concepts in English, the second language, become research shows that bilinguals comprehend code- more accessible than words in Spanish, the first switched words faster when there is an overlap language. Thus, code-switching is not the same for between the two languages’ sound systems. For exam- both languages. Rather, it depends on language ple, Chinese–English bilinguals, where Chinese is the dominance. native language, take longer to recognize English In short, code-switching among bilinguals may be code-switched words in Chinese sentences, but only if indicative of difficulties in retrieval (access) affected the English words begin with consonant–consonant by a combination of closely related factors such as lan- clusters (e.g. block), as opposed to consonant–vowel guage use (i.e. how often the first language is used) clusters (e.g. big), because the Chinese language and word frequency (i.e. how much a particular word words do not begin with consonant–consonant clus- is used in the language). Examination of code- ters. Other important factors reported to influence the switching behavior can contribute to a further under- recognition of code-switched words include context, standing of second-language acquisition, as well as phonetics, homophonic (e.g. words pronounced the language acquisition and development, and general same) and homographic (e.g. words spelled the same) linguistic theory. overlap between the two languages. Another current view suggests that language domi- References nance, which language is used more frequently, plays an important role in code-switching. For example, Lederberg, Amy R., and Cesareo Morales. 1985. Code switch- Spanish–English bilinguals report more linguistic ing by bilinguals: evidence against a third grammar. Journal of Psycholinguistic Research 14. 113–36. interference (code-switching) when they communicate Li, Ping. 1996. Spoken word recognition of code-switched in Spanish, their first language, and little or no code- words by Chinese–English Bilinguals. Journal of Memory switching when they communicate in English, their and Language 35. 757–74. 215 CODE-SWITCHING Macnamara, John, and Seymour Kushnir. 1971. Linguistic Myers-Scotton, Carol. 1993. Duelling languages: grammatical independence of bilinguals: the input switch. Journal of structure in code-switching. New York: Oxford University Verbal Learning and Verbal Behavior 10. 480–87. Press. Heredia, Roberto R. 1997. Bilingual memory and hierarchical ROBERTO R. HEREDIA AND JEFFREY M. BROWN models: a case for language dominance. Current Directions in Psychological Science 6. 34–9. Kolers, Paul. 1966. Reading and talking bilingually. American See also Acquisition; Bilingual Acquisition; Bilin- Journal of Psychology 3. 357–76. gualism Coherence in Discourse Discourse is a communicative event in which language uct (the linguistic construct) and discourse-as-process plays a prominent role. It minimally requires a sender (the communicative event). Coherence can be reserved (writer, speaker), a receiver (reader, listener), and a for the conceptual relationships that comprehenders message that is being communicated. This message is use to construct a coherent mental representation not just a concatenation of clauses; it forms a unified, accommodated by what is said in the discourse. coherent whole. Both the sender and receiver normal- Cohesion is limited to the linguistic markers that cue ly have the implicit agreement that the message being the comprehender on how to build such coherent rep- communicated is coherent. resentations. Cohesion emphasizes discourse-as-prod- Coherence in discourse has been studied in a range uct, and coherence emphasizes discourse-as-process. of disciplines, including linguistics, philology, sociol- Cohesion alone is not sufficient for the interpreta- ogy, philosophy, psychology, and computer science. tion of the discourse. Comprehenders generate infer- Linguists identify and analyze inventories of the lin- ences on the basis of background knowledge and guistic markers of coherence that are available in a discourse constraints. Much of the background knowl- language. Sociologists explore the production and edge is experiential; hence, it involves common proce- comprehension of coherent discourse in naturalistic dures and activities (called scripts), social interactions, conversations that involve different groups and cul- and spatial settings. For instance, a narrative usually tures. Psychologists collect data in experiments that describes a setting, an action sequence with a conflict test hypotheses about the effect of coherence on cog- and plot, and an outcome. A script for eating in a nitive processing and representations. Computer scien- restaurant would furnish inferences and help coherent- tists design and test computer models that attempt to ly tie together the explicit content of a narrative about produce and test coherent text. a bad restaurant experience. Although cohesion alone The term coherence has been defined in various cannot fully account for coherence in discourse, the ways. Some researchers apply the term cohesion to psycholinguistic literature has shown that cohesion the surface structure of the text and the term coher- facilitates coherence. ence to the concepts and relations underlying its Cohesion and coherence can be divided into local meaning. Cohesion has sometimes been applied to (microstructure) and global (macrostructure). Local smaller units of language in the text, and coherence, cohesion and coherence are related to the interrelated- to some general overall interrelatedness in the text. ness between adjacent discourse segments. Global Other researchers have defined cohesion as continuity cohesion and coherence are related to the interrelated- in word and sentence structure, and coherence as con- ness of larger spans of discourse. For instance, script- tinuity in meaning and context. As in the case of ed action sequences are globally coherent. Also, there coherence, discourse has been defined in different are the rhetorical structures of narrative (such as set- ways. Several years ago, the term discourse was ting + conflict + plot + resolution), expository (such as reserved for dialogue, and text was reserved for claim + evidence, problem + solution), and other dis- monologue. In contemporary research, discourse cov- course genres. ers both monologic and dialogic spoken and written Cohesion and coherence can be grammar driven and language. vocabulary driven. Grammar-driven cohesion refers to Somewhat more subtle distinctions are sometimes sentence structure, word structure, and the intonation of made. One can distinguish between discourse-as-prod- the discourse segments. Vocabulary-driven cohesion 216  COHERENCE IN DISCOURSE refers to the lexical vocabulary of the discourse segment. other insofar as they mean the same thing (synonyms) or These cohesion cues activate vocabulary-driven (pre- the opposite thing (antonyms), stand in a superset/subset grammatical, knowledge-based) and grammar-driven relationship (hypernym vs. hyponym, respectively), or (syntax-based) coherence. Vocabulary-driven and gram- have some other conceptual relationship. mar-driven coherence are not necessarily mutually (1) The tax collector sent another letter. exclusive but often support each other, as illustrated (2) I don’t like this guy. below. (3) That monster never leaves us alone. Consider the sentence The dean (i) read the New (4) The sweetheart keeps asking for more each York Times (ii) in his office (iii). A paraphrase with year grammar-driven cohesion would reduce the discourse elements to the grammatical necessities: He (i) always With comparison, a constituent in an expression is reads it (ii) there (iii). A vocabulary-driven para- compared with a constituent in another expression (I phrase, on the other hand, would find meaningful lex- like the oak cabinet. The pine desk is much nicer). ical alternatives, as in The man (i) always reads the Discourse psychologists have extensively investi- newspaper (ii) behind his desk (iii). gated five cohesion and coherence relations that are In addition to the distinctions between local and related to the previously mentioned seven: referential, global and between grammar- and vocabulary-driven spatial, causal, temporal, and additive relationships. cohesion, the types of cohesion discussed below have They answer the questions of the who, where, why, often been recognized. when, and what of the events described by the dis- Conjunctions relate adjacent discourse segments. course. Explicit markers facilitate the comprehension There have been several classifications of these con- process. junctions in virtually every field. Most of these classi- Several classifications of relations have been pro- fications include additive (and, but), temporal (before, posed. Some focus only on the closed set of grammar- until), and causal (because, although) conjunctions driven cohesion, whereas others include vocabulary- that are either extensive (and, before, because) or driven relations and relations that are reconstructed adversative (but, until, although). from world knowledge and the unique situation con- Coreference specifies that two expressions refer to veyed in the text. Those classifications that go beyond the same entity. Often, the coreference is grammar grammar consider the intentions of the producer of the driven by the use of pronouns, both pronominal (he, communicative event. In written monologic discourse, she) and reflexive (himself, herself). The interpretation comprehenders can rely on linguistic cues to a great of these pronouns is determined by their antecedents, extent (although not completely). However, in oral dia- i.e. previously mentioned words referring to the same logic discourse, there are conversational cues that go person or object. The coreference can be both forward well beyond print, such as intonation, gestures, and the and backward. Anaphoric reference is a backward ref- physical environment. erence to an antecedent noun phrase or clause that was A complete theory of discourse coherence requires a introduced earlier in the discourse (John kissed Mary harmonious layering of several levels, including vocab- because he loved her). Cataphoric reference is a for- ulary, sentence structure, meaning, discourse context, ward reference to a noun phrase or clause that will be style, and world knowledge. When these levels lack mentioned later in the text (Because he loved her, coordination, the coherence is more difficult. To get the John kissed Mary). message across, the sender will try to coordinate the With substitution, repeated forms, and ellipsis, a levels. The receiver assumes that the sender’s message constituent of one expression is replaced by a con- is intended to be well formed and will make every stituent of another (substitution), is repeated (repeated attempt to construct a coherent interpretation. forms), or is omitted (ellipsis). The intended meanings can be reconstructed from the preceding discourse and References from world knowledge. Beaugrande, Robert de, and Wolfgang U. Dressler. 1981. Will we make it on time? Introduction to text linguistics. Harlow: Longman. (1) I think so (substitution of we will make it in Brown, Gillian, and George Yule. 1983. Discourse analysis. Cambridge: Cambridge University Press. time by so). Clark, Herbert H. 1996. Using language. Cambridge: (2) Yes, we will make it on time (repeated forms). Cambridge University Press. (3) If we hurry (ellipsis: we will make it on time is Costermans, Jean, and Michel Fayol (eds.) 1997. Processing omitted). interclausal relationships. Studies in the production and comprehension of text. Mahwah: Lawrence Erlbaum. With lexical relationships, the type of cohesion is Gernsbacher, Morton Ann. 1994. Handbook of psycholinguis- vocabulary driven. Two lexical items are related to each tics. New York: Academic Press. 217 COHERENCE IN DISCOURSE Givón, Talmy. 1995. Coherence in the text and coherence in the Knott, Alistair, and Chris Mellish. 1996. A feature-based mind. Coherence in spontaneous text, ed. by Morton Ann account of the relations signalled by sentence and Gernsbacher and Talmy Givón. Amsterdam: John Benjamins. clause connectives. Journal of Language and Speech 39. Graesser, Arthur C., Gernsbacher, Morton Ann, and Goldman, 143–83. Susan (eds.) 2003. Handbook of discourse processes. Mann, William C., and Sandra A. Thompson. 1987. Rhetorical Mahwah, NJ: Erlbaum. structure theory: toward a functional theory of text organi- Graesser, Arthur C., and Gordon H. Bower (eds.) 1990. Inferences zation. Text 8. 243–281. and text comprehension. San Diego: Academic Press. Van Dijk, Teun A., and Walter Kintsch. 1983. Strategies of dis- Halliday, Michael A. K., and Ruqaiya Hasan. 1976. Cohesion in course comprehension. New York: Academic Press. English. London: Longman. MAX M. LOUWERSE AND ARTHUR C. GRAESSER Color Terms Color terms have been used in anthropological and cog- others): white, black, red, yellow, green, blue, brown, nitive linguistics to investigate whether certain aspects purple, pink, orange, and gray. of meaning are universally found in all languages. The Interestingly, these terms occurred in predictable classic study of color terminology, Berlin and Kay patterns, and Berlin and Kay proposed a scheme of 1969, is often cited as proof against the Sapir–Whorf ‘implicational universals’ to explain the evolution of hypothesis of ‘linguistic relativity’ or ‘linguistic deter- color terminology. Some languages have only two minism.’ This hypothesis argues that language affects or basic color terms, and these are always black and even determines cognition—that is, that the categories white. If a third is added, as in the Arawak language of present in the language a person speaks will influence the Caribbean or in Swahili of East Africa, it is always the way that person thinks, the way he or she perceives red. After red, languages with four color categories the world. Languages vary as to which ‘basic color contain either yellow or a term that encompasses both terms’ occur in the vocabulary. Berlin and Kay set out blue and green (Kay devised the term ‘grue’); lan- to test whether the presence or absence of a color cate- guages with five colors include both yellow and grue, gory in the language affects the way in which speakers with differentiation between green and blue occurring see colors. In the course of their research, they found in the six-color system. Brown is added as the seventh, evidence favoring a ‘prototype’ theory of meaning and then purple, pink, orange, and gray, which do not fol- evidence for universal tendencies in the structuring of low a specific ordering. An example of an implica- color vocabulary in the world’s languages. tional universal that follows from this discovery would Berlin and Kay gave color charts to speakers of 98 be that languages with a term for blue will always have different languages and asked them to sort the 329 col- words for black, white, red, yellow, and green. The ored squares into categories that corresponded to the sequence can be depicted as shown below: black “grue” yellow blue purple and > red > or > or > and > brown > pink white yellow “grue” green orange or gray basic color terms in their language. They put three lim- The orderly sequence for the development of color its on what could count as a basic color term: terms leads us to the idea that they might reflect cog- (1) It had to be a simple word; e.g. reddish-brown nitive universals, and this is indeed what Berlin and is not a basic color term. Kay found in their research on the focal meanings of (2) It had to apply generally, not in a specific the color terms. When they asked people to choose the domain (like blonde or sienna). best example of each of the basic color terms from the (3) It could not count as a ‘kind of ___’, in the chart, choices were remarkably similar around the way that emerald and olive are kinds of green. world. For example, people whose language contained no term for orange chose the same focus for yellow as They found that the most basic color terms any lan- those whose language had terms for both yellow and guage has is 11, as in English (also Tagalog, Zuni, and orange; people whose language contained only three 218  COMMUNICATION THEORY color terms had the same focus for red as those whose the set of referents that can be designated by that word, language had all 11 terms. This is the finding that which is determined by how those referents compare refutes the so-called strong version of the to a prototypical exemplar of the category. This pro- Sapir–Whorf hypothesis (a version espoused by nei- duces ‘fuzzy sets’, with some uncertainty about ther Edward Sapir nor Benjamin Lee Whorf), that our whether referents that are marginal really belong to the perception will be determined by the categories of our category or not. Is turquoise a kind of blue or a kind of language. For color, at least, perception seems to fol- green? Is burgundy a kind of red or a kind of purple? low universal human neural response patterns in a way Different people will give different answers. that is unaffected by our language habits, although Researchers continue to debate the ways in which there have been dissenting claims suggesting that meanings of color terms are encoded in our minds. Berlin and Kay’s experimental methods may not accu- There are still questions about how language and cul- rately reflect the way speakers apply color terms to ture relate to cognition, including whether cognitive objects in the real world. images or concepts of color are based solely on neuro- The existence of universals for the core meanings logically determined perceptions that are universal for of color terms does not contradict the fact that culture humans or whether they are molded in part by linguis- does play a role both in determining the wider mean- tic and cultural experience that includes using color ing of color terms for speakers and in the inventory of words in ways quite different from an experimental non-basic terms for colors in one’s speech repertoire. setting. Color terms may be closely associated metaphorically with objects in our experience, for example, the sun with yellow, or fire with red. Color vocabulary may be References highly elaborated in some cultures and for some Kay, Paul, and Chad K. McDaniel. 1978. The linguistic signif- speakers, according to the importance of color in vari- icance of the meanings of basic color terms. Language 54. ous arenas of daily life. A painter or an interior deco- 610–46. rator, for example, will have an extensive lexicon of Kay, Paul, Brent Berlin, and William Merrifield. 1991. Biocultural implications of systems of color naming. color terms. In American society, women have larger Journal of Linguistic Anthropology 1. 12–25. color vocabularies than men, due to the sexual division Berlin, Brent, and Paul Kay. 1969. Basic color terms: their uni- of labor in purchasing clothing and household furnish- versality and evolution. Berkeley: University of California ings and the proliferation of color terms as a form of Press. marketing in American consumer culture. Lucy, John A. 1992. Language diversity and thought: a refor- mulation of the linguistic relativity hypothesis. Cambridge: In applying color terms to objects and to parts of the Cambridge University Press. color chart that are farther away from the focal colors, MacLaury, Robert E. 1991. Social and cognitive motivations of speakers seem to compare the color in question with change: measuring variability in color semantics. Language the focal color and to judge both similarities and dif- 67. 34–62. ferences. The task of drawing boundaries on the color Whorf, Benjamin. 1956. Language, thought, and reality: select- ed writings of Benjamin Lee Whorf, ed. by John B. Carroll. chart around the limits of each color produced more Cambridge, MA: MIT Press. varied answers than determining the best example of a Wierzbicka, Anna. 1990. The meaning of color terms: seman- color did. The focal color seems to act as a prototype tics, culture, and cognition. Cognitive Linguistics 1. 99–150. for that color. Prototype theory views word meaning as ELLEN JOHNSON Communication Theory The contents of the terms ‘communication’ and ‘com- it is generally accepted that not even the learning of munication theory’ are notoriously vague. languages is a simple conditioning process. How could A straightforward psychologist’s definition is the it be possible that after such a complicated process like following: ‘communication is the discriminatory language acquisition human beings use their lan- response of an organism to a stimulus’. Thus, guages for communicative purposes that are as simple Pavlovian conditioning would also be an instance of as Pavlovian conditioning? The stimulus–response communication. This is strange to the linguist, because theory does not seem to be on the right track. 219 COMMUNICATION THEORY Another way to define communication is by infor- Vygotsky, and the philosopher Valentin Voloshinov. mation. Often, the term ‘communication theory’ is Due to political reasons, it took some time before their used in the strict sense of information theory. work reached western research. But since the 1960s Information theory originated in problems around and 1970s, their ideas have influenced western psy- telecommunication techniques and was developed by chology and text linguistics. The main idea of these Claude Shannon and Warren Weaver in 1949. This scholars is that the language sign is inherently dialog- theory of communication suggests the following ic, and that all manifestations of the signs are inher- schema for communicative acts: ently dialogic. Language is formed in the process of social interaction. While the ruling stratum tries to Source ——> Transmitter ——> Channel ——> posit a single discourse as exemplary, the subaltern Receiver ——> Destination classes are inclined to subvert this monologic closure. In effect, the source sends some message to the des- In the sphere of literature, poetry and the epic repre- tination. But let us look at the details between source sent the centripetal forces within the cultural arena and destination. First, the message is encoded into a while the novel is the structurally elaborated expres- signal; this is done by the transmitter. This point is sion of popular ‘ideologiekritik’ (ideology criticism). important in telecommunication; however, in ordinary In Bakhtin’s now famous study Problemy tvorchestva face-to-face talk, source and transmitter coincide. Dostoevskogo (‘Problems of Dostoevskii’s work’), it Leaving the receiver, the signal is sent through the is argued that Dostoevskii’s work is imbued with a channel. In ordinary face-to-face talk, the channel profoundly democratic spirit. Bakhtin argues that would be the air channel. But in telecommunication, it Dostoevskii’s creative method is not Hegelian. In a would be the electric channel. In the channel, signals Hegelian scheme, two positions struggle for ascendan- are usually changed, or, rather, distorted by the noise. cy but are transformed into a synthesis at the end. In ordinary face-to-face talk, the surrounding sounds According to Bakhtin, Dostoevskii does not merge the that are not part of the conversation would be the noise voices into one final, authoritative voice as in the of the channel. After the channel, the signal reaches Hegelian absolute. Instead, Dostoevskii presents an the receiver. There, the decoding of the signal takes unmerged dialogue of voices, each given equal rights. place, and we end up with a message again. This mes- Bakhtin calls this type of dialogue ‘polyphonic’. The sage is finally stored in the destination. In ordinary voice of the narrator resides beside the voices of the face-to-face talk, receiver and destination coincide. other characters. Voices intersect and interact, mutual- As precise as this schema is, and as useful as it is ly illuminating their ideological structures, potentiali- for telecommunication, it neglects several factors of ties, biases, and limitations. Polyphony is also natural communication. First, the message lacks inter- observed across different texts: for instance, literary pretation. It seems odd to say that the message is final- texts show ‘intertextuality’, which means that they ly just ‘stored’ in the destination. This is not even true refer to other texts, thus yielding a huge larger text for the simplest acts of communication. If my neigh- together. ‘Voicing’ is not only a characteristic of liter- bor at the dinner table utters ‘could you please pass me ary texts. In ordinary speech, several factors also lead the salt’, I would not just store this message in my to a multivoiced message: the cultural context, the brain. Instead, I would interpret the message as nonlit- social relations between the persons involved in the eral, and furthermore as an instruction to pass the salt communication, the context of the situation — all to my neighbor. This simple example demonstrates these factors have an effect on the message. that, often, people communicate because they want Linguistics, however, usually concentrates on only others to do certain things. This leads to the next fac- some aspects of communication, leaving cultural and tor of communication that is neglected in the social factors aside. Shannon–Weaver model: the dialogic nature of com- So far, we have concentrated on the two forms of munication. Natural communication is not just one verbal communication: written and oral communica- way as in the Shannon/Weaver schema; normally, tion. But human communication may also be nonver- something goes back to the sender. bal. There are many different channels of nonverbal Considerable research has been done to capture the communication: facial expressions, hand gestures, necessary dialogism of communication. The most body movements (‘kinesics’), and touch (‘haptics’). All interesting schools are the Bakhtin Circle, the Prague these modes of nonverbal communication make use of school, and some versions of functionalism. signs, movement, or touch. In animal communication, Let us focus on the Bakhtin Circle. In the pre-Stalin there are — in addition to nonverbal and (partly) verbal era, a group of three scholars emerged with some com- communication — other ways of communication, mon views on communication and cognition: the liter- namely by smell and taste. Human beings do not make ary critic Mikhail Bakhtin, the psychologist Lev much use of smell and taste in the communicative 220  COMPARATIVE METHOD domain. Humans receive information about the world References through the channels of smell and taste, for instance, Infante, Dominic A., et al. 1997. Building communication the- while eating. But they do not use smell and taste for ory. Prospects Heights, IL: Waveland. communication. Body smells seem to have a commu- Miller, Katherine. 2001. Communication theories: perspectives, nicative function only in sexuality. processes, and contexts. Mountain View, CA: Mayfield We have seen that communication has many Publishing. Neuliep, James W. 1996. Human communication theory: appli- aspects, and as many as communication theory has. If cations & case studies. Boston: Allyn & Bacon. the focus of attention is more on social factors, the Phillipsen, Gerry. 1992. Speaking culturally: explorations in dialogism of communication will be important; if social communication. Albany, NY: State University of New information structure plays a greater role, the theory York Press. will be more technical. Little attention has been paid West, Richard, and Lynn H. Turner. 2000. Introducing commu- nication theory: analysis and application. Mountain View, to integrate the findings on nonverbal communication CA: Mayfield Publishing. into the general frameworks existing for verbal com- MONIKA RATHERT munication. This might be an interesting field of study for the future. See also Text Linguistics Comparative Method The comparative method is the most important of the William Jones, in 1786, that Sanskrit was related to various methods and techniques used to recover lin- Latin, Greek, and German. His famous speech to the guistic history. It is also of interest in language classi- Asiatic Society of Bengal marks the beginning of fication, in linguistic prehistory, and in research on comparative philology. distant genetic relationships, determining whether Furthermore, the year 1816, when the German seemingly interrelated languages descend from a com- scholar Franz Bopp presented the linguistic public mon ancestor after all. It is generally recognized that with language material from Sanskrit compared with 6,000 languages are currently spoken all over the some other Indo-European languages, remains a his- world (one quarter with fewer than 1,000 speakers), torical date in linguistics, because Bopp was the first and they are grouped into more than 30 widely recog- to realize that the question of the mutual relations of nized families. It is estimated that half of those lan- Indo-European languages was worth becoming the guages will die in the next century. The comparative subject of specialized inquiry. Indo-European was method is therefore of much help in clarifying their posited as a hypothetical ancestor language, and com- status and evolution through time. parative studies were almost exclusively focused on The comparative method is a set of principles and the Indo-European language family. a methodology for comparing forms from related lan- Six years later, another German scholar, Jacob guages to reconstruct an earlier (proto-) language or Grimm, included a systematic survey of the relation- an earlier state of a language. The aim of reconstruc- ships between Germanic consonants and their corre- tion is to recover as much as possible of the proto- spondents in Greek, Latin, and Sanskrit in the second language from a comparison of the descendant edition of his comparative grammar of Germanic lan- languages and to determine what changes have taken guages. He established the existence of a circular place in the various languages that developed from ‘soundshift’ (Lautverschiebung) governing these rela- the proto-language. tionships in the prehistory of Germanic. Grimm The comparative method was developed in nine- observed that when other Indo-European languages teenth-century Germany for the reconstruction of such as Latin and Greek have a voiced unaspirated Proto-Indo-European and was afterward applied to the stop (b, d), Gothic has the corresponding voiceless study of other language families. It is considered the unaspirated stop (p, t), and that, when other Indo- most outstanding achievement of linguistic scholar- European languages have a voiceless unaspirated stop, ship in the nineteenth century and was originally stim- Gothic has a voiceless fricative (f,θ). In other words, ulated by the discovery by an English orientalist, Sir the voiced stops inherited from Proto-Indo-European 221 COMPARATIVE METHOD became voiceless, and the voiceless stops became spread of an official standard language, the German fricatives. His discovery is known in linguistics under scholar Johannes Schmidt criticized the family-tree the name of Grimm’s Law (or the Germanic consonant theory as unrealistic and misleading. According to shift), although Grimm himself did not use the term his theory, known in linguistics as ‘the wave theory’ law. Grimm’s Law was probably first worked out in (1872), each sound law has its own territory, and dif- detail by the Danish linguist Rasmus Rask (whose ferent linguistic changes will spread, like waves, studies, being written in Danish, were less accessible from a politically, commercially, or culturally impor- to most European scholars). tant center along the main lines of communication. However, linguists found at least two classes of These innovations will not necessarily cover exactly exceptions to Grimm’s Law with respect to Indo- the same area. European voiceless stops: there are cases in which A real achievement of the method of reconstruction Proto-Indo-European voiceless stops remain voiceless was the Laryngeal Hypothesis presented by the stops in Germanic and cases in which Proto-Indo- famous Swiss linguist Ferdinand de Saussure in his European voiceless stops become voiced stops in 300-page Mémoire sur le Système Primitif des Voyelles Germanic. Grimm’s Law was thus corrected in 1875 dans les Langues Indo-Européennes (1879; Report on by the Danish linguist Karl Adolf Verner, who exam- the Primitive System of Vowels in Indo-European ined words in Vedic Sanskrit and Attic Greek that are Languages): there had to be two hypothetical sounds related to Germanic words that reflect Indo-European for Proto-Indo-European, not attested in any of the voiceless stops as voiced stops. He discovered that in Indo-European languages known at that time, but suf- all such cases, the Proto-Indo-European word had ficient to solve several puzzling irregularities in the intonational stress after the voiceless stop. Verner development of certain vowels in a number of lan- understood that what seems to be an exception to guages. He was fairly conservative about claiming sound laws may be governed by some other regularity what they must have been, but he called them laryn- and that sound changes may be unconditioned or may geals, indicating that these sounds must have been depend on the environment in which the sound occurs. articulated with the larynx (voice box) as the main This was a very important step in linguistics, and articulator, and pointed out the precise locations in ancestral common forms could be reconstructed by words where they must have occurred. Fifteen years using the principle of regular sound change and postu- after Saussure’s death, when the ancient Anatolian lan- lating a number of different sound laws that operated guage Hittite, the oldest attested Indo-European lan- independently in the different branches of the Indo- guage, was deciphered, linguists found with European family. amazement that Saussure was right: Hittite had the The comparative method was often criticized laryngeal sounds, and they were placed exactly where because it is based on a misleading genealogical Saussure had predicted just on the basis of careful metaphor. In the 1850s, the German linguist August reconstruction. Schleicher transferred the Darwinian view of evolu- The history of comparative philology (this was the tion to language and introduced the model of the evo- name for linguistics at that time) shows that the com- lutionary ‘family tree’ to comparative linguistics. As at parative method addresses directly only material in the that time it was customary to concentrate on the recon- related languages that is inherited from the proto-lan- struction of what was supposed to be the source of all guage and has no means of its own for dealing with Indo-European languages, Schleicher ventured an borrowings. All things considered, there are four basic actual reconstruction of a short text, the ‘fable of the assumptions on which the comparative method is sheep and the horses’. based: (1) the proto-language was uniform, with no During the second half of the nineteenth century, a dialect variation; (2) language splits are sudden; (3) group of four young scholars known collectively as the after the split up of the proto-language, there is no sub- Junggrammatiker (‘young grammarians’ or ‘neogram- sequent contact among the related languages; and (4) marians’) proclaimed the ‘absolute exceptionlessness sound change is regular. of sound laws’, i.e. all changes in the sound system of Nowadays, the work of reconstruction usually a language as it developed through time were subject begins with phonology, i.e. with an attempt to recon- to the operation of regular sound laws. The assumption struct the sound system; this in turn leads to recon- that sound laws were in principle absolutely regular in struction of the vocabulary and grammar of the their operation was quite generally accepted and had proto-language. In the nineteenth century, comparative become the landmark of the comparative method. studies were entirely word based: it was words or lex- Realizing that sound changes occur in defined ical word stems that were reconstructed and followed geographical regions that do not coincide with the through history. Only after the knowledge of phonetics 222  COMPARATIVE METHOD had begun to increase did linguists realize that they because they resist borrowing more than other compo- should start reconstruction with sounds, not with lexi- nents of the vocabulary. The final step is to determine cal items. Consequently, historical linguistics has what system of sounds the ancestral language appar- often enjoyed great success in reconstructing ancestral ently had and what the rules were for combining these phonological systems and vocabularies. But for gram- sounds. matical systems, the situation is slightly more compli- The success of any reconstruction depends on the cated: grammatical patterns are often seriously material at hand and on the ability of the comparative perturbed by independent and cointeracting processes linguist to decipher what happened in the history of of change, and grammar usually undergoes more dra- the languages under comparison. Generally, the longer matic changes than the sound system and even the in the past the proto-language split up, the more lin- vocabulary. guistic changes will have occurred and the more diffi- Generally, studies in the comparative method start cult the task of reconstruction with full success is with the assumption that certain languages are relat- going to be. ed and hence descended from a common ancestor. The choice of languages is also very important. If The first step in the work of reconstruction is to find we simply pick some arbitrary languages and com- sets of putatively related words in the languages or pare them, we cannot expect to see systematic corre- dialects being compared, i.e. placing side by side a spondences appear, even if the languages selected are number of words with similar meanings from the indeed genetically related. The comparative method languages under comparison, for example, Sanskrit has to be applied carefully and thoughtfully, and pita, Latin pater, Greek pater, and Gothic fadar, all every piece of information must be taken into consid- of them meaning ‘father’. Such sets of ‘cognate’ eration as possibly relevant. Linguists have to be forms should be examined for what seem to be sys- aware of a number of potential difficulties that might tematic correspondences in their sound structure. lead the reconstruction into error. Perhaps the most Each correspondence must yield a plausible-looking obvious point is that the comparative method cannot sound in the ancestral language, one that could rea- recover any feature of the ancestral language that has sonably have developed into the sounds that are disappeared without a trace in all the attested daugh- found in the several daughter languages; phonologi- ters. Other problems are loanwords, pure coinci- cal change should be taken into account, because dences, nursery words, imitative words, and any difference in the form of cognates is a conse- phonaesthetic words (sound symbolism), because quence of sound change in one or more of the lan- these tend to be somewhat similar even in unrelated guages. Establishing regular correspondences languages. between languages permits the assumption that the languages in question are genetically related; these References correspondences can also be used to formulate like- ly hypotheses about what changes gave rise to those Beekes, Robert S.P. 1995. Comparative Indo-European linguis- correspondences, and to reconstruct antecedent tics: an introduction. Amsterdam, Philadelphia: Benjamins. Brogyanyi, Bela, and Reiner Lipp (eds.) 1993. Comparative- unattested forms in the proto-language from those historical linguistics: Indo-European and Finno-Ugric. changes. Amsterdam, Philadelphia: Benjamins. A set of ‘reconstructed’ sounds can then be postu- Collinge, N.E. 1985. The laws of Indo-European. Amsterdam, lated (marked with an asterisk by the standard conven- Philadelphia: Benjamins. tion) to which the sounds in the attested languages can Meillet, Antoine. 1953. Introduction à l’étude comparative des langues indo-européennes, 8th edition. Paris: Hachette. be systematically related by means of sound laws. Meillet, Antoine. 1967. The comparative method in historical With these results, each word surviving in the various linguistics, translated from the French by G.B. Ford, Jr. and daughters may be put face to face with its form in the H. Champion Paris. Oslo: Instituttet for Sammenlignende ancestral language. The reconstructed Proto-Indo- Kulturforskning. European word for ‘father’ is *pəter. If neither chance Saussure, Ferdinand de. 1968. Mémoire sur le système primitif des voyelles dans les langues indo-européennes (1879; nor borrowing can account for observed similarities, Report on the Primitive System of Vowels in Indo-European shared inheritance (i.e. ‘genetic’ relatedness) is the Languages) (Reprografischer Nachdruck der Ausg. Leipzig only possibility. 1879.) Hildesheim: Olma. Languages replace lexical items over time; the more Sihler, Andrew. 1995. New comparative grammar of Greek and the time, the fewer the items retained from the ances- Latin. New York: Oxford University Press. tral language, and the fewer the potential cognates. LAURA DANILIUC This is why linguists agree that it is convenient to start See also Bopp, Franz; Grimm, Jacob; Historical with cognates belonging to the basic vocabulary, Linguistics; Saussure, Ferdinand de 223 COMPOSITIONAL SEMANTICS Compositional Semantics Compositional semantics includes all approaches to semantic rule to the meanings of the parts of the the meaning of natural language that subsume, in one expression. formulation or another, the principle of ‘composition- Some natural language expressions have been ality of meaning’. Very generally put, this principle claimed to be difficulties for a compositional approach reads: ‘The meaning of a complex expression is a to semantics. One such difficulty is represented by function of the meanings of its parts.’ This is often compound nouns such as bookworm (which is not nor- called Frege’s principle, although it cannot be found in mally a worm but a person) or ashtray (which is not a explicit form in his writings. Almost any theory of tray). However, most of these challenges have been meaning for natural language is based on this princi- successfully met by insightful compositional analyses. ple. There is a psychological motivation for this, Due to its psychological plausibility, the composi- because the principle is able to explain how a speaker tionality principle characterizes not only formal can understand sentences never heard before. approaches to semantics but also one major area of A more explicit formulation of the principle is: ‘The cognitive linguistics. In Adele Goldberg’s Construction meaning of an expression is a function of the meanings Grammar, for instance, syntactic constructions are con- of its parts and of the way they are syntactically com- sidered to be substantive principles of semantic com- bined.’ This formulation makes it explicit that the mean- position: since the constructions per se are taken to be ings of the parts are not sufficient for determining the meaningful, the meaning of linguistic expressions is meaning of the whole, and that the way the parts fit the result of integrating the meanings of the words into together is responsible for differences in meaning (John the meanings of constructions. hit Mary vs. Mary hit John). In this form, the principle is rather uncontroversial, although when one looks at References the way the relation is a function of is spelled out in dif- ferent semantic theories, there is considerable diversity. Goldberg, Adele E. 1995. Constructions. Chicago, IL: This principle is the fundamental one in Montague University of Chicago Press. Janssen, Theo M.V. 1997. Compositionality. Handbook of logic Grammar, an application of the essential methods of and language, ed. by Johan van Benthem and Alice ter formal logic to the study of natural language seman- Meulen. Oxford: Elsevier and Cambridge; MA: MIT Press. tics: it helps in designing a finite system (with a limit- Katz, Jerrold J., and Jerry A. Fodor. 1963. The structure of a ed number of words and ways of combining them) semantic theory. Language 39. with infinite output (meanings for all possible sen- Partee, Barbara H. (ed.) 1976. Montague Grammar. New York: Academic Press. tences). One of the basic tenets of Montague Grammar Partee, Barbara H. 1984. Compositionality. Varieties of formal is the ‘rule-to-rule’ correspondence between syntax semantics, ed. by Fred Landman and Frank Veltman. and semantics: the syntax contains several rules that Dordrecht: Foris. provide several ways to form a complex expression Thomason, Richmond H. 1974. Formal philosophy. Selected from parts, each with a specific semantic effect (i.e. papers of Richard Montague. New Haven, CT: Yale University Press. for each syntactic rule, there is a corresponding ANDREA SANSÒ semantic rule). Put informally, the meaning of a com- plex expression results from the application of a See also Meaning; Montague, Richard; Semantics Compounding Compounding is the process of creating new words tion’) from the two simpler words chat and line. A whose elements are smaller words. For example, the word formed in this way is called a ‘compound’. relatively new English word chatline is created by The ‘words’ that are the constituent parts of com- compounding (a process sometimes termed ‘composi- pounds may be subject to specific formal requirements, 224  COMPOUNDING which depend on the language concerned. In the Latin a bas-bleu (French) ‘bluestocking’ (literally, ‘stocking compound liquefacio, ‘to dissolve’, facio means ‘I blue’) may not be a stocking, but not only does bleu make’, but lique, although clearly related to liquidus, obviously modify bas, the construction as a whole takes ‘liquid’, cannot occur by itself, but only in compounds its gender from bas. and other derivative word-forms. In the Danish com- The same is not so obviously true of another type of pound sovesofa, ‘sleep sofa, sofa bed’, the verb in the compound, sometimes called ‘coordinative com- first element of the compound is in the infinitive. In pounds’. Here, two elements are put together, and the other languages, although it is rare, finite verb-forms compound either denotes the unity of the two elements can be used, as in Arabana-Wangkangurru yanhi·rnda· or is used as the superordinate term for things that tharka·kura ‘speak-pres-stand-past.continuous = they include the two named elements.__ _ In the first case, we were standing talking’. In languages like Finnish and get things such as Marathi ai -ba p, ‘parents’ (literally Sanskrit, both basic and inflected forms of nouns can be ‘mother–father’); in the latter case, we get things like found in the first element of compounds. Compounds Kannada bassu karu, ‘vehicles’ (literally ‘bus–car’). usually consist of two parts: a modifying word and a The borderline between coordinative and endocentric modified one. The modified word is called the ‘head’ of compounds is often blurry, and classifications may the compound because it contributes the dominant vary for different scholars: several distinguishable meaning and determines whether the compound is, e.g. classes of compounds similar to coordinative com- a noun or a verb. Whatever the form of the modifying pounds are found, but they are not always kept clearly element of the compound, it is usually fixed in the com- apart. pound and does not change depending on the environ- All the examples that have been considered so far ment in which the compound occurs. The head element, are examples of what are usually called ‘root’ or ‘pri- however, may be inflected, e.g. for tense (in the case of mary’ compounds. One of the typical characteristics of verbs) or number (in the case of nouns). primary compounds is the wide range of possible Typically, compounds denote hyponyms, or special meaning relations between the elements. For example, cases, of their head. A chatline, for example, is a type a flour mill grinds flour, but a water-mill is powered by of line, not a type of chat. In English, the head is water, whereas a sawmill uses a saw as an instrument almost always the right-hand element in the com- in its work. There is nothing overt in these words to pound. In a language like French, however, the head is show that they have different interpretations, and the frequently on the left: a leçon-cuisine, literally ‘lesson same form may have different interpretations on dif- cookery’ (cooking lesson), is a type of leçon and not a ferent occasions: the Washington flight may be the type of cuisine. Rolls-Royce is quite correctly translat- flight to Washington or the flight arriving from ed into French as Royce-Rolls. A remarkable number Washington. of languages show both patterns of headedness in A contrasting set of compounds, usually called compounds. In either case, compounds are ideal con- ‘synthetic’ or ‘verbal’ compounds (some scholars structions for creating names for subtypes. distinguish between the two labels), has a very restrict- Because compounds are typically hyponyms of their ed interpretation. These are words such as bus driver, head element, they also take their part of speech from in which the head element is a derivative of a verb and their head element: for example, chatline has a noun in in which the modifying element must be interpreted as its head and is hence a noun, user-friendly has an adjec- an argument of that verb. The argument concerned is tive in its head and is thus an adjective, and stage-dive usually the direct object of the verb (as in bus driver, has a verb in its head and is consequently a verb. ‘someone who drives buses’) but may on occasions be Compounds in which this is true are called ‘endo- other arguments, such as instrument in hand-sewn and centric compounds’ because their center of meaning subject in consumer spending. lies within the compound itself (Greek endo, ‘inside’). Precisely where compounding stops and sentence There are also ‘exocentric compounds’ (Greek exo, formation begins is a matter of some dispute. In par- ‘outside’), which are not hyponyms of their head ele- ticular, it is not clear whether structures like Onondaga ment (an egghead is a kind of person, not a kind of waʔ·ha·hwist·ahtu·ʔt·aʔ, ‘he lost money’ (in which the head) or, more radically, whose part of speech is not direct object money is directly attached to the verb deducible from that of their elements (for example, lost) should be regarded as a compound or some other put-down consists of a verb, put, and a preposition, construction. down, but it acts as a noun). Although exocentric compounds are not hyponyms References of their own heads, they can still be seen to be headed, Adams, Valerie. 1973. An introduction to modern English in that one element modifies the other. A redneck may word-formation. London: Longman. not be a type of neck, but it is still clear that red modifies Bauer, Laurie. 2001. Compounding. Language typology and neck and that redneck thus has its own right-hand head; language universals, ed. by Martin Haspelmath, Ekkehard 225 COMPOUNDING König, Wulf Oesterreicher, and Wolfgang Raible, 695–707. Ryder, Mary Ellen. 1994. Ordered chaos. The interpretation of Berlin and New York: de Gruyter. English noun–noun compounds. Berkeley, CA: University Botha, Rudolf P. 1984. Morphological mechanisms. Oxford: of California Press. Pergamon. Williams, Edwin. 1981. On the notions ‘lexically related’ and Coates, Richard. 1999. Word structure. London and New York: ‘head of a word’. Linguistic Inquiry 12. 245–74. Routledge. LAURIE BAUER Computational Linguistics Computational linguistics is a discipline between com- systematic approach to how to perform the task. From puter science and linguistics, for which the main con- the very early years of machine translation studies, we cern is the computational aspects of human language. have two main approaches to the problem of automat- These are aspects that can be put into the form of a ing translation. The first method involves carrying out sequence of instructions that a computer can under- a direct translation from one language into the other. stand. For example, if we have a sentence like John The second method is the ‘transfer’ grammar runs, we can set up some simple grammar rules. These approach, in which everything is translated to a kind of rules state that a sentence is made up of a noun and a intermediate language. From here, it is easier to make verb, a noun is always John, and a verb is always runs. a translation to many other languages. This intermedi- Of course, this grammar only produces our sentence, ate grammar may have the form of a real human lan- but it is a grammar that can be used very easily in a guage (Esperanto or English are examples), or a computer program. We only have to put three instruc- certain symbolic language. A simple example is the tions (Sentence = Noun + Verb, Noun = John, and Verb following: we have the French sentence Marie aime = runs), in a format that the machine understands, to Jean and then we can make a translation to a simpli- produce the sentence. We say then that this grammar is fied English sentence of the form Marie(S)-love- fit for computation. We do not include in our definition Jean(O), with S marking the subject and O the object. those tools that computer science gives us for the study From this intermediate sentence, we translate again, of language. Sometimes these are also presented under now into Spanish: Marie ama a Jean. The technique is the generic label, but they are normally not considered very useful when translating to and from many differ- to be part of this field. ent languages, since we have an intermediate expres- Two main practical goals in computational linguis- sion from which we can easily make translations that tics have been machine translation and natural lan- are faithful to the same extent to the basic structure. guage interaction with computers and machines. When Both techniques have their advantages and drawbacks. we refer to ‘natural language’, we are doing so in con- In addition, even if there should not be any problem trast to ‘artificial language’. Language can be called at all with the translation strategy adopted, there are ‘natural’ when it can be used for communicating with other facts that machine translation has not properly people in the same way as a human being would do. accounted for to date, the most important being the Machine translation, that is, translation carried out resolution of ambiguity in language. The phrase The by a computer, was the first goal of computational lin- cover of the book which I like has syntactic ambiguity, guistics. When scientists started to work in this field, it since we cannot be sure whether what I like is the looked like an easy task. It was thought to be just a cover or the book. Similarly, The seal is on the rock question of making words and expressions in the has semantic ambiguity, since we do not know source language correspond to words and expressions whether what is on the rock is an animal (a seal) or in the target language. Very soon it became clear that some kind of stamp (a seal). We may also have prag- it was not so simple, since language has a structure matic ambiguity, of which a very simple case is the that is more complex than just a lexicon database. difficulty to translate the English pronoun you into lan- There is more to language than just words. guages like Spanish, French, German, or Japanese, However imperfect the process still is, though, where there are several different ways of referring to a automated translation now gets much better results second person according to degrees of formality or to than in the past. And it has achieved them through a the relation between the speaker (or writer) and the 226  COMPUTATIONAL LINGUISTICS hearer (or reader). Obviously, the computer cannot pretations are assigned to them. This is the true com- infer this type of information, which depends on the putational linguistics approach, which has also found communicative situation. problems similar to the ones mentioned for machine Of all these types of ambiguity, the most important translation. It is this approach that shall be explained for machine translation is the last one, since it cannot below. be resolved by looking at the immediate linguistic In computational linguistics, the structure of sen- context. tences needs to be represented in some way. In order Another main area that computational linguistics to do this, the computer has some internal arrange- has always been interested in is the interaction with ment of data that corresponds to the semantic struc- machines using natural language. This can be done tures of the knowledge that we have in our mind, and through the use of the written word or through the use that helps clarify the way in which the linguistic mes- of oral speech. In the latter case, the system needs to sage organizes the information it conveys. This incorporate a module for speech recognition, in addi- arrangement is in the form of a semantic representa- tion to the other modules used for linguistic process- tion. The computer also knows about the rules of the ing that we have mentioned. The usefulness of language for putting words and sounds (or letters) mechanisms that react to ordinary speech in activities together. These are syntactic and phonetic (or where free hands are needed has led to the develop- graphemic) rules. ment of such systems for cars, planes, or control of The semantic representations used in computation- processes in factories, which are typical examples. In al linguistics try to make the most of the resources that all these cases, in addition to the procedures of lan- the machine offers, but they also try to be faithful to guage analysis and generation, there are special calcu- the structure of concepts. Let us take the example of lations that take into account the noise that may the concept ‘bird’ to illustrate how these representa- accompany ordinary language. tions can be made. When we think of a bird, we imme- In particular, there has been a special concern in diately visualize an animal with wings and feathers. In trying to produce language in a manner as human-like our mind, we have a link between the concept ‘bird’ as possible in expert systems. An expert system is a and the concepts ‘wing’ or ‘feather’. If we write the system capable of reasoning in a way similar to a spe- words ‘bird’, ‘wing’, and ‘feather’, and draw two lines cialist in a certain field, and it is therefore necessary, from the label ‘bird’, one to the label ‘wing’ and or at least very convenient, that the queries can be car- another to the label ‘feather’, we will have a represen- ried out in a language as close as possible to natural tation of the above image by using a labeled diagram. human language. This is so because expert systems are If we now add more words to the diagram, such as, for intended for use in many kinds of settings, and may instance, ‘beak’, ‘egg’, ‘nest’, or ‘to fly’, the diagram have to be operated by persons with different degrees becomes more complex, but now the links start to be of expertise. It is also important to try to reproduce the of different kinds. A bird normally ‘has’ a beak, ‘lays’ typical interactive client–expert exchange. eggs, ‘builds’ nests, or just ‘flies’. These different In natural language interaction there are two main types of relation also need to be labeled. Finally, we strategies. The oldest and the easiest is to check for now know that not all these features are necessary and certain words and expressions, ignoring the rest of the sufficient, but rather prototypical. It is difficult to message. This is the direct, crude approach. According imagine a bird without a beak or wings, but certainly to the expressions the system recognizes, it gives cer- not all of them fly or build nests. This is more difficult tain answers. This was the approach used in Eliza, a to put in our diagram, but we can try to use some program created by Joseph Weizenbaum in numbers that denote to which extent these features can 1964–1966. It was designed to take on the role of a be applicable. Rogerian psychotherapist, and it produced responses When our diagram becomes larger, it should look according to the trigger expressions that appeared. like a net. In fact, it is a way of depicting a semantic Thus, to a sentence with the word friends, it could net, a kind of representation that is being used in react saying something like Why are you thinking of computational linguistics since the late 1960s. your friends now? so that to the sentence I have many Semantic nets are incorporated in the data of compu- friends, or My wife has many friends, we would get the tational linguistic applications, so that the semantic question, Why are you thinking of your friends now? structure of sentences is interpreted or generated by The rigidity of this approach is obvious, although it is building small nets that match the connections that the most economical by far. appear in the larger ones. As a formalism, a semantic On the other hand, we have systems where there is net consists of a collection of points (the labels a special module for language recognition, in which ‘bird’, ‘wing’, ‘feather’, etc. in our example) called sentences are parsed (analyzed) and semantic inter- nodes, each representing a concept. A node may be 227 COMPUTATIONAL LINGUISTICS connected to any other node in the net by means of assign several statements as conditions for the truth of relations that in turn carry labels (‘have’, ‘build’, another one, and the resulting structure is called a rule. ‘fly’, etc.). There are some conventions: two very An example will help to understand this. Below there typical relations normally referred to in computation- is PROLOG rule: al linguistics books are the labels IS-A (x ‘is a’ y) and H-A-P (x ‘has as part’ y). But any other that corre- aunt (X, Y) :- sponds to common English verbs or adjectives female (X), (LIKE, COLOUR, etc.) can also be used. brothers (X, Z), Let us look at another example. A semantic net for parent (Z, Y). the sentence Mary likes Peter could be the following: Let us translate these statements into understand- GIRL BOY able English (on the right): IS-A IS-A MARY LIKES PETER aunt (X, Y) :- ‘X (a variable) is Y’s H-A-P (another variable) aunt if...’ EYES female (X), ‘X is a female’ COLOUR brothers (X, Z), ‘X and Z (a third variable) BLUE are brothers/sisters’ parent (Z, Y). ‘Z is a parent of Y’ This net provides a semantic representation for the message conveyed. However, most of the information In plain English, this would be something like is not in the sentence itself, which only carries labels. ‘Some entity X is the aunt of some other entity Y if the We retrieve all the additional information from struc- entity X is a female, the entities X and Z are brothers tures of knowledge that are either in our mind or in the or sisters, and the entity Z is one of Y’s parents.’ computer’s data. There is a matching process in order The computer programmer writes this rule as part to see which parts of the sentence’s semantic structure of a larger program, and we can also add some infor- and the program’s data coincide. It is like having two mation to the program’s database by writing the fol- pictures in two transparencies, A and B, of a face and a lowing statements: nose. The nose has more detail in B, but it coincides partially with what is drawn in A; hence, we know that female (mary) ‘Mary is a female’ it corresponds to that face. In addition, when we put the brothers (mary, john) ‘Mary and John are sister two transparencies together, the face receives a better and brother’ drawing, with the addition of a more detailed nose. parent (john, jim) ‘John is one of Jim’s This matching process is currently referred to in parents’ computational linguistics as unification, and it is con- sidered a very powerful resource for natural language (Notice that the terms ‘female’, ‘mary’, ‘john’, etc. processing. When a computer programmer uses a pro- have no capital letters, because they are constants.) gramming language such as PROLOG, for instance, The computer operator may wish at some stage to there is an extensive use of unification. know who is Mary’s nephew, or to whom she is an We can illustrate this by having a close look at aunt. We can then make use of the ‘aunt’ rule. The how PROLOG works. As has just been stated, PRO- operator writes: LOG is a programming language, and all program- ming languages have a vocabulary and a syntax, ?- aunt (mary, Y) ‘Mary is Y’s aunt’ exactly in the same way as natural languages do. In the vocabulary, we have strings of symbols that As this is a plain statement (not a rule), the PRO- stand for variables (with capital letters, e.g. X, YW, LOG system looks for a suitable term that can change AA) or for constants (without capital letters, e.g. a, the variable Y into a constant. If you look at the infor- bb, person, mother). The corresponding items in a mation provided in the statements that have been natural language for variables would be pronouns incorporated into the program so far, this term is jim. (which stand for other terms). For constants, we Accordingly, it answers: have proper nouns (which, in ideal cases, refer to unique entities). Y = jim Statements in PROLOG normally have a structure in which there is one predicate and several arguments. We then say that jim instantiates the variable Y. If A is the predicate and x, y the arguments, the con- The PROLOG system has carried out a unification vention is to write A (x, y). The symbol :- is used to process in order to come to the right solution. It has 228  COMPUTER-ASSISTED LANGUAGE LEARNING (CALL) unified the rule and the statements in the following nodes were used for events as well as objects, and way: cases appeared as labels in semantic nets. Parallel to the relation IS-A for objects, there is also now ACT- aunt (mary [=X], Y) :- OF. In the sentence John throws the ball, John is female (mary [=X]) linked through an AGENT relation to a node. ‘The brothers (mary [=X], john [=Z]) ball’ is also linked to that same node by the relation parent (john [=Z], jim [=Y]) OBJECT. And the node is linked to ‘throw’ by means of ACT-OF, which therefore links particular instances So that Y = jim. of events to the general node for a certain action in the knowledge database. Unification is a very important procedure in com- An example of how the semantic net paradigm can putational linguistics. It simulates on a computer the be realized is R.C.’s (1973) theory of conceptual processes that take place when we compare informa- dependency. This author’s graphs follow the formal tion from our mental store with the information in the structure of a dependency grammar, and have four dif- linguistic message. ferent types of nodes. These are symbolized by PP, Another well-known programming language for arti- ACT, PA, and AA, corresponding very roughly to ficial intelligence is LISP, of which many different ver- noun, verb, adjective, and adverb, respectively. Schank sions have been created afterwards (schema, xlisp, etc.). makes use of a small set of primitive actions (PRO- LISP stands for ‘List Processing’, and it is a comple- PEL, MOVE, INGEST, EXPEL, GRASP, PTRANS, mentary approach to PROLOG. Whereas in PROLOG MTRANS, ATRANS, SPEAK, ATTEND, and we had a deductive mechanism that unified knowledge MBUILD) and the default action DO. With the help of structures, in LISP we have a powerful engine for pro- a small number of states, this author claims that the cessing lists. This is useful in computational linguistics, meaning of all verbs can be represented. It is interest- where syntactic parsing is an important step in the ing to see how the semantic primitive approach is used analysis of sentences. An additional important feature, here. In addition, we also have rules that must be which is very convenient for many artificial-intelligence applied for well-formed conceptualizations. approaches is that LISP programs are able to self-mod- Sets of semantic primitives have been used in many ify, making it possible to write code that ‘learns’, or that CL systems. However, it is important to clarify here adapts itself to changing circumstances. that there is no claim about their universality or a very PATR (Shieber 1986) is a declarative language that strong psychological justification. The use of a limited has eventually become a standard in computational set has more to do with a question of convenience. linguistics. This is so because it is very easily adapt- able to many different theories. In fact, many current References developments use feature structures and unification in a PATR style. The most prominent characteristics of Schank, R.C. 1973. Conceptual dependency: a theory of natural language understanding. Cognitive Psychology 3. 552–631. the other two languages that we have seen here, unifi- Shieber, Stuart M. 1986. An introduction to unification-based cation (PROLOG) and string concatenation (LISP), approaches to grammar. CSLI Lecture Notes, Vol. 4. are used in this language. Stanford. With respect to semantic nets, computer scientists Wilks, Yorick. 1973. Preference semantics. Memoranda from soon found that representations like the ones shown the Artificial Intelligence Laboratory 206. Stanford, CA: Stanford University. above fail to give detailed information about the situa- CARLOS INCHAURRALDE tions that are represented. Location, time, setting, etc., were notions that did not appear in the structures. This See also Ambiguity; Artificial Intelligence; led to more complex representations, in which concept Machine Translation Computer-Assisted Language Learning (CALL) Computer-assisted language learning (CALL) has been enced and shaped by those occurring more generally in an important aspect of language teaching and learning the fields of education and applied linguistics. The first since the 1960s, and its developments have been influ- applications of computers in language teaching were 229 COMPUTER-ASSISTED LANGUAGE LEARNING (CALL) based on large-scale, specialized mainframe computers, Also since the early 1990s, access to global com- and hundreds of person-hours of programming were munication networks and the World Wide Web, and needed to develop the instructional software. These the availability of point-and-click Web-browsing soft- computer-assisted instruction (CAI) systems were ware, has become increasingly easy to use and inex- meant to be ‘stand-alone’, which meant that all aspects pensive. The Internet has made global of teaching and learning were contained within the communication, through electronic mail (e-mail), hardware and software. It was very difficult (or often commonplace and has led to a variety of tandem lan- impossible) for teachers to tailor the software to their guage-learning projects. For example, English-speak- classroom teaching, and some of the earliest complaints ing students studying French in the United States about CAI were related to this problem. The teaching could be paired electronically with French-speaking approach used in CAI systems was tutorial (drill and students studying English in Canada. Both groups practice) and was based on behaviorist learning theories write to each other in the target language and then that were popular at the time. Among the early CAI send, read, or reply to messages at their convenience; projects, perhaps the two most influential for language they do not need to be on-line at the same time (asyn- teachers, were PLATO (Programmed Logic for chronous communication). Learning is aided as stu- Automatic Teaching Operations) and TICCIT (Time- dents work at their own pace and help each other to shared, Interactive Computer-Controlled Information notice problems in grammatical forms or learn how to Television), developed at the University of Illinois and express ideas in more natural, idiomatic language. In Brigham Young University, respectively. addition, students can provide insights to their e-mail In the 1970s and 1980s, the change to more com- partners about respective cultural practices and municative approaches to language teaching and norms. Such peer-based sharing of ideas to help learn- learning coincided with the development of micro- ers develop a deeper understanding of language is computers and the corresponding proliferation of soft- called ‘scaffolding’ and has been influenced by learn- ware. Many programs were still tutorial (drill and ing theories from education and psychology. practice), but more general learning tools, including, In contrast, synchronous communication, in which for example, word-processing, database, spreadsheet, all people are on-line and writing (typing) at the same and, by the 1980s, communications software, were time, is also used in CALL classrooms. Synchronous adopted from the business environment. In addition, communication, for example, in on-line chatrooms, authoring software (high-level programming lan- requires students to log into centralized ‘rooms’ on the guages) helped teachers write computer programs to Web, where they can adopt pseudonyms or even complement their other classroom activities. assume different on-line identities. The pace of written Consistent with new communicative approaches to exchanges is fast, and the style of language is often language teaching, teachers began to group students so informal. Many language teachers believe that unless that they could work together in teams to write simple there is a well-developed curriculum structure and software programs. Not only did this new instruction- educational purpose to synchronous communication, al approach help students to become more independent students may not learn appropriate linguistic forms. In learners, but also team-programming provided them fact, some researchers believe that on-line chatroom with opportunities to practice language in authentic communication involves a new, hybrid form of lan- communicative contexts. guage that incorporates features of both speaking and Since the 1990s, with the advent of inexpensive writing. memory and other hardware components, for example, Although the Internet supports the exchange of writ- resident CD-ROM or DVD (digital video disk) play- ten text through e-mail or in chatrooms, it also provides ers, teachers have had access to a range of commer- access to a variety of other language-teaching resources. cially developed, multimedia language software. Such These include, for example, access to Web-based soft- software is usually created as a hypertext environment ware (much of which can be downloaded) and special- in which text, sound, video, and graphics are creative- ized Web sites containing organized links to ly linked to one another. Students can point and click language-learning resources around the world. For on different words, ideas, or branches of a program example, students can read daily newspapers, listen to and move through instructional material at their own live radio broadcasts, or watch and listen to television pace, according to their personal interests. However, news from a variety of countries. As a result, access to multimedia software can be expensive and cannot ‘authentic’ language resources is becoming easier for always be tailored to individual classroom learning teachers, but how to evaluate, organize, and integrate on- needs. Some researchers have also questioned whether line materials so that students can learn language in a reading and writing in hypertext environments, dis- structured manner have become important issues in jointed as it can be, helps or hinders language learning. teacher training. Certainly, there is a growing acceptance 230  COMRIE, BERNARD that the role of the teacher and student in CALL class- References rooms is changing, with teachers assuming more facili- Bush, Michael D., and Robert M. Terry (eds.) 1997. tative, guiding roles and students becoming more Technology-enhanced language learning. Lincolnwood, IL: autonomous learners. National Textbook Company. The latest technological changes that are affecting Egbert, Joy, and Elizabeth Hanson-Smith (eds.) 1999. CALL CALL include wireless communication, which relies environments: research, practice, and critical issues. Alexandria, VA: TESOL Publications. on radio signals, rather than physical wiring, to con- Levy, Michael. 1997. Computer-assisted language learning: nect to the Internet, and a variety of new software context and conceptualization. Oxford: Clarendon Press and tools. For example, it is now possible for students to New York: Oxford University Press Inc. use wireless laptop computers, video cameras, and Murray, Denise E. 1995. Knowledge machines: language and word-processing and movie-editing software to write, information in a technological society. New York: Longman. Pennington, Martha C. 1996. The power of CALL. Houston: film, and review their own language materials. Such a Athelstan Publications. project-based approach to teaching and learning could Taylor, Todd W., and Irene Ward (eds.) 1998. Literacy theory in effectively integrate current theories of autonomous the age of the internet. New York: Columbia University learning, peer scaffolding of learning, and authentic Press. communication. Regardless of the technology used, Warschauer, Mark. 1999. Electronic literacies: language, cul- ture, and power in online education. Mahwah, NJ: Lawrence however, the overriding issue for teachers will remain Erlbaum Associates. how to coordinate learning experiences so that stu- Warschauer, Mark, and Richard Kern (eds.) 2000. Network- dents have meaningful access to language—however based language teaching: concepts and practice. Cambridge: sophisticated their computers and whichever software Cambridge University Press. they use. E. MARCIA JOHNSON Comrie, Bernard Bernard Comrie is one of the leading figures in lan- ating interest in these topics, even among linguists guage typology. As a fieldwork typologist, Comrie has who end up disagreeing with his approach. carried out studies on Fula, Tunisian Arabic, Some common linguistic terms are actually used Malayalam, Moroccan Arabic, Khmer, Tsez, Sudanese for the first time in his papers. The term ‘accessibility Arabic, Maltese, and Basque, on which he has taught hierarchy’ came with Keenan and Comrie (1977). His field methods classes. He has also worked in depth in book Aspect played a major role in getting the distinc- the field on two additional languages: Haruai and tion widely accepted between ‘perfect’ and ‘perfec- Bezhta. Comrie has also taught and researched tive’. He is perhaps the first linguist to introduce, Historical Linguistics and Diachronic Syntax. among other terms, the ‘S, A, P’ terminology (as Although typological linguistics is the overall opposed to Dixon’s similar ‘S, A, O’). approach that is closest to his own, his work has very One of the main features in his thinking has been close ties with cognitive linguistics. Even if as a for- the integration of links between language and other mal framework, Relational Grammar is now virtually phenomena, in particular cognitive and social. He is in dead, in terms of formal approaches to grammar, search of correlating linguistic and other (genetic, Comrie believes that linguists have learned a lot about archeological) evidence for human prehistory. He the grammars of different languages and about the insists that these different disciplines provide us with general properties of language from that framework. different windows into human prehistory that must Comrie has been influential in the recent upsurge of then be reconciled with one another, rather than pre- interest in linguistic typology, both through his gener- supposing, for instance, a close connection between al books on the topic and through individual articles. language families and biological populations. Within typology, the areas in which he has had most Comrie’s ideas have not remained static. In some of impact are relative clauses, causative constructions, his later articles (e.g. ‘Rethinking the typology of rela- and grammatical relations. In addition, his books on tive clauses’, 1998), he has argued for substantial revi- tense and aspect have had an important role in gener- sions to the ways in which typologists and other 231 COMRIE, BERNARD linguists look at relative clauses, in particular, paying University of Cambridge, receiving a B.A. (1968) and more attention to semantic and pragmatic factors. With a Ph.D. in Linguistics (1972). His dissertation ‘Aspects regard to aspect, he gives more import to the role of lex- of sentence complementation in Russian’ was written ical properties of verbs than he did in the mid-1970s. under the supervision of Pieter A.M. Seuren. Comrie’s first publications were articles derived Comrie taught linguistics as a University Lecturer from his doctoral dissertation, and thus dealing with in Linguistics at the University of Cambridge generative grammar. However, his first articles in the (1974–1978). In 1978, he joined the Department of field of universals and typology include ‘Noun phrase Linguistics at the University of Southern California, accessibility and universal grammar’, written jointly Los Angeles, and worked there as a Professor of with Edward L. Keenan (1977). He has a remarkable Linguistics until 1998. Since 1997, he is director of the number of publications, mostly on the typology of Max Planck Institute for Evolutionary Anthropology ergativity, aspect, causative constructions, and nomi- in Leipzig, Germany, in addition to being a part-time nalizations. His major publications include Aspect Distinguished Professor of Linguistics at the (1976), Lingua descriptive studies: questionnaire University of California, Santa Barbara (since 2002). (with Norval Smith, 1977), The Russian language In addition, he holds honorary positions as a Research since the Revolution (with Gerald Stone, 1978), Professor of Linguistics, University of Southern Language universals and linguistic typology: syntax California (since 1998) and as a Professor of and morphology (1981), The languages of the Soviet Linguistics, University of Leipzig (since 1999). Union (with B.G. Hewitt and J.R. Payne 1981), Tense (1985), and The Russian language in the twentieth century (with Gerald Stone and Maria Polinsky 1996). References To this list, we should add his several important edited Comrie, Bernard. 1976. Aspect. Cambridge: Cambridge volumes (like Classification of grammatical cate- University Press. gories (1978), Studies in the languages of the USSR Comrie, Bernard. 1979. Noun phrase accessibility and univer- (1981), The world’s major languages (1987), The sal grammar. Linguistic Inquiry 8. 63–99. Comrie, Bernard. 1989. Language universals and linguistic Slavonic languages (with Greville G. Corbett 1993), typology: syntax and morphology, 2nd edition. Oxford: and some others, along with dozens of articles on lan- Blackwell and Chicago: University of Chicago Press. guage typology. Comrie, Bernard. 1998. Rethinking the typology of relative clauses. Language Design 1. 59–86. Dahl, Östen. 1987. Comrie’s tense. Folia Linguistica 21. Biography 489–502. Bernard Comrie was born on May 23, 1947 in Declerck, Rennat. 1986. From Reichenbach (1947) to Comrie (1985) and beyond: towards a theory of tense. Lingua 70. Sunderland, England. By 14 years of age, he knew 305–64. French and had already started learning Latin, Kulikov, Leonid, and Heinz Vater (eds.) 1998. Typology of ver- German, Ancient Greek, and Russian. Coming across bal categories: papers presented to Vladimir Nedjalkov on R.H. Robins’ textbook on linguistics left a deep the occasion of his 70th birthday (Linguistische Arbeiten impression on him, showing him not only that there is 382). Tübingen: Max Niemeyer. Rakhilina, Ekaterina V., and Yakov G. Testelets (eds.) 1999. a field called linguistics, but that it is also a systemat- Typology and linguistic theory: from description to explana- ic, scientific endeavor. tion, for the 60th birthday of Aleksandr E. Kibrik, 319–39. Comrie studied Modern and Medieval Languages Moscow: Jazyki russkoj kul´tury. (mainly French, German, Russian, Linguistics) at the BEHROOZ MAHMOODI-BAKHTIARI Configurationality English is a configurational language because the postverbal object. Of the three main constituents of the grammatical functions subject and object appear com- clause—subject, verb, and object—the syntactic rela- pulsorily in particular syntactic positions: nearly every tionship between verb and object is bounded, as English clause must have an overt preverbal subject opposed to the syntactic relationship between subject and every English transitive clause must have an overt and verb, which is unbounded. The strong syntactic 232  CONFIGURATIONALITY ties between verb and object, which together constitute discontinuity. Syntactic continuity is often referred to in a phrasal unit, the verb phrase (hereafter VP), are evi- the literature as local dependency, as opposed to nonlo- denced by tests like VP deletion (Mark washed his car cal dependency or syntactic discontinuity. and Bill did [wash the car] too), VP pronominaliza- Theoretically speaking, the problem arises of how to tion (Mark washed his car and so did Bill), and VP account for syntactic discontinuity in grammatical phe- fronting (Mark said that he would wash his car and nomena like case assignment and agreement. In gener- wash his car he did). What these tests show is that al, two solutions have been adopted: monostratal subjects appear outside the verb phrase in syntactical- theories recognize a single level of syntactic represen- ly unmarked English sentences (that is, less complex tation, or, more clearly, a single linguistic representa- and more frequent than their unmarked counterparts), tion, in which the order and form of the elements of the whereas objects are immediately contained in the VP expansion (including grammatical case and phonologi- in English. Subjects and objects, therefore, exhibit dif- cal properties of accent, rhythm, and intonation) do not ferent properties of coding and behavior (coding prop- necessarily reflect the order and form of the elements erties are morphological properties such as agreement of the sentence that the speaker utters, whereas multi- with the verb, while behavior properties are distribu- stratal theories recognize more than one level of syn- tional and include, for example, the omissibility of the tactic representation, or, in other words, more than one subject in a coordinate subject construction like linguistic representation, in such a way that every level Matthews complained angrily and left the room): for represents the sentence that the speaker utters in a less instance, in English the verb agrees with the subject— abstract way than the preceding level (abstractness is not with the object—in number and person (He writes understood as the distance between the linguistic vs. They write) and the subject—not the object—of expression and its representation). Multistratal theories dependent clauses can be raised to—becomes—sub- justify the necessity of defining multiple levels of rep- ject or object of the matrix (main) clause under certain resentation on the grounds of the existence of syntacti- conditions (They consider that she is bright/I consider cally marked constructions (that is, more complex and her bright/She is considered bright). less frequent than their unmarked counterparts) like This correspondence between grammatical func- passives, questions, preposition stranding, etc., which tions and syntactic positions or phrase structure con- result from the application of a number of structure- figurations has motivated the configurational changing operation rules or transformations between definition of subject and object according to which the the deep (more abstract) and the surface (less abstract) subject is the noun phrase directly dominated by the levels of syntactic representation. Structure-changing node sentence and the object is the noun phrase direct- operations fall into three types: first, operations of dele- ly dominated by the node VP (subconstituents are tion of specified elements, as in I met a man who was immediately dominated by the constituents where they wearing a yellow coat vs. I met a man wearing a yel- belong). This configurational definition puts aside low coat; second, operations of substitution of one functional relations like subject and object to the ben- specified element by another specified element, as in efit of structural aspects of constituency, including, at These are the shoes [I bought the which in Paris] vs. least dependence (constituents have heads on which These are the shoes which I bought in Paris; and third, subconstituents depend), hierarchy (constituents con- operations of permutations of specified elements, as in sist of compulsory and optional subconstituents), and Jim doesn´t like bananas vs. Bananas Jim doesn´t like. linearization (constituents take up a clausal position The existence of the node VP, which is criterial for relative to the verb). configurationality, has been questioned on a cross-lin- In this approach, case assignment (as in guistic basis. The interlinguistic evidence provided who–whom–whose–to whom) and agreement (as in I eat includes: first, the lack of adjacency between the vs. she eats) crucially depend on the asymmetry object and the verb in VSO and OSV languages; sec- between the noun phrases that bear subject and object: ond, the syntactically unconstrained permutability of the object requires syntactic continuity with respect to the subject and object in languages such as Latin, the verb; the subject does not. Syntactic continuity is where the order of the major elements of the clause defined as a relation of adjacency between a head and can be SVO as well as OVS; and, third, the optionali- one of its dependents or between two heads: in Peter ty of overt subject and object in languages like Navajo, sent them to the Greens, the preposition and the noun a language in which some clauses consist simply of phrase it governs are adjacent to each other in the linear the element V. In consequence, a division into order of the sentence, whereas in Whom did Peter send configurational (with VP) and nonconfigurational them to? the sequence noun phrase-stranded (that is, (without a VP) languages is proposed. Although non- final) preposition is interrupted by the subject, the verb, configurational languages often coincide with relative- and the object, thus constituting an instance of syntactic ly free constituent order languages, this is not always 233 CONFIGURATIONALITY the case: there are languages, like Navajo, with a rela- References tively rigid constituent order that qualify as noncon- Baker, Mark C.2001. The natures of nonconfigurationality. The figurational. Navajo has a rich system of verbal handbook of contemporary syntactic theory, ed. by Mark prefixes, including prefixes of person and number of Baltin and Chris Collins. Oxford: Blackwell. the subject and object, thus allowing for null nominals Dik, Simon C. 1989. The theory of functional grammar. Part 1: (phonologically empty, or nonexistent) in the function the structure of the clause, ed. by Kees Hengeveled. Berlin: Mouton de Gruyter, 2nd edition, 1997. of both subject and object. Dik, Simon C. 1997. The theory of functional grammar. Part 2: Another dimension has been explored in which lan- complex and derived constructions, ed. by Kees Hengeveld guages can be configurational or nonconfigurational: Berlin: Mouton de Gruyter. the discoursive dimension. A language is discourse Hale, Ken. 1983. Walpiri and the grammar of non-configurational configurational if it assigns the discourse functions of languages. Natural Language and Linguistic Theory 1. Hale, Ken. 1989. On nonconfigurational structures. Configura- topic and focus. The discourse function topic is used tionality. The typology of asymmetries, ed. by László for foregrounding a discourse topic about something Marácz and Pieter Muysken. Dordrecht: Foris. that is predicated. Such a discourse topic may or may Kiss, Katherine E. (ed.) 1995. Discourse configurational lan- not coincide with the grammatical subject and is usu- guages, Oxford: Oxford University Press. ally placed in a particular structural position, typically Siewierska, Anna. 1988. Word order rules. London: Croom Helm. Van Valin Jr., Robert D. 2001. An introduction to syntax. clause-initial. The discourse function of focus is used Cambridge: Cambridge University Press. to highlight particular aspects of the predication about Van Valin Jr., Robert D., and Randy LaPolla. 1997. Syntax: the topic. The constituent that bears focus either structure, meaning and function. Cambridge: Cambridge appears in a particular structural position, typically University Press. clause-final, or receives special morphosyntactic cod- JAVIER MARTÍN ARISTA ing or prominence, or participates in marked syntactic See also English; Navajo and Athabaskan-Eyak constructions. Languages; Syntax Connectionism In the 1980s, the connectionist view, an alternative to passes the activation forward to other units; therefore, rule-based approaches to language, gained increasing information in neural networks is represented by pat- acceptance among cognitive scientists. According to terns of activation spread over many neurons and the this view, our knowledge of language is the result of links between them. The networks are said to represent our interaction with the universe of linguistic utter- knowledge in a distributed manner, as a single network ances that we encounter. can accommodate multiple pairs of independent The principal tool of connectionist research to input–output patterns. study the acquisition and use of linguistic knowledge Learning is a natural property of connectionist net- is the neural network. A neural network is a computer works. Learning involves a change of the connection simulation inspired by neural processes that govern weights in the network as a result of a specific training the brain. Neural networks consist of primitive nerve- regimen. Connectionist networks are applied to a wide like processing units, connected at multiple points. variety of linguistic phenomena, including recognition, The strength of these connections depends on past pronunciation, morphology, and syntax, typically with experience. Accumulated experiences shape the con- an eye toward simulating empirical effects observed in nections, using weights that represent the knowledge psychological research. The interactive nature and sen- of the network. Environmental events are registered sitivity to multiple inputs of connectionist networks are via input units as patterns of activation, where the acti- well suited to simulating the recognition of various lin- vation is expressed numerically. Thus, the vector (0, 1) guistic structures. An early interactive model by indicates activations of 0 and 1 for two input units. McClelland and Rumelhart (1986) was capable of sim- Networks vary in complexity in the number of lay- ulating the word superiority effect, the finding that let- ers of units and patterns of connectivity. A unit trans- ters are recognized better when they occur in the forms input signals into an activation function and context of words rather than in isolation. 234  CONSTITUENCY TEST Connectionist researchers have also trained net- great latitude in shaping the architecture and the pro- works to acquire spelling-to-sound correspondences cessing details of a network. Frequently, in a manner by associating orthographic input signals with corre- not apparent to others, a theorist determines all of the sponding phonological outputs. Networks have been features of a model, including the number of input trained to mimic the acquisition of morphological units, hidden units, and output units, as well as the pat- knowledge of speakers, for example, of the regular and tern of connections between units. While the networks irregular past tense of English verbs, indicating that can simulate many empirical findings, there have been children more easily learn those irregular verbs that both failures of prediction and very powerful use the stem for the past tense, as in hit. predictions, for example, in the domain of speech At least two classes of models of syntactic parsing, recognition. or sentence construction are advanced through con- In spite of these criticisms, the neural network nectionist modeling. One interprets sentences via case approach has become a part of the landscape in cogni- role assignments, while the other constructs phrase tive science and—due to its psychological plausibili- structure trees. In both approaches, multiple con- ty—will continue to be a significant player in the field. straints change as a listener progresses through a sen- tence. In the first model, each successive word in a sentence imposes constraints on the interpretation of References the current case assignment and the entire sentence. Anderson, James. 1995. An introduction to neural networks. The second approach uses a recurrent network to Cambridge and London: MIT Press. acquire and to discover the phrase tree structure of Bechtel, William, and Abrahamsen, Adele. 1994. Connectionism and the mind: an introduction to parallel sentences. processing in networks. Oxford (UK) and Cambridge Although increasingly popular for these and other (USA): Blackwell. applications (e.g. the lexicon, language impairment, Ellis, Rob, and Glyn Humphreys. 1999. Connectionist psychol- and language production), neural networks remain ogy: a text with readings. Hove (UK): Psychology Press. controversial. According to Pinker (1999), the models McClelland, Jay L., David. E. Rumelhart, and the PDP Research Group (eds.) 1986. Parallel distributed processing: suffer from an inherent design flaw, as they do not take explorations in the microstructure of cognition, Vol. 2: psy- advantage of the computational device of a variable, chological and biological models. Cambridge, MA: MIT which can stand for a class of words regardless of the Press. particulars of context. The connectionist models are Pinker, Steven. 1999. Words and rules—the ingredients of lan- limited by their contradiction between the averaging guage. New York: Basic Books. Rumelhart, David E., Jay L. McClelland, and the PDP Research mechanism and the value of information about the Group (eds.) 1986. Parallel distributed processing: explo- individual nuances so common in language. rations in the microstructure of cognition, Vol. 1: founda- Finally, although networks acquire their knowledge tions. Cambridge, MA: MIT Press. without programmer intervention, the programmer has KARL HABERLANDT Constituency Test Constituency and Phrase Structure Both NP and VP are units of S, they are constituents of S. The VP in turn can be analyzed as containing the Constituency has played an important role in grammati- verb read, the NP magazines as well as the preposi- cal theory, especially regarding the empirical evaluation tional phrase (PP) in the morning, which are all con- of different approaches to phrase structure. A constituent stituents of the VP. A formal representation in which corresponds to a unit in the syntactic structure. the major constituents of the sentence are labeled Consider the sentence The janitor read magazines in would thus be: the morning. One can account for it as representing a sentence S, containing a noun phrase (NP) the janitor (1) [S [NP The janitor] [VP read [NP magazines ] [PP and a verb phrase (VP) read magazines in the morning. in the morning]]] 235 CONSTITUENCY TEST Constituency Tests English; hence, the embedded clause in (6a) cannot be Many grammatical operations can only target strings of easily topicalized. words that form constituents. This has given rise to a number of widely recognized Constituency Tests. The (7) *[That he is on vacation this week] the janitor rationale is that a string of words that are taken to form said. a constituent has to pass one or more constituency tests. This restrictive character of topicalization yields an The most common tests are summarized below. important conclusion. The fact that a given string of words satisfies a constituency test (i.e. any test) is in Replacement or Substitution Test general sufficient to indicate that this string is a con- A constituent can be replaced by a proform (e.g. a stituent. However, this does not mean that all con- pronominal — it, he — or expressions such as there, stituency tests can equally felicitously be applied, since then, do so). The replacing expression can be inter- each type of sentence structure may be subject to cer- preted in a context so that the meaning of the replaced tain constraints (as, e.g. in topicalization). constituent is completely recovered. For instance, both the underlined subject NP ( a janitor) and the entire Coordination Test VP in (4) can be replaced by proforms (he and did so, respectively), showing that they are constituents. The coordination test shows that only constituents can be conjoined (8a), with the additional requirement that (4) [NP The janitor]j [VP read magazines in the they correspond to the same kind of phrase, as shown library]k. [ He]j [ did so]k everyday. by the ungrammaticality of (8b): On the other hand, there is no proform that can sub- (8) a. The janitor read magazines [PP in the stitute for a string of words that spans over parts of dif- morning] and [PP in the afternoon]. ferent constituents, such as [the janitor read]. b. *Sue drove [NP a car] and [PP to Ottawa]. Displacement Tests Relevance of Constituency A constituent can be realized in a position that is dif- ferent from its canonical position in a sentence, i.e. it Constituency has been a powerful test to evaluate and appears to be ‘displaced’. There are different types of select among different approaches to phrase structure displacement, depending on what additional elements and grammatical theory. For example, it empirically are added to the sentence. In Clefting, a structure like supported the adoption of X’-Theory in the 1980s, (5a) is generated with the addition of it was…that sup- which claimed that the major sentence constituents in porting dislocation of in the morning. However, mag- turn have internal structure. For instance, read maga- azines in the morning, which is not a constituent, zines in (1) can be shown in (9) to behave as a con- cannot be clefted, as (5b) shows (the asterisk shows stituent, excluding in the morning. ungrammaticality): (9) [Read magazines] was what the janitor did in (5) a. It was [PP in the morning] that the janitor the morning. read magazines. This may be taken as evidence for a more fine- b. * It was [magazines in the morning] that grained phrase structure with an additional level the janitor read. between V(erb) and VP, called V’ (hence X’-Theory): In Preposing, the material added to the dislocated (10) [VP [V’ read [NP magazines] ] [PP in the morn- constituent is usually a form of the copula be and a wh- ing]]] word (e.g. what, who, where). The bracketed (embed- ded) sentence in (6a) is a constituent, and thus it can be displaced, as the preposing structure in (6b) shows: References (6) a. The janitor said [S that he is on vacation this Bunt, Harry, and Arthur van Horck. 1996. Discontinuous con- stituency. Berlin: Mouton de Gruyter. week]. Carnie, Andrew 2002. Syntax: a generative introduction. Oxford: b. [S That he is on vacation this week] is what Blackwell. the janitor said. Chametzky, Robert. 2000. Phrase structure: from GB to mini- malism. Malden, MA: Blackwell. In Topicalization, no material is added to the dislo- Chomsky, Noam 1981. Lectures on government and binding. cated constituent. Topicalization is fairly restricted in Dordrecht: Foris. 236  CONTEXT Hawkins, John A. 1994. A performance theory of order and Postal, Paul. 1964. Constituency structure: a study of contem- constituency. Cambridge: Cambridge University Press. porary models of syntactic description. International Journal Huck, Geoffrey J., and Almerindo E. Ojeda. 1987. Syntax and of American Linguistics 30 (1, part 3). semantics, Vol. 20: discontinuous constituency. Orlando: Siewierska, Anna. (ed.) 1998. Constituent order in the lan- Academic Press. guages of Europe. Berlin: Mouton de Gruyter. Kayne, Richard. 1984. Connectedness and binary branching. ACRISIO PIRES Dordrecht: Foris. Lasnik, Howard. 1999. Minimalist analysis. Malden, MA: See also Phrase Structure; Pro-forms; Syntactic Blackwell. Category Context Contexts are texts’ ‘nontexts’ and thus relational phe- new context. Hence, texts, traveling through time, nomena. They are caused by figure/ground mechanisms space, and culture, will always occur in new contexts. where texts are focused figures and their contexts func- The when question thus leads to a where question. tion as their (back)grounds. Contexts are accordingly A simple answer to Where is context? is: internal always changing. Particular academic fields prioritize and/or external. An internal perception implies seeing particular kinds of researched objects, and accordingly contexts as general premises for interpreting utterances, particular kinds of contexts. An overall workable defini- while an external perception perceives the material tion of context consequently may not be found, since it world as constitutive for contexts. However, both views would be too general to cover the diverse and specific tend to be static. A dynamic view implies that contexts needs of subdisciplines. This tension between specific are not just around texts as pre-fixed situations, but are and general is captured in the aphorism Meaning is con- dynamically woven into utterances while uttering. text bound, but context is boundless (Culler 1981:24). Hence, there are contextual elements or hints within Nevertheless, contexts as phenomena have become most texts, which assist the interpretation of the intend- increasingly important within linguistics, and in gen- ed meaning. The merging of given (G) and new (N) in eral. Everyday configurations and common-sense texts is regulated by such hints: Once upon a time there understandings, reflected in different dictionary defi- was a boy [boy N]. The boy [boy G] had a nut [nut N]. nitions, are commonly twofold. Contexts are seen as The nut [nut G] had a little hole [little hole N], etc. textual elements embedding particular utterances, and However, the opening phrase, Once upon a time..., as circumstances surrounding particular situational only seems to imply a fairy tale. Rather, it is an exam- events. Yet, new insights into how language functions ple of just another context, namely the genre and con- challenge these definitions. First, the notion of text is text of an entry in the genre and context of an expanding, incorporating even the meanings associat- encyclopedia. The outcome is an inclusive embedded- ed with pictures, films, and bodies. Thus, an extended ness characteristic for contexts as phenomena [ency- concept of text destabilizes a seemingly safe ground clopedia [entry [example [fairy tale hint]]]]. Hence, for a definition of context. Second, contexts are seen utterances are doubly contextual, relying on existing as interactive and not just as passive ‘embedding’ phe- contexts for their production and interpretation and nomena, a position that triggers new questions. being, in their own right, events that shape new con- When is a context? To report observed situations texts for following actions. implies both decontextualizing and recontextualizing. How ‘given’ and ‘new’ are combined depends on Through decontextualization, focused events are sepa- the kinds of communication, such as discourses, regis- rated from their original environment, and hence their ters, or genres. Concrete texts and utterances focus on original contexts are destroyed. In order to reconstruct the micro level, while discourses, registers, and genres meaning for receivers of reports that are separated from are active as contexts at a macro level. They can create the original context, a text needs to be sufficiently or are presuppositions for contexts of situations. (re)contextualized. However, the report now consists of Some text theories try to specify contextual con- elements of both the ‘original’ text and elements of its stituents more systematically, as M.A.K. Halliday and 237 CONTEXT his followers (e.g. C.M.I.M. Matthiessen and Jim Context can be seen as a relational concept in tran- Martin) do in Systemic Functional Grammar. This sition and hence open to reconceptualization. To approach is contextual and sees language principally rethink something means to recontextualize it. Thus, as meaning-potential for combining text and context reconceptualization is recontextualization. and suggests that language has three interactive main functions: textual, ideational, and interpersonal. References Similarly, Jürgen Habermas (1984) sees each person’s communicative context as a ‘lifeworld’ consisting of a Bakhtin, Mikhail. 1986. Speech genres and other late essays. triad of three basic interactive elements: inner nature Austin: University of Texas, Press. Bateson, Gregory. 1972. Steps to an ecology of mind. New (self), outer nature (world), and others (society). York: Ballantine Books. Although he is not arguing that lifeworld is context, Bourdieu, Pierre. 1989. Outline of a theory of practice. his triad can be linked to Halliday’s so that the textual Cambridge: Cambridge University Press. function is respectively related to communicators, the Culler, Jonathan. 1981. The pursuit of signs: semiotics, litera- ideational to world, and the interpersonal to society, so ture, deconstruction. Ithaca: Cornell University Press. Duranti, Alessandro, and Charles Goodwin (eds.) 1992. that all elements define each other mutually. Rethinking context. Language as an interactive phenome- Following Halliday and Habermas and others on non. Cambridge (UK): Cambridge University Press. this matter, a triadic understanding will imply that Erickson, Fredric, and Jeffrey Schultz. 1981. When is a con- both language (or utterances and texts) and context text? Ethnography and language in educational settings, ed. (genres and discourses) will be triadic. They are all by Judith Green and C. Wallat. Norwood, NJ: Ablex. Fairclough, Norman. 1992. Discourse and social change. interrelated and constitute a systemic contextual life- Cambridge (UK): Polity Press. world. Consequently, contexts are always and simulta- Foucault, Michel. 1972. The archeology of knowledge. neously embodied, world-related, and societal. London: Tavistock. Embodied contexts, however, will in turn function Goodwin, Charles, and Alessandro Duranti. 1992. Rethinking as discursive resources for uttering and as such distrib- context: an introduction. Rethinking context. Language as an interactive phenomenon, ed. by Alessandro Duranti and ute power asymmetrically, even in a context that is Charles Goodwin. Cambridge (UK): Cambridge University seemingly equal for all the participants. Thus, class- Press. room episodes with students and teacher, medical con- Habermas, Jürgen. 1984. The theory of communicative action, sultations between patients and doctor, command Vol 1. Boston: Beacon. situations between soldiers and officer, will represent Halliday, M.A.K. 1978. Language as social semiotic. London: Arnold. not only different embodied communicational Halliday, M.A.K. 1994. An introduction to functional grammar, resources, but different contextual power relationships. 2nd edition. London: Arnold. If contexts are systemic and interrelated, academic Malinowski, Bronislav. 1935. Coral garden and their magic, 2 fields have to find their place in such a system, by devel- vols. London: Allen & Unwin. oping awareness of similarities and differences vis-à-vis Martin, Jim. 1997. Analysing genre: functional parameters. Genre and institutions. Social processes in the workplace other fields. Subdisciplines within linguistics are likely and school, ed. by Frances Christie and Jim Martin. London to focus on textually established contexts. The nearest and Washington: Cassell. contexts for sounds are words, the nearest contexts for Matthiessen, C.M.I.M. 1993. Register in the round: diversity words are sentences, etc. Fields such as applied linguis- in a unified theory of register analysis. Register analysis: tics and anthropological linguistics, for instance, will theory and practice, ed. by Mohsen Ghadessy. London: Pinter. need to take the larger, more general contextual aspects, SIGMUND ONGSTAD such as genres and discourses, into consideration as rel- evant for the (re)production of meaning. See also Halliday, Michael Alexander Kirkwood Conversation Analysis Conversation analysis (CA) involves the study of spo- conversation analysis, however, may be misleading, as ken interactions. In CA, naturally occurring talk is a range of spoken interactions are studied under the recorded, transcribed, and then analyzed. The term heading ‘conversation’, from casual conversation to 238  CONVERSATION ANALYSIS more formal talk such as courtroom interactions or The term ‘context’ is used to refer to the textual medical consultations. As a result, CA is increasingly environment, rather than the sociopolitical environ- referred to as the study of ‘talk-in-interaction’, ment. Talk can be both context bound, in that we although this term has also been criticized for ignoring understand utterances because of their sequential nonverbal issues such as pauses, gestures, or gaze. placement within a conversation, and context free. Traditionally associated with sociologist and others Conversationalists seek to find meaning, interpreting working in the ethnomethodological tradition, CA is utterances in relation to prior utterances or knowledge, used in different ways by different scholars. Scholars and in this sense, all utterances are context bound. from a variety of fields, such as linguists, anthropolo- Conversation is context free in that there are rules of gists, and communication specialists, also analyze conversation that are general and are not constrained spoken interactions, using a range of methods, and by changes in such things as the participants or setting: based on varying assumptions. participants in talk still take turns, repair misunder- Conversation analysis began in the 1960s with the standings, and the like. work of American sociologists Harvey Sacks, Context is also both retrospective and prospective. Emanuel Schegloff, Gail Jefferson, and others, influ- What is said in a given moment emerges in part enced by Erving Goffman’s focus on everyday inter- because of what was said in immediately prior utter- actions and by Harold Garfinkel, the founder of ances, and it also influences what will be said next. In ethnomethodology. Ethnomethodology is concerned this manner, participants mutually construct talk. CA with uncovering participants’ (as vs. researchers’) per- within the ethnomethodological tradition has a number spectives on the activities in which they are involved. of limitations. Much of the work in the field has been Ethnomethodologists seek to understand how people carried out in English. Cross-cultural studies have not produce and understand daily activities, including always concurred with its findings. CA has also been conversation. criticized for ignoring the larger influences of socio- Within this tradition, CA sees talk as social action, cultural context, and for not acknowledging the bene- and attempts to uncover the structural properties of talk fits of quantitative analyses of spoken interactions. in terms of the sequential development of talk. CA pro- Scholars from a variety of fields have undertaken vides a substantive body of evidence indicating that analyses of spoken interactions, often addressing the talk is rule-oriented. One area of investigation is turn- limitations mentioned above. Dell Hymes’s ethnogra- taking. CA revealed that speakers tend to obey a rule of phy of speaking, also known as the ethnography of one-speaker-at-a-time. Several features of talk are communication, is one such area that explicitly exam- related to this preference, including when speakers ines larger sociocultural contexts such as participant chose to begin their turns. Another finding is the adja- roles or setting. One goal of this approach is cross-cul- cency pair, a two-turn sequence of adjacent utterances tural comparisons of interactions. Work on speech acts by successive speakers. The first utterance sets up an focuses on talk as performing social actions such as expectation for the second. An example is greeting– complimenting or apologizing. Analysts in the greeting. When a speaker greets another speaker, there Gricean tradition also tend to use constructed exam- is an expectation that the next utterance will be a return ples of talk in order to determine the cooperative prin- greeting. Other areas of investigation in CA have ciples of conversation. included the concept of repair, or what happens when a Interactional sociolinguistics provides another speaker misspeaks or is not understood, laughter, and avenue into the analysis of spoken interactions, as storytelling. does variation analysis. Much of this work examines A number of assumptions underlie this methodolo- how people vary their talk as a result of changes in gy used within CA, the ideas that data should consist sociocultural contexts. of multiple instances of naturally occurring talk, not Conversation analysis has contributed to linguistics, constructed examples, that talk should be transcribed in part by making talk itself a focus of serious aca- in such a manner as to attempt to capture the sense of demic investigation. We now understand a great deal the sequential development of the talk, of how the about the structure of talk, and therefore how devia- interaction unfolded, and that no item is too small to tions from our expectations of how talk should pro- be investigated. As a result, transcriptions include such ceed lead us to interpret—rightly or wrongly—our items as overlapping speech of multiple speakers talk- speech partners’ intentions. Fail to answer a greeting ing at once, pauses, laughter, and changes in volume. with a greeting, for example, and you may well be In addition, CA tends to dismiss assumed categories of judged as rude in certain speech communities. CA is sociocultural context such as status or gender, claim- also combined with other linguistic disciplines and ing that the relevance of these categories must be approaches, such as second-language acquisition and demonstrated within the data analyzed. syntactic analysis, investigating how syntactic choices 239 CONVERSATION ANALYSIS are affected by the emerging nature of discourse. As Markee, N. 2000. Conversation analysis. Malwah, NJ: CA is incorporated into wider circles of linguistic Lawrence Erlbaum. Psathas, G. 1995. Conversation analysis: the study of talk-in- research, a more complete understanding of the vari- interaction. Thousand Oaks, CA: Sage. ous components and functions of language is possible. Sacks, H., E. Schegloff, and G. Jefferson. 1974. A simplest sys- tematics for the organization of turn-taking in conversation. Language 50. References Tannen, D. 1984. Conversational style: analyzing talk among Drew, P., and J. Heritage (eds.) 1992. Talk at work: interaction friends. Norwood, NJ: Ablex. in institutional settings. Cambridge: Cambridge University ten Have, P. 1999. Doing conversation analysis: a practical Press. guide. London: Sage. Eggins, S., and D. Slade. 1997. Analysing casual conversation. SUZANNE SCOTT London: Cassell. Hutchby, I., and R. Wooffitt. 1998. Conversation analysis: prin- See also Context; Discourse Analysis; Hymes, Dell ciples, practices, and applications. Oxford: Polity Press. H.; Sociolinguistics; Speech Acts Coordination A coordination joins two sentence elements, called Linguists are further concerned with which material conjuncts. In a coordinate structure like cats and dogs, is allowed as a conjunct in a coordinate construction. the conjunction coordinates the conjunct cats with the The second example showed conjoined sentences, but conjunct dogs. In many languages, conjunctions like coordination is also possible for noun phrases as in the and or or can conjoin words or phrases of virtually apples and the pears, verb phrases like run fast or jump every category, under the condition that the categories high and adjectival phrases such as rich and very being conjoined are of the same sort. It might appear famous, etc. Both sentences and phrases intuitively as if coordination was a relatively simple phenome- form meaningful units within a sentence, called con- non. However, coordination is notoriously difficult for stituents; however, not all sentence elements can be linguistic theory to define. constituents. Subject and verb do not form a constituent Although a wide variety of structures can be con- in some frameworks of generative grammar. However, joined, not all coordinations are acceptable. One of the they can occur together as a conjunct in the sentence first generalizations regarding coordination is Ross’s Kim bought, and Sandy sold, three paintings yesterday. Coordinate Structure Constraint (1967). This con- The possibility of this ‘nonconstituent coordination’ straint states that coordination does not allow for has led a number of linguists to relax the notion of con- asymmetrical constructions. For example, the sentence stituency. In categorial grammar, for example, subject This is the man whom Kim likes and Sandy hates Pat and verb can form a constituent. Coordination phe- is unacceptable, because only the first conjunct is rel- nomena can therefore provide a testing ground for even ativized. The sentence This is the man whom Kim likes basic theoretical notions such as constituency. and Sandy hates is acceptable, because both conjuncts Another important question concerns the way coor- are relativized. dination of phrases is interpreted, as phrasal coordina- The Coordinate Structure Constraint might be tion seems strongly related to sentential coordination. explained by the requirement that the conjuncts in a The sentence Kim ran and jumped, in which two verb coordinate construction must be of the same ‘sort’. phrases are conjoined, has the same interpretation as This requirement is sometimes referred to as the Law the coordination Kim ran and Kim jumped, in which of Coordination of Likes. Linguists are uncertain as to two sentences are conjoined. The dominant approach the relationship between ‘sort’ and syntactic category. in generative syntactic theories indicates that phrasal The sentence Pat is stupid and a liar shows that being coordination can be derived from sentential coordina- of the same syntactic category is too strong a require- tion by means of reduction rules. This approach states ment for conjuncts in a coordinate construction, since that Kim ran and jumped is the result of a reduction an adjective phrase (stupid) can be conjoined with a rule having deleted the subject of the second conjunct, noun phrase (a liar). It is therefore unclear what it Kim. Much syntactic research focuses on formulating means for two conjuncts to be of the same sort. appropriate reduction rules. This has turned out to be 240  COPTIC EGYPTIAN quite difficult due to the potential for phrasal con- exactly the same relationship to the conjunction, and structions such as Kim and Sandy are similar, which that the two conjuncts essentially introduce a third lack a sentential source. structural dimension. The work of Richard Montague in the early 1970s Although various solutions have been proposed, ushered in the method of deriving the interpretation of there is not yet a satisfactory explanation for all of the conjoined phrases directly from their surface form. problems discussed here. Coordination occurs both in Semantic explanations of phrasal coordination differ phrase structure and sentence structure, but the rela- from strictly syntactic ones. In semantic analysis, con- tionship between the two remains unclear. The role of juncts are interpreted as functions that require certain coordination in sentence construction is also largely semantic arguments to make a sentence. The semantic undetermined. Coordination therefore remains a phe- type of ran requires a subject to yield an interpretable nomenon that is difficult to explain for any formal lin- sentence. Because jumped is of the same semantic type guistic theory. as ran, ran and jumped can be conjoined according to the Law of Coordination of Likes. If this conjoined verb phrase is applied to the subject Kim, the resulting References interpretation will be that Kim ran and Kim jumped. Goodall, Grant. 1987. Parallel structures in syntax. Cambridge: Under this semantic approach, the interpretation of Cambridge University Press. phrasal coordination is related to, but not derived from, Hartmann, Katharina. 2000. Right node raising and gapping: interface conditions on prosodic deletion. Philadelphia/ the interpretation of sentential coordination. Amsterdam: John Benjamins. Further, a problem particular to syntactic analysis Johannessen, Janne Bondi. 1998. Coordination, New York: concerns how to formally represent coordination, as the Oxford University Press. conjunction structure appears to contradict current Montague, Richard. 1973. The proper treatment of quantifica- models of sentence construction. The term ‘coordina- tion in ordinary English. Approaches to natural language: Proceedings of the 1970 Stanford Workshop on Grammar tion’ implies that the conjuncts are located at the same and Semantics, ed. by Jaakko Hintikka, J.M.E. Moravcsik structural level. However, this view is incompatible and Patrick Suppes. Dordrecht: Reidel. with the assumption in generative syntax that syntactic Oirsouw, Robert R. van. 1987. The syntax of coordination. structures are formed by repeatedly unifying two ele- London: Croom Helm. ments at a time. The argument that a conjunction uni- Ross, John Robert. 1967. Constraints on variables in syntax. Ph.D. Dissertation, MIT, reprinted as Infinite syntax; fies its two conjuncts at the same level entails the Norwood, NJ: Ablex Publishing Corporation, 1986. conclusion that syntactic structures can unify three ele- Steedman, Mark. 2000. The syntactic process. Cambridge, MA: ments. Some linguists attempt to remedy this apparent MIT Press. contradiction by suggesting a subordinating structure Winter, Yoad. 1998. Flexible Boolean semantics: coordination, for coordination, in which the conjunction combines plurality and scope in natural language. Ph.D. Dissertation, Utrecht University. with one conjunct first and the resulting constituent PETRA HENDRIKS then combines with the other conjunct. Others suggest the hypothesis that both conjuncts basically stand in See also Constituency Test; Montague, Richard Coptic Egyptian In the course of its productive language history, which The Copto-Greek alphabet spans a period of over 4,000 years, Ancient Egyptian Coptic, like many other ancient languages of literature, went through several developmental stages. Its latest has been passed down to us through large corpora of stage is Coptic Egyptian (not to be confused with texts. These texts were written in a highly standardized Egyptian Arabic), which is the vernacular of late- notational system of alphabetic signs representing the antique and medieval Christian Egypt (fourth to four- different sounds of the Coptic language. The origin of teenth centuries AD). The modern term Coptic is the Coptic writing system lies in occasional Greek tran- derived from Arabic q_u bt.¯ι, itself a corruption of the scriptions of native words in Egyptian texts of the Greek word (ai)gypt(ios) ‘Egyptian’. Hellenistic and Roman periods. In the first three 241 COPTIC EGYPTIAN centuries AD, the use of such transcriptions became varieties also differ with respect to the time depth of increasingly common and entire corpora of texts with a their attestation. Prior to the Arabic conquest in AD predominantly magical character were written in a 641, Sahidic was the predominant literary dialect of Greek-derived alphabet. The Christianization of the Coptic. Its supremacy became challenged by Bohairic country in the fourth century AD constituted a turning Coptic from the ninth century onward. By the eleventh point: the abandonment of the pagan literary tradition century, Bohairic had replaced Sahidic as the official and culture manifested itself in the replacement of church language and had become the sole representa- hieroglyphic writing and Demotic, its cursive variant, tive of Coptic Egyptian, which survived as the liturgi- by Greek script. In its present form, the Copto-Greek cal language of the present-day Coptic Orthodox alphabet consists of 32 letters, 24 of which are taken Church. The language material of the following typo- from Greek and eight from Demotic writing. With the logical sketch is exclusively drawn from Sahidic exception of the monosyllabic grapheme ti, the Coptic, the main reference dialect. Demotic-based signs represent phonemes that were absent in Greek, but which are part of the native Coptic Phonology Egyptian sound system. The sound system of Sahidic Coptic consists of 20 consonants (p, t, k, d, g, kj, f, s, S, h, B, z, tS, m, n, l, r, Bilingualism and linguistic borrowing w, y, /) and eight vowels (i, e, E, u, o, O, a, ´). Its most The emergence of Coptic was the result of intensive striking property is the absence of voiced consonants language contact in a bilingual (Egyptian–Greek) (i.e. consonants produced with a vibration of the vocal speech community. Greek was not only the language cords) in the class of stops (consonants produced by of the literate elite but also the language of the Holy making a closure at the lips), fricatives (consonants Scriptures and the new religion, and hence a language where the airflow is constricted to form a turbulence of great cultural importance. The impact of this presti- but is not interrupted), and affricatives (sounds pro- gious language on the native vernacular was all-perva- duced by an initial closure that is released gradually, sive. Although no clear statistics are available at making it sound like a fricative), the main exception present, it is estimated that approximately 40% of the being the voiced fricative /B/. It is worthwhile pointing Coptic vocabulary consists of Greek loanwords. The out that the voiced stops /d/ and /g/ as well the voiced transfer of Greek lexical material into the Coptic fricative /z/ have a special status as loan phonemes that vocabulary was not only restricted to content words are by and large restricted to Greek borrowings. The (nouns, verbs, adjectives) with a clear link to glottal stop /// (a sound produced by a complete but Hellenistic and Christian culture (e.g. te-psikhE ‘the brief contraction of the vocal cords) has no alphabetic soul’, klEronomei8 ‘to inherit’, hagios ‘holy’), but also sign of its own. Yet, its presence in the Sahidic sound involved a variety of Greek function words (i.e. gram- system can be deduced from a sequence of two identi- matical words with no descriptive-lexical content), cal vowel letters that indicates ‘broken’ vowels, i.e. such as sentence conjunctions, discourse markers, and long vowels that are pronounced with a brief interrup- some types of adverbs and prepositions (e.g. hOste tion (e.g. maatSe /ma/atSe/ ‘ear’, mEESe /mE/ESe/ ‘such that’, E ‘or’, oude ‘and not’, alla ‘but’). Despite ‘crowd’). Coptic has pairs of vowels: /e/ vs. /E/ and /o/ the massive influx of Greek items, grammatically vs. /O/. The schwa /´/ (a colorless, unstressed vowel) is organized words (pronouns, articles, numbers, and the graphically expressed not by a letter, but rather a spe- like) are all drawn from the native stock. To fit into cial diacritic, the so-called supralinear stroke (a verti- Coptic phrase structure, Greek loanwords underwent cal line above a letter), e.g. n@ /´n/ ‘of’. Stress minor changes in the course of borrowing. (indicated as ") is lodged only on those syllables that either contain a long vowel (…) (e.g. noute /nu…".te/ ‘god’) or a vowel–consonant sequence (e.g. anaS Dialect variation /a.na"S/ ‘oath’). Moreover, there is at most one main Coptic Egyptian is actually a dialect cluster, consisting stress per word, regardless of its length. of at least six regional varieties, two of which gained supraregional importance: Sahidic, the language of the Noun morphology whole Nile valley above the Delta, and Bohairic, the language of the Nile Delta. Sahidic and Bohairic Unlike Pre-Coptic Egyptian, Coptic has only a few Coptic differ significantly from one another in a num- word formation processes for the makeup of nouns. ber of grammatical features (sound system, verbal Thus, there are only a few lexical items where number morphology, syntax) and the amount of lexical and (singular vs. plural) and gender (masculine vs. femi- grammatical borrowing from Greek. Both language nine) are marked on the noun itself (e.g. son ‘brother’, 242  COPTIC EGYPTIAN sOne ‘sister’, sneu8 ‘brothers, fellow monks’). In the tern’. As the terminology suggests, verbal stems are vast majority of cases, gender and number distinctions derived from relatively abstract form-meaning repre- are expressed on the definite article, which is placed in sentations (roots) by the superimposition of particular front of the noun it modifies (e.g. p-rOme ‘the (sing. consonant–vowel patterns with a basic meaning. For masc.) man’, te-shime ‘the (sing. fem.) woman’, ne- example, a verbal root like /m-s / ‘BEAR , GENERATE’ shime ‘the (pl.) women’). The Sahidic determiner sys- may surface in at least four word formation patterns tem offers a three-way contrast between a definite, an that are distinguished from one another by means of indefinite, and a zero article (e.g. p-rOme ‘the (sing. vowel change (Ablaut) and syllable structure. These masc.) man’; u-rOme ‘a man’, hen-rOme ‘men’; rOme are the absolute state mise ‘to give birth’, the nominal ‘man’). Besides the indefinite and definite articles, state mes, and the corresponding pronominal state there are demonstratives articles, so-called because mest ‘to deliver’, which combine with a nominal and they indicate closeness or distance to the location of pronominal object, respectively, and the Stative mose the speaker: pei8-rOme ‘this (sing. masc.) man’ vs. ‘that ‘to be bred’. The absolute state and the construct state man’ (lit. the man who is (over) there). Both demon- (i.e. the nominal and the pronominal states together) strative articles have corresponding demonstrative describe events that change over time, while Statives pronouns, which have a word-like status: pai8 ‘this one’ refer to states or conditions that last for some time. vs. p-et-´mmau8 ‘that one’. The absolute state and the construct states, in turn, dif- Reference to grammatical person (the speaker, the fer syntactically in the way they express the direct addressee, and a third party) is expressed by two sets object relation. In the absolute state, the direct object of pronouns: independent pronouns and bound pro- is encoded as a prepositional phrase (e.g. mise n-u- nouns. Independent pronouns have a word-like status, Se/ere n-shime ‘to give birth to a girl’ (Miracles of while bound pronouns form an inseparable unit with Apa Mêna 10b:33–4)). In the construct state, however, the sentence element they modify. Different forms direct objecthood is indicated by the juxtaposition of exist for pronominal prefixes that precede and the verb and the nominal or pronominal object (e.g. pronominal suffixes that follow their host. Table 1 mes p-Sr-howt ‘to deliver the male child’ (Apocalypse presents the pronominal paradigm of Sahidic Coptic 12:13) vs. mest-f ‘to deliver it’ (Miracles of Apa Mêna (-º indicates a null morpheme, i.e. a grammatical ele- 10b:26)). ment that is not pronounced). Independent pronouns have an emphatic meaning and place the entity or Conjugation system object they refer to at the center of attention. Pronominal prefixes and suffixes, on the other hand, Coptic has more than 20 different verb conjugations lack such a highlighting function. While independent (i.e. the pattern in which verb forms can appear) for pronouns are restricted to peripheral syntactic posi- the expression of tense (i.e. the location of events in tions (e.g. ntos de a-s-onk-s ehrai8 ‘(as for) her (ntos), time), aspect (i.e. the ongoing state, completion, or she leaped up herself’ (Eudoxia 50:17)), bound pro- multiple occurrence of events), and mood (i.e. the nouns are compatible with all nominal positions of the commitment of the speaker toward the truth of the clause (e.g. a-f-ent-s ‘he brought it’, nE-t´n ‘for you’). reported events). At the foundation of this richness of meaning distinctions is the subdivision of the four Verb morphology absolute tenses (the Present, the Habitual, the Future, and the Perfect) into two conjugation classes, tradi- The basic principle of Coptic verb formation, which it tionally known as first and second tenses. The second shares with Semitic languages, is that of ‘root-and-pat- tenses are morphologically derived from the ‘basic’ first tenses by adding the relative markers e- or ´nt- in TABLE 1 Pronominal System front of the verbal cluster (Table 2). First and second have exactly the same temporal Bound Pronouns Independent and aspectual interpretation, but differ from one anoth- Pronouns er with respect to their syntactic distribution. First Prefixes Suffixes 1st sing. ti- -i, -t anok 2nd sing. masc. k- -k ´ntok TABLE 2 Absolute Tense System 2nd sing. fem. te-, ter- -e, -º ´nto First Tenses Second Tenses 3rd sing. masc. f- -f ´ntof 3rd sing. fem. s- -s ´ntos Present f-sOt´m e-f-sOt´m ‘he listens’ 1st pl. ten- -n anon Habitual Sa-f-sOt´m e-Sa-f-sOt´m ‘he (usually) listens’ 2nd pl. tet´n- -t´n, -tEu8t´n ´ntOt´n Future f-na-sOt´m e-f-na-sOt´m ‘he is going to listen’ 3rd pl. se- -se, -su…, -u… ´ntou 8 Perfect a-f-sOt´m ´nt-a-f-sOt´m ‘he listened’ 243 COPTIC EGYPTIAN tenses appear in pragmatically neutral declarative provides prima facie evidence for their categorical sta- clauses (e.g. ei8s hE/Ete anon ne.k-h´mhal t´n-sOt´m tus as auxiliary verbs. Due to their semantic erosion, ‘Look, we, your servants, are listening!’ (Eudoxia these auxiliaries have a fully grammaticalized mean- 62:03)). The derived second tenses are used in the con- ing and function, which is typical of free functional text of relative clauses (e.g. u-hOB [ere p-nu…te moste morphemes. ´mmo-f ]‘a thing [that God hates (it)]’ (Acts of Andrew & Paul 202:126–7), constituent questions Syntax (e.g. e-tet´n-Sine ´nsa nim‘Whom are you (woman) looking for?’ (John 18:4), as well as a range of declar- The basic word order, from which other word-order ative focus contexts (e.g. awo p-woei8n e-f-r woei8n patterns are derived, is subject–verb–object (SVO). h´m p´-kake ‘and (as for) the LIGHT, it is shining in the Word-order alternations are, however, extremely DARKNESS’ (John 1:5)). common and motivated by pragmatic considerations. Apart from absolute tenses, Coptic has several syn- The topic status (i.e. the presupposed, familiar or tactically dependent verbal tenses and moods. Relative known character) of a noun or pronoun is generally tenses locate some event with respect to another event indicated by placing it into the left periphery of the and express three types of temporal relations: prece- clause, its grammatical function being resumed by a dence, subsequence, or simultaneity. In doing so, rela- pronoun with identical person, number, and gender tive tenses often indicate a logical or causal connection specification (e.g. p-aNgelos de m-p-tSoei8s a-f- between two events. In negative tenses, negative wonh-f e-p-arkhiepiskopos ‘the angel of the Lord, meaning (the equivalent of English not) and a particu- he revealed himself to the archbishop’ (Mêna, lar time value are fused together in a single indivisible 4b:6–9)). Verb–subject order has a ‘presentative’ unit. Take for instance the Negative Future ´nne-f- meaning and is used for the introduction of new dis- sOt´m ‘he shall not hear’, where the conjugation ´nne course participants. Notice that the nominal subject combines future tense with negative meaning, which that is removed from its sentence-initial position is are expressed by two separate elements (viz. shall and supplied with the particle nkji (e.g. f-nEu nkji u- not) in the English equivalent. Moreover, Coptic has aNgelos nte p-nu…te ‘(there) comes an angel of God’ two different moods for the imperative (e.g. sOt´m (Budge, Coptic Martyrdoms 214:22)). To highlight ‘Listen!’), which expresses commands, and the opta- the subject or object, a special construction type is tive, which expresses wishes (e.g. mare-f- sOt´m used: the cleft sentence. It is called cleft sentence ‘may he listen’). See Table 3 for further illustration. because it consists of two parts: a sentence-initial The clause-initial or medial position of tense- noun or pronoun and a relative clause (given in aspect-mood markers, their morphological independ- brackets) (e.g. awo m-pe.u8-kjBoi8 an pe [ ´nt-a-f- ence from the verb, as well as their agreement tutSo-u88 ] ‘and (it) is not their arm that has saved behavior (variant forms for nominal and pronominal them’(Psalm 43:4)). More research is needed to clar- subjects, e.g. (Habitual) Sare p-rome sot´m ‘the man ify the relation between pragmatic prominence and listens (usually)’ vs. Sa-f-sot´m ‘he listens (usually)’ sentence form in Coptic Egyptian. TABLE 3 Relative Tenses and Basic Moods Relative Tenses Negative Tenses Temporal ´ntere-f-sOt´m Negative Habitual me-f-sOt´m ‘after he had listened’ ‘he does not listen’ Terminative Sant´-f-sOt´m Negative Future ´nne-f-sOt´m ‘until he listens’ ‘he shall not listen’ Conjunctive n´-f-sOt´m Negative Perfect ´mpe-f-sOt´m ‘and he listens’ ‘he did not listen’ Finalis tare-f-sOt´m Unexpected Negative Perfective ´mpate-f-sOt´m ‘so that he will listen’ ‘he has not yet listened’ Moods Conditional e-f-San-sot´m Negative Conditional e-f-San-t´m-sot´m ‘if he listens’ e-f-t´m-sot´m ‘if he does not listen’ Optative mare-f-sOt´m Prohibitive mp´r-sOt´m ‘May he listen!’ ‘Do not listen!’ Imperative sOt´m ‘Listen!’ 244  CORPUS LINGUISTICS References Reintges, Chris H. 2003. Coptic Egyptian (Sahidic Dialect)—a learner’s grammar. Cologne: Rüdiger Köppe Verlag. Layton, Bentley. 2000. A Coptic grammar with Chrestomathy Reintges, Chris H. 2004. Coptic Egyptian (Sahidic Dialect)—a and Glossary. Porta Linguarum Orientalium N.S. 20. learner’s grammar. Cologne: Rüdiger Köppe Verlag. Wiesbaden: Harrassowitz Verlag. Till, Walter C. 1961. Koptische Dialektgrammatik: mit Loprieno, Antonio. 1995. Ancient Egyptian—a linguistic intro- Lesestücken und Wörterbuch. München: Verlag C.H. duction. Cambridge and New York: Cambridge University Beck. Press. Polotsky, Hans Jakob. 1960. The Coptic conjugation system. CHRIS H. REINTGES Orientalia 27. 392–422. Reintges, Chris H. 2001. Code-mixing strategies in Coptic See also Ancient Egyptian; Auxiliaries; Function Egyptian. Lingua Aegyptia 9. 193–237. Words Corpus Linguistics During the last three decades of the twentieth century, and Altenberg (eds.) 1991, Biber 1995, Leech 1991, computer technology has made it possible to conduct Sinclair 1987, 1991, Stubbs 1995, Svartvik 1990). extensive and complex research on specific linguistic As Biber et al. (1996:115) point out, ‘it is in the features — either lexical items or grammatical struc- area of language use that corpus-based techniques tures—and their systematic associations with other have had the most impact’. These studies complement linguistic and nonlinguistic features. These nonlin- descriptive language structure investigations and pre- guistic features include registers or specific varieties viously neglected aspects of English grammar (Crystal of language (e.g. religious, political, scientific) and 1979). Early studies in corpus linguistics focused on dialects, which are regional or social varieties of a lan- the occurrence of linguistic items (e.g. noun, verb, and guage. This new type of research is part of corpus lin- adjective frequencies), but the development of more guistics, which is the empirical study of language powerful techniques has enabled researchers to identi- using computer techniques and software to analyze fy and analyze complex association patterns. Corpus- large, carefully selected and compiled databases of based research has shown that these linguistic naturally occurring language (Conrad 2000). association patterns generally fall into two major cate- Corpus linguistics represents a departure from the gories: lexical associations and grammatical associa- dominant mentalist approach to linguistic research, tions. In the first case, the goal is to investigate how a which emphasizes the processes taking place in the linguistic feature is systematically associated with par- human mind (e.g. a Universal Grammar-centered ticular words. In the second case, the researcher inves- approach to second language acquisition studies). This tigates how a linguistic feature is systematically type of research is characterized by the empirical associated with grammatical features in the immediate analyses of actual patterns of language use in large and context. principled databases or corpora. Corpus linguistics uti- With respect to lexical associations, a concordancer lizes both quantitative and qualitative analytical tech- shows which words collocate with each target word in niques and relies on computers to perform complex a corpus that is representative of a specific register or analyses. (See Biber, Conard & Reppen (1998) and dialect. For example, in a corpus of working-class Kennedy (1998) for a detailed account of corpus- northern New Jersey cyber discourse, we find that the based investigations of language structure and use.) mental verb know, which appears 15 times in a small sample thread (7,000 total words), collocates primari- ly with personal pronouns I, you, they, and we. This Corpus-Based Investigations of Language Use usage reflects the informal nature of cyber exchanges, Although language structure has traditionally been which characterizes casual face-to-face interaction studied using nonempirical methods and relying on the among peers or friends. researcher’s intuitions, extensive corpus-based studies The results of empirical, large-scale corpus-based describing various aspects of language use were car- research projects (data from a 5.7-million-word sam- ried out in the 1980s and 1990s (Aarts 1991, Aijmer ple from the Longman—Lancaster Corpus) have 245 CORPUS LINGUISTICS as-a-Second Language (ESL) learners will incorporate the KWIC Display (TACT) results of corpus-based research. Conrad (2000) and know (15) Granger (2002) explain that corpus research complements innovations in grammar pedagogy by encouraging (219) That’s how it works and they know it. instructors to design and implement consciousness-raising At least now they’re ¦ activities for second and foreign language learners. (See (227) Are you a total moron? Do you know Conrad (2000) for an overview of the applications of the number of lawsuits ¦ Corpus Linguistics in grammar teaching in the twenty- (314) and hazardous 33 years ago. I know of first century, and Granger (1998a; 1998b; 2002) for a dis- no environmental| ¦ cussion of learner corpora compilation processes.) (329) against Mar...and Cap...You know the condition of the ¦ Corpus Analysis Tools and Corpora Used in (372) DPW does for this town. I don’t know Corpus Linguistics what your job is, but ¦ (372) responsibility at all because they know Since corpus linguistics utilizes large and representa- you couldn’t handle ¦ tive collections of natural texts, there are several types (403) Just Curious about how…We all know of tools that can be used to conduct research: com- that the building is a ¦ mercially available packages or concordancing pro- (418) he supervised his dog. I don’t know if grams (e.g. LEXA, MonoConc, MicroConcord, your post even ¦ TACT, WordSmith, WordCruncher) and computer pro- (419) electrician for years, you don’t know grams developed by researchers for specific types of too much about his ¦ analyses. These latter types of programs are used to (432) you wouldn’t want anyone to know? investigate complex grammatical constructions or Would you? ¦ association patterns, such as Biber etal.’s (1998) study (460) You are there aren’t you? I know it’s not involving the omission of that from that-clauses. election ¦ There are at least 31 commercially and publicly avail- (467) IMPLANT Does Maru… know where able corpora of written and spoken texts. These databas- the D.P.W. ¦ es contain millions of words and are divided primarily (473) Hard Worker #2: How do you know that into three categories: written, spoken, and historical. Caputo visits Kiss ¦ These corpora are further subdivided into American (475) “does Maru…or the mayor know where English, British English, and texts of other varieties of the DPW is?” ¦ English. Corpora also exist for languages other than (479) to above post: how do you know who English. (See Appendix in Biber et al. 1998: 282–7.) visits nortons ¦ Potential Limitations of Corpus Linguistics It is important to understand that meticulous qualitative shown that where certain words are nearly synony- analyses of single texts (e.g. historical, spoken, written, mous in isolation, careful analysis reveals that they and learner-centered) are usually undertaken before tend to be used with very different kinds of words. For embarking on corpus-based research. A careful micro- example, Biber et al. (1998, chapter 2) demonstrate analysis of linguistic features in written or spoken dis- that the word big commonly co-occurs with toe, while course helps us frame new research questions and large commonly co-occurs with number. hypotheses. Thus, corpus-based analysis should be Linguistic features are also systematically associat- seen as a complementary approach to the more tradi- ed with grammatical features in the immediate context tional approaches that have often focused on language as shown by corpus-based research. For example, one structure. The strength of corpus linguistics lies in its of the factors differentiating that-clauses and to- investigations of language use, which necessitate clauses is their lexical associations. ‘The verbs such as empirical analysis of large databases of authentic texts. suggest, conclude, guess, and argue can control a There are basically four potential limitations of cor- that-clause but not a to-clause; the verbs begin, start, pus-based research. The first one is the time-consuming and try can control a to-clause but not a that-clause’ nature of compiling a corpus and tagging parts of (Biber et al. 1998, chapter 3). speech, errors, or other features. The second limitation There are many promising applications of corpus- concerns the nature of the corpus data collected. This based research that have to be explored. For example, it is type of research often involves collecting and storing hoped that new grammar teaching materials for English- large corpora that may need to be continuously updated. 246  COURTROOM DISCOURSE The third limitation is that relying on computers forces Conrad, Susan. 2000. Will corpus linguistics revolutionize linguists to concentrate primarily on written rather than grammar teaching in the 21st century? Tesol Quarterly 34. 548–60. spoken language. Lastly, there is a potential limitation Crystal, David. 1979. Neglected grammatical factors in conver- of a philosophical nature, namely that a corpus is a finite sational English. Studies in English linguistics for Randolph sample of an infinite population. This means that Quirk, ed. by S. Greenbaum, Geoffrey Leech and Jan researchers are often extrapolating from what is found Svartvik, 153–67. London: Longman. in a corpus to what is true of the language or language Granger, Sylviane (ed.) 1998a. Learner English on Computer. New York: Addison Wesley Longman Limited. variety it is supposed to represent (Leech 1998). Thus, Granger, Sylviane 1998b. The computer learner corpus: a versa- one needs to be cautious in drawing general inferences tile new source of data for SLA research. In Granger, 3–18. from the results of corpus-based analyses. Granger, Sylviane, Joseph Hung, and Stephanie Petch-Tyson Despite its potential limitations, almost any aspect of (eds.) 2002. Computer learner corpora, second language linguistics can be studied from a use perspective. acquisition and foreign language teaching. Amsterdam, Philadelphia: John Benjamins Publishing Company. Corpus linguistics provides a variety of tools and meth- Kennedy, Graeme. 1999. An introduction to corpus linguistics. ods that make large-scale research on complex linguis- London, New York: Longman. tic phenomena an extremely productive and challenging Leech, Geoffrey. 1991. The state of the art in corpus linguistics. undertaking. English corpus linguistics, ed. by Karin Aijmer and Bengt Altenberg, 8–29. London: Longman. Leech, Geoffrey. 1998. Preface. In Granger. Sinclair, J. (ed.) 1987. Looking up. An account of the cobuild References project in lexical computing. London: Collins ELT. Aarts, Jan. 1991. Intuition-based and observation-based gram- Sinclair, John. 1991. Corpus, concordance and collocation. mars. English corpus linguistics, ed. by Karin Aijmer and Oxford: Oxford University Press. Bengt Altenberg, 44–62. London: Longman. Stubbs, Michael. 1996. Collocations and semantic profiles: on Aijmer, Karin, and Bengt Altenberg (eds.) 1991. English corpus the cause of the trouble with quantitative studies. Functions linguistics. London: Longman. of Language 2. 23–55. Biber, Douglas. 1995. Dimensions of register variation: a cross- Svartvik, Jan. 1990. The London-Lund corpus of spoken linguistic comparison. Cambridge: Cambridge University English: description and research. Lund, Sweden: Lund Press. University Press. Biber, Douglas, Susan Conrad, and Randi Reppen. 1996. TACT Database is under construction (cyber discourse of work- Corpus-based investigations of language use. Annual ing-class northern New Jersey towns). Contact Susana Review of Applied Linguistics 16. 115–36. Sotillo at Sotillos@mail.montclair.edu. Biber, Douglas, Susan Conrad, and Randi Reppen. 1998. SUSANA M. SOTILLO Corpus linguistics: investigating language structure and use. Cambridge: Cambridge University Press. See also Universal Grammar Courtroom Discourse Studies in courtroom discourse often focus on the sig- Other research has explored the questioning tactics nificance of certain kinds of interactions between spe- used by attorneys during direct and cross-examination cific linguistic features and their pragmatic or of witnesses. The use of open-ended wh-questions sociointeractional functions in the courtroom. (who, what, where, when, how, etc.) tends to elicit nar- Research carried out in North Carolina beginning in rative responses from witnesses, and therefore more the 1970s, for instance, resulted in a series of publica- information, whereas yes/no questions tend and are tions suggesting broadly that lay witnesses are persua- usually intended to limit the amount and type of infor- sive or not depending upon the extent to which they mation that can be offered. A number of studies exam- use “powerful” or “powerless” language. The validity ine such patterns, and it is usual for trial practice of the study’s methodology has been questioned, but courses in law schools to include material on such top- the publications generated by the study have clearly ics, often recommending that wh-questions be used served to advance our understanding of the role of lan- during direct examination and that yes/no questions be guage in judicial process, particularly where the ques- used during cross-examination. For a recent example tioning of lay witnesses is concerned. of such a study, see Sandra Harris’ (2001) examination 247 COURTROOM DISCOURSE of the O.J. Simpson, Louise Woodward, and exact syntactic and semantic bars to juror comprehen- Oklahoma federal building bombing cases. sion of instructions are now well documented by lin- It is usual for defense attorneys to omit mention of guists and psycholinguists. These scholars have the agent in cases of sexual assault or rape; the tactic demonstrated that instructions can be made more com- is thought to take the focus off the accused rapist. The prehensible by simplifying sentence structure and by grammar and prosody of reported speech can be used giving additional information about the meanings of to impeach or cast doubt on a witness’s testimony. abstract terms in both civil and criminal cases. Janet Cotterill examined the O.J. Simpson trial and We know a great deal about the problems faced by explored how metaphors can be a rich source of indi- lay jurors because of the extensive amount of research rect messages, noting that several were used in defense that has been completed. But lay witnesses face simi- attorney Johnny Cochran’s closing arguments. lar problems. They contend with very sophisticated In a more general sense, courtroom discourse question-and-answer tactics on direct and cross-exam- includes attention not only to adversarial questioning ination, usually with little or no preparation. And, of techniques, particularly as used in direct and cross- course, court interpreters and translators face the same examination, but also to all spoken language used dur- problems, even as they deal with the challenge of ing the trial process. In order to look meaningfully at the working with more than one language. full range of courtroom discourse, one must recognize All phases of pretrial, trial, and posttrial (appeal) that a number of speech genres are used during the trial procedure involve judicial procedures in which lay process. Anglo-American lawyers like to talk about the persons must deal with legal language and discourse need for a theory of the case. Once that theory is devel- patterns, including legal genres, with which they are oped, all portions of trial process are theoretically dedi- quite unfamiliar. Jury instructions have been examined cated to developing that theory and persuading the in great detail; research on other patterns and genres judge and/or jury that it is the correct one. In the process remains to be done. of moving through trial process, lawyers make use of It seems likely that much additional research will be opening statements, voir dire, direct examination, cross- reported over the next few years. Increasing numbers of examination, possible redirect and recross, and then linguists have been focusing on language and law issues, closing statements. They also usually participate in the and there are now at least two academic organizations wording of jury instructions, and they make and that regularly focus on the kinds of topics involved in respond to objections during testimony. For each of courtroom discourse. Since 1990, the Law and Society these steps, discourse strategies are developed. Association, an international association that meets out- Based on an examination of a single case, Gail side the United States every third year, has scheduled Stygall (1994) suggests that analysis of legal language sessions on Language and Law at its annual meetings. needs to move away from mere descriptive analysis to The International Association of Forensic Linguists an examination of how the maintenance of legal lan- (IAFL) came into existence as a result of a series of guage serves institutional power and dominance. workshops and conferences in the early 1990s; it meets More specifically, John Gibbons (2003) suggests biannually, usually outside the United States The IAFL that the very real power and dominance problems in publishes a journal, until recently titled Forensic the courtroom arise from the legal genres used in Linguistics: The International Journal of Speech, courtroom discourse and the fact that they are Language and the Law. Future issues will drop the orig- addressed to two very different audiences, lay persons, inal title and use the original subtitle only. often jurors, on the one hand, and of course judges and lawyers on the other hand. Gibbons calls this the two- audience dilemma and cites jury instructions as a par- References ticularly difficult genre, because “[j]urists persist in Coates, Linda, Janet Beavin Bavelas, and James Gibson. 1994. administering these instructions to jurors because they Anomalous language in sexual assault trial judgments. have survived on appeal to a higher court... The view Discourse & Society 5(2 April).189–206. Conley, John M., William O’Barr, and E. Allen Lind. 1978. The that jurors remember and understand any instructions power of language: presentational style in the courtroom. given them and that therefore it is more important to Duke Law Journal 78. 1375–99. focus on an instruction’s survivability on legal appeal Cotterill, Janet. 1998. ‘If it doesn’t fit, you must acquit’: than on its comprehensibility, is ill-informed” (English metaphor and the O. J. Simpson criminal trial. Forensic and Sales 1997. 383). Linguistics: The International Journal of Language and the Law 5(2). 141–58. Where jury instructions are concerned, it is an estab- Cotterill, Janet. 2002. ‘Just one more time…’ aspects of inter- lished fact that since the early 1970s, findings by many textuality in the trials of O.J. Simpson.” Language in the social scientists confirm that lay persons are frequently Legal Process, ed. by Janet Cotterill, Chapter 9, Basingstoke bewildered by the wording of jury instructions. The and New York: Palgrave. 147–61. 248  CRIOULO, GULF OF GUINEA Dumas, Bethany K. 2000a. Jury trials: lay jurors, pattern jury Harris, Sandra. 2001. Fragmented narratives and multiple instructions, and comprehension issues. Tennessee Law tellers: witness and defendant accounts in trials. Discourse Review 67(3 Spring). 701–42. Studies 3 (1 February). 53–74. Dumas, Bethany K. 2000b. U S pattern jury instructions: prob- Matoesian, Greg. 1999. Intertextuality, affect, and ideology in lems and proposals. Forensic Linguistics: The International legal discourse. Text 19(1). 73–109. Journal of Language and the Law 7(1). 76–98. O’Barr, William M. 1982. Linguistic evidence: language, English, P. W., and B.D. Sales. 1997. A ceiling or consistency power and strategy in the courtroom. New York: Academic effect for the comprehension of jury instructions. Press. Psychology, Public Policy, and Law 3. 381–401. Stygall, Gail. 1994. Trial language: differential discourse pro- Gibbons, John. 2003. Forensic linguistics: an introduction to cessing and discursive formation. Amsterdam and language in the justice system. Malden, MA and Oxford: Philadelphia: John Benjamins. Blackwell. BETHANY K. DUMAS Crioulo, Gulf of Guinea Gulf of Guinea Crioulo (GGC) is a cover term for four how many speakers use Sãotomense or Príncipense. creoles lexically based on Portuguese. They are spo- According to a mid-1990s source, virtually everybody ken on the three islands of São Tomé (around 100,000 on São Tomé was capable of speaking Portuguese, inhabitants), Príncipe, and Annobón (about 5,000 while only half of the children were competent in each), of which the former two constitute an inde- Sãotomense, indicating language shift. For reasons yet pendent republic while the latter belongs to Equatorial to be fully understood, this process started earlier and Guinea. has proceeded further on the smaller Príncipe, and in In the following, the languages will be referred to the 1987 census, less than 16% of the population were as Sãotomense, Principense, Annobonese, and reported to speak the language, including virtually no Angolar, although they are also known by other children. names. Sãotomense and Angolar are both spoken on The four GGCs share a number of features on all São Tomé, the latter by descendants of runaway plan- levels, suggesting a common origin. These include, tation slaves in the south of the island. among others things, a circumverbal sentence negation Previously uninhabited, toward the end of the fif- that can be reconstructed as /na … fa/, the second part teenth century the islands were settled by Portuguese of which appears clause-finally. This and other simi- and their African slaves, the latter of which were larities can presumably be traced back to São Tomé, brought partly from what is today the coast of Nigeria, which was the first of the three islands to be settled. and partly from Bantu-speaking areas around the Subsequently, however, the languages have drifted mouth of the Congo. It is in this early period that GGC apart to a degree where mutual comprehension is dif- is assumed to have emerged. ficult, albeit possible. GGC varieties are interesting for a number of rea- Like other creoles, GGCs are heavily analytic, and sons, not least because they probably are among the a participial suffix /-du ~ -ɾu/ seems to be the only oldest creoles known. They are also typologically attested bound morpheme. unusually distant from their lexifier. The Tense Mode Aspect system is mostly preverbal, All four creoles coexist in a diglossic relationship and typically includes a past marker tava (< P estava), with the official languages of the two countries an imperfective element ka (< P ficar, cá or capaz?), (Spanish on Annobón, Portuguese in the case of the indicating habitual aspect in isolation, and progressive three others) and are used neither in media nor educa- when preceded by the copula sa), and completive par- tion. ticles za and kaba (< P já and acabar respectively). Although influenced by Spanish, GGC does not Features that set GGC apart from Portuguese, but seem to be seriously threatened on Annobón, where it make them similar to other creoles in the Atlantic area, is the universal vernacular (although all speakers are include reflexive constructions involving the word for proficient in Spanish). Angolar is believed to have ‘body’, the use of 3pl as a nominal pluralizer, and use about 9,000 speakers, but is giving way to both of verb serialization, for all of which African influence Sãotomense and Portuguese. It is not entirely clear may plausibly be invoked. 249 CRIOULO, GULF OF GUINEA Although being SVO, the GGC languages display better represented in Sãotomense than elsewhere. several rather exotic orderings from a European per- Noteworthy African contributions include 3pl pro- spective, such as N NUM ɔmã dɔs ‘two hands’). nouns, and also a 2pl form in Principense. Principense includes a coarticulated stop //, cor- Not unexpectedly, the Annobonese lexicon nowa- responding to /kw/ elsewhere. It is not clear whether days contains a fair proportion of Spanish loans. Due nasal+stop sequences should be regarded as cluster or to labor migration to Fernando Poo, where pidgin/cre- as single phonemes. In all varieties except ole English is the lingua franca, Annobonese has also Principense, Portuguese /v/ turns up as // in some assimilated some lexical items from this language. items. Principense and Angolar have a rhotic Lexically, Angolar is the odd man out in the GGC phoneme, but otherwise Portuguese rhotics appear as group. While being structurally very similar to laterals or are deleted. Sãotomense, it contains an unusually high proportion All have depalatalized etymological /ʃ, / into /s, of African lexicon, largely derived from Kimbundu. z/, but also sport a productive palatalization rule whereby /t, d, s, z/ are realized as [tʃ, d, ʃ, ] before high front vocoids (exceptions: Principense lacks References [d], and Angolar has /θ, ð/ in positions where its sis- There are five book-length treatments of GGC, listed below. In ters have /s, z/). addition to these, some additional data can be found in scat- Another segmental difference within the group is tered articles by Germán de Granda (mostly Annobonese), Marike Post (Annobonese), Tjerk Hagemeijer (Sãotomense) that in Annobonese the velar plosive /k/ is realized as and Luís Ivens Ferraz, and in Valkhoff (1966). [x] before back vowels (including /a/). Barrena, Natalio. 1957. Gramática annobonensa. Madrid: The GCCs have been argued to be either tonal or Consejo Superior de Investigaciones Científicas. pitch-accent languages. All varieties have a few words Ferraz, Luís Ivens. 1979. The creole of São Tomé. that are etymologically consonant-initial, but that have Johannesburg: Witwatersrand University Press. Günther, Wilfried. 1973. Das portugiesische kreolisch der Ihla been equipped with a prothetic vowel, possibly deriv- do Príncipe. Marburg: Marburger Studien zur Afrika- und ing from the Portuguese definite article. These items Asienkunde. all belong to the core lexicon, and are, interestingly, Lorenzino, Gerardo. 1999. The Angolar Creole Portuguese of far more common in Principense (e.g. ufógo ‘fire’) São Tomé. Munich: Lincom Europa. than elsewhere (fógo). Maurer, Philippe. 1995. L’Angolar: Un créole afro-portugais parlé à São Tomé. Hamburg: Helmut Buske Verlag. With the exception of Angolar (see below), the Valkhoff, Maurius. 1966. Studies in Portuguese and African lexical contribution is limited to a couple of Creole.Johannesburg: Witwatersrand University Press. hundred items at the most. A third of these are derived MIKAEL PARKVALL from languages of present-day Nigeria, most of the remainder being of Bantu origin. The latter group is See also Crioulo, Upper Guinea Crioulo, Upper Guinea Upper Guinea Crioulo is a continuum of Portuguese- Despite not having official status in either of the vocabulary creoles spoken mainly on the Cape Verde three countries (although since 1998 the ‘national lan- Islands and in Guinea-Bissau. A variety of Guinea- guage’ of Cape Verde), Portuguese Creole nevertheless Bissau Creole also spills over into Casamance, the has a strong position in Cape Verde by virtue of it southernmost province of Senegal. The total number being the native language of the entire population, and of native speakers is estimated at 400,000 in Cape in Guinea-Bissau through being the only nationwide Verde, 250,000 in Guinea-Bissau, and 40,000 in lingua franca. Senegal. In addition, there are half a million second- Judgments on how similar Cape Verde and Guinea- language speakers in Guinea-Bissau and some 20,000 Bissau Crioulo really are differ. While some Cape in Casamance. There is also an important diaspora of Verdeans claim to be unable to understand Guinea- Cape Verdeans in both Europe and the United States. Bissau Crioulo, others treat all Upper Guinea 250  CRIOULO, UPPER GUINEA Portuguese Creoles as dialects of one language. In any African items have been attested in each variety, few case, the differences lie more in pronunciation and of which can be considered part of the core vocabu- vocabulary than in structure. Also, Cape Verde Crioulo lary. In addition to this, there are a number of idioms is in itself rather heterogeneous, and the varieties of the of seemingly African origin. northern Sotavento islands are in most respects inter- Conspicuous phonological features include the mediate between those of the southerly Barlavento replacement of /z/ by /s/ in some varieties, and the islands and Guinea-Bissau Portuguese Creole. These realization of Portuguese // as //. Furthermore, three main varieties thus constitute a continuum. Portuguese <x> generally turns out as /s/ in Upper Some authors insist that Cape Verde and Guinea- Guinea Portuguese Creoles, while instances of ortho- Bissau Portuguese Creole emerged independently, graphic <ch> (/tʃ  / in older Portuguese) become /c/ or despite the great similarities. Most, however, seem to  /. Portuguese syllable structures have, especially in /tʃ agree that they are somehow genetically related. Even the mainland varieties, drifted toward a Cape Verde so, there are two possibilities—did the ancestral structure. Creole travel from the mainland to the islands or vice Affixes are scarce, but less so than in most creoles. versa? Both hypotheses have been suggested, without In Guinea-Bissau, verbs are morphologically marked any conclusive evidence being presented for either. for causative and passive, in addition to which there Regardless of the direction of influences during the are about eight or nine other productive affixes. birth of the Upper Guinea Portuguese Creoles, it is Comparatives are basically expressed via independent clear that Cape Verde Crioulo has had some impact on particles across the board, mostly using /mas ADJ (di) the mainland varieties from at least the eighteenth cen- ke/ ‘more ... than’. tury. First, the Portuguese often used Cape Verdean Tense, mood, and aspect are for the most part employees in Guinea-Bissau. Second, famine led to marked by means of free adverbal markers, including migration toward the mainland in the late 1800s, and a progressive seemingly derived from estar and a past third, islanders were prominent in the independence from jà. In addition, there is a suffixed -ba, probably movement. derived from the Portuguese imperfect -va. Anyone accustomed to the better-known Caribbean As opposed to Portuguese, sentences in the Upper creoles will notice the—as yet unexplained—absence Guinea Portuguese Creoles may not have a subject that of many of their characteristic features in the Upper is merely implied, although overt subject pronouns Guinea Portuguese Creoles. frequently attach to the following verb. Broadly speaking, the Sotavento varieties (and even Number is normally not marked on nouns, although more so Guinea Bissau Portuguese Creole) are typo- usually on determiners. Portuguese definite articles logically more distant from Portuguese, and present have also been lost, with demonstratives taking over more features perceived as typical of creole languages. some of their functions. The sentence negator is This is manifested e.g. in the use of (mostly) inde- normally not a reflex of Portuguese não ‘not’, but pendent words to mark tense. Although usually rather /ka/, probably from Portuguese nunca ‘never’. referred to as a creole, it may be that Barlavento Cape In many respects, it is striking, however, how much Verde Crioulo should more properly be designated as like Portuguese Creoles are in their syntax. Cape a semi-creole. Verde Crioulo replicates the Ibero-Romance distinc- Although there is little evidence of this, many an tion between the copulas ser and estar, relatively faith- observer has suggested that the Cape Verde dialects fully and many varieties also display (variable) have in general been significantly more creole-like, adjectival agreement, as in /kaza bnita(a)/ ‘beautiful and that decreolization has brought them closer to house’. Passives are also constructed very much like in Portuguese with time. Today, Many Cape Verdeans are Portuguese. Within the Cape Verde dialects, creole- competent in Portuguese, which remains the official like structures vary with more Romance-like construc- language. An obvious difference between the Cape tions, as in /kes fiu di ʃefi/ vs. /keʃ fi cεf/ ‘the chief’s Verde dialects and the mainland is that the latter creole sons’. varieties are still in close contact with African lan- Relatively few structural features can reliably be guages. At least some change in the direction of related to African languages. The distinction in Cape Portuguese has taken place during the twentieth centu- Verde Criolo between the two habeo verbs /te˜)/ ‘to ry, and is reflected in variation between e.g. more have, to own’ vs. /teni/ ‘to have (at hand)’ has been Portuguese-like Guinea-Bissau /adivia/ vs. basilectal suggested to mirror Wolof, one of the African lan- /dibia/ ‘to guess’ (<adivinhar) /audar/ vs. /uda/ ‘to guages with the greatest influence on Portuguese help’ (<ajudar). Creoles. Other possible African features include the Lexically, all Upper Guinea Portuguese Creoles are replacement of etymological /v/ by /b/ in some vari- overwhelmingly Portuguese, and only about 150 eties, and the use of the word ‘head’ for ‘self’. 251 CRIOULO, UPPER GUINEA References Rougé, Jean-Louis. 1988. Petit dictionnaire étymologique du Kriol de Guinée-Bissau et Casamance. Bissau: Instituto Cardoso, Eduardo Augusto. 1989. O crioulo da ilha de São Nacional de Estudos e Pesquisa. Nicolau de Cabo Verde. Praia: Ministério da Educação. Scantamburlo, Luigi. 1981. Gramática e dicionário da língua Couto, Hildo Honório do. 1994. O crioulo português da Guiné- criol da Guiné-Bissau (GCr). Bologna: Editrice Missionaria Bissau. Hamburg: Helmut Buske Verlag. Italiana. Kihm, Alain. 1994. Kriyol syntax: the Portuguese-based Creole Silva, Baltasar Lopes da. 1957. O dialecto crioulo de Cabo language of Guinea-Bissau. Amsterdam and Philadelphia: Verde. Lisbon: Imprensa Nacional. Benjamins. Wilson, W. A. A. 1982. The Crioulo of Guiné. Johannesburg: Pinto-Bull, Benjamim. 1989. O Crioulo da Guiné-Bissau: Witwatersrand University Press. Filosofia e Sabedoria. Lisbon: Ministerio da Educação. Quint-Abrial, Nicolas. 1998. Dicionário Caboverdiano- MIKAEL PARKVALL Português. Variante de Santiago, CD-ROM edition. Lisbon: Coputação e Linguagem & Priberam Informática. See also Crioulo, Gulf of Guinea 252  D Dakota and Siouan Languages The Siouan languages are referred to in more recent Siouan-Catawban languages. The distinction among literature as the Siouan-Catawban family, as the older these three languages was based on what is called the term often really meant only the Western Siouan lan- d/l/n correspondences; where Dakota has d, Lakota guages, which consists of most of the languages, has an l, and Nakota has an n. However, further study whereas Catawban and Woccon comprise the Eastern has shown that this distinction is inadequate and not Siouan group. The Siouan-Catawban languages, of necessarily accurate for these languages. Furthermore, which the Dakota languages are one branch, were spo- Assiniboine and Stoney have been shown to be related ken, at the time of European contact, as far north as to the Dakota languages despite the relationship not Alberta and Saskatchewan, stretching southward having been recognized initially as these two tribes through the middle of the United States into had already split from the Dakotas prior to contact in Mississippi and Arkansas. Those languages still spo- 1640. ken today are found mainly on reservations in North The Siouan-Catawban languages consist of four and South Dakota, Nebraska, Montana, Minnesota, main branches: a Mississippi Valley group, also called Manitoba, Saskatchewan, and Alberta. There are 20 Siouan Proper or Mississippi River; a Missouri Valley known languages in all. According to J.W. Powell group; an Ohio Valley group or Southeastern group; (1891), in 1890, there were approximately 43,400 offi- and an eastern group, Catawban. The only language in cially recognized Siouan tribe members. According to dispute over its placement is Mandan, which is listed 2000 US Census Bureau statistics, 153,360 people in the Ethnologue as part of the Mississippi group, but identified themselves as Siouan tribe members, and in others believe that it may stand alone as a subgroup Canada, 3,880 people reported membership in a since similarities to either Mississippi group languages Siouan/Dakota tribe in 2001. or Missouri Valley languages are probably due to bor- The term ‘Siouan’ seems to have originated from rowing. the shortened colonial French, Nadouessioux, an inter- pretation of an Ojibwa term meaning ‘a small rattle Mississippi Valley Languages snake’ or ‘the snake-like ones’. In either case, it was a term for a traditional Ojibwa enemy. The Mississippi Valley languages are divided into four The designation ‘Sioux’ at the time of first contact subcategories: the Dakota languages, the Dhegiha lan- by Europeans referred to a rather loose confederation guages, the Chiwere languages, and Hochunk. The of seven tribes—the Sioux nation. The people of these Dakotan languages were once thought to consist of tribes called themselves the Dakota, the Lakota, or the three or four languages, Lakota (also called Teton), Nakota, which meant ‘allies’. These three groups are Dakota (also called Santee or Santee-Sisseton), and usually referred to as ‘Dakota’ and their dialects make Nakota (also called Yankton or Yanktonai), but recent up the Dakota languages, which in turn, are part of the work suggests that there are five dialects, including 253 DAKOTA AND SIOUAN LANGUAGES Assiniboine and Stoney. Total speakers in both the in the late 1700s, so that their dialect, which is United States and Canada number about 19,000. believed to have been another language of the Dakota (also Dakhota) has approximately 4,000 to Mississippi Valley group, is no longer distinct. The last 5,000 speakers, including children and about 30 fluent speaker of Oto died in 1996 and only half a monolinguals. There are 250 speakers of Nakota and dozen fluent speakers of Iowa remain. about 6,000 to 7,000 speakers of Lakota (also The last language of the Mississippi Valley group Lakhota) out of a population of 20,000 according to was formally known as Winnebago, now Hochunk, the 1990 US Census. Assiniboine (also called Hohe) ‘people of the sacred language’. Winnebago was the has about 150 to 300 fluent speakers in Canada and name used by the Algonquin, which meant ‘foreign- Montana out of a population of 3,500 tribal members. ers’ and adopted by the Europeans. The Winnebago Assiniboine shares nearly 90% lexical similarity to changed their name officially to the Hochunk Nation Lakota and Stoney, and 90% to 94% similarity to due, in part, to the association of the name to the RV Dakota. Stoney is spoken by 1,000 to 1,500 people out brand as well as a desire to name themselves. There of a population of 3,200 in Canada. Although Stoney are about 250 speakers left in a population of about shares approximately 90% lexical similarity with 6,000, but the tribe has undertaken measures to reverse Assiniboine, the two languages are not immediately their language loss, including building a Language and mutually intelligible. Culture Center in Wisconsin with a mission to teach Most of the Dakota tribes have undertaken some Hochunk to all tribe members, especially the children, measures to preserve their languages or increase the from kindergarten through college. number of speakers. Some Lakota communities are making an effort to teach their children the language The Missouri Valley Languages since they estimate that by 2013, less than 10% will be able to speak the language if they cannot find a way to The Missouri Valley group includes two closely relat- stop the language loss. The tribe is working with ed languages, Crow and Hidatsa. Crow is a growing Indiana University to establish language revitalization language through efforts of the tribe, including estab- projects. The Stoney tribe has also initiated a lishing their own two-year college where approxi- Language Project, and Dakota and Lakota are taught mately 75% of students speak Crow as their first in schools in North and South Dakota, Montana, language. While classes are not conducted in Crow, Nebraska, Oregon, and Minnesota. the business functions are. In total, there are about There are four Dhegiha languages: Quapaw, Osage, 4,500 Crow speakers. Hidatsa, on the other hand, is Kansa, and Omaha-Ponca. Quapaw (also called spoken by just over 100 speakers, all elderly, who live Arkansas) became extinct sometime in the 1970s or in western North Dakota. There are no efforts to revi- 1980s although there are still nearly 2,000 tribe mem- talize this language; hence, it will probably become bers. Kansa is also nearly extinct, with just a couple of extinct in the next generation. elderly speakers left. Both languages were last spoken in northern Oklahoma. Omaha-Ponca is considered The Ohio Valley Languages one language that has barely perceptible differences to nonspeakers. Estimates of numbers of speakers are The Ohio Valley Languages include the extinct lan- consistent for Ponca at about 25 fluent speakers, all guages Tutelo, Ofo, and Biloxi. Ofo and Biloxi were last over 65, but varies widely for Omaha from 20 or so spoken during the first half of the twentieth century and fluent speakers over 60 years old to 1,500 speakers. the last speaker of Tutelo died in the latter half of the Most sources agree with an estimate of a couple of century. Tutelo is also sometimes referred to as Saponi, dozen speakers. Only a handful of speakers remain for but only a list of Saponi words has survived. They were the last language, Osage. The 2000 US Census asserts probably dialects as it has been reported that Saponi and that over 15,000 people claim membership in the Tutelo speakers could understand one another. Osage tribe. Speakers and linguists working on the Dhegiha languages have preserved these languages in Catawba and Mandan grammars, dictionaries, and/or texts and a few classes are taught in some of the languages, but it is doubtful The remaining languages of the Siouan-Catawban that any of these languages will be able to revitalize. family are the most distant. The Catawban group is The third branch of the Mississippi Valley group, usually not listed as a group but a single language, Chiwere, is sometimes listed as two languages, Iowa Catawba. However, it is believed that there once were and Oto, or as one, Iowa-Oto. The dialects are nearly a number of languages that fit into this branch. indistinguishable and are divided on tribal lines rather Adequate materials exist on Catawba, but only a than linguistic criteria. The Missouria joined the Oto vocabulary list of 143 words in Woccon, a language 254  DEEP STRUCTURE that seems to be related by both morphological and These pronominal prefixes occur in two sets: one for lexical properties to Catawba, but these properties are semantic agents (the entity performing an action) and not found in other Siouan groupings. Catawba became the other for patients (the entity affected by an action). extinct when the last speaker died in 1996. Particles on the verb may encode a host of other mean- The last language, Mandan, has been thought to ings, including plurality and negation. Some lan- stand alone as a branch of the Siouan-Catawban lan- guages have such particles for marking word forms guages while others feel it belongs with the that are exclusively used by males or females in com- Mississippi Valley group. The resemblances to Crow mands, assertions, or in giving permission. In these and Hidatsa are probably due to borrowing, while languages, it is also common for nouns to become an properties linking it to Mississippi Valley languages integral part of the verb. were inherited. Mandan is nearly extinct with only six speakers left, all of whom are also Hidatsa speakers. References Boas, Franz. 1991. Introduction to handbook of American Indian Siouan-Catawban Features languages (1911). Lincoln: University of Nebraska Press. Chafe, Wallace L. 1976. The Caddoan, Iroquoian, and Siouan The Siouan-Catawban languages are generally charac- languages. The Hague: Mouton. terized by Subject–Object–Verb word order. The sound Mithun, Marianne. 1999. The languages of native North America. Cambridge: Cambridge University Press. system of the languages is generally typified by the fact Powell, J.W. 1991. Indian linguistic families of America North that stops ([p, t, k]) have several distinct manners of of Mexico (1891). Lincoln: University of Nebraska Press. articulation—unaspirated, aspirated, as well as ejec- Rankin, R.L. 1994. Siouan languages. The encyclopedia of lan- tive. The languages have a full inventory of voiceless guages and linguistics, ed. by R.E. Asher. Oxford: Pergamon fricatives, two nasals, m and n, and the glides w and y. Press. Rankin, Robert, John Boyle, Randolph Graczyk, and John Most of the languages have three nasal vowels plus five Koontz. 2002. A synchronic and diachronic perspective on oral vowels, and the vowel length is distinctive. ‘word’ in Siouan. Word: a cross-linguistic typology, ed. by Morphologically, the Siouan-Catawban languages R.M.W. Dixon and Alexandra Y. Aikhenvald. Cambridge: make extensive and complex use of affixes. Verbs are Cambridge University Press. the most highly inflected category, taking locative, Rood, D.S. 1979. Siouan. The languages of native America: his- torical and comparative assessment, ed. by L. Campbell and instrumental, reflexive, and pronominal affixes. Case M. Mithun. Austin and London: University of Texas Press. marking is also a feature. Pronominal prefixes are used Swanton, J.R. 1943. Siouan tribes and the Ohio valley. to indicate first, second, and inclusive persons, while American Anthropologist 45. the third person is generally left unmarked or implied. LYNN ANN BURLEY Deep Structure The distinction of the deep (‘underlying’) and the ‘sur- (Jim was seen by Mary). The distinction of deep and face’ structure of a sentence reflects the fact that the surface structure, already discussed by H.B. Curry actual structure of the sentence, in which the verb can (1961), was systematically elaborated in the 1960s, in be understood as a predicate accompanied by one or connection with constituency-based grammars (espe- more arguments (and adjuncts, adverbials), is cially with the transformational description of lan- expressed more or less regularly by the means of end- guage, initiated by N. Chomsky). Later, this distinction ings, function words (especially prepositions and con- became characteristic of some of the linguistic junctions), and word order. Thus, e.g. the English approaches based on dependency syntax. We present prepositions in, into, and from carry the functions of here first a brief survey of the position of this level in Locative, Goal, and Origin, respectively; the surface different stages of theories based on constituent struc- subject (preceding the verb) expresses the deep sub- ture (and transformations), and then we outline ject (Actor) with a verb in the active voice (Jim saw the present-day role of deep structure in descriptions Mary) and the deep object with a verb in the passive using dependency syntax, with illustrations taken from 255 DEEP STRUCTURE the Praguian approach of Functional Generative prefers Description. / \ Jim girls Chomsky’s (1965) Standard Theory of linguistic / \ description was later extended to contain not only a friend have division of the sentence into constituents (such as the / / \ verb and its arguments) but also the opposition of ini- my who hair / tial Phrase Markers (as the representations generated dark by a context-free phrase structure grammar) and enriched D-structures. The transformational compo- Figure 1. A simplified scheme of a dependency tree for sen- tence (3). nent, which took the D-structures as its input, then consisted of the single, general rule ‘move alpha’, and transferred these structures into S-structures. The word structure in the minimalist framework is Logical order in an S-structure may differ from the underlying Form, which reflects the semantically relevant pat- order due to the operation of the rule ‘move alpha’, terning of the sentence. i.e. to some of the constituents moving (usually to The concept of deep structure (partly even with the the left). Traces of the source positions of such use of Curry’s term ‘tectogrammatics’) was further movements are indicated in the S-structures by the elaborated in the context of Dependency Grammar, symbol ‘t’. which analyzes the composition of the sentence from Let us illustrate the framework of the Extended word forms, and shows that the sentence is patterned Standard Theory by the following examples: in a highly perspicuous way. The verb can be under- (1) (a) You told me that Bill saw who? stood as the center (‘head’) of the sentence, on which (b) Who did you tell me that Bill saw t? its arguments (such as subject, direct and indirect (2) The police know who the FBI discovered that object, etc.), adjuncts (adverbials of place, time, man- Bill shot t. ner, cause, etc.) depend, since every verb determines a set of such (obligatory and optional) complements. While in (1)(a) the wh-pronoun occupies its pri- Dependency Grammar emerged in European linguis- mary position as the direct object of the verb form tics in the 1830s (with K. Becker) and penetrated into saw, in (1)(b) this pronoun has undergone two move- school teaching in several countries of the Continent. ments determined by the cyclic application of ‘move It was systematically elaborated in the 1930s by L. alpha’. In both cases, it moves to the beginning of the Tesnière from a viewpoint closely related to the given clause. Also in (2), the pronoun who has moved Prague School of functional and structural linguistics. in a similar way, although here it has not been placed While some of the dependency-based approaches at the initial position of the whole sentence. render just a kind of ‘surface structure’ of the sentence Substantial contributions to the discussion on the (often found useful in computational linguistics), other levels of the system of language can be seen in C. such theories work with deep structure; cf. especially Fillmore’s account of ‘deep cases’ (such as Agentive, the frameworks of I. Mel’chuk, J.D. Apresjan, and the Experiencer, Theme, Instrument), as well as in the Praguian Functional Generative Description (FGD, see controversy between N. Chomsky’s ‘interpretative Sgall et al. 1986; HajiJová et al. 1998). The sentence semantics’ and the ‘generative semantics’ of G. is then described as consisting of pairs of lexical heads Lakoff, J. McCawley, and others, or in J. Bresnan’s and their dependents (modifiers, complements). A typ- distinction between constituent structure and function- ical sentence can be patterned as a dependency tree (a al structure. finite graph in which every pair of nodes is connected In the recent approach, formulated by N. Chomsky by a single sequence of edges, and a single node is in the 1990s and called Minimalism, the derivation of determined as the root of the tree); every edge con- a sentence is characterized as starting by the choice of nects a head with one of its dependents, cf. Figure 1. lexical units, which may then undergo the effects of the ‘merge and move’ rules (combining words in (3) My friend Jim prefers girls who have dark hair. accordance with restrictions defined in the lexicon, In FGD the tree can be written in a linearized see Abraham et al. 1996). To a certain degree, the form, with pairs of parentheses used to embrace position occupied in the preceding versions of every dependent: Chomskyan linguistic description by the deep struc- (3') (((my) friend) Jim) prefers (girls ((who) have ture can be compared to the ‘spellout’ point of the der- ((dark) hair))) ivation; at this point, the representation of the sentence reaches the form from which its phonologi- The underlying syntactic functions are rendered by cal form is derived. Another counterpart of deep labels of the edges of the tree, or by indices at the 256  DEEP STRUCTURE parentheses (at those oriented towards the head), and more-dimensional networks can be handled by limited the values of grammatical categories (tense, aspect, means, similar to those of the propositional calculus. number, etc.) are represented by indices of the labels This may be useful for specifying how the core of lan- of nodes. Example (4), whose deep-structure represen- guage can be acquired by children. tation (4') displays a small subset of the values, illus- A formal definition of deep structure has been for- trates the handling of the relations of coordination and mulated by PetkeviJ (1995) both in the form of a gen- apposition in FGD: erative procedure and in that of a declarative definition. Other components of a complex description (4) Mary and Jim, her husband, who live in Boston, can handle the relationships between underlying syn- have two sons. tax, morphemics, and phonology, including the more (4') (((Mary Jim)Conjunction ((she)Appurtenance hus- or less regular expression of grammatical values by band.Sing)Apposition (Descriptive (who.Plur)Actor morphemes, function words, and specific word order live.Present.Indicative (Locative Boston))Actor positions, as well as the possibility of a zero expres- have.Present.Indicative (Objective (two)Restrictive sion and other specific phenomena. son.Plur) The importance of the interactive communicative This account reflects the fact that the deep structure conditioning of language makes it necessary to ana- requires a descriptive framework of more dimensions lyze the sentence with due regard to its position in the than the two that characterize a tree; cf. the groups context, paying attention to its topic-focus articula- connected by and (conjunction) and by the first tion. This articulation, which includes the specific syn- comma (apposition) in (4). Such a view differs from tactic positions of focus-sensitive operators such as that of I. Mel’chuk, who understands coordination as a negation or only, even, also (see HajiJová et al. 1998), specific case of dependency. reflects the degree to which the structure of natural The syntactic properties of a lexical unit are han- language has been influenced by the ‘given-new’ strat- dled on the basis of its possible dependents, such as egy (starting a prototypical utterance by referring to Actor (deep subject), Addressee (deep indirect object), items activated in the hearer’s memory). Objective (deep direct object), Means, Locative. With nouns, possible dependents are the Restrictive or Nonrestrictive (Descriptive) adjunct, Appurtenance, References etc. These kinds of dependencies are expressed by Abraham, Werner, Samuel David Epstein, Höskuldur function words, endings, or word order. The syntactic Thráinsson, and C. Jan-Wouter Zwart (eds.) 1996. Minimal valency of a word (its ‘valency frame’) comprises its ideas. Syntactic studies in the minimalist framework, Vol. obligatory and optional dependents, their properties, 12. Amsterdam/Philadelphia: John Benjamins. restrictions on their combinations, on their movement, Chomsky, Noam. 1965. Aspects of the theory of syntax. Cambridge, MA: MIT Press. etc. Lexical entries thus contain much of the grammat- Chomsky, Noam. 1975. Reflections on language. New York: ical information. Pantheon. The distinction between a head and a dependent can Chomsky, Noam. 1995. The minimalist program. Cambridge, be specified on the basis of an operational criterion: MA: MIT Press. the dependent is syntactically omissible, if not in a lex- Curry, Haskel B. 1961. Some logical aspects of language struc- ture. The structure of language and its mathematical aspects, ically specified pair of words, then at the level of word ed. by R. Jakobson, 56–68. Providence, RI. classes: e.g. in ((very) slow) progress, the syntactic HajiJová, Eva, Barbara H. Partee, and Petr Sgall. 1998. Topic- potential of the heads prototypically is identical to that focus articulation, tripartite structures, and semantic con- of the whole groups. In Jim met Sally nothing can be tent. Dordrecht: Kluwer. deleted, but other cases show that the verb is never Hudson, Richard. 1984. Word grammar. Oxford: Blackwell. Kahane, Sylvain (ed.) 2000. Les grammaires de dépendence— deletable (without a specific context), whereas object Dependency grammars. Revue T.A.L., 41, No. 1, Paris. can be absent not only with objectless verbs (sleep, Mel’chuk, Igor. 1988. Dependency syntax: theory and practice. emerge) but also with verbs like read. Verbs such as New York: State University of New York Press. rain even occur without a lexical subject (it as in it PetkeviJ, Vladimír. 1995. A new formal specification of under- rains is not a lexical subject, but rather a filler without lying structures. Theoretical Linguistics 21. 7–61. Sgall, Petr, Eva HajiJová, and Jarmila Panevová. 1986. The semantic relevance). meaning of the sentence in its semantic and pragmatic The linearized form of the underlying representa- aspects, ed. by Jacob L. Mey. Dordrecht:Reidel and tions is made possible thanks to their projectivity (non- Prague:Academia. projective, i.e. discontinuous sentence parts are PRAGUE PETR SGALL handled by rules mediating between underlying and morphemic representations). The possibility of lin- See also Bresnan, Joan; Chomsky, Noam; Lakoff, earizing the prototypical structures shows that the George 257 DEFINITENESS Definiteness Primarily a pragmatic, essentially deictic (‘pointing first is less clear-cut and the utterance does not end as at’) function, definiteness is expressed cross-linguisti- abruptly as the second, whose intonation contour is the cally by different devices: phonological, morphologi- one characteristic of the affirmative sentence, with a cal, syntactic, and lexical. The most characteristic such clear descent of tone and ending in a clear-cut pause. device is the definite article (the), i.e. a bound mor- In English, if an abstract noun is definite, it is actu- pheme operating on a noun or noun phrase. When it alized, cf. ø Truth is what we should stand for, but the operates on a nonnominal element, the latter is nomi- truth is that we don’t. Other nouns whose referents too nalized—it is turned into a noun. Conversely, all deic- are seen as nonindividuated, i.e. mass nouns, are tics and nominals that function deictically (i.e. all incompatible with the indefinite article, cf. The / *a linguistic elements that ‘point out’ a referent), includ- sand. Compatibility is obtained through individuation ing proper nouns, are definite intrinsically. by numeral classifiers, cf. a grain of sand. When a Definiteness is a scalar opposition, i.e. definite- member of a set is definite but nonreferential, nonspe- ness/indefiniteness are two poles between which there cific, nonindividuated, it is generic, i.e. stands for the are multiple intermediate points. Nonreferential indef- whole set and is equivalent to the indefinite nonrefer- initeness and denominalization are iconically bound to ential, nonspecific, nonindividuated plural, cf. The be marked by zero (indicated below by ø), intermedi- bear hibernates ≈ ø Bears hibernate. A bear hiber- ate degrees are cross-linguistically marked by several nates, in contrast, would be indefinite nonreferential, devices, e.g. indefinite articles (a), a clitic deictic nonspecific, individuated. Unique elements are defi- demonstrative (this-), etc. nite, e.g. the sun, although they may not be, if they are Definiteness is a multidimensional notion that can seen as part of a set, cf. love under another sun. There combine referentiality, specificity, identification, actu- are languages that devote a special form or syntactic alization, genericity, individuation, familiarity, and structure to mark the indefinite partially referential, cf. shared knowledge. Some combinations are: French Je cherche du pain ‘I’m looking for some bread’. —definite referential, specific, identifying, cf. The Negative constructions are hardly compatible with book I am reading is Tom Sawyer; definiteness since most of its dimensions are absent, —indefinite referential, specific, nonidentifying, cf. Fr. Je veux de la soupe ‘I want some soup’ vs. Je cf. Tom Sawyer is a book I am reading; ne veux pas de ø soupe ‘I do not want ø soup’, Russian —definite referential, specific, shared knowledge, Ivan kupil komputer ‘Ivan bought a computer’ (accu- cf. I’m looking for the book [I was reading] # sative) vs. Ivan ne kupil komputera ‘Ivan did not buy —indefinite referential, specific, nonshared a(ny) computer’ (genitive).This is valid for ergative knowledge cf. I’m looking for a book [bu:k]… languages too, cf. Basque Nik dut baratze bat ‘I have (≈ that was here a minute ago] # a garden’ (absolutive) vs. nik ez dut baratzerik ‘I do —indefinite nonreferential, nonspecific, nonshared not have a garden’ (partitive). If negation is identified knowledge cf. I’m looking for a book [buk] # contrastively, definiteness is possible, cf. Je ne veux (≈ any book). pas la soupe, je veux la salade ‘I do not want the soup, I want the salad’. The last two utterances clearly differ by content and Nouns that are incorporated into a verb are incom- context. The first of the two may answer a question of the patible with definiteness, cf. to go hunt a bear vs. to go type What are you looking for [on the table/in the ø bearhunting, and so are other denominalized nouns, room/...]?, or: Have you lost anything?, etc. The person e.g. adverbialized ones, cf. take ø fire. answering has a specific book in mind. The second utter- A particular effect is obtained when definiteness ance, on the other hand, may represent the first sentence operates on nouns definite by nature, e.g. proper of a client entering a store, who does not necessarily have names (of which the definite article is not a permanent a specific book in mind. As far as form is concerned, constituent) both utterances are likely to differ as well, by means of vowel length, intonation and prosody. In the first one, the —referential, specific, cf. I’m looking for ø (Mr.) accentuated vowel of the indefinite noun is likely to be Jones slightly longer than in the second utterance, where it is —referential, specific, identifying, cf. I’m looking non-marked for length. The intonation contour of the for the Mr. Jones who was here yesterday 258  DEFINITENESS —referential, specific, nonidentifying, cf. I’m sauere ‘The cherry is of the sour type’, Fr. Il est ø psy- looking for a Mr. Jones who is supposed to live chologue ‘He understands people’ vs. C’est un psy- here (when an explicit article is present, chologue ‘He is a psychologist’. The link between (1), prefixed civility classifiers (Mr...) or suffixed (2), and (3) is apparent in Spanish, where estar (‘be’, human classifiers (...boy/girl), cf. a/the Mr. punctual-dynamic-accidental) is incompatible with the Jones/Jones boy/guy/Beth girl, etc. block the indefinite article, while ser (‘be’, durative-stative- reifying effect of the article). inherent) is compatible with it, cf. respectively *Está / Es una cereza amarga/(un) sicólogo; (4) noun class, In English, the definite article allows also to plural- including sex gender. In languages displaying this cat- ize and actualize proper nouns, e.g. last names: The egory, its marks coalesce with those of deixis and often Smiths. In Córdoba (Argentina) Spanish, in rural definiteness so that the class prefixes in Bantu; French, etc., it is first names that are actualized by the Guaykuru (Amerind); etc., function as definite articles. definite article in all functions to convey familiarity. On Diachronically, a definite article is descended from a the other hand, when an indefinite article is used with a deictic demonstrative. Discursively, the definite article proper name, it strips it of its individuality and makes is an anaphoric i.e. an intradiscursive deictic device par it the representative of a class, cf.: It would take a excellence, i.e. it always points to something men- Henry James to describe that man’s psychology. There tioned, either previously or afterwards, or given/infer- are languages in which the definite article operates on able from context (including general truths). This is proper nouns that are the topic of the utterance. accomplished either explicitly, cf. We reached a river Topicality (old information) and definiteness are nar- nearby. The river was majestic, or implicitly, cf. We rowly correlated, as are focality (new information) and reached a river nearby. The other bank was too distant indefiniteness. In Nêlêmwâ (Melanesian), /-xe/ func- to be seen. Deixis is also the first function cast upon the tions both as a definite article and a topicalizer. Topics definite article by the child acquiring language. These tend to be subjectal, agentive, human, deictic, and first facts illustrate that definiteness is essentially deictic, actants of transitive verbs; topical nouns with any or and hence of a communicative-pragmatic nature, which some of these properties tend to be definite. Focal (new is why it is conveyed in all tongues, albeit not necessar- information) ones tend to be predicative, objectal, pati- ily by a specific morpheme. Quintilian’s (born AD 35) entive, nonhuman, nondeictic, second actants of transi- words: Noster sermo articulos non desiderat, ideo in tive verbs and indefinite. If definite and/or human, they alias partes orationis sparguntur ‘Our language does are discursively marked, and often formally as well, cf. not want articles; hence, thei(r functions) are cast upon Sp. Vi la casa ‘I saw the house’ vs. Vi a la mujer ‘I saw other parts of the sentence’ apply cross-linguistically; the woman’ Contemporary Hebrew (CH) [ra?iti ø languages not having developed a specific form of a dira] vs. [ra?iti ?et ha-?i∫a], Guaraní [ahe∫a oga-ø] vs. deictic demonstrative to work as a definite article apply [ahe∫a kuña-me]. Hence, existential constructions to other mechanisms to perform this task. Classical (There is...) in which the noun is the focus are cross-lin- Latin is an example, cross-linguistically current, of def- guistically incompatible with definiteness, cf. Spanish initeness marking in a tongue with no articles: a definite *Hay el libro# * ‘There is the book#’ (the asterisk noun is placed in sentence initial position (which often marks ungrammaticality), CH *[ye∫ ha-sefer#], Fr. *Il coincides with subject position). There are languages y a le livre# ‘id.’ One way to override this contraint, viz. that developed a definite article, then lost it as such to actualize or topicalize an indefinite noun, is to use a either by phonological or by semantic attrition, and then deictic demonstrative, cf. There was a guy# vs. There developed a new one. This includes, among others, was this guy, who… or to focalize the existential, cf. African languages of various stocks. In Ethiopian-spo- CH [ye∫ ø-sefer#] vs. [ye∫ no ha-sefer#]. Conversely, a radically in Ge’ez and widely in Amharic-a 3rd p. pos- means to focalize a definite noun is the presentative sessive suffix (of deictic origin) is used as a definite construction, cf. Here is the book, Fr. Voilà le livre, CH article. In Nahuatl, the deictic-nominalizer /in/ func- [hine ha-sefer], Sp. He aquí el libro. Accordingly, the tions as definite article when prefixed to the noun; this definiteness gradient correlates with (1) aspect: bound- is corroborated by the fact that when a noun does not ed action ~ definite agent vs. unbounded action ~ bear such a prefix, it is predicative. The suffix /-tl/ indefinite agent; note that genericity blocks the actual- marks a vast majority of nouns (except in incorporation, izing aspect, cf. ø The bear hibernates /* is hibernat- in the plural, when the noun is possessed and in quanti- ing; (2) dynamicity: active verb ~ definite agent vs. fiers, indefinites, and interrogatives); Neo-Aramaic /-a/ stative verb/adjective/nominal predicate ~ indefinite behaves similarly. Those are erstwhile deictics that cliti- actant; (3) inherency: operating on a nominal predicate, cized into definite articles, and then spread to all nouns the indefinite article assigns the subject to a set estab- in all positions and became mere nominalizers. lished by that predicate, cf. German Die Kirsche ist ø The numeral ‘one’ often develops a clitic form to sauer ‘The cherry is sour’ vs. Die Kirsche ist eine mark an indefinite noun as referential, and the process 259 DEFINITENESS starts by marking it as specific: CH [?exad/?axat] ‘one’, Gregersen, E. 1967. Prefix and pronoun in Bantu. IJAL, respectively, m. and f., evolved a clitic form [-(?e)xad/- Memoir 21. Baltimore: LSA. Guillaume, G. [1919] 1975. Le problème de l’article et sa solu- (?a)xat], cf. [cipor(?a)xat ?amr-a li] ‘a (certain) bird tion dans la langue française. Paris: Librairie A.-G. Nizet told me’ vs.[ha-xasida hi cipor-nod ø] ‘the stork is a and Québec: Presses de l’Université Laval. migrating bird’. At present, an anteposed, concording Harris, M. 1979. The marking of definiteness: a diachronic per- and often stressed form of /?ejze/ ‘which’, followed by spective. Papers from the 4th International Conference on the relative particle /∫e/ and a third person deictic is Historical Linguistics, ed. by E. Closs Traugott, 75–86. Amsterdam: John Benjamins. spreading to focalize not the noun itself but its being Karmiloff-Smith, A. 1979. A functional approach to child lan- indefinite referential, specific-, cf. [je∫ ?ejzo∫ehi hitkad- guage: a study of determiners and reference. Cambridge: mut] ‘there is some [undoubted] progress]’. Both recent Cambridge University Press. marks are incompatible with each other as well as with Kirtchuk(-Halevi), P. I. 1993. Deixis, anaphore, accords, classi- the definite article /ha-/ and with a free deictic, which fication: morphogenèse et fonctionnement. Lille: ANRT. Kirtchuk(-Halevi), P. I. 2000. Deixi and Noun Classification in confirms that (in)definiteness is a scalar opposition. Pilagá and Beyond. Between Grammar and Lexicon, E. Contini Morava and Y. Tobin (eds.), 31–55. Amsterdam- References Philadelphia: John Benjamins. Kolde, G. 1996. Nominaldetermination. Tübingen: Niemeyer. Bello, A. [1847] 1984. Gramática de la Lengua Castellana. Krámsky, J. 1972. The article and the concept of definiteness in Madrid: Edaf. language. The Hague–Paris: Mouton. Bril, I. 1994. Indéfini et degrés de définitude en nêlêmwâ. Faits Launey, M. 1996. Retour au /-tl/ aztèque. Amerindia 21. 77–92. de Langues 4. 211–20. Leiss, E. 2000. Artikel und Aspekt - die grammatischen Muster Corblin, F. 1987. Indéfini, défini, démonstratif. Genève-Paris: von Definitheit. Berlin–New York: Walter de Gruyter. Droz. Lyons, C. 1999. Definiteness. Cambridge: Cambridge Givón, T. 1981. On the development of the numeral ‘one’ as an University Press. indefinite marker. Folia Linguistica Historica—Acta Rosén, H. 1991. The definite article in the making, nominal Societatis Linguisticae Europaeae I(2). 295–304. constituent order and related phenomena. Linguistic studies Goldenberg, G. [1991] 1998. On predicative adjectives and on Latin; selected papers from the 6th International syriac syntax. Studies in Semitic Linguistics, 579–90. Colloquium on Latin Linguistics, ed. by J. Herman, 129–50. Jerusalem: Magnes Press. Amsterdam–Philadelphia: John Benjamins. Grasserie, R. de la. 1896. De l’article. Mémoires de la Société Wackernagel, J. 1928. Vorlesungen über Syntax. Basel: de Linguistique de Paris 9. 285–322, 381–94. Birkhäuser. Greenberg, J. 1978. How does a language acquire gender mark- PABLO I. KIRTCHUK-HALEVI ers? Universals of human language, 47–82. Stanford: Stanford University Press. See also Deixis; Determiner; Reference Deixis Deixis concerns the ways in which languages encode by the German philosopher Karl Bühler in his book or grammaticalize information regarding the extralin- Sprachtheorie (1934; Theory of language). Three main guistic context in which utterances occur, and the ways categories of deixis have been traditionally distin- in which the interpretation of utterances depends on guished: person, time, and place. It is generally accept- the analysis of that context. For instance, the sentence ed that deixis is an egocentric system in that the speaker ‘Peter likes me’ cannot be properly interpreted unless chooses his own perspective when integrating personal, we know certain details about the context in which it temporal, and spatial information into the message: the was uttered, namely, the identity of the speaker. central person is the speaker, the central time is the time The study of deixis goes back to the work of the at which the speaker encodes the message, and the cen- ancient Greek grammarians (the term ‘deixis’ is the tral place is the speaker’s location at coding time. Greek word for ‘pointing’ or ‘indicating’). Although Person deixis involves the identification of the par- deixis has proved itself as a central linguistic concept, ticipants in the communicative situation as well as during the twentieth century there has been a consider- other individuals referred to in the utterance. Among able philosophical interest in deictic phenomena. In the linguistic expressions that fulfill this identifying fact, the topic was reintroduced into modern linguistics function, personal pronouns and inflectional affixes 260  DEIXIS attached to nouns and verbs are the most usual ones. itself. Since discourse develops in time, discourse The basic grammatical distinctions in person deixis deixis makes use of terms borrowed from time deixis. are the categories of first, second, and third person: the For example, the expression the last paragraph would first person refers to the speaker himself, the second be analogous to last week. In a similar way, discourse person refers to the addressee, whereas the third per- can be thought of as having a spatial extension since son refers to a person or persons who are part of the any point within it can be located either behind or conversational group but who are neither the speaker ahead of a central reference point. Thus, place deictic nor the addressee. In addition to this contrast, person expressions like the adverbs above and below or the deictic items typically include information about num- demonstratives this and that are commonly used to ber and gender. The singular/plural distinction is the refer to a preceding or following portion of discourse. primary number opposition in all languages. As Social deixis involves the encoding of the social sta- regards gender, person deictic items may convey infor- tus and rank of the participants in the communicative mation about the sex (semantic gender) or arbitrary situation and the social and personal relationship class (grammatical gender) of their referents. between them or between one of them and persons Place deixis deals with the specification of the spa- referred to. One of the most obvious manifestations of tial location of the entities referred to in the utterance. social deixis is the choice of specific pronominal forms Although locations are most commonly established by (known as ‘honorifics’) to refer to the addressee in reference to that of the speaker, some languages also many European languages: French and German, for take the addressee’s position as the central place. The example, make a distinction between the familiar terms basic spatial contrast in most languages, illustrated by tu and du and the polite terms vous and Sie, respective- the English pair here vs. there, is that of ‘proximal to ly. Social status can also be encoded throughout parti- the speaker’ and ‘distal to the speaker’. Many lan- cles and affixes that indicate, as in Japanese, respect or guages, however, exhibit a three-fold opposition; deference. Titles of address (Mr. President or My Lord Spanish, for example, distinguishes three locative in the courtroom) also fulfill this social deictic function. adverbs: aquí (‘near the speaker’), allí (‘far from the Many languages also make a division between two or speaker’), and ahí (‘a little way from the speaker’). more registers in which different vocabulary and syn- Spatial relations may be expressed linguistically by a tactic constructions are used depending on whether the wide range of items. The most common perhaps are communicative situation is seen as formal or informal. locative adverbs (here/there), demonstrative pronouns In short, the essential feature of deictic expressions (I will take that), and adjectives (this house). is that their semantic values depend on the real-world Time deixis involves the identification of the time context in which they appear. Thus, deixis constitutes of the events described in the utterance by reference to a natural link between semantics and pragmatics. the time of the communicative situation itself. Time However, deixis is also bound to nonlinguistic aspects deixis is commonly treated as a metaphorical exten- such as the speaker’s attitude or the interaction of sion of place deixis: time is seen as a unidirectional grammar and culture and thus it offers itself as a com- stream in which events are located before (or behind), mon meeting ground for investigations undertaken after (or ahead), or simultaneous with the time at from different perspectives such as those of philoso- which the utterance is produced. Probably, the most phy, cognitive psychology, psycholinguistics, sociolin- common temporal deictic device is tense, although guistics, or anthropology. some languages, like Chinese, lack this verbal catego- ry. Languages also rely on temporal adverbs (now/then) and demonstratives. Demonstratives are References said to be imported from the local domain and used Anderson, Stephen R., and Edward L. Keenan. 1985. Deixis. metaphorically in expressions like this week, which Language typology and linguistic description, Vol. III, ed. by implies proximity to the utterance time, or that Timothy Shopen. Cambridge: Cambridge University Press. evening, which implies remoteness. Finally, most lan- Bühler, Karl. 1934/1982. Sprachtheorie. Jena/Stuttgart: Gustav guages have lexical items that specify temporal rela- Fischer Verlag; as Theory of language. The representational function of language, translated by Donald Fraser Goodwin. tions, like the English words today or yesterday. Amsterdam and Philadelphia: John Benjamins, 1990. In the early 1970s, Fillmore’s works brought a Fillmore, Charles J. 1971. Santa Cruz lectures on deixis, repro- renewed interest in the study of deixis and new types duced by the Indiana University Linguistics Club. were added to the traditional inventory. Among them, the Bloomington; Indiana. most widely accepted are discourse and social deixis. Fillmore, Charles J. 1997. Lectures on deixis. Stanford, CA: CSLI Publications. Discourse deixis concerns the use of certain expres- Jarvella, Robert J., and Wolfgang Klein (eds.) 1982. Speech, sions within the discourse to refer, not to an extralin- place, and action: studies in deixis and related topics. guistic entity, but to some portion of the discourse London and New York: John Wiley and Sons. 261 DEIXIS Lakoff, Robin. 1974. Remarks on this and that. Chicago Perkins, Revere D. 1992. Deixis, grammar, and culture. Linguistic Society 10. 345–56. Amsterdam and Philadelphia: John Benjamins. Levinson, Stephen C. 1987. Pragmatics. Cambridge: Rauh, Gisa (ed.) 1983. Essays on deixis. Tübingen: Narr. Cambridge University Press. LUISA GONZA´ LEZ-ROMERO Lyons, John. 1977. Semantics, 2 vols. Cambridge: Cambridge University Press. See also Bühler, Karl; Fillmore, Charles Determiner A determiner is a member of a syntactic category that Both interpretations are indefinite. Specificity is thus a combines with a noun (e.g. house, book) to express different concept from definiteness, although both are information about the reference, quantity, proximity, encoded in the same determiners in English. In other or gender of the noun. It is a function word that adds languages, specificity may be more clearly encoded in grammatical information to the basic meaning of a determiners that are different from definite and indef- noun. It may appear in different languages as an inde- inite articles. Or, unlike in English, the concept of pendent word (English a, the), an affix to a noun (Farsi specificity, rather than definiteness, may be the key kitob-ro, ‘book-the’), a combination of both independ- factor in how articles are interpreted. ent word and affix (Swedish den vita bilen, ‘the white Demonstratives (this, that, these, those) are deter- car-the’), or a whole phrase (as the dog’s in the dog’s miners whose main function is deictic, or ‘pointing owner). There are several subtypes of determiners, out’. They indicate the proximity or closeness of a each with its own particular function. noun to the speaker. A proximal demonstrative identi- Basic determiners, or articles (a, an, the), give infor- fies a noun as close to the speaker (this chair), where- mation about the definiteness of a noun. Their use as a distal demonstrative identifies it as distant from the depends on the discourse context. In general, a speaker speaker (that chair). The distance can be metaphorical will use the indefinite articles a or an to introduce a or temporal rather than physical, as in that situation or noun into a conversation or to name one that does not this week. Although systems of two degrees of proxim- have any specific reference. The definite article the ity are the most common, some languages may express introduces a noun that has a unique reference and is more. Maori, for instance, identifies three degrees of familiar or readily identifiable from context. The arti- proximity: close to the speaker, distant from the speak- cles thus play an important role in discourse in signal- er, and distant from both speaker and hearer. ing the appearance of new referents (with indefinite A secondary use of the proximal demonstratives this articles) and maintaining their status as current topics of and these is to introduce an indefinite but specific noun conversation (with the definite article). Not all lan- that will figure prominently in the conversation that guages make use of an overt indefinite article, however. follows—for example, There was this car blocking the Often there may be only one basic definite article, with intersection, or This guy called for you. In such con- bare nouns interpreted as indefinite. In English, the structions, the demonstrative is a stylistic alternative to absence of an overt article signals either a mass noun the indefinite article a. As these examples suggest, this (rice, butter) or a generic use of a countable noun, i.e. a is a colloquial use of the demonstrative, and it typical- noun used to identify a class of objects of a certain type, ly appears in openings to anecdotes and jokes. It is an as students in Students cannot afford to live here. exceptional indefinite use of the demonstrative, which The use of the indefinite article a in English is com- is otherwise inherently definite when used deictically. plicated by the fact that it is ambiguous with respect to Quantifiers are determiners that express the quanti- specificity. The sentence I am looking for a woman ty of a noun, for example, some, any, most, and the with red hair has two interpretations. The article a negative quantifier no. Quantifiers interact in complex could signal that the speaker is searching for a type of ways with articles and demonstratives. Some are woman identifiable only by the characteristic of hair compatible only with definite interpretations (all, color. Or, it could signal that he is in fact searching for both, every, most), whereas others are inherently indef- a particular woman, whom he already knows. The first inite (some). All and both are the only determiners that interpretation is a nonspecific reading of the article a, can combine with other definite determiners in whereas the second interpretation is a specific reading. English, as in all the students. None of the quantifiers 262  DETERMINER may combine with indefinite determiners: *all some with the determiner the acting as the head of the con- students, *each a student. All and both are sometimes struction. Most determiners are like the in that they referred to as predeterminers and might be better ana- must combine with a following nominal category, such lyzed as a separate category of word. as house. However, personal pronouns may be ana- Possessive determiners include possessive pro- lyzed as independent DPs that do not necessarily need nouns such as my and your and entire noun phrases to combine with any other category. marked with the possessive inflection -’s, as in every The DP analysis is supported by facts about the man’s rights. Possessive determiners imply definite- form, word order, and meaning of determiners in many ness and give specific information about the relation- languages. In particular, the DP analysis is a fruitful ship of one noun with another. Usually, this is a matter means of representing similarities between nominal of ownership, but it could also be a matter of origin or phrases and full sentences. Noam Chomsky pointed inclusion. In some languages, possessive determiners out in the 1970s that a nominal construction such as can encode a distinction of ‘alienable’ vs. ‘inalienable’ the Romans’ destruction of Carthage is in many ways possession. In inalienable possession, a noun (typical- analogous to a corresponding sentence The Romans ly a body part or kinship term) cannot be separated destroyed Carthage. The information carried by the from the possessor. In alienable possession, a noun can determiner is central to the grammatical interpretation be separated from the possessor, and the relationship of the noun, in the same way that the information car- of possession can be terminated. This is usually the ried by tense and inflectional elements is central to the case with inanimate objects. This very specific infor- interpretation of a sentence. Because tense and inflec- mation can be encoded in the form of the possessive tion are assumed to be the head categories of the sen- determiner. tence, it follows that the determiner should be Wh- determiners can take the place of any of the analyzed as the head of the nominal construction. The other determiner types if the identity, quantity, or own- possessive determiner (the Romans’) is the structural ership of a noun is unknown and must be questioned: parallel of the subject of a full sentence. Which chair? What book? Whose rights? Wh- deter- In some languages, a determiner may agree syntac- miners thus express unspecific reference in English. tically with the noun it accompanies. That is, it may In older grammars, the determiner is not always rec- show inflection to indicate that it shares grammatical ognized as a separate syntactic category, or the term features such as number (singular or plural) or gender may be used only to refer to the basic indefinite and with the noun. Agreement between determiners and definite articles. Instead, the subtypes of determiners nouns is similar to the agreement of the subject and listed above are often treated as completely separate verb in a sentence. Extensions of the DP analysis have word classes. It is fairly recent that these have been therefore suggested that the structure of nominal given a unified analysis as members of the more gener- phrases is very similar to that of sentences. al category of determiner. This change is motivated by the fact that they all precede a noun and may substitute References for one another but in effect never co-occur in the same phrase. Thus, *the these papers is ungrammatical. A Abney, Steven. 1987. The English noun phrase in its sentential unified analysis is also supported by the similarities in aspect. Cambridge, MA: MIT. Baltin, Mark, and Chris Collins (eds.) 2001. The handbook of their meanings and the way in which some imply the contemporary syntactic theory. Oxford: Blackwell. definiteness usually encoded by the. Given this view, Bernstein, Judy B. 2001. The DP hypothesis: identifying clausal even personal pronouns such as you and we may be properties in the nominal domain. In Baltin and Collins. classified as determiners, because they may combine Hawkins, John A. 1978. Definiteness and indefiniteness: a with nouns and signal definite and deictic meanings in study in reference and grammaticality prediction. London: Croom Helm. phrases such as you hypocrites and we taxpayers. Longobardi, Giuseppe. 2001. The structure of DPs: some prin- The determiner’s status as a function word was re- ciples, parameters, and problems. In Baltin and Collins. evaluated during the mid-1980s, when syntactic theo- Lyons, Christopher. 1999. Definiteness. Cambridge: Cambridge ry began to reinterpret all types of function words as University Press. the most important (i.e. ‘head’) words in a phrase. Ritter, Elizabeth. 1991. Two functional categories in noun phras- es: evidence from modern Hebrew. Perspectives on phrase Before this time, function words were viewed as being structure, ed. by Susan Rothstein. San Diego, CA: Academic dependent on lexical categories such as nouns and Press. verbs. Steven Abney argued in 1987 that the determin- Siloni, Tal. 1997. Noun phrases and nominalizations: the syntax er was the head of its own determiner phrase rather of DP. Dordrecht: Kluwer. than simply a word appearing in a noun phrase headed Van der Auwera, Johan (ed.) 1980. The semantics of determin- ers. London: Croom Helm. by a noun. By this analysis, which is known as the DP SHEILA DOOLEY COLLBERG (Determiner Phrase) Hypothesis and is now widely accepted, a phrase such as the house is a DP structure, See also Definiteness 263 DEVELOPMENTAL STAGES Developmental Stages Identifying the stages of language development cannot relationships. For example, a child will not be able to be done independent of the notions of ‘comprehen- advance in developing negation in English without the sion’ vs. ‘production’ and ‘competence’ vs. ‘perform- development of the copular verb and the auxiliary sys- ance’, because, when we claim that a child has tem, including insertion of ‘do’, since the negative par- acquired ‘the past tense’ or ‘how to tell a structured ticle (n’t) suffixes to these tensed verbs in most story’, we need to know whether we mean that the instances. (Thus, an early expression of negation such child shows evidence of comprehending, say, the dif- as ‘Not fit’ becomes ‘It doesn’t fit.’ or ‘It can’t fit.’) The ference between past and present tense, or whether the issue of what is related to what is extremely important child accurately codes the distinction between them in in discussions of the modularity of language, both in his or her speech or writing. Beyond this, it is possible terms of the relationship of language development that data from neither comprehension nor production with nonlinguistic development, and in terms of the accurately reflects the child’s mental representations internal connectedness of developments within lan- (competence). Performance, in other words, always guage. This is an area where linguistic theory and first risks being an inadequate reflection of competence. language acquisition research are vitally connected. Despite these problems, most child language researchers are confident in crediting a child with hav- Language in Infancy ing acquired a particular aspect of language when they demonstrate it in their production in around 90% of the Although infants under a year old do not yet produce required contexts. Hence, if a child is using the regular language, they are building mental representations for past tense ‘-ed’ ending in nine out of every ten cases language and are beginning to comprehend the lan- where it is required, then we can say the child has guage(s) around them. In addition, the ability to com- acquired it. Given that comprehension almost always municate desires and beliefs by vocalization precedes productive capacity, this conservative (including intonation) and both hand, face, and body approach is generally justified. There are, though, situ- gesture combined with appropriate eye gaze, develops ations in which children (some more than others) pick rapidly in the prelinguistic phase. For those who see up chunks of language, sometimes quite large ones, pragmatic development as part of language develop- without really understanding what they mean or how ment (some do, some do not), the emergence of ges- they are constructed. In these cases, production pre- ture and nonverbal vocalization, along with the ability cedes comprehension, and crediting acquisition needs to get and direct the attention of another person and to wait for the child’s internal analysis of the chunks. engage in reciprocal turn-taking, are all important Notions of general, across-the-board stages of lan- milestones of pragmatic competence that are in place guage development are difficult to maintain, even prelinguistically. when variations in individual children’s rate of devel- Much of the language used to a preverbal child is opment are taken into account. Although writers often comprehensible from accompanying nonverbal clues refer to ‘the one-word stage’, ‘the multiword stage’, (e.g. ‘Get me the ball’ accompanied by a pointing ges- etc. these provide little substantive information, actual- ture), and in most cases it is the context, rather than the ly only serving to trigger the association that, in the words spoken, which allows the child to understand. In first case, we are probably talking about (normally their own productions, children begin with cries and developing full-term) children in their second year of coos, which give way to ‘babbling’ at around six life, and in the second, slightly older children, but still months. Babbling is characterized by increasingly sta- preschoolers. The problem with identifying across the ble vocalizations that, over a six-month period, board stages is, firstly, that the range of normal lan- approximate the word structure of the language(s) they guage development is large and shades imperceptibly are exposed to. Toward the end of the babbling stage, into language disorder. Secondly, the stages are not ‘proto-words’ emerge. These are relatively stable discrete, one-word utterances co-occurring with two- vocalizations that seem to have a recognizable mean- word utterances for some time. Thirdly, while some ing. By the time first words emerge, around the end of aspects of language development are relatively inde- the first year, infants show significantly more pendent of others (e.g. the acquisition of adjectives, or advanced responses to language than their own pro- of consonant clusters), others reflect highly dependent ductions would indicate. 264  DEVELOPMENTAL STAGES Deciding when the prelinguistic phase has given at even the single-word stage that bound morphemes way to the linguistic depends on cultural norms, chil- such as verbal inflections are beginning to emerge, and dren’s control over their vocal apparatus, and normal as the single-word stage gives way to the multiword individual variation. Whether a proto-word is recog- stage, children acquiring such languages advance faster nized as a word depends on how like a recognizable in morphology than those learning languages like adult word it is, and the extent to which the adults English (even though full mastery of these systems around the child are willing to interpret them as such. may take the learner of Finnish well into adolescence). In some cultures, it is the emergence of a particular Researchers have noted that there is a relatively pre- word (e.g. the word for ‘breast’) that is seen as mark- dictable order of acquisition of functional morphemes ing the emergence of language. Finally, some children in English, from the progressive ‘-ing’, prepositions ‘in’ seem to experiment freely with early words, while oth- and ‘on’, and plural ‘s’ through to the contractible cop- ers hang back until they are capable of clearer, more ula and auxiliary, as in ‘He’s happy’ and ‘He’s walking’, recognizable speech. respectively. This order, first explored by Roger Brown (1973), is often used to estimate a child’s stage of devel- opment. The mean length of utterance (MLU), a meas- Language in the Preschool Years ure of productive capacity proposed by Brown, is also Many observers have noted that lexical development used. An MLU of between 1.50 and 2.0 represents starts slowly (about one new word per week), but that Stage 1 of development and is usually attained during somewhere around the 50-word mark, many children the second year. MLU then rises through four succes- (but not all) experience a sudden increase in vocabu- sive stages to Stage 5 when MLU reaches 4.0 or above, lary acquisition. They then add words at a rate that normally at around three-and-a-half years of age. translates into an average of nine new words per day However, children may vary individually in terms of for the next ten years or more. In production, children both the pace and overall rate of development. generally begin by acquiring names for people and As MLU rises, children’s utterances not only objects in their environment as well as useful interper- include more content words but also come to include sonal expressions such as ‘hi!’ and ‘bye-bye’, with both function words such as ‘he’, ‘not’, and ‘a’, and some children having a greater proportion of such bound morphemes such as ‘-ing’, ‘-ed’, and ‘-s’. In expressions in their early vocabularies than others. adding bound morphemes in their productions, chil- Pronunciations are often simplified through the use of dren also try to rationalize the system by overgeneral- strategies such as regularizing the consonant–vowel izing endings to produce things such as ‘breaked’, syllable structure (‘wawa’ for ‘water’), deleting final ‘goed’, and ‘deers’. Gradually, however (and it is a consonants (‘du’ for ‘duck’), simplifying consonant slow process), irregular forms such as ‘broke’ and clusters (‘tain’ for ‘train’), and avoiding difficult (and ‘went’, which tend to make a brief appearance early in typologically rare) sounds such as /r/ and /θ/. Words development and then disappear as the overgeneraliza- also start out restricted in their reference (e.g. ‘teddy’ tions take over, reappear and shut out the earlier forms. may start out being applied to only the child’s own As complexity increases, children’s capacity for teddy, or only to teddies when they are in the child’s expressing basic sentence types (affirmative declara- home), only gradually coming to have the usage of the tives, negative declaratives, interrogatives, etc.) adult language, and not infrequently passing through a becomes more sophisticated in ways that are difficult stage of being overextended in their use (e.g. ‘teddy’ to describe independent of a particular theory of syn- applies temporarily to all soft toys). Words overex- tax. Whether acquisition is seen as the accumulation of tended in production in this way are not usually constructions or the genetically powered interplay of overextended in comprehension. In comprehension, abstract features and the lexicon will affect how the during the second year of life, children begin to be increasing complexity of a child’s utterances is viewed. able to understand a wide range of words for people, Gradually, though, the auxiliary system emerges, games and routines, familiar objects, animals, body adverbs are put in the right places, subjects of sen- parts, and action words. They pick up some of these tences are expressed, etc. Then, at around three years of words on only one exposure. age, complex sentences begin to emerge with the suc- In languages such as English where word order is cessive appearance of complements of verbs such as the main indicator of who does what to whom, early ‘think’ and ‘know’ (‘I think you’re funny’, ‘I know how development is marked by the absence in the produc- many legs a dinosaur has’), infinitive complements (‘I tion of inflectional and derivational bound morphemes want to go to school’), relative clauses (‘That’s the and of function words such as prepositions, auxiliaries, teddy that Mummy bought me’), conditionals (‘I can if and articles. However, in more richly inflected lan- I want to’), and conjunctions (‘I have to go ‘cos it’s guages, such as Spanish and Finnish, there is evidence time’, ‘I went to Grandma’s and we had tea’.). 265 DEVELOPMENTAL STAGES Semantic development is intimately bound up with interests and educational expectations. The capacity to cognitive/conceptual development and it is not clear read and write and to engage in abstract thought drives whether language acquisition introduces new semantic both late syntactic development and unlimited vocab- possibilities, or whether it provides coding opportuni- ulary development. ties for concepts already developed through observation Around the age of six or seven, the capacity to reflect of the world, including the language of others in con- on language itself, to be metalinguistically aware, text. Researchers have suggested that certain overarch- emerges. This capacity culminates in the ability, among ing semantic relationships appear early as soon as other things, to define abstract nouns, explain the mean- words are combined. These include ‘nonexistence’, ing of idioms and proverbs, and solve complex verbal argued to be characterized by the expressions ‘no more’ analogy problems. It is also reflected in verbal humor and ‘allgone’; ‘recurrence’ (‘more noise’ and ‘nother and, coupled with developments in the ability to ‘read’ raisin’); and ‘attribution’ (‘microphone hot’ and ‘animal the minds of others, the capacity to exploit complex book’). Later, more mature semantic notions, such as social uses of language, for example, in narratives. existence (‘There dog’) and negation (‘No fish’), devel- op followed by quantificational notions coded by words such as ‘some’, ‘lots of’, ‘many’, ‘few’, and ‘several’. References The preschool period is when conversations can be Bloom, Lois. 1993. The transition from infancy to language: reliably started and maintained linguistically, when acquiring the power of expression. Cambridge, UK: children begin to adapt their messages to different Cambridge University Press. audiences, when miscommunications can be repaired, Brown, Roger. 1973. A first language. Cambridge, MA: Harvard University Press. and when stories begin to have proper beginnings, de Boysson-Bardies, Benedicte. 1999. How language comes to middles, and ends. Words such as ‘here’ and children. Cambridge, MA: MIT Press. ‘there’,‘this’ and ‘that’, together with ‘I/me’, ‘you’, Fletcher, Paul, and Brian MacWhinney (eds.) 1997. The hand- and ‘us’, whose references shift depending on who is book of child language. Oxford: Blackwell. talking, also begin to be used more accurately. Foster-Cohen, Susan. 1999. An introduction to child language development. Harlow, Essex: Longman/Pearson Education. Hoff-Ginsberg, Erika. 2000. Language development, 2nd edi- Language in the School Years tion. Pacific Grove, CA: Brooks/Cole Publishing Company. Karmiloff, Kyra, and Annette Karmiloff-Smith. Pathways to Although much of language development is complete language: from fetus to adolescent. Cambridge, MA: by age five, some syntactic constructions do not Harvard University Press. Nippold, Marilyn. 1998. Later language development: the school- appear for many children until they enter school. The age and adolescent years, 2nd edition. Austin, TX: Pro-ed. passive construction in English (‘The book was writ- Owens, Robert. 1996. Language development: an introduction, ten by the teacher’) is one such. However, it might be 4th edition. Boston: Allyn & Bacon. noted that in languages such as Sesotho, where the Pinker, Steven. 1994. The language instinct: how the mind cre- passive is used extensively in speech, it appears much ates language. New York: William Morrow and Company Inc. Pinker, Steven. 1999. Words and rules: the ingredients of lan- earlier. Irregular past tense forms such as ‘fell’ and guage. New York: Basic Books. ‘brought’ may not win out over ‘falled’ and ‘bringed’ Ritchie, William, and Tej Bhatia (eds.) 1999. Handbook of child or ‘brang’ until at least the age of eight or nine, if ever. language acquisition. New York: Academic Press. Complex sentences with relative clauses and other Slobin, Dan (ed.) 1985–1997. The cross-linguistic study of lan- types of complement begin development in the pre- guage acquisition, Vols. 1–5. Mahwah, NJ: Lawrence Erlbaum Associates. school years, but are not consolidated until middle SUSAN H. FOSTER-COHEN childhood. The extent to which children learn advanced literary constructions is driven by individual See also Acquisition; Acquisition Theories Dialectology Dialectology investigates regionally and socially con- history of the discipline of dialectology, the following ditioned linguistic variation, called regional dialectol- motivations can be distinguished: normative interests, ogy and social dialectology (or sociolinguistics). In the which prevailed in the sixteenth and seventeenth 266  DIALECTOLOGY centuries (correct standard language vs. incorrect isoglosses or heteroglosses, but rather symbols that dialect); antiquarian interests, starting in the second reveal transition zones much more clearly and are thus half of the seventeenth century, when dialect lexicog- more suited to linguistic reality. In delimiting dialect raphy began; documentary interests in the nineteenth areas, a qualitative approach must be distinguished and twentieth centuries (written records of spoken lan- from a quantitative one. In the first case, a few features, guage, including the collection of fairy tales and or often only a single feature, are chosen to differenti- phonographic records); linguistic interests, also in the ate dialects from each other. The more important the nineteenth and twentieth centuries (language history, chosen feature, the more reliable the dialectal structure local dialect monographs, dialect geography, semantic obtained. Phonemic features carry more weight than field theory, and structural and generative dialectol- phonetic or lexical ones. Phonemes are the smallest ogy); interests in cultural geography in the twentieth units differentiating the meaning of a language. They century (extralinguistic interpretations of sound and are structural linguistic features and thus very stable. word geography); and psycholinguistic and sociolin- In contrast, phonetic features are nonstructural; they guistic interests, also dating from the twentieth centu- show a great deal of variation. The least stable element ry. Although the psycholinguistic approach tries to of a language is its lexicon. A lexical feature is there- discern from the dialect the view of life underlying it, fore least reliable in differentiating dialects. In a limit- and vice versa—this line of research has not had much ed quantitative approach, a dozen or more features are of an effect—the sociolinguistic approach deals with selected from any linguistic level—whether it is from broad matters such as language and society, dialect the sound, word formation, sentence structure, or and education, and dialect as a language barrier. vocabulary systems—their borders are calculated, and Finally, an interest in nonlinguists’ views of areal lin- thicker or thinner lines are drawn according to the guistics arose in the last decades of the twentieth cen- number of features they distinguish. Then, the results tury; it is called perceptual dialectology. of the various linguistic levels are compared with one In the beginning of the nineteenth century, linguists another to see whether they are in substantial agree- often referred to as Neogrammarians tried to prove ment or not. In recent years, the use of the computer that the sound system of the languages develops and has made it possible to quantify enormous amounts of changes according to rigid and highly regular ‘sound data and to present the results on linguistic maps. This laws’, the term law being understood as in the natural branch of linguistic geography, which makes use of sciences. The perceived regularity of sound changes, highly sophisticated statistical methods, is called they thought, could best be observed in dialects, but as dialectometry or dynamic dialectology. Whereas the it turned out later, this was not the case. procedure involving isoglosses/heteroglosses is a sam- The comparative method of the nineteenth century ple of a sample, the dialectometric approach is more had produced two theories with regard to linguistic objective and exact because it is capable of taking into relationships. One was the naturalistic concept of the account all the available linguistic data collected in a family tree theory, advocated by August Schleicher, certain area. Dialectometry sets off the linear approach and—in contrast to this theory—the wave theory of of traditional dialectology with an areal one. Its Johannes Schmidt, a student of Schleicher’s who methodological procedure is based on the question of maintained that adjacent regions resembled each other identity or nonidentity of two linguistic forms. most and that differences increased with distance. Dialectology has also contributed to linguistic theo- Research in dialect geography or geolinguistics later ry. In the field of structural dialectology advocated first revealed that this is largely the picture presented by by Uriel Weinreich in 1954, mention must be made of dialects. In modern terminology, these situations are the principle of maximum differentiation between two called geographical dialect continua. At their farthest phonemes and the description and comparison of dialec- geographical points, dialects may no longer be mutu- tal systems, to which members of the Prague Linguistic ally intelligible, but they will be linked by a chain of Circle have also greatly contributed. By formulating mutual intelligibility. diafeature rules, as attempted by, e.g. Lawrence M. An important aim of dialectology is to delineate Davis in 1973, taking into account both regional and dialect areas. Traditionally, this has been done—and social variation in language, it has been shown that still is—by drawing lines on linguistic maps, enclosing dialectology added an important dimension to genera- one linguistic phenomenon each time and isolating it tive phonology, too (see, e.g. Rudolph C. Troike 1971). from its surroundings. These lines are called isogloss- Results in dialect geography are normally present- es. Because isoglosses often do not overlap, Hans ed in linguistic atlases. Apart from regional and Kurath coined the term heterogloss, which reflects the national linguistic atlases, there are atlases of language conceptual content more accurately. However, more groups, such as the Panslavic linguistic atlas. The most and more modern linguistic atlases no longer show comprehensive atlas so far is the Atlas Linguarum 267 DIALECTOLOGY Europae, the European linguistic atlas, which covers References all six language families and their dialects spoken in Bloomfield, Leonard. 1933. Language. New York: Holt. the whole of Europe. Also, Sprachbund phenomena Carver, Craig M. 1987. American regional dialects: a word can be treated in a similar way. A Sprachbund is geography. Ann Arbor: The University of Michigan Press. formed by a group of related or nonrelated languages Chambers, J.K., and Peter Trudgill. 1998. Dialectology, 2nd that show systematic similarities in grammar and lexi- edition. Cambridge: Cambridge University Press. Davis, Lawrence M. 1973. The diafeature: an approach to struc- con, which cannot be accounted for by direct relation- tural dialectology. Journal of English Linguistics 7. 1–20. ship with a common underlying language. An (not Goossens, Jan. 1969. Strukturelle Sprachgeographie (Structural undisputed) example of a Sprachbund comprises the linguistic geography). Heidelberg: Winter. Balkan languages. The production of linguistic atlases Preston, Dennis R. (ed.) 1999. Handbook of perceptual dialec- today involves the most recent information technolo- tology, Vol. 1 (Vol. 2, 2001). Amsterdam: Benjamins. Troike, Rudolph C. 1971. Overall pattern and generative gy, including speaking linguistic atlases. phonology. Readings in American dialectology, ed. by Schmidt’s wave theory was also applied to social Harold B. Allen and Gary N. Underwood, 324–42. New dialects. As early as in 1933, Leonard Bloomfield York: Appleton Century Crofts. noted: ‘The higher the social position of the nonstan- Weinreich, Uriel. 1954. Is a structural dialectology possible?. dard speaker, the more nearly does he approach the Word 10. 268–80. standard language’ (Language, p. 49). Such situations WOLFGANG VIERECK are called social dialect continua. See also Sociolinguistics; Sprachbund Diglossia In his seminal article in Word (1959), Charles dialect poetry and advertising. Fourth, given its superi- Ferguson defined diglossia as: or status as the language used in prestigious domains, ‘a relatively stable language situation in which, in the H variety also tends to be more standardized than addition to the primary dialects of the language (which the L variety—grammars and dictionaries are written may include a standard or regional standards), there is a for the H variety, but not usually for the L variety. Fifth, very divergent, highly codified (often grammatically while the L variety is the language of the home, the more complex) superposed variety, the vehicle of a large H variety is not spoken natively by anyone in the com- and respected body of written literature, either of an munity and has to be learned through schooling. Finally, earlier period or in another speech community, which is although the L variety may gradually replace the H vari- learned largely by formal education and is used for most ety due to factors such as more widespread literacy and written and formal spoken purposes but is not used by broader communication among different regional and any section of the community for ordinary conversation.’ social groups, a diglossic situation is relatively stable— Using the examples of Greek, Arabic, Haitian it can persist for centuries or even millennia. Creole, and Swiss German, Ferguson pointed out sever- A diglossic situation is likely to come into existence al characteristics that are common across all diglossic under the following conditions: (1) when literacy in situations. First of all, in terms of function, there is a the community is restricted to a small elite; (2) when strict division of labor between the two language vari- there is a sizeable body of literature in a language eties: while the superposed variety or the H(igh) variety closely related to the vernacular of the community, and is used mostly in prestigious domains such as education this literature embodies some of the fundamental val- and government, the vernacular or the L(ow) variety is ues of the community; and (3) when a considerable restricted to informal domains (e.g. family, neighbor- period of time elapses following the establishment of hood). Second, although the two varieties are genetical- these two conditions. Nevertheless, although these ly related to each other, the H variety is structurally social contexts favor the emergence of diglossia, they more complex than the L variety. Third, the H variety is do not necessarily lead to its development. In fact, it is more highly valued than the L variety: although there is possible that diglossia may develop from other origins. a sizeable body of literature written in the H variety, the Diglossic situations are different from other com- L variety is rarely used in the written form except in monly found language situations in several respects. 268  DISCOURSE ANALYSIS In contrast to a diglossic situation, a bilingual situation diglossia. Although the Spanish-Guaraní situation in does not maintain a clear functional compartmental- Paraguay resembles, say, the diglossic situation in ization of the two varieties. In many Arabic-speaking Arabic-speaking Syria, the two differ from each other countries (one of Ferguson’s canonical examples), for in terms of the social history of the code matrix and instance, although colloquial Arabic serves as the the social processes that led to their emergence. While basic medium of interaction, Standard Arabic is still the former came into being as a result of colonial con- the preferred variety for formal purposes. However, in tact—in other words, the confluence of two independ- a bilingual community such as Flemish- and French- ent sociolinguistic traditions, the latter was derived speaking Belgium, both varieties are used to perform from the internal functional differentiation within a similar functions in formal and informal domains. single sociolinguistic tradition. Furthermore, when Furthermore, while a diglossic situation involves two language shift occurs in a bilingual community such as genetically related language varieties (at least accord- Paraguay, it is usually the more prestigious variety or ing to Ferguson’s (1959) definition), two or more the H variety that replaces the L variety. On the other unrelated languages often coexist in a bilingual or hand, in the terminal stages of diglossia (in Ferguson’s multilingual community (e.g. English, Chinese, sense), the L variety often displaces the H variety. Malay, and Tamil in Singapore). Synchronically speaking, the two language situations A diglossic situation is also different from a stan- may appear the same. Diachronically speaking, how- dard-with-dialects situation. In a community with stan- ever, they are rather different. dard-with-dialects variation, there are those who use the In recent years, scholars have called for a shift in standard language (which is usually the more highly focus in research on diglossia. Some suggest more valued variety) in everyday conversation and speak it as cross-community studies that examine the origin and their mother tongue. In a diglossic community such as the development of various diglossic situations. German-speaking Switzerland (another one of Another kind of productive research, as Ferguson Ferguson’s examples), the H variety (Hochdeutsch) is pointed out in a more recent article, would be studying only learned through formal schooling and is not used a given diglossic situation during a period of rapid as the medium of everyday interaction. On the other social and linguistic change. hand, in Italy (a standard-with-dialects situation), a considerable number of people speak standard Italian natively and use it in formal as well as informal settings. References Ferguson’s article has spurred interest in and stim- Britto, Francis. 1986. Diglossia: a study of the theory with ulated research on this particular type of linguistic application to Tamil. Washington, DC: Georgetown phenomenon. In particular, there have been numerous University Press. efforts over the years to rework Ferguson’s definition Ferguson, Charles. 1959. Diglossia. Word 15. 325–40. Ferguson, Charles. 1991. Diglossia revised. Southwest Journal of diglossia. In particular, while maintaining the strict of Linguistics 10. 214–34. functional compartmentalization of the two varieties, Fishman, Joshua. 1967. Bilingualism with and without diglos- Joshua Fishman broadened the definition of diglossia sia; diglossia with and without bilingualism. Journal of to include genetically unrelated varieties. Thus, Social Issues. 23. 29–38. according to this broad definition, Spanish- and Hudson, Alan. 1991. Toward the systematic study of diglossia. Southwest Journal of Linguistics 10. 1–22. Guaraní-speaking Paraguay would be classified as a Timm, Lenora. 1981. Diglossia old and new: a critique. diglossic community, in that the two genetically unre- Anthropological Linguistics 23. 356–67. lated varieties function like H and L varieties in ANDREW WONG diglossic situations. However, some have criticized that this definition dilutes the original meanings of See also Ferguson, Charles Albert Discourse Analysis The term discourse analysis was introduced by Zellig beyond the sentence, examining what language ele- Harris in 1952. In an effort to analyze connected ments might occur next to each other, or in the same speech and writing, Harris took language description linguistic environment. 269 DISCOURSE ANALYSIS Originating in the field of sociology in the early and ‘description’, ‘procedure’, and ‘exposition’-type 1960s, conversation analysis explores norms and pat- texts. In particular, researchers use typical organiza- terns in spoken interactions. Conversation analysis tional structures and the language that is typically used pays particular attention to everyday interactions such in the realization of particular genres, to determine the as casual conversations, ‘chat’, and ‘ordinary narra- relationship between grammatical and lexical items in tives’. Conversation analysts have, however, also texts. More recent research considers how relations of examined other interactions such as doctor–patient power and ideology shape discourse, and the effects consultations, legal hearings, news interviews, psychi- that discourse has upon social identities, relations, atric interviews, and interactions in courtrooms and knowledge, and beliefs. Critical contrastive rhetoric classrooms. One of the aims of the conversation ana- examines cultural differences in language and commu- lyst is to avoid prior assumptions or speculations about nication, viewing these differences as dynamic and sit- analytical categories in the analysis of their data. uated in relations of power and ideologies. Rather, they look for phenomena that regularly occur A number of aspects of language use considered and then make that the point of further investigation. under the heading of discourse analysis are also often Researchers are particularly interested in the sequen- discussed under the more general heading of pragmat- tial structure and coherence of conversations. ics. Pragmatics is especially interested in the relation- The areas of research known as ‘contrastive rheto- ship between language and context. It includes the ric’ and ‘contrastive discourse analysis’ compare writ- study of how the interpretation of language depends on ten and spoken genres in different languages and knowledge of the world, how speakers use and under- cultures. Kaplan’s work in the mid-1960s is especially stand utterances, and how language use is influenced important here. Kaplan analyzed the organization of by relationships between speakers and hearers. In a rel- English as a second-language students’ essays and atively short time, the field of Discourse Analysis has found different patterns of organization for students examined issues of structure and coherence, while also from different cultural backgrounds. While many stud- placing the discussion within a broader social context. ies in this are focused on academic writing, studies Research provides insights into the organization and have also been carried out that examine genres such as interpretation of spoken and written discourse. economic reports, business letters, and the use of lan- However, what each of these schools of thought reveals guage in everyday interactions from the point of view is, in part, a result of the perspective the researchers of discourse structuring, politeness strategies, and have taken, and the questions they have asked. reader/writer/speaker orientations. The Birmingham model of discourse analysis (see Sinclair and Coulthard 1975) proposed a ‘rank struc- References ture’ view of discourse, where discourse is made up of Chimombo, Moira, and Henry Roseberry. 1998. The power of a number of lower level parts, much as a sentence is discourse: an introduction to discourse analysis. Mahwah, made up of ranked elements such as the clause, phrase, NJ and London: Lawrence Erlbaum. and word. Researchers recorded a number of British Connor, Ulla. 1996. Contrastive rhetoric: cross-cultural aspects primary school classrooms based on the categories of of second language writing. Cambridge: Cambridge University Press. lesson, and proposed a rank structure for the lesson: Coulthard, Malcolm. 1985. An introduction to discourse analy- transaction, exchange, move, and act. They then drew sis, new edition. London and New York: Longman. up rules based on the data to show how the acts com- Coulthard, Malcolm (ed.) 1992. Advances in spoken discourse bined together to form moves, and the moves formed analysis. London and New York: Routledge. various kinds of exchange. The Birmingham approach Fairclough, Norman. 1995. Critical discourse analysis. London and New York: Longman. has since been applied to many different interactional Georgakopoulou, Alexandra, and Doinysius Goutsos. 1997. settings, such as medical examinations, TV quiz shows, Discourse analysis: an introduction. Edinburgh: Edinburgh and everyday conversation. The notion of moves has University Press. also been used in the area of genre analysis where Halliday, M.A.K., and Raquiya Hasan. 1976. Cohesion in researchers such as Swales and Dudley-Evans have English. London and New York: Longman. Halliday, M.A.K., and Raquiya Hasan. 1989. Language, con- examined the organizational structure of academic gen- text and text: aspects of language in a social-semiotic per- res such as research articles, theses, and dissertations. spective. Oxford and New York: Oxford University Press. The Australian ‘genre school’, which originated in Jaworski, Adam, and Nikolas Coupland (eds.) 1999. The dis- Sydney in the 1980s, utilizes the notion of language as course reader. London and New York: Routledge. a system of choices and views on the social functions Kubota, R., and Leaner, A. 2004. Toward critical contrastive rhetoric. Journal of second language writing 13(1). 7–27. of language are important here. Research in this area Swales, John. 1990. Genre analysis. English in academic and has examined spoken and written texts such as service research settings. Cambridge and New York: Cambridge encounters, academic essays, casual conversations, University Press. 270  DISCOURSE STRATEGIES ten Have, Paul. 1998. Doing conversation analysis: a practical See also Coherence in Discourse; Conversation guide. Thousand Oaks, CA and London: Sage. Analysis; Discourse Strategies; Genre; Harris, Zel- Wodak, Ruth. 1996. Disorders of discourse. London and New York: Longman. lig Sabbetai; Politeness; Pragmatics; Relevance in BRIAN PALTRIDGE Discourse; Speech Acts; Text Linguistics Discourse Strategies The concept of discourse strategy is based on a view fairly obligatory, move. From this point of view, and of communication as a purposeful social activity. The given the difficulty in having access to the workings of success of this activity will depend on the application speakers’ minds, it might be better to avoid the issue of of a series of tactics involving linguistic resources (e.g. consciousness and to consider a discourse strategy as sounds, words, grammatical structures, topics, func- a systematic way of using language, which is general- tions) as well as nonlinguistic resources (e.g. gestures, ly recognized by the members of the speech commu- dress, proximity). According to this view, discourse nity as aiming at a particular communicative goal. strategies are all those linguistic moves that competent One way of classifying discourse strategies is to language users make, from several possible choices, in take into account the communicative goal of the order to achieve their aim in what they consider to be speaker. Thus, it is possible to conceive that human the most efficient, effective, and appropriate way. The communication is the result of the speakers’ mutual linguistic choices that language users make in adopt- agreement to abide by two types of constraints related ing a particular communicative move may involve to (1) their social self and (2) the mode of communi- aspects related to form, i.e. how to say something, as cation being used. In this sense, speakers’ deployment well as aspects related to content, i.e. what to say. of a particular discourse strategy should be interpreted Formal choices include pronunciation, grammar, as an attempt to achieve their goal and, at the same wording, and textual organization. Content choices time, attend to one or more communicative con- basically involve the introduction of particular topics straints. Thus, an utterance like Can I have a cup of and functions. coffee? can be interpreted as a means for the speaker Taking into account the cognitive process involved, to obtain a drink while attending to the social con- discourse strategies may be approached from two points straint of avoiding the impression that the conversation of view: (1) as an individual enterprise or (2) as a result partner is being imposed on to carry out the action. of a social, interactional process involving the active Likewise, the use of a signal like yeah or uh-huh can participation of more than one individual. According to be understood as a discourse strategy aimed at coping the first approach, a discourse strategy is an individual’s with the constraint of reciprocity that characterizes implementation of a mental plan in response to an inter- face-to-face conversation. The constraints proposed nal signal that a communicative problem needs to be below should be understood not as precise maxims of solved. The second approach considers that discourse communicative behavior but rather as scales along strategies can be studied as a result of the joint effort which individuals locate their discourse strategies (represented by specific verbal moves) of two or more depending on their cultural values and the situation in discourse participants to reach an agreement on their which they find themselves. communicative and social intentions. The constraints related to the social self that have Besides being problem-based and intentional, dis- been studied in greater detail are (1) presentation of course strategies are often defined as conscious. This self, (2) size of imposition, (3) social distance, and (4) is not a characteristic on which everybody agrees. power. In order to attend to each of these constraints, Thus, it could well be that although at some point a language users can adopt a series of discourse strate- speaker may have consciously decided to make a par- gies that are conventionally accepted as effective ticular verbal move (for instance, responding to a moves. Let us look at some examples. To ‘present’ greeting like How do you do? with the same question), ourselves, we tend to avoid assertiveness by preceding this expression forms part of such a frequent commu- our statements with parenthetical verbs like I think... nicative routine that it has become an automatic, and or I guess... . In order to mitigate the ‘imposition’ of a 271 DISCOURSE STRATEGIES command like Give me a cigarette, we may express it the passive construction to change the information by means of the question Can I have a cigarette? To focus of the utterance or the choice from a wide array attend to the ‘social distance’ or ‘solidarity’ constraint, of connecting words or phrases to express the meaning we can address someone by their first name instead of relationship between one utterance and the following their surname. Finally, it is possible to interpret a one. Other systems of communication (e.g. letters, e- request for permission and its subsequent granting or mail, telephone, visual signals) may have different refusal as a discourse strategy to acknowledge the constraints and consequently may demand different ‘power relationship’ between the discourse partici- discourse strategies. pants. Depending on the context or their personality, the speakers may decide on the way they want to References respond to the constraints and, consequently, on the ‘discourse strategies’ they will adopt. For instance, Brown, Penelope, and Stephen Levinson. 1987. Politeness: given a particular situation, it is not at all impossible to some universals in language usage. Cambridge: Cambridge University Press. think of a speaker who at some point decides to show Calsamiglia, Helena, and Amparo Tusón. 1999. Las Cosas del himself or herself as assertive, imposing, distant, and Decir. Manual de Análisis del Discurso. Barcelona: Ariel. equal in power to the other speaker. Gumperz, John. 1982. Discourse strategies.Cambridge: As for the constraints imposed by the mode of com- Cambridge University Press. munication, in the case of face-to-face conversation, Jaworski, Adam, and Nikolas Coupland (eds.) 1999. The dis- course reader. London and New York: Routledge. we can mention (1) topical coherence, (2) turn-taking, Kasper, Gabrielle, and Eric Kellerman. 1997. Communication and (3) linear structure. To attend to the first con- strategies: psycholinguistic and sociolinguistic perspectives. straint, speakers make use of discourse strategies like London and New York: Longman. by the way, like I said, or anyway to introduce, rein- Stenström, Anna-Brita. 1995. An introduction to spoken inter- troduce, or change a topic, respectively. The turn-tak- action. London and New York: Longman. Titscher, Stefan, Michael Meyer, Ruth Wodak, and Eva Vetter. ing constraint requires the speakers’ deployment of 2000. Methods of text and discourse analysis. London: strategies for taking the turn (e.g. yeah but ..., let me Thousand Oaks and New Delhi: Sage. just ...), holding the turn (e.g. fillers like um or well, Van Dijk, Teun A (ed.) 1997. Discourse as social interaction. repetitions), and yielding the turn (e.g. questions, London: Thousand Oaks and New Delhi: Sage. appealers like you know, right?). Finally, the constraint Van Dijk, Teun A (ed.) 1997. Discourse as structure and process. London: Thousand Oaks and New Delhi: Sage. imposed by the fact that we produce and process lan- JOSEP MARIA COTS guage linearly, that is, one word after another, requires speakers to adopt discourse strategies like the use of See also Discourse Analysis Dravidian A group of 24 languages spoken primarily in the is spoken in the province of Andhra Pradesh (over Indian subcontinent make up the Dravidian family of 75,000,000 speakers), Tamil, the oldest of all Dravidian languages. It is the second largest group of languages languages, is spoken in the province of Tamil Nadu spoken in India, Indo-Aryan, a branch of Indo- (70,000,000 speakers) and Malayalam is spoken in the European, being the largest. The other language province of Kerala (30,000,000 speakers). Irula, groups of the subcontinent are Afro-Asiatic and Sino- Kurumba, Tulu, Badaga, Toda, Kota, Koraga, and Tibetan. Kodagu are minor languages of the southern branch. Out of 24 languages of the Dravidian group, four The other minor languages are spoken in the northern major literary Dravidian languages of the Southern and central parts of the Indian subcontinent. Dravidian group are spoken in the southern Indian Northern Dravidian languages, Brahui, Kurukh, provinces of Karnataka, Andhra Pradesh, Tamil Nadu, and Malto are spoken in the province of Baluchistan in and Kerala. Kannada is the language of Karnataka Pakistan, Bengal, and Orissa in India, Nepal, and (40,000,000 speakers), Telugu, numerically the largest, Bangladesh. 272  DRAVIDIAN Central Dravidian languages are represented by Dravidian words are easily segmented into its con- Gondi, Konda, Kuvi, Pengo, Manda, Parji, Kolami, stituent morphemes; e.g. Kannada maadisikondanu+ + Naikri, Gadaba and Naiki, and Kui. These are nonlit- ‘(he) made (him) (do) (it)’ is composed of maad ‘to + erary languages spoken in the provinces of Andhra do, make’ + is ‘causative’ + kol ‘reflexive’ + d ‘past + + Pradesh, Madhya Pradesh, and Orissa in India. tense’ + an ‘third-person singular masculine’ + u The Dravidian languages form completely a sepa- ‘default vowel’. Dravidian words mostly have only rate group of languages, although Indo-Aryan lan- suffixes and no prefixes. guages have influenced the development of these Nouns in Dravidian inflect for case, number, gen- languages primarily in the domain of phonology and der, and person. Verbs in Dravidian are either finite or lexicon. However, there have been attempts made ear- nonfinite. Finite verbs are inflected for both tense and lier to link Dravidian languages to Finno-Ugric and subject–verb agreement while nonfinite verbs are not. Altaic group of languages. Finite verbs are marked either for past or nonpast Speakers of Dravidian languages have lived in the tense. Indian subcontinent for more than 3,500 years and The default word order in a simple sentence is they have recorded history of more than 2,000 years. It Subject–Object–Verb. Word order is free except that was once thought that Dravidians were native to the the verb has to be in the final position. In a typical Indian subcontinent. But, recent archeological, complex sentence, the matrix clause follows rather anthropological, and ethnographic studies dispute this than precedes its complement; e.g. Kannada meeriyu claim. Based on the types of agricultural and naviga- awanige sullu heelidalu endu (complement clause) ++ + + tional instruments used by these people, it is now jaananu tilididdanu (matrix clause) ‘John thought that + believed that Dravidians must have come from West Mary lied to him’. Similarly, relative clauses precede Africa and the Mediterranean region, and settled in the the noun they modify. Indus river basin in present-day Pakistan around 4000 Dravidian languages use ‘Brahmi’ script, which has BC. The Dravidian civilization that flourished in and an alpha-syllabic writing style with diacritics used for around Harappa and Mohenjedaro region on the banks vowels occurring in postconsonantal position. of the Indus River came to be known as the ‘Indus val- ley civilization’. The period of this civilization is broadly placed between 2500 and 1700 BC. References Excavations done at this site led to the discovery of Andronov, M.S. 1970. Dravidian languages. Moscow. Nauka over 3,000 seals. These seals have logo-syllabic and Publishing House. pictographic writings on them. Attempts have been Bhat, D.N.S. 1978. Pronominalization. Poona: Deccan made to decipher these writings with very little College. success so far. Bhat, D.N.S. 1998. The adjecival category. Amsterdam and Philadelphia: John Benjamins Publishing Company. Burrow, T., and M.B. Emeneau. 1961. A Dravidian etymologi- Typological Characteristics cal dictionary [DED]. Oxford: Clarendon Press. Caldwell, R. 1961. A comparative grammar of the Dravidian or Phonologically, the Dravidian consonant system con- South Indian family of languages. Madras: University of sists of obstruents (p. t. t. J, k); nasals (n, n, ñ, m); lat- + + Madras. erals (l, l); the flap (ɾ (D)); semivowels (y, w); glottal Emeneau, M.B. 1962b. Dravidian and Indian linguistics. + Berkeley: Center for South Asian Studies, UC (h), and voiced retroflex continuant (ɾ). The presence + (Miimeographed). of retroflex series ( t, n, l, r) is characteristic of the + + + + Emeneau, M.B. 1967b. Dravidian linguistics, ethonology and Dravidian consonant inventory. folktales: collected papers. Annamalainagar: Annamalai Dravidian has five vowels (i, e, o, u, a) and the University. vowel length is distinctive. Word stress is predictable. Krishnamurthy, Bh. 2003. The Dravidian languages. London: Cambridge. If a word has short vowels, the initial syllable receives Masica, Colin Paul. 1976. Defining a linguistic area: South the primary stress. However, if a word has a syllable Asia. Chicago: Chicago University Press. with long vowels (VV), then the syllable with long Steever, Sanford B. 1988. The serial verb formation in the vowels receives the primary stress. The syllable pat- Dravidian languages. tern is basically of Consonant–Vowel (open syllable) Steever, Sanford B. (ed). 1988a. The Dravidian languages. London and New York: Routledge. type except that sonorants can close the syllable in Zvelebil, K.V. 1990a. Dravidian linguistics: an introduction. some languages. If a root word/stem ends with an Pondicherry Institute of Linguistics and Culture. obstruent, a default vowel (u) is added to the end of the Zvelebil, K.V. 1970b. Comparative Dravidian phonology. The word; e.g. Kannada kaa d. root word for ‘forest’, is + Hague: Mouton. pronounced as kaadu. + CHANDRA SHEKAR 273 DUTCH Dutch Dutch is one of the Western Germanic languages, a survived. Even from the time before 1300, more than branch that belongs to the Germanic subgroup of the 2,000 texts have been preserved. A large number of Indo-European family of languages. It stands about these texts are to be found in the Corpus Gysseling, midway between English and German (both geo- which is a standard edition of Middle Dutch texts from graphically and lexically) and is the closest to English the period before 1300. of all the major languages of the Germanic family. In the sixteenth and seventeenth centuries, a new Dutch is an official language in the Netherlands and linguistic consciousness developed in the Netherlands. Belgium. It is the mother tongue of more than 21 mil- During this period, Modern Dutch began to form on lion people in the Netherlands, in several provinces of the base of the dialect of the Amsterdam region after it Belgium called Flanders (Ghent, Brabant, Antwerpen, had become the capital of an independent nation. In and Limburg), in the former Dutch colonies in 1584, the first Dutch grammar appeared, written by America (Surinam) and Asia (Indonesia), and in the the rhetorician H.L. Spieghel, the so-called Twe- Netherlands Antilles in the Caribbean Sea. spraack vande Nederduitsche Letterkunst, ‘Dialogue In Belgium, the language is called Flemish of Dutch Letter Craft’, presented in the form of a dia- (Vlaams), although there is practically no difference logue (tweespraak). compared with Dutch in the Netherlands. Dutch and Modern Standard Dutch is the direct heir of the Flemish are actually the same language, but because of dialects spoken in the provinces of North and South long-lasting cultural and religious reasons, there are Holland and Utrecht. Starting in the nineteenth centu- two separate terms for one and the same language. The ry, this standard Dutch increasingly replaced the Flemish areas have been subject to French influence, dialects of other provinces, even in oral use, especial- especially from 1830 onward, when the state of ly among the more cultivated classes, and it was also Belgium came into being with French as the sole offi- adopted in Belgium. cial language. In 1921, Belgium was divided into two Despite the rather limited geographical distribution linguistic sectors. Flemish became the official lan- of the Dutch language, there are a wide variety of guage in East Flanders, West Flanders, Antwerp, regional dialects whose mutual intelligibility is often Limburg, and eastern Brabant, and French became the low. Traditionally, linguists divided Modern Dutch official language in all the other provinces. This lin- dialects into six large groups: (1) the central-western guistic frontier was adjusted in 1962 and 1963, and the dialects, including all those in the provinces of North two principal areas were made formally unilingual, but and South Holland and Utrecht, large parts of the Brussels region has remained bilingual (the Gelderland, and the Zeeland Isles; (2) the northeastern Belgian capital was originally a Dutch-speaking dialects (often called ‘Saxon’) in Groningen, Drenthe, town). However, the requirement for unilingualism Overijsel, and the eastern part of Gelderland; (3) the applied only to the so-called buitendienst ‘foreign central-southern dialects in the Netherlands province affairs’, i.e. it became compulsory in communications of North Brabant and adjacent parts of Limburg, and with officials and in official life in general, with a free in the Belgian provinces of Antwerp, Brabant, and choice for ‘domestic affairs’. East Flanders; (4) the southwestern dialects in the The ancestor of Dutch was the West Low Belgian province of West Flanders, French-Flanders, Franconian dialect of Germanic tribes. The history of Zeeland, and the islands of Goeree and Overflakkee in the Dutch language (whose origin has been set by lin- the province of South Holland; (5) the southeastern guists at approximately 700 CE) may be divided into dialects (often called ‘Eastern Low Franconian’) in three main periods—Old, Middle, and Modern Dutch. Belgian and Dutch Limburg and some villages in north Old Dutch extends to approximately 1150; there are Brabant; and (6) the northwestern dialects in North few direct records of this language, but the only Holland above the IJsselmeer, the non-Frisian Wadden important extant monument of this period is a transla- islands, the coastal stretch of Holland province, and tion of the Psalter. Middle Dutch extends from 1150 to the South Holland islands apart from Goeree and 1500 and is characterized by a significant influence Overflakkee. from Latin and French and by the first Dutch diction- Dialects that are geographically widely separated aries. In contrast to Old Dutch, a large number of from each other sometimes show very significant dif- Middle Dutch texts, both literary and official, have ferences: a speaker of the Groningen dialect will 274  DUTCH scarcely be able to communicate with a speaker of pronunciation’; i.e. generally Dutch orthography West Flemish, unless they both speak standard Dutch. closely reflects pronunciation. For most Dutch and Flemish speakers, the most evident Most loanwords come from English, not just as a division is into northern and southern dialects—to the matter of new technical concepts, but also of loans in South of the ‘great rivers’ (the Rhine and the Meuse), varied areas of everyday life. In many cases, the people use a ‘soft g’ that is not found in the North. For spelling and pronunciation of the English words are dialectologists, the contrast between East and West is adapted to fit Dutch usage. There are also many more important because the western dialects still show German loanwords in Dutch. Some of them are a number of Coastal Germanic characteristics. already so old and so adapted to Dutch usage that most The Modern Dutch vowel system is the result of a people do not even realize that the words are of great number of changes, but the Proto-Germanic con- German origin (e.g. tijdschrift ‘newspaper’). Other sonant system has generally survived in Dutch. German loan-words have managed to retain their orig- The history of the Dutch language witnessed con- inal sound and spelling, e.g. sowieso ‘anyway’. stant simplification in the domain of word structure. Nowadays, the German influence has weakened The noun system in Middle Dutch displayed three because of the strong influence of English. Alongside genders: common, neuter, and feminine. Today, Dutch many French loanwords, there are Dutch equivalents nouns may be either common gender nouns or neuter that could not be driven out completely and are now nouns, with the former taking the definite article de used in parallel. However, the French word is often and the latter het: de man ‘the man’, de vrouw ‘the preferred: jus d’orange/sinaasappelsap ‘orange juice’. woman’, het kind ‘the child’, het huis ‘the house’. The Dutch had contacts mainly within the Germanic feminine gender has been lost. group of Indo-European languages. Contact of Dutch In the pronominal system, most pronominal cate- and native languages gave rise to creoles in the Dutch gories make the distinction between two forms accord- East Indies (as the so-called language of the Sinyos ing to the nature of the concepts they stand for. and Nonahs ‘gentlemen and ladies’), in the former Generally, there is a ‘neutral’form, used for singular Danish Antilles (whose colonists were mainly nonhuman referents, and one used for human refer- Dutchmen), and in South Africa. ents, either plural or singular. Personal pronouns are A language that has evolved directly from Dutch is the only Dutch words that still display a clear opposi- Afrikaans, one of the two official languages of the tion between subject and object forms, i.e. Dutch pro- Republic of South Africa, the other being English. It is nouns distinguish between nominative and accusative spoken by more than 4 million people of the various case. However, there is no difference between ‘accu- ethnic groups found in the country. Afrikaans is also sative’ (direct object) and ‘dative’ (indirect object) one of the two official languages of Namibia, a former uses of the object forms. protectorate of South Africa. Dutch word order is generally subject–object–verb; Like Dutch, Afrikaans is also a member of the i.e. in basic main clauses, the verb is always in the sec- Western Germanic languages. It developed in the seven- ond position, and it usually follows the subject. The teenth century from Dutch brought to South Africa by verb appears at the end of the clause in subordinate the first settlers from Holland. Originally, it was a pop- clauses. It comes at the beginning of the clause in ular dialect composed of Dutch with several borrowings yes/no questions and imperative clauses and in main from aboriginal languages of Africa (especially from the clauses preceded by a subordinate clause. Khoisan family and the Niger-Congo family). By 1900, The Dutch language has known a long literature tra- Afrikaans was emerging as an independent language, dition, and its lexicon remains quite rich. Moreover, it and in 1925 it was finally recognized as an official lan- is quite pure, because Dutch did not borrow many guage, having been used in the schools and church since words from other languages. The languages that influ- 1914 and permitted for use in universities since 1918. enced Dutch over the centuries were French and its At the level of pronunciation, there is an apprecia- northern dialects (especially Picardian), mainly in ble difference between Afrikaans and Dutch. Where Belgium, and classical and medieval Latin. Words of Dutch uses oo, Afrikaans often has eu, and Dutch a Germanic origin form the overwhelming majority of and aa were replaced by e. Afrikaans is almost the the noncomplex part of the basic vocabulary. only Germanic language that uses nasal vowels, espe- The Dutch spelling system (orthography) is based cially before s. on the rules devised and published by the Dutch lin- The Afrikaans noun, displaying a pattern similar to guists De Vries and Lammert Allard Te Winkel in the Dutch noun, fully lost the gender distinction. 1864 and was adopted officially both in Belgium and The Afrikaans verbal system is characterized by the in the Netherlands in the second half of the nineteenth complete loss of person and number. In order to indi- century. Its authors had used the criterion of ‘received cate tense, Afrikaans uses auxiliary verbs and adverbs. 275 DUTCH During its history, Afrikaans was significantly influ- reference to Middle Dutch. Amsterdam; Philadelphia: John enced by the Zulu, Bushmen, and Gottentote lan- Benjamins. Donaldson, Bruce. 1997. Dutch: a comprehensive grammar. guages of Africa, mainly with respect to vocabulary. London, New York: Routledge. The drastic simplifications in word structure may also Konig, Ekkehard, and Johan van der Auwera (eds.) 1994. The be the result of lengthy contact with African, Germanic languages. London, New York: Routledge. Portuguese, and Malayan languages as well. Koster, J. 1975. Dutch as an SOV language. Linguistic Analysis 1.111–36. Lippi-Green, Rosina L., and Joseph C. Salmons (eds.) 1996. Germanic linguistics. Amsterdam,Philadelphia: John Benjamins Publishing Company. References Shetter, William Z. 1984. Introduction to Dutch: a practical Brachin, Pierre. La langue néerlandaise. 1977. Brussels: Didier; grammar, 5th edition. Leiden: M. Nijhoff. as The Dutch language: a survey, translated from the French LAURA DANILIUC by Paul Vincent. Cheltenham: Thornes, 1985. Burridge, Kate. 1992. Syntactic change in Germanic: a study of See also Belgium; Indo-European 2: Germanic some aspects of language change in Germanic with particular Languages Dyslexia There are two broad categories of dyslexia: acquired lower than their IQ scores (about a 75% difference) and and developmental. ‘Acquired dyslexia’ is a failure of who have at least the minimal intelligence needed to the ability to read that sometimes follows a stroke or benefit from instruction in reading (e.g. an IQ of 80). other condition that causes injury to the brain. Lesions However, discrepancy definitions are problematic. For of the parietal lobe and occipital regions, in particular, example, a reader who has a high IQ but reads only sometimes result in severe reading and/or writing dif- moderately well would deviate from expectations of ficulties. Often, there is some recovery, particularly if what should have been achieved, given his or her remedial training is given. The more common use of strong intelligence. Such a person would be classified the term ‘dyslexia’ refers to developmental reading as dyslexic by this formula in spite of being function- disability. ‘Developmental dyslexia’ is the failure to ally literate. Further, definitions based on discrepancies learn to read despite receiving instruction that is nor- between reading achievement and IQ have little theo- mally adequate. ‘Dyslexia’ has sometimes been retical or practical significance. It is adequate to define applied to people who are nearly incapable of reading dyslexia only in terms of the reading performance itself the simplest text. However, the term is more often used relative to reading performance norms for the reader’s inclusively, of individuals who may read but do so age (omitting any reference to an IQ discrepancy). with difficulty in comparison to their peers. A dyslex- A persistent myth is that dyslexics see letters and ic, then, is someone whose level of reading ability, as words backward. This belief is without foundation. measured by standardized tests, is substantially below Although reversals can be observed in many beginning the expected level for his or her age and level of readers, these reversals are caused by faulty learning, schooling. For example, a child whose spoken lan- not by disordered visual perception. Few of these chil- guage is English and who is of adequate intelligence dren become dyslexic. Current understanding of and education but who reads below the 25th percentile dyslexia is that it is not primarily a visual but a lan- on a standardized reading test may be considered guage-based disorder, particularly a disorder of sound dyslexic. perception and production. In order to read print in an A precise definition of dyslexia depends on whether alphabetic writing system (such as English, Russian, or it is based only on a reader’s performance relative to Arabic—but not Chinese or Japanese), good readers his or her peers (as in the example above), or on a dis- must be able to read words they have never seen crepancy between the individual’s reading perform- before. In order to do this, they must understand the ance and his or her general intellectual functioning. As alphabetic principle: that letters correspond to sound an example of the latter, a conservative definition of units or segments, the fundamental components of dyslexia has been suggested for individuals whose speech. To do this requires, in turn, an ability to under- reading test scores are more than 1.5 standard errors stand that spoken words are sequences of those sound 276  DYSLEXIA segments. With this understanding, a reader can ‘sound believed to be involved in sound processing, visual out’ (or ‘decode’) a printed word, thereby identifying processing, and the coordination of the two. In spite of it. Children who are dyslexic nearly always lack an congenital individual differences in neurobiology, ini- ability to understand consciously the segmental nature tial studies suggest that children who are identified of the spoken word. Tests designed to assess a child’s early with neurological deficits in brain areas affecting ability may ask questions like, ‘Can you say monkey their reading ability can be helped. Most useful is an without the mmmm?’ or ‘Which word has more sounds intensive program of reading instruction that empha- in it, car or cart?’ The child who cannot respond sizes the segmental nature of the spoken word and its appropriately is not able to consciously analyze the correspondences to the writing system. This general spoken form. Such a child is at an obvious disadvan- approach is popularly called ‘phonics’, although, with- tage in learning spelling-to-sound correspondences. in that rubric, there are many different methods that Dyslexia is less prominent when the written lan- differ in detail and effectiveness. guage’s spelling represents the spoken word in a sim- ple, unambiguous manner. For example, the writing References systems for Finnish, Russian, and Spanish (to name only a few so-called ‘transparent’ alphabetic orthogra- Fletcher, Jack M., Robin Morris, G. Reid Lyon, Karla K. phies) have an unambiguous relationship between each Stuebing, Sally E. Shaywitz, Donald P. Shankweiler, Leonard Katz, and Bennet A. Shaywitz. 1997. Subtypes of sound unit and each letter of the alphabet. Nearly all dyslexia. Foundations of reading acquisition and dyslexia, children who have adequate intelligence learn to ed. by Benita A. Blachman, Mahwah, NJ: Lawrence decode in these languages. In contrast, so-called Erlbaum Associates. ‘opaque’ writing systems like English and Hebrew have Katz, Leonard, and Ram Frost. 1992. The reading process is a great deal of ambiguity in letter-to-sound relation- different for different orthographies: the orthographic depth hypothesis. Orthography, phonology, morphology, and ships. In English, for example, there are many letters (in meaning, ed. by Ram Frost and Leonard Katz. Amsterdam: particular the vowels), each of which corresponds to Elsevier North-Holland Press. several sounds. For example, the consonant letter g has Liberman, Isabelle Y., Donald Shankweiler, and Alvin M. three different pronunciations in the words giant, get, Liberman. 1992. The alphabetic principle and learning to and night. Moreover, there are many sounds that can be read. In Shankweiler and Liberman. Olson, Richard, Helen Datta, Javier Gayan, and John DeFries. represented by different spellings, e.g. compare the 1999. A behavioral–genetic analysis of reading disabilities identical vowel sound in the words chief, beef, and leaf. and component processes. Converging methods for under- However, even if a child is able to sound out print- standing reading and dyslexia, ed. by Raymond Klein and ed words, she or he will not necessarily become a Patricia McMullen. Cambridge, MA: MIT Press. skilled reader who can read words fluently (i.e. in a Rayner, Keith, and Alexander Pollatsek. 1989. The psychology of reading. Englewood Cliffs, NJ: Prentice-Hall. rapid, automatic manner) and comprehend well what Shankweiler, Donald P., and Isabelle Y. Liberman (eds.) 1992. is read. Even in writing systems that are not opaque or Phonology and reading disability. Ann Arbor: The even alphabetic, nonfluent readers exist. Good com- University of Michigan Press. prehension in reading depends critically on being able Shankweiler, Donald P. 1992. How problems of comprehension to recognize printed words rapidly. Slow word recog- are related to difficulties in decoding. In Shankweiler and Liberman. nition requires so much of the reader’s attention and Shaywitz, Sally A., Bennett A. Shaywitz, Kenneth R. Pugh, short-term memory that there is little left over to Pavel Skudlarski, Rob K. Fulbright, R.Todd Constable, devote to the cognitive processing demands of com- Richard A. Bronen, Jack M. Fletcher, Alvin M. Liberman, prehension (which includes understanding both gram- Donald P. Shankweiler, Leonard Katz, Cheryl Lacadie, matical information and meaning). For alphabetic Karen Marchione, and John C. Gore, 1996. The neurobiolo- gy of developmental dyslexia as viewed through the lens of writing systems, slow decoding is seen as the primary functional resonance imaging technology. Neuroimaging, bottleneck that blocks good comprehension. ed. by G. Reid Lyon and Judith M. Rumsey. Baltimore: Current evidence from brain imaging research indi- Brookes Publishing. cates that dyslexics are deficient in certain brain areas LEONARD KATZ 277  E Emeneau, Murray Barnson Growing up in Lunenberg, a linguistically unique Thoroughly versed in Greek, Latin, Sanskrit, and small town in Nova Scotia, studying French, German, comparative Indo-European at Dalhousie University, Latin, and Greek, it is easy to comprehend how a Halifax, Nova Scotia, Oxford University, and Yale, bright Canadian lad was destined to become a linguist. Emeneau brought with him a superb background for his Combining a background in the classics, general lin- (1931) Ph.D. dissertation, an edition of a Sanskrit folk guistics, and anthropology, Murray Barnson Emeneau tale text. His rigorous training under teachers Franklin became one of the greatest linguistic Indologists of all Edgerton and Edgar H. Sturtevant, together with his time. A prolific author, no linguist interested in India experience in linguistic fieldwork, led to his innovative can afford not to read him, while every general linguist research of the Indian linguistic area — a topic that was will profit from his lucid prose, no matter what the totally neglected. Although the notion of the field of specialization. Among his most noteworthy Sprachbund was known from the Caucasus, the contributions are his work on the Indian Sprachbund Balkans, and the northwest coast of the United States or ‘linguistic area’, bilingualism and structural bor- (American Indian languages), Emeneau’s pioneering rowing, linguistic prehistory, as well as Dravidian investigation of the Indian scene describes a situation in comparative linguistics and etymology. which ‘languages belonging to more than one family Emeneau’s overall approach to linguistics was show traits in common which do not belong to the other influenced by the Boasian and Sapirian traditions, members of (at least) one of the families’ (Emeneau which consider it a major branch of cultural anthro- 1980:127). Franz Boas and Edward Sapir had discussed pology. His anthropological background was largely the problems of the diffusion of linguistic traits across self-taught through voluminous reading, auditing genetic boundaries, using data from various American courses taught by Edward Sapir while still at Yale, and Indian phyla, and Emeneau used their methodology to associating with him from 1931 to 1935, and through investigate the historical relationships of the three major three years of fieldwork on nonliterary languages in subcontinent language families coming in contact: British India. Since academic positions were hard to Indo-Aryan, Dravidian, and Munda (Austro-Asiatic). secure during the years of the Great Depression and Lexical borrowing has long been recognized, and were virtually nonexistent for Indologists, Emeneau all the Dravidian languages have loan words from went off to India on a fellowship from Yale, the Sanskrit and later Indo-Aryan languages. The Munda American Council of Learned Societies, and the languages have been influenced by Indo-Aryan vocab- American Philosophical Society. There, he did field- ulary, and Indo-Aryan languages have borrowed work from 1935 to 1938 in the Nilgiris on three little- Dravidian words. The relationship among the three known Dravidian languages: Toda, Kota, and Kodagu, families is a direct reflection of the linguistic history as well as in central India, where he studied Kolami. and prehistory of India. 279 EMENEAU, MURRAY BARNSON Two phonological features used by Emeneau are neglected family, Emeneau traveled to England in illustrative of the linguistic diffusion described above. 1949 and proposed a comparative Dravidian diction- Most Indian languages have a set of retroflex conso- ary to England’s leading Indologist, the late Thomas nants that contrast with dentals. A situation of pre- Burrow of Oxford University. In many ways, Indo-Aryan and pre-Dravidian bilingualism allowed Emeneau’s magum opus, the coauthored Dravidian pre-Indo-Aryan allophones to diffuse as redistributed etymological dictionary (Oxford: Clarendon Press), retroflex phonemes. Furthermore, Emeneau demon- first published in 1961, 2nd ed. in 1984, capped a long strated that the Peshawar dialect of (Iranian) Pashto and as yet unfinished distinguished career. has been Indianized in the occurrence of the retroflex- es as well as in borrowing many words from modern Biography Indo-Aryan. Since the retroflexes are Proto-Dravidian and not Proto-Indo-European, this is a clear case of Murray Barnson Emeneau was born in Lunenburg, phonological influence, or as Emeneau put it, ‘the Nova Scotia on February 28, 1904. He received his BA Indianization of the Indo-European component’ (1923 Dalhousie University and 1926 Oxon.), MA (1980:111). (1935 Oxon.), and Ph.D. (1931) from Yale University. Another example of phonological influence con- He was a Rhodes Scholar in 1923–1926; Governor- cerns the palatals. In Marathi, the palatals of Old Indo- General’s Gold Medallist 1923; Guggenheim Fellow Aryan affricated into  and  before front vowels and in 1949–1950, 1956–1957; Faculty Research Lecturer, ts and dz before back vowels. As a similar develop- UC Berkeley in 1955–1956; and Wilbur Lucius Cross ment occurs in Indo-Aryan Oriya, in two Dravidian Medallist, Yale Graduate School in 1969. He was an languages, Telugu and Kannada, and in the Munda instructor of Latin in Yale in 1926–1931; he carried language Kurky, one can conclude that all of these lan- out anthropological linguistic research in India on guages are in contact and form a continuum across Dravidian languages in 1935–1938. He was Assistant central India. Professor of Sanskrit and General Linguistics at UC One of the most interesting phenomena used by Berkeley in 1940–1943; Associate Professor, Emeneau to illustrate areal diffusion is the case of 1943–1946; Professor, 1946–1971; and Founder and numeral classifiers. The research demonstrates that Chairman, Dept. of Linguistics, UC Berkeley, 1953. Indo-Aryan is closer to Dravidian than to Indo- He received honorary degrees (L.H.D.) from the European. Further, the Magadhan languages of mod- University of Chicago in 1968; LL.D. from Dalhousie ern Indo-Aryan seem to be the origin of all Indian University in 1970; D.Litt. (Honoris Causa) from classifier systems. University of Hyderabad, India in 1987; and an Two other areal traits deserve mention. First, Honorary Degree, Vidyavachaspati from K.S.D. Dravidian and Indo-Aryan have similar onomatopoeic Sanskrit University, Darbhanga (India) in 1999. He systems, each with stems that occur in reduplicated was President, Linguistic Society of America in 1949; and nonreduplicated forms and verbal derivations Editor, Journal of the American Oriental Society therefrom. Since it is doubtful that the Indo-Aryan (JAOS)1941–1951; President, AOS, 1954–1955; system derives from that of Proto-Indo-European, Honorary Fellow Royal Asiatic Society in 1969; Dravidian influence is likely, which would account for Fellow of the American Academy of Arts and Sciences the Rigvedic occurrences. Second, the North Indo- in 1970; Corresponding Fellow, British Academy in Aryan verbal system of intransitive, transitive, and 1993; Honorary Member Linguistic Society of India causative is reminiscent of the Dravidian causative 1964 (winner Citation and silver plate, Golden Jubilee system. of Society, 1978); and Honorary Member, The As for Emeneau’s comparative Dravidian work, the Philological Society in 2000. He also received a Medal genetic relationship among the members of this close- of Merit, AOS, in 2000. ly related family (more like Romance than like Indo- European [1980:74]) was of interest to him from the References very beginning, and etymologies were easy to discern. Emeneau, Murray B. 1944–1946. Kota texts. Emeneau correctly theorized that the relationship ———. 1951. Studies in Vietnamese (Annamese) grammar. among Latin, English, and Proto-Indo-European was ———. 1955. Kolami: a Dravidian language. mirrored by that of Kolami, Telugu, and Proto- ———. 1971. Toda songs. Dravidian. Emeneau’s research on the Kolami lexicon ———. 1980. Language and linguistic area. may serve as illustrative: 720 words had good ———. 1984. Toda grammar and texts. ———. Coauthor (with Thomas Burrow). 1961, 1968. A Dravidian etymologies; however, 166 of these were Dravidian etymological dictionary, 2nd edition 1984. found to be borrowings from Telugu. Desiring to pro- ———. 1994. Dravidian studies: selected papers. duce a definitive etymological dictionary of this ———. 1980. Author’s postscript. Emeneau, pp. 350–4. 280  EMOTION AND LANGUAGE ———. 1991. A Nova Scotian becomes a linguistic Indologist. Van Nooten, Barend. 1988. Sanskrit Studies of M. B. Emeneau: First person singular, Vol. II, ed. by E. F. K. Koerner, Selected Papers. 85–101. Amsterdam: John Benjamins. Krishnamurti, Bhadriraju. 1969. Comparative Dravidian stud- ALAN S. KAYE ies. Current trends in linguistics, Vol. V, ed. by Thomas A. Sebeok, 309–33. See also Dravidian; India; Sprachbund Emotion and Language There is no established definition of emotion. The latter type can be illustrated by examples such as term is usually used to refer to feelings, such as joy, mother, brother, friend, or hometown. Furthermore, jealousy, or disappointment, but it is also used to refer utterances may evoke an emotion in the addressee; e.g. to more diffuse types of moods, such as irritability or a derogatory remark may cause an outburst of fury. stress. Furthermore, it is not clear how many emotions Emotions can also be expressed paralinguistically. exist. Many disciplines have addressed the issue, and Relevant paralinguistic features involve higher pitch, results diverge considerably. Some accounts from psy- increased tempo, a trembling or a creaky voice, groan- chology list more than 500, but other accounts assume ing, giggling, and laughing. It is worth noting that dif- a much smaller number. ferent ways of laughing may express different Expressing emotions is one of the most fundamen- emotions, e.g. an embarrassed laugh, a derisive laugh, tal functions of language. This function has been or relieved laughter. termed ‘emotive’ (or ‘expressive’). Although it may Problems in verbal planning and delivery can also not be the dominant function in many cases, the emo- signal emotion. Examples include hesitation, repeti- tive function affects virtually all utterances. tion, stammering, and unfinished sentences. On the Emotions may be expressed at all linguistic levels. level of turn-taking and discourse organization, phe- Semilinguistic noises, such as wow, ouch, or tut-tut, are nomena (potentially) indicative of emotion involve traditionally classified as interjections. Interjections are interruptions, intentional overlap, simultaneous considered purely emotive signs. They do not have any speech, and—as an extreme form—shouting some- cognitive meaning. A conventional feature on the pho- body down. On the level of topic selection and topic netic level is the emphatic prolongation of a vowel, as treatment, emotions may motivate the choice of taboo in [bi:g] for big, as opposed to neutral [big]. Intonation topics, or speaking the plain truth. and other phenomena, such as word stress, also play an Emotions can be communicated not only linguisti- important role. On the word level, the emotive function cally and paralinguistically but also nonverbally. is expressed, for example, by the so-called diminutive Nonverbal behavior of this type includes, first and suffixes and familiarity markers, as in doggie, weirdo, foremost, facial expression. Contrary to popular belief, fats, and babykins. Other word-formation devices facial expression conveying a particular feeling is not include reduplication, e.g. teeny-weeny, okey-dokey. universal, but culture specific. Emotion can also be expressed by stylistically marked Apart from studying verbal manifestations of emo- word choices, e.g. bean for head. Further lexical means tions, linguistics also investigates the way we speak include terms of endearment and swear words. On the about emotions: when, how, and why humans talk sentence level, we find intensification and emphatic about emotions and how emotions are negotiated in repetition (e.g. This is very very ugly, He worked and discourse. Emotion words and emotion concepts are worked) and exclamations such as What a sight! On the the focus of semantic analysis. discourse level, speech acts such as insults are impor- Every language has an emotion lexicon, i.e. a set of tant means of communicating emotions. Ritualized words that function as labels for emotions. These are communicative patterns are used to react to social stan- used to name and communicate about emotions. The dard situations concerning illness, death, birth, etc.; i.e. emotion lexicon consists predominantly of nouns, e.g. they are expressions of grief, sadness, commiseration, fear, disgust, and pride, and adjectives, e.g. sad, or happiness. angry, and happy. A distinction can be made between words that Emotions can be positive, e.g. joy, love, or negative, express emotions and words that evoke emotions. The e.g. anxiety, shame. Many emotion words, both nouns 281 EMOTION AND LANGUAGE and adjectives, form antonymous pairs, i.e. pairs of The way we speak about emotions reveals the way opposing emotions. Examples include joy–sorrow, we think about them. Cognitive semantics demonstrates love–hate, happy–sad. Some emotion words have more that we use metaphors for conceptualizing emotions. than one antonym, e.g. joy–sorrow/grief, and some Emotion concepts are reflected in proverbs, idioms, and emotion adjectives have two opposite partners, one that collocations, such as, for instance, You make my blood is related in form, e.g. happy vs. unhappy, and one that boil or He’s just letting off steam. Such expressions is related only in meaning, e.g. happy vs. sad. illustrate that the central metaphor for anger is ‘heat of Antonyms related in form are the result of negation, i.e. fluid in a container’. Metaphorical expressions of other of adding a negative prefix (e.g. un-). In this respect, the emotions show that the human body is generally seen as adjectival emotion lexicon displays certain asymme- a container for emotions, as in She was filled with love, tries. For instance, happy can be negated (unhappy), or Her feelings overflowed. To a large extent, however, but sad cannot be negated (*unsad). As a rule, only emotion concepts are culture specific. positive emotion adjectives can be negated. In summary, linguists examine how we communi- Emotion words form so-called lexical fields. These cate emotions and how we communicate about them. are subsets of the vocabulary of a language whose Emotions can be expressed linguistically, and also par- members are close or related in meaning. Emotion alinguistically, at all levels of language. The emotive words of the same field name the same emotion or very function is one of the most fundamental functions of similar emotions. For example, the word field ANGER human language. The emotion vocabulary is struc- comprises, among other words, anger, rage, fury, wrath, tured in lexical fields, involving basic vs. peripheral ire, irritation, frustration, annoyance, and indignation. words as well as antonymous pairs. There is no one-to- The most typical representative of this field is anger. one equivalence of emotion vocabularies across lan- Less typical members include annoyance and indigna- guages. Emotion concepts are reflected in tion, whereas bitterness and resentment are considered culture-specific metaphorical expressions. only peripheral members. Some members refer to dif- ferent intensities of the same emotion; e.g. rage is more References intense than anger, and fury is even stronger than rage. Athanasiadou, Angeliki, and Elzbieta Tabakowska (eds.) 1998. Annoyance, on the other hand, is weaker than anger. Speaking of emotions. Berlin: Mouton de Gruyter. In the emotion lexicon, some words are considered Fussell, Susan (ed.) 2002. The verbal communication of more basic than others, because some emotions are con- emotions: interdisciplinary perspectives. Mahwah, NJ and sidered more fundamental than others. This has been London: Erlbaum. Kövecses, Zoltán. 1990. Emotion concepts. New York and established empirically in both linguistics and psycholo- Berlin: Springer. gy. The most basic emotions seem to be love, fear, anger, Niemeyer, Susanne, and René Dirven (eds.) 1997. The language happiness, and sadness. When asked to enumerate emo- of emotions. Amsterdam: Benjamins. tions, native speakers of English are most likely to men- Russell, James, Fernández-Dols, José-Miguel, Anthony tion these five. There is a high degree of similarity Manstead, and Jane Wellenkamp (eds.) 1995. Everyday con- ceptions of emotion. Dordrecht: Kluwer. concerning these emotions across languages, including Volek, Bronislava. 1987. Emotive signs in language and seman- not only Western European languages but also typologi- tic functioning of derived nouns in Russian. Amsterdam and cally distant languages. At the same time, no direct cor- Philadelphia: Benjamins. respondence exists between the emotion vocabularies of Wierzbicka, Anna. 1999. Emotions across languages and cul- different languages. For instance, the same lexical field tures. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Sciences de l’Homme. may have a different number of members in different KLAUS P. SCHNEIDER languages. For one English emotion word, there may be two or more possible translations in another language. See also Metaphor Empty Categories The elements of a sentence do not always have to be or because they are easily recoverable from the con- pronounced. This is true in elliptical structures, in text. After a statement such as She did something which elements have been left out for stylistic reasons weird, a possible follow-up question would be She did 282  EMPTY CATEGORIES something weird?, but most speakers would prefer the its interpretation on the subject of the main clause, truncated form She did? Something weird remains as in John promises his mother [PRO to study]. In implied. Some linguists claim that sentences may also the second type, PRO has the same referent as the contain covert or null elements that are not the result object of the main clause: Mary forces her son [PRO of optional ellipsis, but always remain unpronounced. to eat]. Whether PRO is controlled by the subject or These are called empty categories. The reason for by the object depends uniquely on the properties of assuming the existence of such empty categories is the main verb. that they contribute to the meaning of the sentence The fact that PRO is also pronominal means that the and that they can also play an important syntactic interpretation of PRO can be the same as a pronoun. In role. The task of the listener is to identify all these It is better [PRO to take the car], PRO may have the empty categories in a given sentence and to determine same interpretation as they in It is better [that they take their proper function and interpretation. The task of the car]. In sentences like It is impossible [PRO to the linguist goes beyond mere identification: a proper learn the secret code], the meaning of PRO is some- formal theory of language has to determine the spe- what more vague and may be similar to somebody in It cific properties of the contexts where these empty cat- is impossible [that somebody learns the secret code]; egories can appear. PRO is said to have an arbitrary interpretation in such In the early 1980s, Noam Chomsky proposed a types of sentences. In its pronominal use, PRO is not classification of empty categories that are important controlled by an antecedent. for sentence structures. He identified several different The second empty category is ‘small’ pro, the types of empty categories and argued that they can be subject of an inflected verb. In languages such as differentiated according to whether or not they are Spanish and Italian, the pronoun subject can be left anaphoric and whether or not they are pronominal. unexpressed, as in the Italian Ha mangiato ‘(He or Anaphoric elements refer back to elements that have she) has eaten’. There, the implicit subject is repre- been mentioned before and require an antecedent, typ- sented by pro: pro ha mangiato. ‘Small’ pro has to ically in the same sentence. Pronominal elements be present in such sentences because, just like large function in a manner similar to pronouns such as I, PRO, it represents the subject of the sentence and is she, or they, i.e. they may refer to entities that must be thus crucial for its interpretation. Furthermore, it can recoverable from the context but do not need to occur act as an antecedent for anaphors. For example, the within the same sentence. Italian reflexive si ‘self/each other’ in pro si vedono Consider first an infinitive subordinate clause, indi- spesso ‘(They) often see each other’ is anaphoric and cated by square brackets: I want [to leave]. Like all has to rely on another sentence element for its inter- clauses, these must have a subject, i.e. someone who pretation. The only possible antecedent for the leaves. In this case, it is clear to the listener that the reflexive si in this sentence is the null subject of subject of the main clause (I) is also the subject of the verb. leave. However, it is not the case that the speaker The pro phenomenon is far from being unitary: could optionally insert I into the subject position of the sometimes pro acts exactly like a pronoun; it has a subordinate clause; I want I to leave would be ungram- very specific interpretation, as in the above examples. matical, and I want me to leave sounds somewhat Sometimes, however, pro is used nonreferentially, in strange. Thus, the empty subject is not the result of a contexts where the subject appears in postverbal posi- stylistic type of ellipsis. Rather, Chomsky argued, this tion: pro arriva un uomo ‘(There) arrives a man’. In is the case of an empty subject that he called ‘PRO’. these cases, it acts as the nonovert counterpart of the The empty subject of the infinitive has to exist in the English expletive there, i.e. a sentence element that is sentence structure, because it plays a part in the inter- almost void of meaning. The notion of expletive pro pretation of the clause. has been criticized. The most important objection The element PRO has both pronominal and against this element is that it is postulated on purely anaphoric properties. The fact that PRO is anaphor- theory-internal grounds, namely, the assumption that ic means that, for its interpretation, it depends on sentences require a subject in front of the verb. another nominal phrase. In other words, PRO is However, there is no empirical evidence to justify the ‘controlled’ by another nominal phrase. For exam- existence of expletive pro. ple, in John wanted [PRO to leave immediately], the Not all languages use pro. The languages that allow referential properties of PRO depend on those of the the pronominal subject to be unexpressed are called controller, namely John. John can thus be considered ‘pro-drop languages’ or ‘null subject languages’. Since the antecedent of PRO, and PRO is ‘controlled’ by the 1970s, considerable work has been done to deter- John. There are two types of control: subject control mine how pro is formally licensed in languages that and object control. In the first type, PRO depends for use it. Many different hypotheses have been advanced, 283 EMPTY CATEGORIES but the idea present in almost every hypothesis is the nouns do not allow an antecedent within the same ‘richness of the verbal morphology’. It has often been clause. said that pro appears in those languages that have a Another type of trace is that left by question words, rich verbal morphology, i.e. in those languages in which are fronted in English (see Figure 3). What, which the verb inflects for person, number, and gender. here, is clearly not the subject of the sentence: you is If the form of the verb already provides such clues the subject. Rather, what must be interpreted as the about the subject, an empty subject is interpretable. object of done. Thus, in Spanish pro hemos comido ‘we have eaten’, Thus, Chomsky assumed that question formation pro is interpretable because the verbal inflection -emos involves moving the question word to the front of clearly indicates that the subject must be first-person the sentence, again leaving a trace. Because what plural, i.e. ‘we’. acts as the antecedent of the trace, the trace again The third type of empty categories depends on the cannot be pronominal. However, it cannot be assumption that the construction of certain (or all) anaphoric either, because it allows its antecedent to sentences involves moving some elements around. occur beyond the immediate sentence boundary, as For example, sentences like The man has been bitten in Figure 4. by the dog are often claimed to involve movement of The embedded sentence that the famous painter the man from the object position of the verb (see will paint t thus contains a trace without an antecedent Figure 1). within the same clause, something that typical The reasoning is that even though the man is in the anaphoric elements do not allow. grammatical subject position, the passive verb requires In short, there are four types of empty categories, it to be interpreted as the object. Chomsky proposed PRO, pro, traces of regular noun phrase, and traces of that such movement leaves a ‘trace’, usually indicated question words. These four categories can be formally by t (see Figure 2). distinguished as follows: (1) PRO is anaphoric and This type of movement trace is anaphoric, pronominal; (2) pro is pronominal, but not anaphoric; because its interpretation depends on the man. (3) traces of regular noun phrases are anaphoric, but Because the man acts as the antecedent within the not pronominal; and (4) traces of question words are same clause, these traces are not pronominal: pro- neither anaphoric nor pronominal. Twenty years of linguistic research have thus provided a systematic classification of empty cate- gories, and empty categories belong to the stan- has been bitten the man by the dog dard descriptive tools of linguists working in the Chomskyan tradition. However, one must keep in mind that competing theoretical frameworks Figure 1 do without empty categories or movement. Empty categories are thus theoretical metaphors that may or may not have a correspondence in the actual processing of sentences by human speakers and The man has been bitten t by the dog listeners. Figure 2 References Chomsky, Noam. 1981. Lectures on government and binding. Dordrecht: Foris Publications. Chomsky, Noam. 1982. Some concepts and consequences of What have you done t? the theory of government and binding. Cambridge, MA: MIT Press. Haegeman, Liliane. 1991. Introduction to government and Figure 3 binding theory. Oxford and Cambridge, MA: Blackwell; 2nd edition, 1994. Roberge, Yves. 1990. The syntactic recoverability of null argu- ments. Kingston: McGill-Queen’s University Press. Ruzicka, Rudolf. 1999. Control in grammar and pragmatics: a cross-linguistic study. Amsterdam: John Benjamins. Whom do [ you think t [ that the famous painter will paint t ] ]? KAREN LAHOUSSE Figure 4 See also Chomsky, Noam; Generative Grammar 284  EMPTY MORPHEMES Empty Morphemes The concept of morpheme was defined by the first struc- signifier (problems 1–4) and meaning (problem 5). In turalists as the minimum linguistic unit containing signi- order to try to solve them, the linguistic theory has for- fier and meaning, that is, as a formal element with a mulated diverse ones. Among them, we are now inter- content that cannot be divided into smaller segments ested in the following concepts: that, in turn, also possess meaning: preexistence contains On the one hand there is the established distinction three meaningful morphemes (pre-, -exist-, and -ence) between morpheme as the unit of a language, and and cannot be divided into smaller meaningful parts. morph, its phonetic realization. The separation of The morphemes are recognized by their recurrence, these two concepts allows to postulate that a biunivo- that is to say, for their presence with the same mean- cal correspondence among them does not have to ing and the same form in different words: in the previ- exist, in such a way that a morpheme can be repre- ous example, pre- also appears in pre-history or sented by several morphs (called allomorphs; problem pre-establish; exist appears in exist-ent or exist-ential; 1); a morph can comprise more than one morpheme (a and -ence in prefer-ence or compet-ence. phenomenon denominated syncretism; problem 2); the However, some linguists observed that this concep- morphs of two morphemes can merge phonically tion is not always sufficient, because not all the seg- (something usually known as amalgam or morphs ments obtained by means of formal word analysis portmanteau; problem 3); or a morphologic content meet these characteristics. This can be observed in may not correspond to any morph (it is said to be a different situations: zero morph; problem 4). With respect to problem 5, which affects the mean- ing of the morpheme, two positions have been adopted 1. The same content can be present in phonetically that, although they are not always distinguished in the different segments: the regular English plurals linguistic bibliography, is convenient to differentiate: have the forms /s/, /z/ o /iz/ (hits /´hits/, boys In the first place, some authors, adhering to the /´boiz/, houses /´hausiz/), depending on the cir- statement of Leonard Bloomfield that all production of cumstances. the language is completely composed of morphemes 2. A segment may not be formally divided in to (principle of total accountability), would likely affirm smaller units, but it may seem to contain more that elements like those that appear in the series re-mit, than one meaning: in the lat. amo, the segment - de-mit, com-mit, trans-mit; re-fer, de-fer, pre-fer, in- o cannot be formally divided, but it contains the fer; or cran-berry, boysen-berry, huckle-berry, in spite meanings for ‘first person’, ‘present’ and of lacking a meaning definable in linguistic terms, ‘indicative’. they should be considered morphemes, because they 3. Some morphemes are linked to the extreme such appear in a recurrent way in diverse words (the prefix- that the boundary between them disappears, like es de-, re-, con-, etc. and the roots mit and fer are pres- it is the case of the Spanish al (a  el) or the ent in many words of the English language) or simply French au (à  le). because they appear linked to forms that appear in 4. Some morphemic contents appear in certain some other constructions (cran -, boysen-, and huckle- words, although they cannot be assigned to any only appear in these three cases, but berry appears part of the words: the difference between ‘pres- scattered and it combines with other morphemes in ent’ and ‘past’ observed in I beat John every day straw-berry, blue-berry or black-berry). Consequently, vs. I beat John yesterday, does not correspond to these would be morphemes lacking meaning, or empty any formal segment of beat. morphemes. 5. The formal analysis sometimes uncovers recur- Such an interpretation comes from Bloomfield him- rent segments that are not necessarily assigned a self, who would affirm that if a complex linguistic meaning, such as the case of -mit - in the series form contains a residual, it should necessarily be formed by remit, demit, commit… another linguistic form (as in the case of cran-berry), and that forms such as remit or conceive should be These problems affect the two faces that the mor- interpreted as complex words despite the fact that, in pheme, as a linguistic sign, possesses: it affects its them, the meaning of the morphemes is impossible to 285 EMPTY MORPHEMES describe. This affirmation would be adopted by later However, not all the linguists accept these interpre- linguists, as the case of Zellig Harris, Eugene Nida, tations, and there are many who prefer to think of Robert Robins, or Mark Aronoff. words like the ones mentioned in this article as The main consequence of this method of analyzing monomorphic, since the analysis of, for example, data is a substantial modification of the concept of remit into two morphemes can only be based on exclu- morpheme. According to many modern authors, this sively formal considerations. However, in approaches unit can only be defined as the minimum unit of gram- that emphasize the meaning and mechanisms that the matical analysis, that is, a unit that appears recurrent- speaker triggers when using the language, purely for- ly in the words and that allows the differentiation of mal generalizations may not be of interest. some words from others. A paradigmatic example of this idea is the definition by Aronoff: ‘A morpheme is References a phonetic string which can be connected to a linguis- Anderson, Stephen R. Morphological theory. 1988. Linguistics: tic entity outside that string. What is important is not the Cambridge survey (I. Linguistic theory: foundations), its meaning, but its arbitrariness.’ ed. by Frederick J. Newmeyer. Cambridge: Cambridge Opposing this interpretation, Charles Hockett University Press. would speak of the empty morph, a term with which Aronoff, Mark. 1976. Word formation in generative grammar. he designates the portion of phonic material morpho- Cambridge, MA: MIT Press. Bloomfield, Leonard. 1926. A set of postulates for the science logically irrelevant, that is to say, a morph that does of language. Language 2. 153–64. not correspond to any morpheme of the language. For Bloomfield, Leonard. 1933. Language. New York: Holt, example, in the form of the Fox language poon-i- Rinehart & Winston. meewa ‘he stops speaking to him’, the segment -i - Gleason, Henry A. 1961. An introduction to descriptive linguis- lacks semantic content; hence, it should be considered tics. New York: Holt, Rinehart & Winston. Harris, Zellig S. 1942. Morpheme alternants in linguistic analy- a morph that does not correspond to any morpheme of sis. Language 18. 169–80. that language. Harris, Zellig S. 1951. Structural linguistics. London and This interpretation seems identical to the previous Chicago: The University of Chicago Press. one at first glance. In fact, it also alludes to the appear- Hockett, Charles F. 1947. Problems of morphemic analysis. ance in certain words of segments that cannot be Language 23. 321–43. Matthews, Peter H. 1974. Morphology: an introduction to the assigned a meaning, and, as before, it also allows theory of word structure. Cambridge: Cambridge University maintaining the principle of total accountability. Press. However, the conceptual difference among both posi- Nida, Eugene A. 1949. Morphology. The descriptive analysis of tions is large: the first one explicitly refuses that mor- words. Michigan: The University of Michigan Press. phemes necessarily possess meaning; the Robins, Robert H. 1964. General linguistics. An introductory survey. London: Longmans. interpretation of Hockett, on the other hand, defends JOSÉ CARLOS MARTÍN CAMACHO the definition of the morpheme as a unit with meaning, because the segments without meaning are not mor- See also Bloomfield, Leonard; Empty Categories; phemes, but simply, morphs. Harris, Zellig Sabbetai; Morpheme; Structuralism Endangered Languages There are some 6,700 known languages in the world, 20 languages, and eight languages (Mandarin, and linguists estimate that 40–90% of them will dis- Spanish, English, Bengali, Hindi, Portuguese, appear during this century, a situation generating Russian, and Japanese) have over 100 million speak- alarm among linguists and the speakers of languages ers. Ninety-six percent of all languages are spoken by threatened with disappearance. There is a strong cor- just 4% of the population, and approximately one relation between a language’s vitality, or likelihood of fourth of all languages have less than 1,000 speakers. being transmitted from generation to generation, and In North America alone, fewer than 200 languages the number of speakers it has. In this respect, the remain, although there were certainly hundreds of world’s languages are unevenly distributed: approxi- distinct languages several centuries ago. Today, only a mately half the world’s population speaks one of just handful of these (e.g. Cree, Dakota, Ojibwa, Navajo) 286  ENDANGERED LANGUAGES have a hope of survival, and even their security is Members of a language community often believe that doubtful. knowledge of the dominant language is necessary to be In a few cases language loss occurs due to the loss of successful in the dominant culture, and that knowledge the speaker population itself. Language change and loss of the heritage language is a hindrance to socioeconomic are natural, ongoing processes. Latin, for example, was advancement. This is closely linked to the prestige fac- ‘lost’ in the sense that its various dialects evolved over tor: the heritage language often suffers from a general time into the modern Romance languages, such as lack of prestige. There may be no conscious decision to Spanish, French and Italian. Some languages disappear shift from one language to another; rather, the youngest entirely, and it is this absolute language loss that is generation learns the dominant language through contact occurring at an unprecedented rate in modern times. The and, frequently, in the schools. Despite the fact that primary cause is language shift, when speakers cease to speakers abandon their native tongue for what they deem speak their own native tongue in favor of the language of to be pragmatic and/or prestige factors, the majority of what is usually the dominant culture, dominant political- communities that have lost their language subsequently ly, and/or economically. Such a shift from the heritage to experience this as a deep and regrettable loss. the dominant language can occur over several genera- Native communities and linguists alike are respond- tions, or even as quickly as over the course of a single ing to the threat of language endangerment in a number generation. In many cases, the oldest generation, the of ways. Language revitalization programs are being grandparents, speak the heritage language as their first implemented in a number of areas; these aim to (re)edu- and primary language, the middle generation has some cate speakers in their heritage language and promote its knowledge but uses the dominant language primarily, use. Technical support includes the development of ped- and the youngest generation has little to no knowledge of agogical materials, dictionaries, the introduction of the heritage language, and may at most know a few radio and television broadcasts, and the use of the words or phrases. In cases of rapid language shift, how- Internet. Linguists are involved in the recording and ever, these changes occur across a single generation. documentation of endangered languages and the writing Language vitality is usually ranked in terms of num- of grammatical descriptions of the endangered lan- bers and generations of speakers. On the one end of the guages. They often play a key role in the development scale are extinct languages that are no longer spoken, of alphabets and literary standards for previously and on the other end are viable languages in no current unwritten languages. One crucial and often controver- threat of endangerment. In between a number of stages sial issue in the development of many literary languages can be recognized. A healthy language with strong is the selection of the dialect or variety that will serve as vitality is used with a variety of functions and in a the basis for the standard language. Revitalization pro- range of settings (or domains) of language use. This is grams face a number of obstacles, such as a lack of ped- true even for bilingual speakers, where one language or agogical materials and trained teachers, and the lack of linguistic variety is used in personal and home life (e.g. financial resources to implement any of these. at home, with friends, at social events, etc.) and a dif- ferent language or variety is used at the workplace, or References in more public or official settings. In cases of language attrition, however, the native language is used in Crystal, David. 2000. Language death. Cambridge: Cambridge University Press. increasingly fewer domains with fewer functions. Dorian, Nancy C. (ed.) 1989. Investigating obsolescence. The causes of language attrition and loss are com- Cambridge: Cambridge University Press. plex and vary with individual situations. Although sig- Fishman, Joshua. 1991. Reversing language shift: theoretical nificant, the size of the speaker population is not the and empirical foundations of assistance to threatened lan- only reliable predictor of language vitality. Languages guages. Clevedon: Multilingual Matters. Grenoble, Lenore A., and Lindsay J. Whaley (eds.) 1998. with relatively few speakers can be more stable than Endangered languages: current issues and future prospects. languages with a larger speaker base. Navajo, a Native Cambridge: Cambridge University Press. American language, provides an example of a lan- Krauss, Michael. 1992. The world’s languages in crisis. guage long considered to be stable because of its rela- Language 68. tively large number of speakers. It is now considered Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices. Oxford: Oxford University Press. endangered, due to rapid language shift stemming Robins, R. H., and E. M. Uhlenbeck (eds.) 1991. Endangered from a combination of social and economic factors. languages. Oxford and New York: Berg. Two key factors—relative isolation and a strong sense The most up-to-date information about languages and numbers of community identity—can help a language’s vitality. of speakers can be found in The Ethnologue, ed. by Barbara The opposite influences—heavy language contact and Grimes, <www.sil.org/ethnologue> a weakened sense of ethnic identity (and pride) can LENORE A. GRENOBLE increase the likelihood of language shift. See also Language Death 287 ENGLISH English In 1780, English was a language with fewer than 15 Germanic (NGmc) family, whose linguistic distance million speakers, most of whom lived in England, was, at the time of the formation of E, not so large as Ireland, Scotland, the United States of America, to inhibit communication—a fact that was relevant Canada, and the Caribbean. Yet, John Adams foresaw during the Old English (OE) period. E lost many a world where English would play a larger role: WGmc traits during the OE and Middle English (ME) English is destined to be in the next and succeeding cen- periods and has become a peripheral WGmc member. turies more generally the language of the world than Historically, E is the result of the transplantation of Latin was in the last or French in the present age. The the dialects of the Angles, Saxons, Jutes, and Frisians, reason for this is obvious, because the increasing popu- who invaded England in AD 449, forcing the Celts to lation in America, and their universal connection and the fringes of the islands or to assimilate. The Anglo- correspondence with all nations, aided by the influence Saxon or OE phase of E (to c. 1100) witnessed the for- of England in the world, whether great or small, force mation of several kingdoms, three of which their language into general use. (quoted from McCrum (Northumbria, Mercia, and Wessex) gained enough et al. 1986:239) power to create dialects. A national form of E did not Today, English has an estimated 350 million native develop. The Anglo-Saxons became victims of the and 400 million second-language speakers on all conti- Viking invasions from the middle of the eighth centu- nents. The number of foreign language users easily ry. King Alfred of Wessex was able to defeat them at tops the 400 million mark and is rising as globalization the end of the ninth century and forced them north of and modern technologies continue to impinge on all a line from London to Chester (Danelaw). England areas of modern life worldwide. In terms of political was partitioned into a NGmc/WGmc north and a status, English is the national language in native, WGmc south. Sociolinguistically, the Viking invasions anglophone countries (United States of America, UK, were an obstacle to the formation of a national form of Canada, Australia, New Zealand, Republic of Ireland, E at first, but after their defeat the sociopolitical and Bahamas, etc.), although none of these countries has linguistic prestige of Wessex promoted the rise of a declared English as its official language. English has quasi-OE standard. Christianization (from the late (co-)official status in close to 30 countries, most of sixth century) and the political support it had made which had been colonies (e.g. Ghana, Philippines, Latin the language of religion and writing and intro- India, Western Samoa, Costa Rica). It is the dominant duced Latin, Greek, and Hebrew loanwords into E. language in international institutions (UN, IMF, APEC, The Norman Conquest (Battle of Hastings, 1066) EU) and outdoes other languages in international brought about the most significant change in the histo- media, companies, conferences, air and maritime traf- ry of E. For over two centuries, England was under for- fic, publications, and the web. Sociolinguistically, the eign rule. E was demoted to a low, rural language, and norms of English (E) no longer emanate from a single French and Latin were promoted to the high languages source, even British (BrE) and American English in a triglossic situation. French gained the status of a (AmE) are only two of several norm-setting epicenters national, public language, as well as being the native in a pluricentric world of E. National norms also exist language of the French aristocracy; Latin was the lan- in many second-language countries. Foreign language guage of learning, where it was dominant up to the contexts, too, have been claimed to develop norms. It early Modern English period (EModE). There was has been said that E is best described in terms of three some bilingualism among the Norman elite and the overlapping circles: native, second, and foreign E. A urban English; but the shift to E occurred when the traditional view, which is adopted here, recognizes a Normans, facing the loss of their French possessions in common core across all varieties. the thirteenth century, decided to stay. As E regained status, it was no longer a mere continuation of its OE form. E was a second language for a transitional peri- History od, it absorbed thousands of French and Latin expres- As for genetic affiliation, E belongs to the West sions, and, under the congruent influence of NGmc and Germanic (WGmc) branch of Indo-Aryan languages French, lost most inflections and became a fixed word- and is related to German, Dutch, and Frisian. Danish, order language. From 1400, a written standard devel- Swedish, and Norwegian are members of the North oped in the Chancery, the oldest national institution; an 288  ENGLISH educated spoken form developed only from the six- ing. Again, it was only during the nineteenth century teenth century. Standard BrE and the Received that the United States of America assumed its current Pronunciation (RP) gradually rose to the position they extension and it was as late as at the turn of the twen- enjoy today, with the former having greater national tieth century that the United States of America export- and international prestige than the latter, while the for- ed their E to the Philippines, Guam (1896), and the mer has never ceased to be socially controversial. Caribbean. AmE was coming into its own and H.L. As for expansion, E included Cornwall, when Mencken (The American language, 1919) could Wessex conquered that region during the OE period. declare it to be a ‘language’, separate from BrE, an The Anglo-Normans founded settlements in coastal, allusion to the debates about a ‘Federal English’ in the eastern Ireland (The Pale) in the twelfth century, late 1770s. The mid-twentieth century witnessed fur- which were soon Gaelicized. The decisive steps for- ther expansion. World War II took American soldiers ward were made during the Age of Exploration. At the to the theaters of war in Asia, Africa, and Europe. Its beginning of the seventeenth century, the East India postwar economic, technological, and cultural power, Company transported E to India (1600). In 1607, the Britain’s enduring role in Europe and the foundation of the first American colony, Virginia, Commonwealth, the problems that former colonies transplanted E to America, although some experts experienced in nation building, and the collapse of argue that whaling had taken E speakers to communism supported the further expansion of E, Newfoundland and Nova Scotia earlier. Slave ships which was now fed mostly by the appeal of the United brought E to West Africa and the Caribbean islands. States of America and, logically, AmE. The Bermuda settlement, for instance, was founded in Today’s position of E is, thus, not entirely due to the 1609. As for the British Isles, Ireland came under migratory or imperialist expansion of Great Britain and British domination during the reign of Elizabeth I and the United States of America. There was a confluence James I. The Union with Britain in 1801 was a blow to of factors and it is worth emphasizing that the econo- her linguistic independence, leading to the decline of cultural, technological impact of E was possible only at Gaelic, the shift to E, and the development of IrE, a particular juncture, as David Crystal explains: which incorporated some Gaelic features. Gaelic has [But] international language dominance is not solely the survived in the west and made a recent recovery, part- result of military might. It may take a militarily powerful ly due to the European Union’s language policies. nation to establish a language, but it takes an economi- Scotland’s southeast had been English from the OE cally powerful one to maintain and expand it. This has period, but Scotland succeeded in developing a nation- always been the case, but it became a particularly critical al variety, Scots, during ME. The Union of Crowns factor early in the twentieth century.... Any language at the (1603), the Union of Parliaments (1707), and the centre of such an explosion of international activity would defeat of the Highland uprising in the mid-eighteenth suddenly have found itself with a global status. And century intensified the E influence so much that English ... was in the right place at the right time. (1997:8) Scotland was anglicized by the late eighteenth centu- Just as postcolonial nations have been willing to ry. Scots declined, but revived as a literary form in the retain E, so has the rest of the world where it is a for- nineteenth century, Gaelic was forced to the Highlands eign language. Real or perceived benefits attached to and western islands, Norn to the Shetlands and E act as powerful incentives and an E language indus- Orkneys, where it died out in the eighteenth century. try guarantees easy access. The seventeenth century paved the way for global The expansion did not remain without consequences expansion, and the eighteenth and nineteenth centuries for the formal system and lexis of E, norms of use, and witenessed the most significant geographic and attitudes. Up to the seventeenth century, expansion may sociopolitical growth. The East India Company, for have amounted to no more than the formation of a sin- instance, had been no more than a commercial ‘play- gle and contiguous British language community. The er’, with factories (trading outposts) in coastal cities. post-Renaissance period witnessed the formation of Only by the middle of the eighteenth century did it more — and distinct — communities: first, the gain political control over Bengal. It took another 100 American one (see below), then an (Anglo-) Indian years for the subcontinent to become a Crown Colony one, and then others. English absorbed foreign ele- (1857) and for an Indian elite to emerge that was will- ments, norms were diluted and others recreated, and ing to accept, and later to oppose, E. Similar develop- attitudes of inferiority gave way to those of pride in ments occurred elsewhere in Asia and Africa. one’s own E. This development went unnoticed for a Although Anglophone from the beginning, Australia long time; it was welcomed in the early United States and New Zealand were fully known only by the end of of America, when the notion of a Federal (American) the nineteenth century. The story of the western fron- English was promoted. In the early twentieth century it tier in America is too well known to require recount- was, however, perceived as a dangerous disintegration 289 ENGLISH of the language: ‘In our study of the growth of the closed syllables. ScE is, in other words, more system- Empire we forget that this territorial expansion of our atic than RP. The role of phonetic context on length is language sowed the seeds of disintegration.’ (Lloyd worth noting: word boundaries and voiced consonants James, secretary, BBC language committee 1928; lengthen vowels (fee/feed/feet). Centering diphthongs, quoted from Leitner 1989:12). e.g. in pair or tour, mainly occur in nonrhotic accents. E was a native (ENL), second (ESL), and foreign An interesting characteristic is that vowels maintain a (EFL) language in widely different contexts, and constant qualitative articulatory difference between assumed pluricentric characteristics, with American, them. Thus, all of RP’s short vowels are fairly open; all British, Australian, Indian, etc., epicenters. Malaysia, long vowels (and rising diphthongs) start with a rather Singapore, Hong Kong, and other nations face an low vowel. In AusE and NZE, in contrast, short vow- uneasy choice between condoning the facts, i.e. local- els are closer; long ones and rising diphthongs have an ization, or endorsing an ideal, i.e. foreign, model. open (or back) onset. (Some southern E accents, e.g. Many countries insist on native-like E but start accept- London’s Cockney, are similar in this respect.) These ing mixed BrE–AmE forms. Endonormative countries properties have given rise to speculations about a show signs of internal pluricentricity, with national Second Great Vowel shift in ‘southern hemisphere’ E, norms coexisting with those of ethnic communities, but some experts argue for the conservative mainte- such as Afro-American E or Australia’s Aboriginal E. nance of older features of the East Anglia accent. The Although localization became a powerful factor, the reduction of unstressed vowels is another area of vari- disintegration of E was checked by the rise of educat- ation. RP and many E accents use the bit vowel, AusE, ed, British-oriented middle classes in many countries NZE and—more advanced RP—have a schwa. during the nineteenth and twentieth centuries. Thus, Homophony between word pairs like boxes and boxers pluricentricity has not completely diluted family rela- is the result. Accents differ most perhaps in lexical tionships between E varieties. And internationalization phonology. Thus, either has /aI/ in RP, /i:/ in London’s is a new unifying factor. Cockney and genAmE; AusE may have either. From a comparative angle, one should add that E lacks front rounded vowels, which occur in WGmc languages. As Structure for spelling, the AmE and BrE systems differ in a few A formal structural survey can exploit this situation. It areas (center/-re, dialog/-ue, program/me, fueling/- will be based on a common core and point to areas of lling) and more numerously in the spellings of indi- variation. Starting with phonology, vowel and conso- vidual words (check/cheque). Other varieties follow nant systems are quite similar, variation being limited one or the other system, or attempt to develop a system to a few phonemic contrasts between BrE and AmE of their own, as AusE does. and, otherwise, being confined to phonetic detail. E differs most from contemporary WGmc lan- Scottish E (ScE) and some forms of AmE still distin- guages in its paucity of inflectional morphology. guish witch from which. ‘Flaps’, short, voiced taps of Articles lack inflection, nouns have lost most inflec- the tip of the tongue against the alveolar ridge, are tional contrasts (e.g. gender), leaving behind number realizations of /t/ and /d/, e.g. letter, in AmE, AusE, and the possessive case, which has an analytic variant New Zealand E (NZE), less so in BrE. Rhoticity dif- (the roof of the car for the car’s roof). The genitive can ferentiates a southern and northern type of E. The for- mark an NP (A friend of y brother’s favorite car) and mer — exemplified by southern BrE, AusE, NZE, and fulfill a syntactic, phrasal function. As for number, the Indian E (IndE) — are nonrhotic and lack postvocalic OE dual has been lost except semantically in both, /r/ in car. Rhotic accents, i.e. northern EE , ScE, IrE, either and neither, irregular forms, such as umlaut AmE, and CanE, retain it. In RP, /l/ has a clear and a (goose/geese), -en (ox/oxen; child/children), and zero dark articulation, cf. leaf and fall; other varieties, e.g. (sheep/sheep) reflect OE or ME patterns. Verbs have AusE, may have a single articulation in all positions. no inflectional contrasts for voice and mood. Were is a /l/-vocalization in postvocalic position is typical of peripheral remnant of the subjunctive, which is southern EE, especially popular London, and results in expressed by means of should (mainly in BrE) or the a dark vowel, e.g. school [skuo]. Vocalization occurs bare infinitive (mainly AmE) (e.g. I suggest he in AusE and NZE. RP has an exceptionally complex (should) resign from office). What remains is a third- and imbalanced vowel system with its 21 mono-, di-, person singular in the present tense (he goes vs. and triphthongs. For instance, fit/feet, foot/food, I/you/they go), a simple present and past tense (I like/I abound/purr illustrate the only long/short vowel pairs. liked), and the progressive, which is marked by -ing Words like pot and port differ qualitatively, not in (in combination with a form of be). A striking aspect length. ScE has no phonemic length, length being is a small number of (‘strong’), but frequent, irregular determined by whether a vowel occurs in open or verbs. Some verbs have three forms (go, went, gone), 290  ENGLISH others two (buy, bought), and still others one (put) for empathy, friendliness, etc., without the intention of the infinitive, simple past and participle, respectively. asking a question (My name is Betty ´Smith). There are other ‘irregularities’, e.g. vowel changes Turning to grammatical relations, E contrasts with (take, took, taken), remnants of old past tense or par- many languages, including WGmc ones, in ruling out ticiple endings (taken, leapt). As for adverbs, those subjectless sentences (G mich friert ‘I am cold’), with initial stress and two syllables have the -ly suffix although spoken registers like sportscasting have them, (e.g. beautifully). Others are formed analytically (e.g. if the subject is retrievable. E, however, permits a wide in the most friendly fashion instead of *friendlily). range of semantic roles to occur in subject position, Nonstandard, dialectal varieties, and speech can have while some of them are more marked in other lan- unmarked adverbs (e.g. right). guages where, as in German, semantic roles are encod- There is more inflection in pronouns. Personal and ed as adverbials. Consider the following examples: possessive pronouns have a nominative, possessive, John can certainly act wonderfully well (Agent) and objective case (e.g. I, my/mine, me), and natural This advertising campaign will open up new gender (in the third-person singular). Wh-pronouns markets, we hope (Instrument) have a type of (natural) gender (human/nonhuman, Berlin can be a rather cold city in winter e.g. in who and which), case (genitive whose, objective (Location: Place) whom). A widely discussed gap is the lack of a second- Money is not the problem, really (Object, Theme) person singular/plural, which is available in languages like German (du, ihr) or French (tu, vous). OE had that Verbal syntax is another area where E has charac- distinction and, moreover, a dual. Most varieties, espe- teristics of its own. Thus, E has a causative/processual cially the standard ones, have lost both, but some have contrast with transitive and intransitive constructions, recreated it; consider AmE (you/you all or y’all, you provided the object of the transitive construction is guys) or IrE (you/youse), where it is a calque on the identical to the subject of the intransitive one: Gaelic tu:sibh contrast. John broke the glass The trade-off between inflectional morphology and The glass broke syntax is a good starting point for a survey of syntax. E is essentially an SVO language; other WGmc lan- Ergative constructions contrast with the passive voice, guages permit more variation. Thus, German has two which implies agency without expressing it, while the standard orders, i.e. ‘verb-second’ and SOV, in main intransitive-ergative construction attributes some kind and dependent clauses, respectively. ‘Verb-second’ of propensity to the subject, a feature that is clearer if means that any order (SVO, OVS, AVS/O, etc.) is cor- manner adverbs occur, as in the following examples: rect as long as V comes second. There is less flexibil- The window pane breaks easily ity in E, but it is not entirely absent, as these examples The line washes well and fast in this washing show: The young student I like for his sense of humor machine (OSV), Down he came (ASV), Here comes the train Potter’s latest book sold extremely well (AVS); SOV is impossible, except in verse. Mention must be made of cleft (It is his sense of humor I like), Two characteristics, i.e. particle verbs and nominal- pseudocleft (What I like is his sense of humor), and verbal complexes, deserve mention as very frequent existential sentences (There is an interesting program and productive markers of colloquial E. As to the for- on TV). Non-SVO constructions are closely related to mer, several types of such prepositional verbs (run up the information structure in discourse. Interrogatives (a hill)), phrasal verbs (run up (a bill), (a plan) takes manifest subject-auxiliary inversion, which is regular off), and phrasal–prepositional verbs (come up with (a in polar questions (Can he do that? Wasn’t that a great good idea)) can be identified. Nominalization creates show!). A note on so-called guided questions, a side ‘particle nouns’, which are often marked by stress (to issue, is called for, as there is a good deal of variation. take ‘off - ‘take-off). Take a bath, give a smile, have a They arise when a high rising intonation contour is chat illustrate nominal-verbal complexes that can superimposed on a declarative construction (So, you often be replaced by the verb they contain formally are having an ´affair?) and permit the inference that (e.g. to bathe, to smile, to chat). There is, one might the proposition contained is assumed to be true. ESL add, substantial variation with regard to the choice of varieties, in contrast, often extend this pattern to ordi- e.g. take, have, etc. E, in contrast to, say, German also nary polar questions (He came home late?) without has a dynamic have (have a baby). that inference. A different pattern has been reported Let us close with tense, aspect, and modality. The from AusE, NZE, IrE, some northern E dialects, and future is expressed analytically by will (There will be varieties of AmE: they combine a high rising tone with rain in the afternoon). But there are other ways, e.g. declarative word order to express social meanings of the chair is going to dip over, be careful; the Foreign 291 ENGLISH Secretary is about to go to Kuwait; I’m coming, don’t clerk, and demon (the latter two of Greek origin). worry, We leave for London tomorrow, which all French has exerted a pervasive and ongoing, if express subtler shades of future time reference, often decreasing, influence toward the nineteenth century. merged with modal meanings. A sentence like ‘I will With Romance words being so frequent, they barely come and see you later, John’ is both a promise and a require illustration, except to mention that Romance prediction. Apart from being speech act markers, (and Greek) introduced a differentiation of style. Gmc modals can express three types of meaning, i.e. deon- words tend to be common and/or informal, Romance tic meaning (I must be gone  ‘I am obliged to’), epis- ones formal and/or technical: consider begin-com- temic meaning (He can’t be there, the light’s not on  mence. Romance also increased the potential of E from what I know’), and what has been described as word-formation patterns that derived from Wgmc, e.g. ‘ability’ meaning (I could swim when I was a small kingly, royal, and regal, and introduced new ones, e.g. child  ‘I was able’). These meanings interact with in -ify (beautify), -ate (nominate), or describe-descrip- negation and question formation in ways that cannot tive-description. The latter cannot be analyzed in be detailed here (std BrE You mustn’t ( deontic) be terms of morpheme sequencing like kingdom. As a too strict with the boy vs. He *mustn’t/cannot ( epis- result of borrowing, the Gmc word stock is now a low temic) be at home yet), but which suggest that there 30% and the Romance one is 50%. will be considerable of variation, as there is; in ScE We now illustrate the absorption of expressions and AusE, for instance, that sentence is grammatical. from languages that E in its standard E (ENL) and in Apart from the ‘instantaneous present’ in sports ESLs varieties has come in contact with during its commentaries (Beckenbauer kicks the goal), etc., the global expansion: American-Indian (caucus fr. simple present has to be paired with the progressive to Algonquian; hickory), Arabic (zero; assassin), Hindi refer to present time (*Billy goes to school (now) for (sari; bandh call ‘call for a strike’), Telugu (bandi- Billy is going to school (now)). On its own, it express- coot), Australian Aboriginal languages (boomerang; es habitualness, etc. The simple past refers to a definite cooee ‘type of call’), Malaysian/Singaporean E (la past time, even if it is unexpressed (Billy went to ‘emphatic particle, fr. Chinese), African languages school), while the present perfect expresses indefinite (baobab), Japanese (rickshaw; karaoke), German past time and notions like current relevance (I know (angst; -friendly, see ‘user-friendly), Malay (jubah ‘in New York, I have been there [ indefinite past]; Islamic culture, a kind of clothing’), Spanish (provin- Shakespeare has written important plays [ current ciana/o ‘woman/man from provinces, considered relevance]. A sentence like *Shakespeare has lived in unsophisticated’, current in Philippine E). Add to this Stratford, in contrast, calls for the simple past, as there loan translations, e.g. acupressure [ Jap.], walkabout is no conceivable current relevance. Here, too, there is (‘walk in public’,  Australian Aboriginal languages), a great deal of variation. AmE, for instance, has a so- or the productive use of some patterns in e.g. crash- called colloquial simple past; the perfect is more for- friendly (cars). mal (in contrast to, say, German and French, where the The most important feature of all is the lexical cre- opposite would apply). There are constraints on the ativity of new varieties. The impact of AmE on sci- combination of the perfect with adverbs, such as yet, ence, technology, the media, sports, informal E, and already. Once again, varieties differ a great deal. Std slang is too well known to require illustration. To men- BrE, for instance, disallows *Did John come in yet; tion a few older words, e.g. to trade ‘to shop’; to ship AmE, AusE, BrE do not. ‘to send’; log-rolling; on the fence, some of which Lexis is especially interesting from a pluricentric echo the moving frontier society. Generally speaking, perspective. The quite homogeneous, original WGmc AmE is more informal and its dominant form is word stock has been enriched by tens of thousands of defined as general AmE, in contrast to the more elitist expressions that E has absorbed from languages it has definition of standard BrE, especially RP. Well-known been in contact with. As a result, such words have been expressions in AusE are mate ‘pal, friend’, to barrack integrated; others have been lost or specialized to new for ‘to support (e.g. a sports team)’, bushie ‘someone meanings, uses, or stylistic patterns. Contact with who lives in the bush’, in IndE carlifter ‘car thief’, Celtic and NGmc dialects provided the first nonnative handphone in Asian E. Localized idiomatic expres- influence. NGmc introduced some frequent words like sions, such as We must meet again (AmE), Can I leave law, fellow, they, and both. Some of them became ref- now, or Have I your permission to leave, Sir (IndE, erentially distinct (skirt–shirt), others dialectal (kirk, PakE ‘it’s been nice talking to you’), are highly visible ScE). Brogue, whisk(e)y, and Tory are Celtic in origin. and invite misunderstanding, as do words like Latin has influenced E from as early as OE, even prior Australian AborE women’s business ‘knowledge con- to the WGmc tribes settling in England, as street and fined to (eligible) women’. Others such as to jump the kitchen suggest. Late OE loans from Latin are sign, dragon gate in Chinese E may not be understood at all. 292  ENGLISH Many expressions remain confined to local varieties, sider the unrounding of the lot-vowel, yod-dropping in hence the pluricentricity of E; but many have become tune, to genAmE. The dialect divide is difficult to date, common and have made E a lexically rich language, as but Webster’s American dictionary of the English lan- the Oxford English dictionary and Webster’s third guage (1828) is probably as good a marker as any. It international dictionary and national dictionaries in marks differences in lexis, spelling, lexico-grammar, Australia, India, etc., demonstrate. and grammar. Apart from the national element in With respect to word formation, three processes American debates, a strong, shared, prescriptive tradi- will be mentioned. The first is conversion or zero der- tion was visible during the late eighteenth and nine- ivation, the most productive pattern today, which per- teenth centuries and counteracted separation. Bishop mits words to change word class membership without Lowth’s, Murray’s, or Cobbett’s grammars, etc., a formal marker (chair—to chair). Often, words occur catered to markets on both sides of the Atlantic. in several word classes (right as a preposition, noun, Prescriptivism continued, for instance, with Henry verb, adjective, or adverb). Similar to conversion is the Alford’s The Queen’s English (1864), H.W. and F.G. derivation of noun–verb pairs by means of a stress Fowler’s The King’s English (1906), and H.W. shift, as in ‘import—to im’port, ‘take-off—to take ‘off. Fowler’s A dictionary of modern English usage The second is the deletion of a preposition in preposi- (1926). Although they explicitly denied writing for an tional verbs, which results in an ordinary transitive American market, which they perceived as different, verb. Thus, to appeal against a decision becomes Fowlers’ books were used there. Debates about appeal a decision. The inverse process inserts a prepo- Webster’s third international dictionary and English sition (to meet with). The third is clipping (bod  curricula in the United States of America show the body) and suffixation in -i(e) and -o (pollie/y  ‘politi- strength of prescriptivism today. It is slowly giving cian’, garbo  ‘garbageman’). Phonological adapta- way to descriptive approaches. The animosity of the tions may be necessary, e.g. mossies ( mosquitoes) British against AmE, which became a public topic at with a /z/, like Aussies, instead of /s/. An interesting the end of the eighteenth century, has survived all aspect of clipping is the formation of first names to intellectual shifts and extended its reach to practically express solidarity, acquaintance, etc.; consider Kez ( all varieties of E of BrE descent. Kerry), Baz ( Barry) These processes are common in Although they are considered distinct languages, E- E, but their application may result in words that are based pidgins and creoles must be mentioned as they specific to some variety of E. ESL varieties do not show, among others, the power of the language in seem to add processes not found in native E either. intercultural situations. Pidgins typically emerged in Hybridization may be more frequent (IndE bandh call colonization scenarios in Africa (e.g. Krio), the ‘call for a strike’), although it also occurs in ENL con- Caribbean (e.g. Jamaican Creole), southern states of text when a borrowed word may be difficult to under- the United States of America (e.g. Gullah), Australia stand (e.g. koala bear in AusE). (e.g. Kriol), India (e.g. Butler English), and the Pacific (e.g. Tok Pisin in New Guinea). Controversies about what factors gave rise to them, the formative role of Formation of ‘Englishes’ superstrate E vis-à-vis local, substrate languages, the This tour de force has shown that the transplantation of diffusion of pidgins through maritime traffic, and the dialects into new habitats, dialect contact, and contact role of universal processes of language development with local languages have been the major factors that remain unresolved. The fact remains that pidgins and have shaped the development, texture, and status of E. creoles are similar, whether one looks at North Diversification has been the predominant outcome, the America, Africa, or the Pacific, where genetically and loss of a variety being rare. But it has happened, as the typologically different languages are used. The avail- decline of Scots shows. Centrifugal and centripetal ability of E and upward social mobility draws them forces have always been active. Two themes must be into the realm of mainstream E and they tend to added to this account, viz. the dating of important BrE become the (socially) low, nonstandard, end of a con- and AmE bipolarity and English-based contact lan- tinuum that goes up to (national or international) stan- guages, especially pidgins and creoles. dards of E. Ethnic forms, e.g. Afro-American E or The ‘great divide’ between the two main ‘epicen- Aboriginal E, are in a middle range. Speakers can ters’ of E, viz. BrE and AmE, has been a slow one, exploit the continuum for communicative purposes, although signs of separation had been seen early. With but it would be wrong to assume a simple ‘upward’ respect to phonology, the divide has been dated at orientation, with the highest variety having the highest around 1750. The loss of postvocalic /r/ and of sec- prestige. The ‘high’ variety may be merely used in out- ondary stress in suffixes (‘secretary vs. AmE side communication or selected domains. Contact ‘secre,tary) have remained confined to RP; others con- varieties often symbolize ethnic identities and show 293 ENGLISH verbal repertoires that differ significantly from main- Graustein, Gottfried, and Gerhard Leitner (eds.) 1989. Reference stream E, such as Afro-American E or Jamaican grammars and modern linguistic theory. Tübingen: Niemeyer. Kachru, Braj. 1986. The alchemy of English. Oxford: Creole. The dialect continuum may be less real, soci- Pergamon Press. olinguistically, than a formal analysis suggests. Leitner, Gerhard. 1982. The consolidation of “educated south- John Adams’s prediction of a global role of E has ern English” as a model in the early 20th century. IRAL come true. But it is not the E that he may have had in 10(2). 91–107. mind. It is not bicentric, it is pluricentric, subject to Leitner, Gerhard. 1986. The English reference grammar. Language and linguistics, writers and readers. Tübingen: conflicting forces that emphasize local and global Niemeyer. trends. The future is hard to predict, but it seems Leitner, Gerhard. 1989. BBC English und Englisch lernen mit unlikely that it will encourage disintegration. der BBC. München: Langenscheidt. Leitner, Gerhard (ed.) 1991. English traditional grammars: an international perspective. Amsterdam: Benjamins. References Leitner, Gerhard. 1992. Pluricentric English. Pluricentric lan- guages, ed. by Michael Clyne, 178–237. Berlin: Mouton de Bailey, Richard W. 1991. Images of English. A cultural history Gruyter. of the language. Cambridge: Cambridge University Press. Maurais, Jacques, and Michael Morris (eds.) 2003. Languages in Bailey, Richard W., and Manfred Görlach (eds.) 1983. English a globalising world. Cambridge: Cambridge University Press. as a world language. Ann Arbor: University of Michigan McArthur, Tom (ed.) 1992. The Oxford companion to the Press. English language. Oxford: Oxford University Press. Brutt-Griffler, Janina. 2002. World English. A study of its McArthur, Tom. 1998. The English languages. Cambridge: development. Clevedon: Multilingual Matters. Cambridge University Press. Burchfield, Robert. 1985. The English language. Oxford, New McCrum, Robert, William Cran, and Robert MacNeil. 1986. York: Oxford University Press. The story of English. London: Faber & Faber. Conrad, Robert, Joshua Fishman, and Robert Cooper. 1977. Milroy, John, and Lesley Milroy. 1985. Authority in language. The spread of English. Rowley, MA: Newbury House. London: Routledge and Kegan Paul. Crystal, David. 1997. English as a global language. Cambridge: Schneider, Edgar (ed.) 1996. Focus on the USA. Amsterdam: Cambridge University Press. Benjamins. Dillard, J.L. 1992. A history of American English. London: Trudgill, Peter, and Jean Hannah. 1994. International English. A Longman. guide to the varieties of standard English. London: Arnold. Fisher, John. 1996. The emergence of Standard English. Wells, John. 1982. Accents of English, 3 vols. Cambridge: Lexington, KY: The University Press of Kentucky. Cambridge University Press. Görlach, Manfred. 1991. Englishes: studies in varieties of English. Amsterdam: Benjamins. GERHARD LEITNER Görlach, Manfred. 1999. English in nineteenth century England. An introduction. Cambridge: Cambridge See also Celtic Languages; Indo-European 2: University Press. Germanic Languages; Old English Epenthesis and Syncope Syncope (from Greek ‘cutting off’) and epenthesis Further extension to sounds (both vowels and conso- (from Greek ‘insertion’) are cases of the more general nants) in word-initial syllables (also known as phenomena of the deletion and insertion of segments, ‘aphaerisis’) is less common, no doubt matching the respectively. Initially utilized in the description of his- lesser frequency of the phenomenon itself given the torical phonological change, use of the terms has been usual salience of initial syllables. However, it does extended to characterize synchronic alternations as occur, as reflected in the loss of initial /k/ in such well. Most frequently, ‘syncope’ refers to the dropping English words as knife and knee, the initial vowels in of a word-internal unstressed vowel, as in Latin bishop (from Latin episcopus), possum (from CALIDU(M) ‘hot’, NOBILE ‘noble’, VIRIDE(M) Algonquian opossum), special (from Old French espe- ‘green’, which became Spanish caldo, noble, verde, cial), or the disappearance of initial /h/ in English respectively. Expanded usage of the term ‘syncope’ dialects (Cockney being the typical example). covers dropping of a final vowel (also called ‘apoc- ‘Epenthesis’, too, designates a variety of phenome- ope’) as in Latin BLANCU(M) ‘white’, MURU(M) na involving the insertion of both vowels and conso- ‘wall’ becoming French blanc, mur, respectively. nants. Consonantal epenthesis (where the appropriate 294  EPENTHESIS AND SYNCOPE conditions are created by the previous syncope of described here — syncope for loss of a word-internal Latin E) is illustrated in (1): unstressed vowel; epenthesis as a more general term for vowel or consonant insertion, again word-internally — (1) Latin Old French seem to reflect the most widespread usage (see Lass CIN(E)REM cendre ‘cinder’ 1984:183–8 for a thorough survey). Both syncope and epenthesis (in either their more *ESS(E)RE estre ‘to be’ limited or wider senses) are interesting because of the MOL(E)RE moldre ‘to grind’ general theoretical issues they raise. Both are linked to NUM(E)RU(M) nombre ‘number’ phonotactic considerations — the constraints on sequences of segments or on syllable structure. Note that, historically, English empty and thunder Consonantal epenthesis often results in the addition of show epenthetic /p/ and /d/, respectively. Moreover, syllable-initial sounds and strengthens, as a conse- the English words (French loans all) humble — humil- quence, syllable onsets (cf. cendre above). Vocalic ity, cinder — incinerate, number — enumerate, etc., epenthesis has the effect of separating consonants show both the remnants of historical epenthesis at the whose proximity violates the restrictions on consonan- Old French stage and a minor synchronic alternation. tal combinations and leads to an improved syllable This Old French epenthesis underlies a synchronic structure. Thus, in athlete (athalete), the rare /θl/ alternation occurring in a number of irregular verbs in sequence is interrupted by the insertion of schwa and Modern French: connaître — connaissons ‘to know’, a VC.CVC syllable structure (where V  vowel and C coudre — cousons ‘to sew’, craindre — craignons ‘to  consonant) becomes the more acceptable fear’, falloir — faudra ‘to be necessary’, tenir — tien- V.CV.CVC. Syncope, in addition to phonotactic con- dra ‘to hold’, valoir — vaudra ‘to be worth’, venir — siderations, is also sensitive to accentual and rhythmic viendra ‘to come’, among others. patterns (accented vowels and vowels in initial sylla- Consonantal epenthesis has been investigated using bles are rarely involved). Speech rate and word fre- both phonological considerations and phonetic quency also play a role: deletion of vowels is fostered approaches. In phonological (syllable structure) terms, by rapid speech, as we see in pronunciations such as epenthesis often reinforces the onset of the second syl- /ptéjDow/ (or even taters) for potato, and the greater lable. Thus, in cendre, earlier /n-r/ becomes /n-dr/ frequency of memory ( [mεmɹij]) vs. mammary where syllable-initial /dr/ provides a stronger onset. In facilitates deletion of the penultimate vowel in the for- phonetic terms, the relative timing of nasalization or mer. Finally, if the operation of syncope were to create fricative cessation relative to the following segment a prohibited sequence, it is frequently the case that results in the insertion of a transitional or epenthetic deletion is blocked or delayed. Consider the Latin consonant. forms in (2) where the vowel preceding stress would ‘Epenthesis’ can also refer to the insertion of a normally be expected to delete: vowel in various positions in the word. Initial vocalic epenthesis (also known as ‘prothesis’) is seen in Latin (2) Latin Old French SCHOLA ‘school’, SCRIBET ‘he writes’, STELLA CAPRIFOLIU chevrefeuil ‘honeysuckle’ ‘star’ becoming Spanish escuela, escribe, estrella. *PUTRITURA porrëure (Mod. French pourrit- Word-internal vocalic epenthesis (also called ‘anaptyx- ure ‘rottenness’) is’ or ‘svarabhakti’, the latter from Sanskrit where the QUADRIFURCU carreforc (Mod. French car- refour ‘crossroads’) process is common) occurs variably in such nonstan- dard English pronunciations as athalete ["θəlijt], In each case, the highlighted vowel in Latin is retained artheritis [aɹθəɹajDis], fillum [filəm] (standard film), as schwa in Old French because deletion would have while final epenthesis, normally of a consonant and produced the unacceptable consonant clusters PRF, much rarer, is seen in such English forms as amongst, TRT, and DRF, respectively. (Later simplification of whilst and nonstandard acrosst, heightth (with analog- the consonant groups obscures the original conditions.) ical pressures from length, width, and depth being the Syncope and epenthesis, as characteristic of dele- obvious source in the latter case). Historically, the final tion and insertion in general, are directly involved with stops in pheasant and sound ( Old French fesan and a wide variety of central phonological issues. They are son, respectively) are also epenthetic. As the foregoing relevant to considerations of syllable structure and discussion indicates, the addition or deletion of seg- phonotactic constraints in general, of the rhythmic ments can occur in a variety of positions in the word, properties of speech, rule naturalness, and the use of with a correspondingly complex set of terms reflecting phonetic explanations in phonology, and no doubt oth- this behavior. Not all scholars use all terms, moreover, ers. As such, the two processes merit wide-ranging and and their use is not always consistent. The practice detailed attention. 295 EPENTHESIS AND SYNCOPE References Trask, Larry. 1996. Historical linguistics. London: Arnold. Walker, Douglas. 1978. Epenthesis in Old French. Canadian Lass, Roger. 1984. Phonology. Cambridge: Cambridge Journal of Linguistics 23(1). 66–83. University Press. Murray, Robert. 1989. On epenthesis. Folia Linguistica XIII DOUGLAS C. WALKER (3–4). 293–316. See also Phonetics; Phonology; Syllable: Structure Eskimo-Aleut The Eskimo-Aleut language family is spoken from the history of the language family even further back in time. Chukotkan Peninsula on the Siberian coast in the West While there is little solid linguistic evidence of a genet- to Greenland in the East. It includes two major branch- ic relationship between Eskimo-Aleut and other lan- es, Aleut and Eskimo. Eskimo consists of at least two guage families, there are strong suggestions of very further subgroups, Yupik and Inuit, and there is some early contact, particularly with Uralic (for a thorough question as to the proper classification of Sirenikski, discussion of possible linguistic affinities and contact, which may be a Yupik language or may constitute a see Fortescue 1998). third branch of Eskimo. The terms ‘Eskimo’ and Within the Eskimo branch, the split between the ‘Aleut’ both have unclear origins, and they are not uni- Yupik and Inuit branches must have taken place about versally in use within the respective areas. Thus, it is 2,000 years ago, and there is good linguistic evidence commonly held that ‘Eskimo’ was derived from a (e.g. shared phonological and prosodic features derogatory Algonquian term meaning ‘eaters of raw between Yupik and neighboring Inuit speakers in areas meat’; the negative association has led to the general now no longer predominantly Yupik) for positing the replacement of ‘Eskimo’ with Inuit, the native term for prehistoric (i.e. prior to European contact in the eigh- ‘people’, in most of Canada, and there are various des- teenth century) presence of a continuum of Yupik ignations for the language, including Inuttun, Inuktitut, speakers around the Bering Strait. From their original and others. The term ‘Aleut’ was bestowed on the peo- homeland in Alaska around the Seward Peninsula, ple native to the Aleutian Islands and their language by Yupik speakers gradually occupied southwestern Russians in the eighteenth century; the native term in Alaska and, moving westward across the Bering Strait, use today for self-designation is Unangan or Unangas reoccupied the Chukchi Peninsula in Siberia. Central for the people and Unangam Tunuu for the language. Siberian Yupik and neighboring Chukchi have gradu- Because there is still no other general term to describe ally displaced Sirenikski, and the latter is now extinct. all of the languages and dialects encompassed by the If Sirenikski is a separate branch of Eskimo, then it term ‘Eskimo’, and for reasons of linguistic tradition, may have split off about 2,500 years ago and its ori- both terms are still commonly used within the field of gins may be on the Chukchi Peninsula. About 1,000 Eskimo-Aleut linguistics. years ago, there was another eastward migration out of It is generally accepted that the Eskimo and Aleut Alaska; known as the Thule culture, this very rapid people were part of a relatively late migration from Asia migration is associated with the spread of the present- across the Bering Land Bridge, sometime between 4,000 day Inuit language. The present dialect differentiation and 6,000 years ago. The development and differentia- is possibly as recent as the past 500 years (see Dorais tion of Eskimo-Aleut are believed to have taken place in 1996). There is also evidence for the spread of Inuit Alaska because of the linguistic diversity found on the speakers into Yupik-speaking territory in the Seward Alaskan side. From Alaska, there was continued migra- Peninsula in the past two centuries. tion down to the Aleutians and eastward to Greenland. There are certain features that, taken together, are The split between Aleut and Eskimo is thought to have particularly characteristic of the Eskimo-Aleut lan- taken place around this time, although this figure is not guage family. Thus: undisputed, and there is archeological and skeletal evi- dence to suggest that the divergence between Aleut and ● All languages have three basic vowels (i, u, a), Eskimo may be at least 9,000 years old. Various attempts which are derived from an original four-vowel have been made to link Eskimo-Aleut with other lan- system (i, u, a, and schwa, represented by [e] guage families on the Asian continent, and thus trace the in Yupik, which maintains the four-vowel 296  ESKIMO-ALEUT distinction; [e] will be used to represent schwa in ● certain restrictions on syllable and other phono- this discussion). logical structures, ● All of the languages are almost exclusively suf- ● up to four or five nongrammatical cases, in addi- fixing, the only exception being an anaphoric tion to ergative and absolutive cases (locative, prefix ta- on demonstrative stems (e.g. pan- modalis, ablative, allative, vialis, and equalis), Eskimo una, tauna ‘this one’). ● transitive and antipassive structures, whose • All are polysynthetic: to a greater or lesser choice appears to be partially determined by def- degree within the language family, they allow for initeness, as in the Central Alaskan Yup’ik exam- very complex verb structures that encode mean- ples below: ings for which other languages need whole sen- tences. This is particularly developed in Canada Angute-m neqa ner-aa and Greenland; e.g. Holman Island Inuinnaqtun man-ERG fish-ABS eat-he:it (example from Lowe, in Tersis and Therrien ‘The man eats the fish’ 2000:167): Angun neq-mek ner’-uq iqaluk-hiu-riaqtu-qati-gi-tqiktaqpak-kalua- man-ABS fish eat-he ramiung ‘The man eats a fish’ fish-hunt-go.to.do-partner-have-several.times- In the descriptions of the languages, there are also even.though-he:him certain features that are commonly used to differenti- ‘even though he went fishing several times with ate the individual languages or dialects. Below, Aleut, him’ Yupik, and Inuit are described in terms of their major ● All are clause-chaining: sentences typically con- languages or dialects, the major linguistic features that sist of a series of dependent clauses headed by an differentiate them from each other, and their current independent clause. In the following relatively status and outlook. simple example from West Greenlandic, there is Aleut is a language with two clearly distinct, extant one indicative clause and two subordinated par- dialects, and at least two other dialects historically ticipial clauses; clause chains can be quite exten- attested. Eastern Aleut is spoken east of Amukta Island sive (example from Fortescue 1984:39): to the Alaskan Peninsula, as well as on the Pribilof Islands. Atkan, also variously called Western or irn-i qajartur-tuq Central Aleut, is today spoken on Atka Island, and a son-his.own be.out.in.kayak-his version of it is spoken on Bering Island. A third qinnguar-paa dialect, Attuan, is essentially extinct, and it had close see.through. binoculars-he:him affinities with Atkan; a mixed language known as natsirsu-up sursuk-kaa Copper Island Aleut, consisting of Attuan stems and hooded.seal attack-it/his Russian inflection, is still spoken on Bering Island. In ‘hei saw his sonj through his binoculars being very early descriptions, there was apparently a sepa- attacked by a hooded seal while in hisj kayak’ rate dialect spoken on the Rat Islands in between Attu and Atka, although almost no extant information is ● All have SOV (subject–object–verb) word order available on this form of Aleut. Characteristic of Aleut to a more or less fixed degree (Aleut has essen- are tially fixed word order). Note that in the previous example, the object clause irniq qajarturtuq pre- ● phonological features such as its lack of a [p], cedes the verb, and the subject of each of the and its aspirated nasals, subordinated clauses precedes its respective ● consonant clusters that differ from those in verb. Eskimo in their distribution (e.g. they are per- ● All have, or originally had, a largely mitted word-initially, as in sngaxsix ‘to dream’), ergative–absolutive case-marking system (with in the combinations of consonants possible (e.g. extreme modifications in Aleut), where the sub- velar–uvular fricatives), and in their complexity ject of intransitive verbs receives the same mark- (allowing up to three consonants, as in Eastern ing as the object of transitive verbs, namely Aleut ixchxingin ‘his neck’), absolutive case, while the subject of transitive ● use of independent pronouns, as opposed to the verbs receives a different marking: ergative case. complex verb structures of Eskimo languages: The Eskimo languages are much more closely relat- Aleut txin achix-ku-qing ed to each other than to Aleut. In addition to the com- you teach-presently-I mon features listed for Eskimo-Aleut, they share ‘I am teaching you’ 297 ESKIMO-ALEUT Greenlandic ilinniar-tip-pakkit (and on the Kenai Peninsula these include children), learn-cause-I:you and with immersion programs in the schools and ‘I am teaching you’ active production of learning materials. Sirenikski is considered related to the Yupik lan- ● its typologically unusual agreement system, in guages; however, it has also been regarded as a separate which ergative case marking is only used if a branch of Eskimo. It is seen as an important link to transitive object or an object of posession is not Proto-Eskimo because of particularly conservative fea- overtly expressed: tures such as a retention of consonants between vowels, Tayagˆu-xˆ qa-xˆ qa-ku-xˆ which were lost in all other Eskimo-Aleut languages man fish eat-presently (e.g. Proto-Eskimo *ataRuciR, Sirenikski ategesegh, ‘The man eats the/a fish’ Central Alaskan Yup’ik atauciq, Iñupiaq atausiq ‘one’). It has, however, undergone sound changes quite differ- Tayagˆu-m qa-kuu ent from other Eskimo languages. For example, in all man-ERG eat-presently:he:it nonstressed (essentially noninitial) syllables (as in the ‘The man eats it’ example given above) the vowel changed to schwa. Modern Aleut also has a large proportion of Unfortunately, it was first discovered and described at Russian loanwords, the result of extensive contact the end of the nineteenth century, when it was already with Russian traders and colonizers in the eighteenth highly moribund; the last speaker died in the year 2000. and nineteenth centuries. As a result of early decima- Inuit is generally described as a language with four tion of the people and later suppression of the lan- distinct dialect groups, each of which have their own guage in schools, the language is severely endangered recognizable subdialects; these groups include today, with at most 150 speakers in the Aleutian Alaskan Iñupiaq, with four major subdialects spoken Islands, the Pribilof Islands, and Anchorage, and per- from the Seward Peninsula and north; Western haps eight to ten on Bering Island. With the exception Canadian Inuit, with four major subdialects and spo- of Atkan, the speakers are elderly. ken over a vast area of Central Arctic Canada from The Yupik languages include Naukanski, spoken MacKenzie Coast to Repulse Bay; Eastern Canadian around East Cape on the Chukchi Peninsula; Central Inuktitut, with six major subdialects, spoken in Baffin Siberian Yupik, also spoken on the Chukchi Peninsula as Island, Arctic Québec, and Labrador; and Greenlandic well as on St. Lawrence island; Central Alaskan Yup’ik, Kalaallisut, with three major subdialects and spoken in spoken from Norton Sound to Bristol Bay in Alaska; and Greenland (there is also a sizeable population of Sugpiaq, also known as Sugcestun or Alutiiq, spoken speakers in Denmark). Characteristics of Inuit are around Prince William Sound, the Alaskan Peninsula, ● lack of intonational stress as compared with Kodiak and Afognak Islands, and the tip of the Kenai Yupik (with noticeable difference in intonational Peninsula. Yupik languages are characterized by patterns between the dialect groups), ● their retention of a fourth vowel that presumably ● loss of the fourth vowel, with various important stems from the ancestral (proto) Eskimo lan- phonological traces, guage (cf. Proto-Eskimo *neqe became Central ● various degrees of consonant and vowel cluster Alaskan Yup’ik neqa ‘food’, Iñupiaq niqi ‘food’) simplifications (cf. Iñupiaq aglaun ‘pen’, ● more or less complex effects of intonational Greenlandic allaat ‘pen’, in which gl became ll stress; in stressed syllables, for example, the and au turned into aa), and vowel is often lengthened (for more on Yupik ● a tendency to merge parts of the tense system prosody, see Krauss 1985). most important in narration, with most extensive merging in Alaska and least in Greenland. There are some nonnegligible syntactic differences between the languages, although these have not yet In Alaska and Canada, large numbers of English loans been well described. Siberian Yupik languages have a are found; in Eastern Canada, there are also a few loans number of English loan words, from contact with nine- from French and German through the influence of mis- teenth-century whalers, and Alaskan Yupik languages sionaries. In Greenland, on the other hand, loans are pre- have a large number of Russian loans from eighteenth dominantly from Danish. The status and viability of the to nineteenth century Russian colonization, as well as language are quite different in the different regions. In twentieth-century English loans. Most Yupik lan- Alaska and Western Canada, the language is severely guages are severely endangered today, with numbers endangered, with only about 3,000 speakers, almost of speakers ranging from 70 (Naukanski) to 1,300 none of whom are children. In Eastern Canada, there are (Central Siberian Yupik). The notable exception is about 20,000 speakers, but there is widespread bilingual- Central Alaskan Yup’ik, with about 10,000 speakers ism in almost all age-groups and a growing tendency for 298  ETHNICITY AND LANGUAGE English to replace Inuktitut. Active efforts are under way Jacobson, Steven A. 1995. A practical grammar of the Central to reverse this process, including the encouragement of Alaskan Yup’ik Eskimo language. Fairbanks: University of Alaska, Alaska Native Language Center. Inuktitut programs in schools. In Greenland, however, Kleinschmidt, S. 1851. Grammatik der Grönländischen Sprache over 95% of the native population of some 50,000 are (Grammar of the Greenlandic Language). Berlin: G. Reimer. speakers of Kalaallisut, and the language is thriving. Krauss, Michael (ed.) 1985. Yupik Eskimo Prosodic systems: descriptive and comparative studies. Alaska Native Language Center Research Papers 7. Fairbanks: ANLC. References MacLean, Edna Ahgeak. 1986. North Slope Iñupiaq grammar: first year, 3rd edition revised. Fairbanks: ANLC. Bergsland, Knut. 1997. Aleut Grammar/ Unangam Tunuganaan Menovshchikov, Georgiy Alekseevich. Yazik Sirenikskikh Achixaasix, Research Paper Number 10. Fairbanks: Alaska Eskimosov: Fonetika, Ocherk Morfologii, Teksty i Slovar Native Language Center. (The language of the Sireniki Eskimos: phonetics, morpho- de Reuse, W.J. 1994. Siberian Yupik Eskimo. The language and logical sketch, texts, and dictionary). Moscow and its contacts with chukchii. Salt Lake City: University of Leningrad: Isdatelstvo Nauk. Utah Press. Tersis, Nicole, and Michèle Therrien (eds.) 2000. Les Langues Dorais, Louis-Jacques. 1996. La Parole Inuit: Langue, Culture Eskaléoutes: Sibérie, Alaska, Canada, Groënland (The et Société dans l’Arctique Nord-Américain (The Inuit Word: Eskimo-Aleut Languages: Siberia, Alaska, Canada, language, culture, and society in the North American Greenland). Paris: CNRS. Arctic), Paris: Peeters. Veniaminov, Ioann. 1846. Opyt grammatiki aleutsko-lis’evska- Fortescue, Michael. 1984. West Greenlandic. London: Chroom go yazyka (An essay on the grammar of the Fox dialect of Helm. the Aleut language). St. Petersburg: Academia Nauk. Fortescue, Michael. 1998. Language relations across Bering ANNA BERGE Strait: reappraising the archaeological and linguistic evi- dence. London and New York: Cassell. See also Canada Ethnicity and Language The words ‘ethnicity’ and ‘ethnic’ were in common people think you are. We find it convenient in certain use at the end of the 20th century in contexts so wide- contexts to use the phrase ‘ethnic group’ for a wide ly disparate that no common definition will suffice to variety of minorities in America and other countries: unite the variety of meanings. We speak of ‘ethnic Irish Americans, Italian Americans, Asian Americans, studies’, ‘ethnic groups’ and ‘ethnic neighborhoods’, immigrant South Asians and their children, Native ‘ethnic and racial groups’, of ‘ethnic revivals’, and Americans, African Americans, Roma (Gypsies) in the ‘ethnic cleansing’—a euphemism for genocide that many countries where they live, French Canadians in came into use in the 1990s when certain countries tried Canada and Maine, Kurds in Turkey, Basques in Spain to drive minority ‘ethnicities’ from their territory and France. In recent years, advocates of the interests through terror and murder. In the 2000 census of the of gay and deaf communities in America have argued United States, although the Census Bureau of the for the benefits of using the term ‘ethnic’ in referring United States used traditional categories of ‘race’ such to these groups. ‘Ethnicity’ in current usage is so elas- as White, Black, Asian, or Pacific Islander, American tic and so convenient a term of reference that it may Indian or Alaska Native, Other, and Multiracial in perhaps most usefully be defined simply as the gathering census data, any of these might equally well ‘Other’—a member of any minority group in a coun- be regarded as ‘ethnic’ classifications in today’s lan- try who retains or is thought to retain the distinctive guage usage. It is unclear where ethnicity leaves off characteristics of that group. and race begins. ‘Hispanic’, in the 2000 census, was Otherness links ethnicity and language. Language stipulated by the Census Bureau as ‘may be of any always has a setting. If this setting is the majority of a race’—which indicates the extent to which ‘race’ and country’s population, then the language of that major- ‘ethnicity’ overlap in contemporary discourse. ity will be the ‘language of the country’. If this setting Ultimately, as we use the word today, ethnicity is is an ethnic minority of a country’s population, then not a matter of strict definition. It is a matter of identi- the language of that ethnicity will either be a language ty: you are what you say you are and what other different from the language of the country or a dialect 299 ETHNICITY AND LANGUAGE of the majority language that nevertheless differs sig- dents of ethnicity now have reservations about the nificantly from it. Language does not exist apart from melting pot theory. Nathan Glazer and Daniel Patrick ethnicity; it is a visible badge of ethnicity. A different Moynihan, in their influential book Beyond the melt- ethnicity therefore often implies a different language, ing pot (1963), set the tone for contemporary views or at least a different variant of a majority language. when they insisted that ‘the point about the melting (We sometimes refer to ethnic dialects as ‘ethnolects’, pot is that it did not happen’. Ethnic minorities have modeled after the pattern of ‘dialect’ and ‘idiolect’, the proved to be far more tenacious in clinging to aspects speech of a single individual.) of their Otherness—their ethnic identities—than the From the end of the Civil War to the end of World melting pot theory assumed. Some minorities, the War I (1865–1918), the population of the United Amish, for example, or Chasidic Jews, have success- States grew rapidly through immigration. There were fully resisted any but superficial assimilation into prejudices and laws against immigration of members mainstream American life. of ‘races’ thought to be inferior or ‘incompatible’ with However, to say that the melting pot is a faulty the majority population of America—against Asians metaphor for the realities of American assimilation is and Africans for example—but for others, gaining less true of language than it is of such qualities of eth- admittance to America was not difficult. This led to nicity as religion, cuisine, marriage patterns, and massive immigration from Ireland, Germany, Italy, social customs. Foreign languages do not normally Poland, Russia, Scotland, Sweden, England, and other survive in America in the second generation except in predominantly European countries. Jews, who had areas where a large number of immigrants are concen- faced discrimination and pogroms in the old country, trated, notably Spanish- speakers from many different emigrated from eastern Europe in large numbers to Spanish-speaking countries in Miami, New York City, America during this period. southern Texas, and southern California. There are Initially, the new immigrants tended to settle where pockets here and there in America where foreign lan- earlier immigrants from their country and members of guages defy the rule of second-generation loss, but their family had settled before them. The Lower East there are always special circumstances at play in such Side in Manhattan became a camp of crowded tene- cases. Examples are the preservation of the Basque ments divided along ethnic boundaries: Italian, language among shepherds in northern California and Jewish, Irish, German. Every great city of the north- Québec French in northern Maine. The most common eastern United States—Philadelphia, Pittsburgh, complaint of immigrants who remain in the United Boston—had similar concentrations of immigrants States is that their children lose the language; they joined by ethnicity. learn English and rarely have native fluency in the lan- In most cases, the first wave of immigrants retained guage of their parents. the language they had brought with them—Jews When the first language of immigration declines, Yiddish, Italians Italian, Germans German, and so what is often left behind as a residue is an ‘ethnic on—at least until their sons and daughters, born and accent’—a pattern of speech influenced by the immi- educated in the United States of America, grew up grant language and noticeably different from standard without fluency in the ancestral tongue. The life of the American English. Thus, the pronunciation of Long Lower East Side of Manhattan at the turn of the twen- Island, which in standard American pronunciation is tieth century was vibrant with newspapers, magazines, [lɔŋ ajlənd] (without a [g]), is often pronounced social clubs, and theaters in the foreign languages of [lɔŋgajlənd] (with a [g]) in the ‘Jewish American the immigrants. German-language newspapers, accent’ of the northeast. This is of course, a stereotype, German-medium schools, and German-language since by no means do all or most ‘Jewish Americans’ churches flourished in areas of heavy German immi- even in the northeast have this pronunciation; but to gration—Wisconsin and central Texas, for example— the extent that it does occur, it is a carryover from the down to World War I, when war with Germany made sound patterns of the Yiddish language. Similarly, dese pro-German attitudes unpopular and even dangerous. [diz] and dose [doz] for these [ðiz] and those [ðoz] are Nonetheless, even into the 1960s one could hear found in many ethnic accents in which the source lan- German spoken in America by people who had never guage does not have the fricatives [θ] and [ð]. set foot in those German-speaking countries from Not all immigrants who came to America were which their grandparents and great-grandparents had speakers of languages other than English. The Irish emigrated. began coming to America in large numbers during the America is often referred to as a ‘melting pot’, period of the Great Famine (1846–1851) when the implying that different ethnicities merge into one uni- Irish potato crop failed through disease, and going to form ‘American identity’ after the first generation of America was often the only alternative to death by immigrants has assimilated to American life. Most stu- starvation. Relatively few of these immigrants were 300  ETHNICITY AND LANGUAGE monolingual in the Irish language, the Celtic language such as New York, Philadelphia, Chicago, Detroit, and indigenous to Ireland (sometimes called ‘Gaelic’, Los Angeles. The sociolinguist William Labov, in pio- although ‘Irish’ is the preferred designation). Most neering research undertaken in the 1960s and 1970s, spoke English, but the variety of English they brought established that AAVE, far from being the ‘corrupt, with them was ‘Irish English’, which is different from degenerate, ungrammatical, bad’ English that linguis- more ‘standard’ versions of British or American tically naive observers had thought it to be, is as legit- English. imate a dialect of ethnic English as any other. AAVE is Characteristics of Irish English pronunciation rule-governed and has its own precepts of correctness include: (1) retention of historical /r/ after vowels in and incorrectness—it is therefore a language. words like car (as in standard American but not stan- Some of the major characteristics of the pronuncia- dard British pronunciation); (2) the use of ‘clear’ /l/ in tion of AAVE are the following: (1) deletion of /r/ all positions (i.e. the alveo-palatal [l] of leaf vs. the except before a vowel (dough and door rhyme); (2) velar [%] of full); (3) retention of the contrast between simplification of final consonant clusters (he picks is /w/ and /&/ (so that weather and whether, witch and he pick’, she found is she foun’); (3) metathesis (inter- which differ in their initial sound); (4) monophthongs change) of consonants in words such as ask (aks); (4) /e: o:/ in place of diphthongs [ej ow] in the vowels of neutralization of the contrast between [ε] and [I] words like face, take and goat, soak; (5) replacement before nasal consonants (compare Irish English of the fricatives /θ ð/ by stops and affricates (thin [θIn] above); and (5) replacement of postvocalic [θ] and [ð] pronounced as tin [tIn] or tthin [tθIn]); (6) retention of by [f] and [v] ([wIf] for with, [brəv ] for brother). vowel distinctions before /r/ (so that the vowels in Syntactically, a major systematic difference between words such as bird, learn, beard, turn, in Irish English, AAVE and standard American English is the presence are all different); (7) neutralization of the opposition or absence of the verb be to distinguish between habit- between /ε/ and /I/ before nasals (pin and pen are ual and momentary state (compare Spanish ser and homonyms, as in much of southeastern American estar): English). Syntactically, there is the occurrence of Lisa be sick. ‘Lisa is always sick.’ reflexive pronouns in sentences such as: And it’s him- Joe sick. ‘Joe is sick at the moment.’ self that told me … and … they were paying no atten- He be late. ‘He is chronically late.’ tion to anything at all as long as themselves were well. He late. ‘He’s late this time.’ Then, there is the curious matter of the ‘after perfect’: I’m after doing it already, She understands; she’s after There is scholarly dispute about the origins of havin’ children herself, and They seemed pretty cool, AAVE. The leading opinion at present is that most of for what they were after goin’ through (in place of the ‘nonstandard’ features of AAVE can be traced back standard I’ve done it, She understands; she’s had chil- to the influence of the African languages spoken by dren herself, and They seemed pretty cool after all they the slaves, who were taken from many different tribes had gone through). and languages of mostly western Africa and acquired As above in the case of the stereotypical ‘Jewish English only after they were forced into slavery in American accent’, most Americans of Irish descent America under English-speaking masters and over- have long since accommodated their speech to seers. According to this theory, AAVE arose as a pid- General American, and none of these Irish-English gin language based on English in which slaves traits remain in the way they talk. But some of the speaking different tribal languages could communi- ‘Irishness’ would be apt to persist for several genera- cate. This pidgin language then became a creole lan- tions, especially in areas of big cities such as Boston, guage as the children of the slaves grew up speaking it, New York, and Chicago where the Irish have been res- and from this creole AAVE of the present day has ident for a long time and where in many cases the pop- developed. ulation has been steadily replenished by immigration It must be emphasized that individual differences from Ireland. To the extent that Irish-English speech matter. African Americans, or members of any ethnic patterns do persist, they make for what can be called group who grow up in contact with standard American an ‘Irish ethnic accent’. English, will speak standard American English. A African American Vernacular English (AAVE) has blindfolded listener could not distinguish their speech been the most intensively studied ethnic dialect of from anyone else’s. It is a matter of educational oppor- English. (The terminology is not settled. What is here tunities, mobility, and one’s language contacts. called AAVE is variously referred to as Black English, Another ethnic dialect of English is Chicano African American English, and Ebonics.) AAVE is the English, the language of Mexican Americans in the dialect of English spoken by African Americans living southwestern United States and in urban areas to which in the ‘inner city’—the ghetto—of large urban centers its speakers have migrated. Characteristics of Chicano 301 ETHNICITY AND LANGUAGE English are: (1) substitution of tense vowels for lax for the dissolution of a country. It remains to be seen vowels (sit and seat are both pronounced as seat); (2) whether Québec will remain a part of Canada or substitution of ch [tʃ] for sh [ʃ] (chew and shoe are secede. In Belgium, a virtual language war between both pronounced like chew [tʃu]) and vice-versa French and Flemish (similar to Dutch) threatens this ([ʃεk] for check); (3) consonant cluster simplification, country with its weakly fused identity. Here, as as in par’ for part and He like’ her for He likes her); always, ethnicity is bound up with language, econom- and (4) nonstandard patterns of stress and intonation. ic grievances, and demands for power. The lines of A key ingredient of ethnicity is language. guerilla warfare in Sri Lanka are drawn between Tamil Language, that is to say, helps define our place in the Hindus and Sinhalese Buddhists—and between the world; it can serve either as a sign of membership in Tamil and Sinhalese languages. The demands for inde- the community or as a reason for exclusion from mem- pendence of the Baltic states, Latvia, Lithuania, and bership in that community. In many parts of the world, Estonia, formerly ‘republics’ of the Soviet Union, ethnic unity and cultural identification are often were intimately bound up with fears for the future of defined by language rather than by geography or reli- their respective ethnicities—and languages—in a sea gion. This is notably true of Arabic, whose speakers— of Russianness. the Arabs—base their identity in large measure on the One final aspect of the relationship between ethnic- use of a common tongue. The locus of Bengali ethnic ity and language deserves attention: the role of lan- identity resides in language despite the division of the guage in preserving ethnic identity, especially in speakers of Bengali between two countries, India and diaspora settings. The world has seen countless Bangladesh, the number of speakers living in instances of peoples forced from their homelands Bangladesh being about 100,000,000 and 68,000,000 because of their ethnicity. In most cases, diaspora leads in India. The Bengali language—a language with an to loss of language and ultimately of ethnic identity. ancient and much-revered literary history—is the prin- The celebrated counterexample to this is the role of the cipal basis of ethnic unity. This is clear from the name Hebrew language in the maintenance of Jewish ethnic- for itself taken by the new nation of eastern Bengal, ity. Hebrew was traditionally the language of the formerly East Pakistan, following the 1971 war of Jewish people; however, it had become extinct as their secession from Pakistan. The noun Bangladesh is spoken language by the beginning of the Common Era. composed of bangla plus desa, the latter meaning It was maintained as the language of ritual, prayer, and ‘country’. The first part of the compound does not disputation among rabbis. After the destruction of the mean the Bengali people or the territory of Bengal; the Temple in 70 CE, Jews were dispersed throughout the term bangla refers specifically to the Bengali lan- then known world. Nevertheless, they preserved their guage: Bangladesh = ‘land of the Bengali speakers’. identity, Otherness, and distinctive ethnicity. The The case of the French language in Québec demon- Hebrew language, which was reborn as a spoken lan- strates a particularly strong association between eth- guage in Palestine in the nineteenth century, was part of nicity and language. Six million French speakers, five the glue that held Jewish ethnicity together through million of them in Québec, compose about one-quar- almost 2,000 years of diaspora. ter of the population of Canada. The Québecois To paraphrase the great linguist Edward Sapir, we (speakers of French resident in the province of should never make the mistake of confusing a lan- Québec) regard themselves as a distinct island in a sur- guage with a dictionary and a grammar. Both the effect rounding sea of Anglophones (monolingual English and the affect of language go well beyond words and speakers). They resent the English language and fear rules of grammar. Language touches us in the deep its spread in Québec; they resent the historical domi- places of our being—in our identity, in our sense of nation of their economy and culture by Anglophones; where we belong. One of the most sensitive of these and they resent immigrants who want their children to places is our ethnicity. In ethnicity begins the true be educated in both English and French. Language is study of language as a badge of identity. intimately tied to Québecois ethnicity—to Québecois identity. Worship of the French language is the almost sacral force that fortifies and unifies the movement for References an independent Québec. Chiswick, Barry R. (ed.) 1992. Immigration, language and eth- Language is a major symbol of ethnicity, often the nicity: Canada and the United States. Washington, DC: AEI major and most tangible symbol. In this role, language Press. has always been a force both for unity and division in Dow, James R, (ed.) 1991. Language and ethnicity. Amsterdam, Philadelphia: J. Benjamins Publishers. the world. It has helped to unify countries—as English Fishman, Joshua A. 1989. Language and ethnicity in minority makes it possible for Indians from all parts and eth- sociolinguistic perspective. Clevedon, Avon, England, nicities of India to communicate—and it can be a force Philadelphia: Multilingual Matters, Ltd. 302  ETHNOGRAPHY OF COMMUNICATION Glazer, Nathan, and Daniel Patrick Moynihan. 1963. Beyond Shukla, Hira Lal. 1985. Language, ethnicity and history: the melting pot. Cambridge, MA: MIT Press. dimensions in anthropological linguistics. New Delhi, India: Gleason, Phillip. 1992. Speaking of diversity: language and eth- B.R. Publishing Corporation. nicity in twentieth-century America. Baltimore: Johns Van Horne, Winston A. (ed.) 1987. Ethnicity and language. Hopkins University Press. Milwaukee: University of Wisconsin System, Institute on Haarman, Harold. 1986. Language in ethnicity: a view of basic Race and Ethnicity. ecological relations. Berlin, New York: Mouton de Gruyter. ROBERT D. KING Lourie, Margaret A., and Nancy Faires Conklin (eds.) 1978. A pluralistic nation: the language issue in the United States. See also African American Vernacular English; Rowley, MA: Newbury House Publishers. Sociolinguistics Ethnography of Communication Ethnography of communication lies at the intersection work by using concepts like ‘speech community’, of anthropology and linguistics. Until the 1960s, ‘communicative situation, event, and act’, and ‘compo- anthropology and linguistics had existed largely in iso- nents of speaking’. lation from each other. Although they devoted much of A speech community consists of members who not their time to the study of mythology, folklore, and ritu- only speak a common language but also interpret the als, in which language clearly played an important role, interaction in a similar way. Besides language, mem- few anthropologists studied language in a systematic bers of a speech community share other aspects of manner. In the meantime, few linguists examined the common ground, such as social likeness or bonds, relationship between language and culture systemati- nationality, ethnicity, religion, and personal interests. cally. It is against such a background that ethnography These common grounds can partly account for the fact of communication came into being. The primary con- that most speech communities are bounded geographi- cern of this cross-disciplinary approach is to describe cally, although living in the same region is not a suffi- and explain the patterns of communicative behavior of cient prerequisite of belonging to the same speech people from a specific culture, across different cultures, community. The way that speech community members or both. The findings of the study are used to illuminate share a common language or norms of interpretation is theories of language, as well as theories of culture. largely a matter of degree, and it is up to an individual Patterns of communicative behavior are regularities to decide how much commonality is required. A person that occur so repeatedly that they become predictable. may simultaneously belong to several discrete or over- These patterns can be discerned in ritualized events, lapping speech communities. Britain and the United such as funerals and wedding ceremonies, in which States may, as a whole, be regarded as one speech com- people are often unwittingly compelled to repeat munity if a researcher is interested in the comparison of stereotyped expressions. In those not-so-ritualized Western and Eastern cultures. However, when it comes events, patterns also exist and underlie what people say to the comparison of British and American ways of and how they interact, despite the fact that on the sur- speaking, they have to be treated as two different face participants have every right to choose what to say. speech communities rather than as a single unit. However, patterns are not necessarily universal. The second essential constituent in the ethnography Different cultures may have divergent norms governing of communication framework is communicative situa- the same communicative behavior. A typical Chinese tion, event, and act. This is a hierarchical concept, with way of greeting (e.g. Qu na? [literally, ‘Where are you the communicative situation being the general setting going?’] or Chi le ma? [literally, ‘Have you eaten?’]) (such as a dinner party) and creating the broad context of may sound prying or impolite to an English speaker communication. A communicative situation may consist brought up in a European culture. With a view to of one or several communicative events, which refer to recording and analyzing culture-sensitive patterns specific activities (such as a joke). Communicative of communication, proponents of ethnography of events further consist of communicative acts, the mini- communication, such as Dell Hymes, John Joseph mal unit of analysis (such as greeting). Gumperz, Muriel Saville-Troike, and Richard Bauman In ethnography of communication, the key compo- and Joel Sherzer, have developed an analytic frame- nents of analysis are often subsumed under the 303 ETHNOGRAPHY OF COMMUNICATION mnemonic code SPEAKING, in which each letter rep- With its unique analytic framework, ethnography of resents one aspect of a communicative event. These communication contributes to the understanding of the are the following: interrelationship between culture and language by highlighting the role of culture in communicative • Setting and Scene: the concrete physical (e.g. behaviors. This has important implications for other location, time, and size of room) and abstract fields of study, such as psycholinguistics, second-lan- psychological (e.g. cultural definition of the guage acquisition, and cross-cultural communication. occasion) circumstances in which communica- For psycholinguistics, one of the most significant con- tion takes place. tributions made by ethnography of communication is • Participants: the role of each participant in the that it draws attention to how a child develops a par- interaction, their age, sex, social background, ticular way of speaking, apart from innate linguistic and their relationships with one another. knowledge, in the context of a particular culture. For • Ends: the purpose and expected outcome of an second-language acquisition, the SPEAKING model interaction. identifies what a speaker needs to know to communi- • Act sequence: order of actions, message form, cate appropriately within a particular communicative and message content. situation. Learning a second language involves not • Key: manner and the general tone of interaction. only acquiring general linguistic knowledge and inter- • Instrumentalities: medium of communication action skills but also social and cultural knowledge (i.e. spoken or written). that governs what to say to whom and when and how • Norms: norms of interaction and interpretation, to say it appropriately in relation to communicative i.e. what communicative behaviors are regarded goals. For cross-cultural communication, the diversity as appropriate by a speech community and how in the norms of interaction and interpretation in differ- communicative behaviors are construed by a ent cultures, emphasized by ethnography of communi- speech community. cation, highlights potential sources for • Genres: categories of communication, e.g. poet- misunderstanding and nonunderstanding in cross-cul- ry, prayer, or lecture. tural interaction. This is also an area that makes Among these components, norms of interaction and ethnography of communication interesting for busi- interpretation are the most culture specific in that peo- ness consultants and market researchers. ple of a particular culture may share different values or assumptions from other cultures. Consequently, the norms constitute the focus of studies in ethnography of References communication. Moreover, the eight components in Bauman, Richard, and Sherzer Joel. (eds.) 1989. Explorations the list do not hold equal positions in a communicative in the ethnography of speaking. Cambridge: Cambridge event. Some elements, such as setting, participants, University Press. Gumperz, John Joseph, and Dell Hymes (eds.) 1986. Directions and key, may determine what norms of interaction and in sociolinguistics: the ethnography of communication. interpretation are to be followed and what act Oxford: Blackwell. sequence an interaction may take. However, act Hymes, Dell. 1962. Foundations in sociolinguistics: an ethno- sequence may tell what norms of interaction and inter- graphic approach. Philadelphia: University of Pennsylvania pretation participants have adopted in conversations, Press. Saville-Troike, Muriel. 1989. The ethnography of communica- because unlike other variables, norms cannot be tion: an introduction. Oxford: Blackwell. directly observed, and their identification has to rely ZHU HUA on act sequences, as well as language content, in a conversation. See also Psycholinguistics; Sociolinguistics Etymology The field of etymology studies the origin and the earli- mon) is the word at the time and in the language in est meaning of semantic units, i.e. morphemes, words, which it emerged, i.e. when and where we can break it and phrases. The source (or to use the Greek term, ety- down into its constituent parts, understand its formation 304  ETYMOLOGY and the reasons that led to the emergence of its earliest phoneme y-, in light of sound changes established by the meaning, as well as understand its semantic motivation. comparative method, indicates the existence at an earli- The word soldier is borrowed from Old French sol- er stage of the sound i in the second syllable, which was dier, which in turn developed from Medieval Latin subsequently lost before the historical period. Thus, we solidarius ‘one serving in the army for pay’. This is its can establish that ‘fist’ was pronounced *fu¯sti in the origin, as in this form the word can be explained as a Proto-West-Germanic period. derivative with the suffix -arius, which among other A word from a prehistoric period, the form and things formed terms for individuals (cf. Medieval meaning of which are established through the applica- Latin commissarius  English commissary or tion of the comparative method, is called a reconstruc- Medieval Latin officarius  (Old) French officier  tion. An asterisk before the reconstructed form English officer), of Medieval Latin solidus ‘gold coin’, indicates that it is not attested in written documents. which, after applying the respective historical sound Despite the fact that the reconstructed *fu¯sti is not change rules is preserved in French sou and Italian found in other Germanic languages, this form is not yet soldo ‘wage’, pl. soldi ‘money’. The etymological the origin, as the word cannot yet be analyzed into its explanation of the word soldier is thus resolved, as the original morphemes and thus its original meaning can- mathematical product of the meanings, i.e. ‘paid (mil- not be established. The comparative method allows itary) individual’, along with knowledge of the comparison between the West German reconstruction Medieval military situation, corresponds to its formal and Proto-Slavic *pe' stь ‘fist’, which can be recon- construction: solidus and -arius. It is not important for structed from materials in modern and historical Slavic the etymological analysis of the word to indicate that languages, e.g. Slovene pest, Russian pjast’, Polish solidus in turn is an ellipsis for the collocation num- pie' ´sc´, and Old Church Slavic pe' stь. In the comparison, mus solidus ‘solid coin’, nor that the suffix -arius was the following comparative-linguistic sound laws are originally adjectival, cf. Latin manuarius ‘of, or applied: (1) Germanic f developed from Indo-European belonging to the hand’ from manus ‘hand’. p (Grimm’s Law); Proto-Slavic preserves Indo- The discovery process for etymological explanation European p. (2) Germanic u¯ before h developed from of individual words (morphemes or phrases) normally un, which in turn goes back to Indo-European n• , which consists of three parts. The first is a synchronic docu- in Proto-Slavic normally develops into the nasal vowel mentation of the word in the language in question and its e' . The development of Germanic long u¯ thus proves philologically verifiable development, e.g. fist ‘the hand that the word is Proto-Germanic and that it goes back closed tightly with the fingers bent against the palm’, to *funhsti-. The sound h in this position could have which is attested with the same meaning in its Old developed either from Indo-European k or kw, which in English predecessor fy-st. Because the word is attested in the position before another consonant in Proto-Slavic the earliest Old English texts, it can be assumed that it would have disappeared. (3) Proto-Germanic and existed in the preliterate period, i.e. it is prehistoric. The Proto-Slavic s both come from Indo-European s. (4) second step in the etymological process is to establish Proto-Germanic and Proto-Slavic t both come from the direction of further analysis. The question must be Indo-European t. (5) The fronting of Old English y- asked: did the word arise in a continuous linguistic indicates an older stem ending in -i, which is directly development—which for English is understood as the affirmed by the Slavic material. The mathematical development from the prehistoric, preliterate layers intersection of all possible reconstructions offered by Proto-Indo-European  Proto-Germanic  Western the Germanic and Slavic material results in the possi- Germanic through the historical stages of Old English to bilities *pn• ksti- and *pn• kwsti-. At this point the com- Modern English—or was it at some point borrowed parative-linguistic part of the etymological process from another language, a substratum, adstratum, or ends and the discovery of its origin and analysis into its superstratum? In view of the fact that the appropriate constituent parts begins. This is done in the framework source form in the languages from which the word might of Proto-Indo-European word-formational processes. have been borrowed—Romance, Celtic, and Northern It turns out that the second of the possibilities men- Germanic, from which the majority of borrowings tioned above, *pn• kwsti-, is more likely, as it can be came—is lacking, and, on the other hand, the correspon- derived from Indo-European *pn• kw-stHi-, which can ding Western Germanic cognates are attested (cf. Old be segmented into a compound with the first element High German fu¯st [ G. Faust], Old Frankish fest), it *penkwe ‘five’ in the zero-grade and a noun derived can be posited that the word arose at least in the Proto- from a verb with the root *staH- ‘to stand’, also in its West-Germanic period if not earlier. The comparisons at zero-grade variant. The etymological process is this stage can be affirmed or refuted by the comparative completed when it is discovered that the original form method, which is the fundamental tool of the etymolo- must have meant ‘a standing of five (fingers)’. The gist. In the example presented here, the Old English explanation is all the more likely as it is typologically 305 ETYMOLOGY paralleled by *mu-stH-i- ‘(small) fist’ (attested in occurs in borrowed words with minor ‘corrections’ that Vedic mus

tí- and Tocharian B mas´ce), in which the seem to speakers to be native derivatives, e.g. German first element comes from Indo-European *meue ‘four’ Abenteuer ‘adventure’ is a borrowing of French aven- (Luwian mauwa-), and the rest is identical to the ture. However, it has been connected with German Germanic-Slavic word for ‘fist’. Abend ‘evening’, apparently on the view that adven- Folk etymology is a change in a word or phrase, the tures are as a rule evening events. Further, sparrow constituent parts have become incomprehensible, and grass is found in the eastern part of the United States as as such is reanalyzed with morphemes that are mean- a ‘correction’ of the unintelligible plant name aspara- ingful. The potential for a folk etymology arises when gus, itself a borrowing via Latin of Greek aspáragos. there is a coincidental similarity between the actual morphemes and those—usually with a minor phono- References logical or morphological ‘correction’—that speakers Malkiel, Yakov. 1976. Etymological dictionaries. A tentative see in the word on the principle that ‘it makes sense, typology. Chicago and London: The University of Chicago therefore it must be true’. For example, the common Press, ix +144pp. name of a bird wheatear (Oenanthe oenanthe) is folk Malkiel, Yakov. 1993. Etymology. Cambridge: Cambridge etymologized as a compound of wheat and ear. This University Press, xii + 223pp. Onions, C. T. 1966. The Oxford dictionary of English etymolo- form in turn is a singularized form of an earlier form gy. Oxford: Clarendon Press, xvi + 1025pp. that had been felt to be a plural, wheatears. In fact, the Poruciuc, Adrian. 1991. Folk etymology (in English and else- word is a compound from Old English hwι¯t ( Modern where). Kansas Working Papers in Linguistics 16, pp. 53–8. English white) and arse, a reference to the bird’s MARKO SNOJ prominent white rump. Folk etymology frequently See also Morpheme; Semantics Euphemism Meanings of words and phrases may be divided into discuss the body’s excretory processes. It has only been denotation (the referent of the term) and connotation in the last 30 years that ‘shit’ was admitted as an entry in (the associated properties, including social import). If most unabridged dictionaries. Dictionary notes on usage either the denotation or connotation of a word or phrase point out that ‘shit’ is considered ‘vulgar’. In polite com- is offensive to social sensibilities, a ‘softer’, more pany, one is expected to refrain from using this word. polite form may be used. Such forms are euphemisms. Should it slip out as an exclamation or as a part of an When the possible offense in a word lies in its con- utterance, one is expected to apologize for the breach of notation, a word with the same or a similar referent, but etiquette. In settings, such as medical interviews, where lacking this connotation, may be sought. The word excretory processes must be discussed, euphemisms ‘ofay’ refers to Caucasians, but carries with it connota- appear. Biomedical labs refer to ‘stool samples’. Nurses tions of opprobrium. If a speaker wishes to denote this inquire about the patient’s ‘bowel movements’. group of people, but does not wish to insert negative Euphemisms change with register shifts, as formality ‘racial’ stereotypes, distancing as ‘out-group’, hatred or levels and statuses of interlocutors vary. Children who dislike, then she or he will use a euphemistic word, must discuss their excretions with the parents during the such as ‘White’ or ‘Caucasian’. While a given speech period of their lives euphemistically known as ‘potty’ community may largely share connotations and can training learn to use terms acceptable to their caregivers, thus count on words ‘meaning’ the same thing to all in- often including ‘b.m.’, ‘doodoo’, and ‘caca’. group interlocutors, connotations are not always shared Euphemisms are subject to pejoration. Constant use among all speakers of a language. For some speakers, in reference to something that is itself considered dirty ‘Aryan’ might serve as a euphemism for ‘ofay’; for oth- or vulgar pollutes the terms. As long as there is social ers, the connotations of ‘Aryan’ are also negative, stigma attached to the referent, some (or all) of it may although the non-in-group scope of this term is larger. transfer to the euphemistic term. In the United States, While societies differ in their ascription of positive ‘nigger’ has long been considered a ‘hate’ word, not apt and negative values, all societies have rules of appropri- for polite conversation. In the early 1900s and through- ateness of topics for particular groups and settings. In out the civil rights movement in the 1950s, ‘colored American society, it is generally considered bad form to people’ was the phrase of choice. It is preserved today 306  EUROPEAN TRADITIONAL GRAMMAR in the name ‘National Association for the Advancement embarrassment. The term ‘woman’ itself carries sexu- of Colored People’, which denotes a social justice al connotations; whereas the term ‘lady’, being more organization founded in 1909 by Ida Wells-Barnett, refined, masks these connotations. Assuming that W.E.B. DuBois, Henry Moscowitz, Mary White ‘Jane’ is a 13-year-old girl, compare the following: (a) Ovington, Oswald Garrison Villiard, and William Jane has become a young lady. (b) Jane has become a English Walling. In the decade following the 1950s, cul- young woman. When speakers say that a woman who tural revitalization groups adopted the term ‘black’. It is a doctor is a ‘lady doctor’, this indicates not so much spread from slogans such as ‘black is beautiful’ to com- her manners, but the fact of her gender, euphemistical- mon use as a group label. In the 1970s, Afro-American ly stated, and calls attention to it as a ‘marked’ case, came to be used. In the next decades, the phrase African not the cultural norm. As cultural norms shift, the ref- American replaced the hyphenated form. erences also shift. While a society may agree that a topic is ‘taboo’ or As cultural strife is resolved, euphemisms fall out socially loaded, so that reference to it must be oblique, of play, or lose their sense of masking. As tensions the euphemism of choice may vary, regionally or with- between earlier immigrants and Irish newcomers to in social strata. Euphemisms for public places where the United States subsided, ‘Gael’ and ‘Irish’ have fall- one may go to ‘shit’ include: the loo, the w.c., the toi- en out of a euphemistic relation to ‘Mick’ and ‘potato- let, the restroom, the washroom, the lavatory, the eater’. As new stresses arise, new lexicon masks and jensen, the john, the ‘ladies’, and the ‘gents’. outlines the divisions. Ageism, youth-centrism, and When the discomfort with a term stems from social legislation against forced retirement are reflected in stresses and when those social schisms become politi- the conversion of ‘old-age homes’ to ‘retirement vil- cized, the lexical choice is politicized as well. Thus, lages’, ‘nursing homes’, and ‘assisted living’ places euphemisms are used to achieve ‘politically correct’ for our ‘senior citizens’ or ‘golden agers’. Language parlance. Feminism has painted the use of generic ‘he’ forms and is formed by society. Euphemisms are the and ‘man’ as indicators of reactionary male hegemon- fresh coat of paint over the stress fractures in the social ic ideology, so that sentences such as ‘Man nurses his structure. young’ are sexist, as well as semantically odd. In those areas where pejoration has proliferated terms, cultural revitalization movements may seek to References reclaim elements. In many Latin American countries, Allan, Keith, and Kate Burridge. 1991. Euphemism & dys- the term indio ‘Indian’ has been an insult for centuries; phemism: language used as shield and weapon. New York: in the last 30 years, some groups in highland Mexico, Oxford University Press. Dickins, James. 1998. Extended axiomatic linguistics. Berlin Guatemala, and Peru are beginning to eschew the and New York: Mouton de Gruyter. euphemistic indígena and to self-denominate as indio, Enright, D.J. (ed.) 1985. Fair of speech: the uses of euphemism. as a marker of cultural pride. In North American, Oxford and New York: Oxford University Press. indigenous groups have been referred to by a variety Lakoff, Robin Tolmach. 1975. Language and woman’s place. of euphemistic terms: ‘Indian’, ‘AmerInd’, ‘Native New York: Harper and Row. Neaman, Judith S., and Carole G. Silver. 1999. The Wordsworth American’, ‘First Americans’, and ‘First Nations’. book of euphemism. Lincolnwood, IL: Wordsworth Editions Amid this multiplicity, one person’s euphemism is Ltd., NTC Publishing Group. another’s insult or effete affectation. Spears, Richard A. 1981. Slang and euphemism: a dictionary of One can trace areas of conflict, stress, and unease oaths, curses, insults, sexual slang and metaphor, racial within a society by mapping its euphemisms. Robin slurs, drug talk, homosexual lingo, and related matters. Middle Village, NY: David Publishers. Lakoff in her 1975 book Language and woman’s place notes that female sexuality is one such area of cultural JUDITH M. MAXWELL European Traditional Grammar The Greek term grammatike (Latin grammatica) of human knowledge—constituted our only sources of meant no more at first than the understanding of let- grammatical progress. ters. For the early periods of European history, philo- Plato (c. 429–347 BCE) provided a fundamental divi- sophical works—examining virtually the whole realm sion of the Greek sentence into a nominal and a verbal 307 EUROPEAN TRADITIONAL GRAMMAR component, and this remained the primary grammatical sections on syntax, i.e. sentence structure. Syntax was distinction underlying sentence analysis and word clas- dealt with extensively by Apollonius Dyscolus (sec- sification in all subsequent European linguistic descrip- ond century CE). He sharply distinguished form and tions. Aristotle (384–322 BCE) maintained this meaning, and he justified grammatical classifications distinction, but added a third syntactic component, the by reference to meaning rather than to form. His work, syndesmoi, a class covering what were later to be distin- however, is a case study of Greek, rather than a gener- guished as conjunctions, articles, and pronouns. al theory of syntax. The Stoic philosophical school, founded by Zeno Varro (first century BCE) is the first important Latin (c. 315 BCE), developed new methods and doctrines author with linguistic concerns whose work has been in philosophy and rhetoric and gave separate treatment transmitted. In the surviving part of his De Lingua to grammar, with great achievements: the number of Latina, he distinguishes word formation processes in word classes was increased, and more precise defini- which the class of the word is changed (e.g. derivation of tions and additional grammatical categories were a noun from a verb via the addition of an affix, such as introduced to cover the word and sentence structure the noun ‘runner’ from the verb ‘to run’) and class-pre- associated with these classes. Case in its modern usage serving inflections (e.g. tense markings on the verb, such as an inflectional category of nouns (and other words as ‘moved’ for the past tense of ‘move’). This distinction inflected like nouns) was the creation of the Stoics. between derivational and inflectional processes was not Case inflection became the fundamental division commonly made in antiquity. Varro also proposed his between noun and verb. The Stoics also introduced the own classification of Latin-inflected words by setting up distinctions between transitive and intransitive verbs, a system of four contrasting classes of inflections. i.e. verbs that require an object and verbs that do not. Aelius Donatus was the most influential Latin Building on the Stoic ideas, the Greco-Roman grammarian of the fourth century CE and dominated world devised technical grammars with an effective grammatical doctrine in Europe at least until the and lasting influence on the future. The framework of twelfth century. His Ars minor deals with the eight technical grammar in antiquity was the word-and-par- parts of speech in a question-and-answer format. His adigm model. A word-based grammar involves three Ars maior is more comprehensive and includes style main procedures: the identification of the word as an guidelines for Latin. isolable linguistic entity, the establishment of a set of Priscian’s voluminous grammar (c. 500 CE) was word classes to distinguish and classify the words in the most important late Latin grammar. He was under the language, and the working out of adequate gram- the spell of Greek grammar, accepting Apollonius and matical categories to describe and analyze the struc- his son Herodian as authorities and using the closest ture and formation of words (morphology) entering Latin translations for the Greek technical terms. into paradigms of associated forms. Priscian organized the morphological description of The first piece of technical grammar in Europe was the forms of nouns and verbs and of the other inflect- the Tekhne Grammatike of Dionysius Thrax (c. 100 ed words by setting up canonical, or basic, forms; BCE). Its first part is considered genuine; the rest is from these, he derived the other forms by a series of probably a later addition. The grammarian classified the letter changes, the letter being for him, as for the rest consonants of Greek and distinguished eight word of Western antiquity, both the minimal graphical unit classes: noun, verb, participle, article, pronoun, preposi- and the minimal phonological, or sound, unit. tion, adverb, and conjunction. The number of word In the first centuries after the collapse of the Western classes, with one change necessitated by the absence of Roman Empire, grammar served mainly practical and articles in Latin, remained constant until the end of the normative purposes. Later writings took the form of Middle Ages in the grammatical description of Greek commentaries and glosses, mainly based on Priscian. and Latin, and it had a very marked influence on the Donatus and Priscian shared the place of the principal grammatical analysis of several modern European lan- grammarians of the Middle Ages as teachers’ textbooks. guages. Each word class defined in the Tekhne In England, Bede and Alcuin wrote grammars of Latin Grammatike is followed by a statement of the categories in the seventh and eighth centuries, following Donatus. applicable to it: gender (masculine, feminine, or neuter), Until the twelfth century, Greek and Latin were type (primary or derived), form (simple or compound), regarded as the pinnacle of culture, which explains the number (singular, dual, or plural), and case (nominative, grammarian’s exclusive attention to these languages. vocative, accusative, genitive, or dative). The classifica- In the latter part of the Middle Ages, however, descrip- tion categories introduced for the verb are mood, voice, tions of other European languages appeared, serving type, form, number, person, tense, and conjugation. the ends of literacy, popular literature, and education- The main omission of the Tekhne from the stand- al standards. Irish grammarians showed a great deal of point of modern linguistics was the absence of any originality in their treatment of Irish; descriptions of 308  EVOLUTION OF LANGUAGE 1: OVERVIEW Provençal started appearing after 1240; an anonymous published in 1560, 1571, and 1640, respectively. The grammarian of the period demonstrated great insight seventeenth century witnessed the publication of into the description of Icelandic, especially in phonol- grammars of Japanese and Persian, and the first com- ogy and its graphemic representation. prehensive grammars of Chinese published in The Renaissance saw the first European grammars European languages appeared at the beginning of the of Hebrew and Arabic. Interest in these languages and eighteenth century. Confidence in the traditional the separate scholarly traditions in which they had Greco-Roman grammatical categories was weakened been treated contributed to further loosening the strict when it became known that Chinese generally uses focus on Latin and Greek. The first known native independent words where Greek and Latin use affixes, grammars of Spanish and Italian appeared in the fif- whereas Native American languages often use the teenth century, followed closely by French, Polish, and opposite strategy of expressing the equivalent of Old Church Slavonic grammars. The first printed Greek or Latin sentences in highly complex words. grammar of English appeared in 1586. The ‘discovery’ of Sanskrit and the emergence of The new grammars of modern languages paid great the so-called Neogrammarian theories in the nine- attention to the relations between spelling—now being teenth century, however, represented the most serious standardized in printing—and pronunciation. challenges, both to the concepts of European tradi- Although the confusing equation of letter and spoken tional grammar and to the idolization of Greek and sound continued, phonological inadequacies of exist- Latin. The fact that Sanskrit was found to be closely ing spelling were noted and criticized. The system of related to Greek and Latin stripped them of their eight word classes was challenged, with systems of attributed uniqueness, and the Neogrammarian focus fewer and more classes being proposed. on the development of languages over time put them English grammarians of the sixteenth, seventeenth, into historical perspective. and eighteenth centuries usually started with the Latin system of the eight Priscianic classes. They either fol- References lowed it or felt the need to express and justify their Keil, Heinrich (ed.) 1857–1870. Grammatici Latini, 8 vols. disagreements with it. The English articles, a, an, and Leipzig: Teubner. the, having no Latin counterpart, were not given the Kemp, Alan. 1987. The Tekhne Grammatike of Dionysius status of a part of speech, but were merely referred to Thrax: English translation with introduction and notes. The as notes or signs set before nouns to identify them as history of linguistics in the Classical Period, ed. by Daniel J. nouns. Others treated the articles as adjectives (still a Taylor. Amsterdam and Philadelphia: John Benjamins. Law, Vivien, and Ineke Sluiter (eds.). Dionysius Thrax and the subcategory of nouns), and Ben Johnson (1640) Techne Grammatike. The Henry Sweet Society studies in assigned them to a class of their own. The influence of the history of linguistics 1. Münster: Nodus Publikationen. the Latin tradition is seen in the retention of the adjec- Lepschy, Giulio (ed.) 1994, 1998. Storia della linguistica, tive within the noun class, although there are fewer Bologna: Il Mulino; as History of linguistics: Vol. II: classi- reasons for this theoretical choice in English than in cal and medieval linguistics, Vol. III: renaissance and early modern linguistics. London and New York: Longman. Latin. The preoccupation of most of the grammarians Robins, R.H. 1967. A short history of linguistics. London and with the participle, treated either (purely traditionally) New York: Longman, 4th edition, 1997. as a class in its own right or as a noun adjective hav- Taylor, Daniel J. (ed.) 1987. The history of linguistics in the clas- ing particular derivational associations with the verb, sical period. Amsterdam and Philadelphia: John Benjamins. is also a carryover from Latin grammar. Schneider, R., G. Uhlig, and A. Hilgard (eds.) 1867–1910. Grammatici Graeci. Leipzig: Teubner. In the New World, the first grammar of a Native American language, Tarascan (Mexico), was pub- CHRISTOS G. NIFADOPOULOS lished in 1558; thereafter, grammars of Quechua See also Grammar, Traditional; Grammar, Theo- (Peru), Nahuatl (Mexico), and Guarani (Brazil) were ries; History of Linguistics: Overview; Morphology Evolution of Language 1: Overview Although the details are disputed, it is apparent that of language must take into account the principles human linguistic ability has a biological basis. That noted by Charles Darwin; they remain central tenets of being the case, any meaningful study of the evolution evolutionary biology. 309 EVOLUTION OF LANGUAGE 1: OVERVIEW The evolutionary mechanism most often associated limited lexical and syntactic ability were present in the with Darwin is natural selection, which acts on ‘any earliest stages of the evolution of human language and variation, however slight and from whatever cause argues against theories that posit an abrupt transition proceeding, if it be in any degree profitable to an indi- that yielded human syntactic ability. Natural selection vidual of any species, in its infinitely complex rela- could have gradually enhanced preexisting rudimentary tions to other organic beings and to external nature…’ lexical and syntactic ability. In contrast, the fact that (1859; On the origin of species, p. 61). Because human chimpanzees totally lack voluntary speech suggests that language enhances virtually all aspects of our interac- the speech capabilities of early hominids were limited, tions with each other and external nature, selection for and this is consistent with studies of the skulls of extinct the biological characteristics that make language more hominids. These skulls suggest a chimpanzee-like vocal effective is to be expected. anatomy incapable of producing the full range of Comparative studies of primates suggest that lan- human speech sounds. guage may foster group cohesion and the formation of The archeological record also preserves artifacts large social groups working cooperatively by means of that hint at the cognitive and linguistic abilities of seemingly useless ‘gossip’. Other studies note that ancestral and related extinct hominid species. language allows information to be shared by many However, we must take into account the effects of the individuals, thereby fostering the solution of problems accelerating pace of human culture. Although we can through ‘collective insight’. Robust experimental data fly through the air, our cognitive and linguistic ability show that human speech itself is one of the keys to is not superior to that of our ancestors during the tens human linguistic ability. The process by which human of thousands of years in which horses provided the speech is produced and perceived overcomes the slow fastest means of land transportation. Moreover, genet- information transfer rate of the human auditory sys- ic data must temper conclusions based on the archeo- tem. As we talk, phonemes (speech sounds roughly logical record. Although tool technologies seem to equivalent to the letters of the alphabet) are melded have abruptly changed 40,000 years ago in Europe, it into syllables that are transmitted below the rate at is improbable that modern human linguistic ability which the human auditory system would have fused first evolved at that date. Genetic studies and the fos- individual sounds into an incomprehensible buzz. sil record show that modern human beings left Africa Studies of sentence comprehension reveal another role about 100,000 years ago and settled in most parts of for speech. Words are silently ‘rehearsed’ by using the the world between 60,000 and 40,000 years ago. neural mechanisms that regulate overt speech, to Because any child from any region of the world can maintain them in the neural buffer, verbal working acquire any language with native proficiency, the bio- memory, in which sentence comprehension takes logical capacities for language must have existed place. Other avenues of inquiry that use computer before humans left Africa. modeling techniques suggest that the complex syntax But some linguists question how human language of human language derives from limited neural mem- could have evolved by means of natural selection in ory and sentence-processing capacities. the brief 5-million-year period separating humans and Darwin also introduced the comparative method, living apes. As Darwin himself noted, natural selection which can yield reasonable inferences concerning the can ‘act only by the preservation and accumulation of evolution of the ‘derived’ characteristics that differenti- infinitely small inherited modifications’ (1859:95). ate a living species from its extinct ancestors. For exam- How then could he account for major adaptations to ple, comparative studies of living apes and humans new environments, such as the transition from aquatic show that upright bipedal locomotion involves certain to terrestrial life? Darwin’s solution, often termed derived anatomical features that can be traced back in ‘exaptation’ or ‘preadaptation’, rested on the observa- time by examining the fossil record. In this light, stud- tion that ‘an organ might be modified for some other ies of chimpanzees shed light on the evolution of human and quite distinct purpose’ (1859:190). Converging language. Although present-day chimpanzees are not evidence suggests that preadaptation played a part in members of the extinct 6- to 7-million-year-old species the evolution of the unique characteristics of human that was the common ancestor of humans and apes, they language, such as voluntary speech and complex syn- preserve many of the skeletal features of early tax. The anatomy of the face and tongue was modified hominids, including small brains and similar vocal to allow the production of sounds that make human anatomy. However, chimpanzees raised in human-like speech more efficient. Brain structures initially adapt- environments can acquire approximately 150 words and ed for motor control may have been the starting point the ability to comprehend sentences that have a simple for neural networks that regulate syntax and cognition. syntax with manual languages (American Sign In studying the evolution of linguistic ability, gener- Language or computer keyboards). This suggests that al evolutionary principles, such as Von Baer’s ‘law’, 310  EVOLUTION OF LANGUAGE 2: COGNITIVE PREADAPTATIONS provide another avenue of inquiry. Von Baer’s law Gibson, Katherine R., and Thomas Ingold. 1993. Tools and holds that the ontogenetic development, from fetal life language in human evolution. Tools, language and cogni- tion in human evolution. Cambridge: Cambridge University onward, of the derived anatomical and related behav- Press. ioral characteristics of a living species generally fol- Hauser, Marc D. 1996. The evolution of communication. lows a sequence that mirrors their occurrence in related Cambridge, MA: MIT Press. living and extinct species. Inferences on the evolution Kirby, Simon, and Morton Christiansen (eds.) 2001. Language of the anatomy of speech production, syllable structure, evolution. The states of the art. Oxford: Oxford University Press. vocabulary ability, and sentence structure have fol- Lieberman, Philip. 1984. The biology and evolution of lan- lowed from studies of the development of children. guage. Cambridge, MA: Harvard University Press. The exact course of the evolution of human lan- Lieberman, Philip. 1998. Eve spoke; human language and guage will probably never be known, because detailed human evolution. New York: W. W. Norton, London: timetables cannot be subjected to test. However, Picador, Macmillan. Lieberman, Philip. 2000. Human language and our reptilian because the mark of evolution is apparent when we brain: the subcortical bases of speech, syntax, and thought. study the human brain and body, further insights into Cambridge, MA: Harvard University Press. the evolution of human language will become apparent Mellars, Paul. 1996. The Neanderthal legacy: an archaeological as our knowledge of the biological basis of human lan- perspective from Western Europe. Princeton, NJ: Princeton guage and cognition advances. University Press. Mellars, Paul, and Christopher B. Stringer. 1990. The human revolution: behavioral and biological perspectives on the References origin of modern humans, Vol. 1. Edinburgh: Edinburgh University Press. Darwin, Charles. 1859. On the origin of species, London: Savage-Rumbaugh, Sue, and Roger Lewin. 1994. Kanzi: the Murray, Cambridge, MA: Harvard University Press, ape at the brink of the human mind. New York: Wiley. Facsimile edition, 1964. PHILIP LIEBERMAN Deacon, Terrance W. 1997. The symbolic species: the co-evolu- tion of language and the brain. New York: W. W. Norton. See also Animals and Human Language 4: Pri- Donald, Merlin. 1991. Origins of the modern mind. Cambridge, mates; Evolution of Language 2: Cognitive MA: Harvard University Press. Gardner, R. Allan, Beatrix T. Gardner, and Thomas A. Van Preadaptations; Evolution of Language 3: Physical Cantfort. 1989. Teaching sign language to chimpanzees. Preadaptation; Evolution of Language 4: Social Albany: State University of New York Press. Preadaptations Evolution of Language 2: Cognitive Preadaptations A preadaptation is a change in a species which is not a receiver’s mind about states of affairs in their shared itself adaptive but paves the way for subsequent adap- world. Preadaptations for language involved the fol- tive changes. For example, bipedalism set in train lowing capacities or dispositions: anatomical changes, which culminated in the human pre-phonetic capacity to produce speech sounds or vocal tract. Although speech is clearly adaptive, manual gestures. bipedalism is not itself an adaptation for speech; it is a pre-syntactic capacity to organize longer sequences preadaptation. This example involves the hardware of of sounds or gestures. language, the vocal tract. Many changes in our pre-semantic capacities to: (a) form basic concepts, species’ software, our mental capacities, were neces- (b) construct more complex concepts (e.g. proposi- sary before we became language-ready; these are cog- tions), and (c) carry out mental calculations over com- nitive preadaptations for language. plex concepts. A human language is a system mapping combina- pre-pragmatic capacities to: (a) infer what mental cal- tions of speech sounds or manual signs onto meanings. culations others can carry out, (b) act cooperatively, The use of language is a cooperative, social activity. (c) attend to the same external situations as others, and When humans use language, they cause inferences in (d) accept symbolic action as a surrogate for real action. 311 EVOLUTION OF LANGUAGE 2: COGNITIVE PREADAPTATIONS elementary symbolic capacity to link sounds or ges- imitation has been found in monkeys, in the form of tures arbitrarily with basic concepts, such that percep- ‘mirror neurons’, which fire both when an animal is tion of the action activates the concept, and attention carrying out a certain action, such as grasping, and to the concept may initiate the sound or gesture. when it observes that same action carried out by If some capacity is found in species distantly relat- another animal. A recurrent theory in phonetics is the ed to humans, this indicates that it is an ancient, prim- ‘motor theory of speech perception’, which claims that itive capacity. Conversely, if only our nearest relatives, speech sounds are represented in the brain in terms of the apes, possess some capacity, we can conclude that the motor commands required to make them. it is a more recent evolutionary development. Twin Although they cannot speak, our ape cousins have recurring themes in the discussion of many of these no trouble in recognizing different spoken human abilities are learned, as opposed to innate behavior, words. The capacity to discriminate the kinds of and voluntary control of behavior. Human languages sounds that constitute speech evidently preceded the are largely learned systems. The more ways in which arrival of speech itself. a species is plastic in its behavior, the more complex are the cultural traditions, including languages, that Presyntactic Capacity can emerge. Our nearest relatives, the chimpanzees, are plastic in a significantly wider range of behaviors Syntax is the stringing together of independent sub- than any other nonhuman animals; their cultural tradi- units into a longer signal. The oscine birds, in which tions are correspondingly more multifaceted while learned complex song is observed, are extremely dis- falling far short of human cultural diversity and com- tant relatives of humans. Many other birds, and more plexity. Combined with plasticity, voluntary control closely related species, including most mammals, do adds more complexity, and unpredictability, to pat- not, as far as we know, produce calls composed of terns of behavior. Much of the difference between independent subunits. Our closest relatives, the apes, humans and other species can be attributed to greatly do produce long calls composed of subunits. The long increased plasticity and voluntary control of these calls of gibbons are markers of individual identity, for preadaptive capacities. advertising or defending territory. The subunit notes, used in isolation, out of the context of long calls, are used in connection with territorial aggression, and it is Prephonetic Capacity not clear whether the meanings of these notes can be Chimpanzees cannot speak. They typically have little composed by any plausible operation to yield the iden- voluntary breath control. To wild chimpanzees, volun- tity-denoting meaning of the whole signal. tary breath control does not come naturally. On the Male gibbon singing performances are notable for other hand, chimpanzees have good voluntary control their extreme versatility. Precise copies of songs are over their manual gestures, although they are not as rarely repeated consecutively, and the song repertoires capable as humans at delicate manual work. A of individual males are very large. Despite this vari- preadaptation that was necessary for the emergence of ability, rules govern the internal structure of songs. modern spoken language was the extension of volun- Male gibbons use a discrete number of notes to con- tary control from the hands to the vocal tract. struct songs. Songs are not formed through a random Learning the controlled use of physical actions assortment of notes. The use of note types varies as a entails an ability to imitate. Imitation involves an function of position, and transitions between note impressive ‘translation’ of sensory impressions into types are nonrandom. (Mitani and Marler 1989:35) motor commands. Think of a smile. Without mirrors Although it is fair to call such abilities in apes or language, one can have no guarantee that a certain ‘presyntactic’, they are still far removed from the set of muscle contractions produces the effect one per- human ability to organize sequences of words into com- ceives in another’s face. Given the required voluntary plex hierarchically organized sentences. Little is yet control, and the anatomical hardware, imitation of known about the ability of apes to learn hierarchically speech sounds should be easier than imitation of facial structured behaviors, although all researchers seem to gestures, because one can hear one’s own voice. A expect apes to be less proficient at it than humans. capacity for imitation is found in a perplexing range of species. Some birds can imitate human speech, and Presemantic Capacities many other types of sound, as well. Dolphins can be trained to imitate human movements. A capacity for Basic Concept Formation imitation can evolve separately in different species, Many distantly related species lead simple lives, com- with or without the other necessary preadaptive pared to humans, and even to apes, and so may not pos- requirements for human language. A neural basis of sess very many concepts, but they do nevertheless 312  EVOLUTION OF LANGUAGE 2: COGNITIVE PREADAPTATIONS possess them. ‘Perceptual categorisation and the reten- problem-solving by animals, leading to ranking of var- tion of inner descriptions of objects are intrinsic charac- ious species according to how well they perform in teristics of brain function in many other animals apart some task involving simple inference from recent from the anthropoid apes.’ (Walker 1983:378). The dif- experience. Apes and monkeys perform closest to ference between humans and other animals in terms of humans in problem-solving, but their inferential abili- their inventories of concepts is quantitative. Animals ty falls short of human attainment. have the concepts that they need, adapted to their own physiology and ecological niche. What is so surprising Prepragmatic Capacities about humans is how many concepts they have, or are capable of acquiring, and that these concepts can go well Mind-reading and Manipulation beyond the range of what is immediately useful. Basic When a human hears an utterance by another person, concrete concepts, constituting an elementary preseman- he often has to figure out what the speaker intended; tic capacity, were possessed by our remote ancestors. this is mind-reading. When a human speaks, he or she Something related to voluntary control is also rele- usually does so with some estimation of how the hear- vant to presemantic abilities. We do not need to be er will react; this is social manipulation. Humans have stimulated by the actual presence of an object for a especially well-developed capacities for social manip- concept of it to be brought to mind. Some animals may ulation and mind-reading, and these evolved from sim- have this to a limited degree. When an animal sets off ilar abilities in our ancestors, still visible in apes. Social on a journey to its usual foraging ground, it knows intelligence, a well-developed ability to understand and where it is going, because it can get there from many predict the actions of fellow members of one’s group, different places, and even take new routes. Hence, the was a necessary prerequisite for the emergence of lan- animal is entertaining a concept of a place other than guage. Recent studies amply demonstrate these manip- where it currently is. But for full human language to ulations and mind-reading abilities in chimpanzees. have taken off, a way had to evolve of mentally reviewing one’s thoughts in a much more free-ranging Cooperation way than animals seem to use. People are able to understand the intended import of statements whose literal meanings are somehow inap- propriate, such as ‘It’s cold in here’, intended as a Complex Concept Formation request to close the window. To explain how we cope The ability to form complex conceptual structures, with such indirectness, traditional logic has to be sup- composed systematically of parts, is crucial to human plemented by the Cooperative Principle, which stipu- language. Logical predicate-argument structure under- lates that language users always try to be helpful in lies the messages transmitted by language. The words various specified ways. The use of language requires comprising human sentences typically correspond to this basis of cooperativeness. No such complex com- elements of a conceptual/logical representation. While munication system could have evolved in the absence apes may perhaps not be capable of storing such com- of assumed cooperativeness between members of the plex structures in their heads as humans, it seems cer- community. tain that they have mental representations of the Humans are near the top of the range of cooperative- predicate-argument form. Simply attending to an ness, compared with other species. The basis of cooper- object is analogous to assigning a mental variable to it, ation in social insects is entirely innate, and the range of which functions as the argument of any predicate individual cooperative behaviors is small. In humans, expressing a judgment made by the animal. The two building onto a general natural disposition to be coop- processes of attending to an object and forming some erative, cooperation on a wide range of specific group judgment about it are neurologically separate, involv- enterprises is culturally transmitted. Children are taught ing different pathways in the brain. This is true not only to be ‘team players’. No concerted instruction in coop- for humans but also for apes and closely related mon- eration exists outside humans, but there are reports of keys as well. It seems certain that all species closely cases where an animal appears to be punished for some related to humans, and many species more distantly transgression of cooperativeness. So the basis for coop- related, have at least this representational capacity. This erative behavior, and for instilling such behavior in oth- capacity is a presemantic preadaptation for language. ers, exists in species closely related to humans. Common Chimpanzees and bonobos, in particular, fre- Mental Calculation quently engage in many different types of reconciliation Humans are not the only species capable of reasoning and peace-making behavior. Dispositions to coopera- from experienced facts to predictions about nonexpe- tion and maintenance of group cohesion are pragmatic rienced states of affairs. There is a large literature on cognitive preadaptations for language. 313 EVOLUTION OF LANGUAGE 2: COGNITIVE PREADAPTATIONS Joint Attention asociations between signals and their meanings, to be Cats are hopelessly inept at following a pointing fin- discussed in the next section. ger; dogs are somewhat better. Language is also used to point at things, both directly and indirectly. Elementary Symbolic Capacity Linguists and philosophers call this ‘reference’. When a speaker refers to some other person, say by using a The sound of the word tree, for instance, has no icon- personal pronoun, such as he, the intention is to get the ic similarity to any property of a tree. This kind of hearer to attend to this other person. Successful use of arbitrary association is central to language. Human language demands an ability to know what the other linguistic symbols are entirely learned. This excludes person is talking about. A mechanism for establishing from language proper any possible universally instinc- joint attention is necessary. Human babies and chil- tive cries, such as screams of pain or whimpers of fear. dren are especially adept at gaze- and finger-follow- In the wild, there are many animals with limited reper- ing. The fact that humans, uniquely, have whites in toires of calls indicating the affective state of the ani- their eyes, probably helps us to figure out what other mal. In some cases, such calls also relate people are looking at. systematically to constant aspects of the environment. Primates more closely related to humans are better The best-known example is the vervet monkey alarm at following the human gaze than those less closely system, in which there are distinctive calls for differ- related. Chimpanzees follow human gaze cues, while ent classes of predator. There is no evidence that such non-ape species such as macaques fail to follow calls are ever learned to any significant degree. Thus, human gaze cues. But experiments on rhesus no animal calls, as made in the wild, can, as yet, be macaques interacting with other rhesus macaques taken as showing an ability to learn an arbitrary map- show that members of this species do follow the gaze ping from signal to message. of other members of their own species. Spontaneous Trained animals, on the other hand, especially apes, pointing has also been observed in captive common have been shown to be clearly capable of acquiring chimpanzees (who had not received language training) arbitrary mappings between concepts and signals. The and in young free-ranging orangutans. It thus appears acquired vocabularies of trained apes are comparable that animals close to humans possess much of the cog- to those of five-year-old children, with hundreds of nitive apparatus for establishing joint attention, which learned items. An ape can make a mental link between is the basis of reference in language. an abstract symbol and some object or action, but the circumstances of wild life never nurture this ability, Ritualized Action and it remains undeveloped. Short colloquial greetings such as Hello! and Hi! are The earliest use of arbitrary symbols in our species just act-performing words; they do not describe any- was perhaps to indicate personal identity. They thing, and they can not be said to be true or false, as a replaced nonsymbolic indicators of status such as description can. We can find exactly such act-perform- physical size, and involuntary indexes such as ing signals in certain ritualized actions of animals. The plumage displays. In gibbons, territorial calls also classic example of a ritualized action is the snarling bar- have features that can indicate sex, rank, and ing of the teeth by dogs, which need not (now) precede (un)mated condition. an imminent attack, and is a sign of hostility. In order to The duetting long call behavior seen in chim- bite, a dog must get its lips out of the way. Originally, panzees and bonobos, in which one animal matches its parting the lips was only a preparatory movement call to that of another, indicates a degree of transferra- before biting, but observers became aware of the baring bility of the calls between individuals, and an element of teeth as an index of an attack and took evasive action. of learning. But it seems likely that such duetting is Teeth-barers soon noticed that simply baring the teeth ‘parrot-like’, in that the imitating animal is in no way brought about the desired effect of scaring someone off. attempting to convey the ‘meaning’ (e.g. rank, identi- Now, baring the teeth is a ritualized signal with a mean- ty) of the imitated call. The duetting behavior is not ing something like ‘Beware!’. Human ritualized expres- evidence of transfer of symbolic behavior from one sions such as Hello are relics of ancient animal individual to another. Probably, the duetting behavior behavior, now mostly clothed in the phonemes of the itself has some social/pragmatic significance, perhaps relevant language. But some human ritualized expres- similar to grooming. sions, such as the alveolar click, ‘tsk’, indicating disap- In humans the ability to trade conversationally in proval, are not assimilated into the phonology of their symbols comes naturally. Even humans have some dif- language (in this case English). The process of dissoci- ficulty when the symbol clashes with its meaning, as, ation between the form of the signal and its meaning for example, if the word ‘red’ is printed in green. can be seen as the basis of the capacity to form arbitrary Humans are able to overcome such difficulties and get 314  EVOLUTION OF LANGUAGE 3: PHYSICAL PREADAPTATIONS a response to the symbol to take precedence over the Hauser, Marc D. 1996. The evolution of communication. response to the thing. But chimpanzees apparently can- Cambridge, MA: MIT Press. Hurford, James R., Michael Studdert-Kennedy, and Chris not suppress an instinctive response to concrete stimuli Knight (eds.) 1998. Approaches to the evolution of lan- in favor of response to symbols. With few exceptions, guage: social and cognitive bases. Cambridge: Cambridge even trained apes usually only indulge in symbolic University Press. behavior to satisfy immediate desires. The circum- Jolly, Alison. 1985. The evolution of primate behavior, 2nd edi- stances of wild chimpanzee life have not led to the evo- tion. New York: Macmillan. Knight, Chris, Michael Studdert-Kennedy, and James R. lution of a species of animal with a high readiness or Hurford (eds.) 2000. The evolutionary emergence of lan- willingness (as with humans) to use symbols, even guage: social function and the origins of linguistic form. though the rudiments of symbolic ability are present. Cambridge: Cambridge University Press. Mitani, J.C., and Peter Marler. 1989. A phonological analysis of male gibbon singing behavior. Behaviour 109. References Walker, Stephen. 1983. Animal thought. London: Routledge and Kegan Paul. Byrne, Richard W., and Andrew Whiten. 1988. Machiavellian intelligence: social expertise and the evolution of intellect in JAMES R. HURFORD monkeys, apes and humans. Oxford: Clarendon Press. See also Animals and Human Language 3: Parrots; de Waal, Frans B.M. 1989. Peacemaking among primates. Animals and Human Language 4: Primates; Evo- Cambridge, MA: Harvard University Press. Donald, Merlin. 1991. Origins of the modern mind: three stages lution of Language 1: Overview; Evolution of Lan- in the evolution of culture and cognition. Cambridge, MA: guage 3: Physical Preadaptation; Evolution of Harvard University Press. Language 4: Social Preadaptations Evolution of Language 3: Physical Preadaptations We owe the concept of preadaptation (sometimes chimpanzees must have evolved during that period. termed ‘exaptation’) to Charles Darwin (1859; On the Studies of the communications of chimpanzees in sit- origin of species: 190), who observed that ‘an organ uations where they have been raised by humans and might be modified for some other and quite distinct pur- exposed to manually transmitted languages (American pose’. Preadaptation explains many seemingly unac- Sign Language or computer keyboards that code countable transitions in the evolutionary process. For words) show that they possess the ability to acquire example, the bones of the mammalian middle ear derive about 150 words and can comprehend spoken English from the hinged jawbones of reptiles, and the retina of sentences that have a simple syntax. In contrast, chim- the eye derives from normal skin tissue. Preadaptation is panzees cannot talk, nor can they comprehend or pro- responsible for many of the characteristics that differen- duce (by using sign language) utterances that have tiate human linguistic ability from the communications complex syntax. We can conclude that speech produc- of closely related living species, such as chimpanzees. tion is a derived feature of human language. In this light, studies of the communicative abilities Preadaptation is involved in the evolution of virtual- of chimpanzees are of special interest. Approximately ly all aspects of speech production. Speech production 98% of the genetic code is common to humans and involves three components: (1) the lungs, which gener- chimpanzees; the skeletal remains of early, extinct ate the expiratory airflow that is the power source; (2) hominids (species close to or in the line of human evo- the larynx, which serves for phonation, i.e. it converts lution) are similar in many ways to those of chim- alveolar airflow through vibration into a quasiperiodic panzees. Therefore, any aspect of language shared by series of puffs of air containing acoustic energy at audi- humans and chimpanzees is most likely a ‘primitive’ ble frequencies; and (3) the airway between the larynx characteristic that characterized the linguistic ability and lips, the supralaryngeal vocal tract (SVT), which of our common ancestors some 6 to 7 million years acts as an acoustic filter, allowing maximum acoustic ago. And features that differentiate humans from energy to pass through it at certain formant frequencies. 315 EVOLUTION OF LANGUAGE 3: PHYSICAL PREADAPTATIONS Speech has an overt linguistic function closely viding a sealed nose-to-lung air path during drinking. linked to syntax. The flow of air from the lungs can The distance between the back of the hard palate (the segment the flow of speech into sentences or into roof of the mouth) and vertebral column is relatively shorter clause-length segments. Human speakers can long to accommodate the raised larynx. During the mark out the units for syntactic analysis, usually sen- first year of human life, the face moves backward tences, by producing each one with a single breath. A relative to the rest of the skull, and the larynx constellation of acoustic cues marks the end of each descends. The larynx continues to descend until age breath group. The fundamental frequency of phonation five to six years, when the length of the mouth (the (F0), the rate at which the larynx produces puffs of air oral cavity) and the pharynx (the vertical section of (perceived as the pitch of a speaker’s voice), and dis- the SVT) are almost equal. The pharynx and oral cav- tinctions in syllable duration and amplitude, signal the ities are almost at right angles. During the course of end of a breath group. These acoustic cues are the pri- this process, the posterior contour of the human mary physical markers of sentence intonation. Long tongue becomes almost round, forming the anterior sentences can be linked by using two or more breath (front) wall of the pharynx. During the production of groups. The fact that speech is produced and segment- speech, the human tongue is moved about, almost ed by expiratory airflow follows from the preadaptive, undeformed, in the right-angle space defined by the evolutionary history of our lungs. As Charles Darwin oral cavity and pharynx. noted in 1859, the lungs of terrestrial animals derive In contrast, apes follow a different growth process. from the swim bladders of fish. The elasticity of the As they mature, their faces restructure and project for- lungs, which reflects their original function, lends ward. They retain the high laryngeal position, long itself to the production of expiratory airflow patterns mouths, and SVTs of newborn primates. The skulls of that have a relatively steady air pressure. During spon- early extinct hominids, such as Australopithecines, taneous speech, human speakers generally anticipate seem to have retained the ape-like SVTs of the com- the sentence that they intend to speak and execute a mon ape and human ancestral species. Examination of complex set of motor gestures to produce a breath the skulls of juvenile and adult Neanderthal hominids, group that has the appropriate length. The neural who diverged from modern human beings 500,000 capacity to execute breath groups and intonation keyed years ago, show that they followed a modified nonhu- to sentences may be a unique human derived capacity. man growth trajectory. Preliminary studies of the nerve channels of one spec- The human SVT maintains this approximately 1:1 imen of Homo erectus suggest that this capacity did oral/pharynx proportions throughout life. The formant not exist 1 million years ago—perhaps not until the frequency patterns that differentiate vowels and con- epoch of Neanderthals and modern humans. sonants are determined by the shape of the SVT. The The larynx’s linguistic functions likewise show the formants of the vowel [i] necessitate a wide pharynx mark of preadaptation. The original function of the lar- and constricted oral cavity, and those of the vowel [a] ynx was to seal off the lungs of lungfish when they necessitate a constricted pharynx and wide oral cavity. were underwater. In frogs and many species ultimate- However, the length of the SVT, which varies from ly related to frogs (including humans), the larynx has person to person, results in different absolute values been modified for phonation. Variations in the funda- for the formant frequencies of the ‘same’ vowel or mental frequency (F0) convey emotion and affect in consonant produced by different individuals. Human humans and in many other species. In many lan- listeners unconsciously take this into account by esti- guages, systematic F0 variations alone serve to differ- mating the length of a speaker’s SVT from the speech entiate words, for example, [ma] produced with the signal itself. four different tones of Mandarin Chinese yields four We, in effect, normalize formant frequencies, different words. implicitly taking into account the anatomical con- The human SVT generates the formant frequencies straints of speech production. The vowel [i], which that differentiate vowels and consonants. The human cannot be produced by nonhuman SVTs that lack a 1:1 SVT also shows the mark of preadaptation. In all pri- oral/pharynx proportion, is the optimal signal for this mates and most mammals, including human beings process. Because infants are able to normalize vowels older than three months, the airway above the larynx is between the ages of three and six months, this neural adapted for simultaneously ingesting liquids and swal- ‘computation’ seems to be innate. But the neural basis lowing small chunks of solid food during breathing. In for normalization is another example of preadaptation; newborn primates, including humans, a relatively thin monkeys and many other animals (perhaps even alli- tongue is positioned almost entirely within the mouth; gators) seem to use formant frequencies to estimate the larynx is positioned close to the opening to the the size of another animal. Preadaptation from this nasal cavity and can be raised (periscope fashion), pro- function to linguistic communication clearly occurred 316  EVOLUTION OF LANGUAGE 4: SOCIAL PREADAPTATIONS and may have driven the restructuring of the human mammals is to regulate adaptive motor control, SVT to optimize the process. preadaptation is evidently responsible for their role in The neural system of the human brain that regulates the neural system that confers human linguistic ability. speech production is closely tied to the comprehension of language; its evolution again seems to involve References preadaptation. The meaningful vocalizations of chim- Darwin, Charles. 1859. On the origin of species. London: panzees and monkeys often have different formant fre- Murray, Cambridge, MA: Harvard University Press, quency patterns. However, the vocal signals of Facsimile edition, 1964. nonhuman primates are tied to specific emotions or Gibson, Katherine R., and Thomas Ingold. 1993. Tools and lan- affectual states, whereas humans produce arbitrary guage in human evolution. Tools, language and cognition in vocal signals that convey referential information. human evolution. Cambridge: Cambridge University Press. Hauser, Marc. D. 1996. The evolution of communication. Humans are able to freely sequence the motor com- Cambridge, MA: MIT Press. mands that generate vocal signals to form words. Kimura, Doreen. 1993. Neuromotor mechanisms in human Although Broca’s and Wernicke’s areas of the cortex communication. Oxford, New York: Oxford University have been traditionally identified as the traditional lan- Press. guage areas of the human brain, subcortical basal gan- Lashley, Karl S. 1951. The problem of serial order in behavior. Cerebral mechanisms in behavior, ed. by L.A. Jefress, glia structures, which also sequence the motor 112–46. New York: Wiley. commands necessary for walking and other aspects of Lieberman, Philip. 1967. Intonation, perception and language. motor control, are critical elements of a complex dis- Cambridge, MA: MIT Press. tributed neural system that confers cognitive flexibili- Lieberman, Philip. 1984. The biology and evolution of lan- ty, allowing humans to comprehend sentences with guage. Cambridge, MA: Harvard University Press. Lieberman, P. 2000. Human language and our reptilian brain: complex syntax and adapt thought processes to chang- the subcortical bases of speech, syntax, and thought. ing circumstances. Recent studies show that the basal Cambridge, MA: Harvard University Press. ganglia perform similar sequencing operations, both in Mayr, E. 1982. The growth of biological thought. Cambridge, motor control and in cognitive and linguistic opera- MA: Harvard University Press. tions. Moreover, the neural structures that generate Negus, Victor E. 1949. The comparative anatomy and physiol- ogy of the larynx. New York: Hafner. overt speech maintain words in short-term verbal working memory to comprehend the meaning of a sen- PHILIP LIEBERMAN tence by the use of syntactic (arrangement of words), See also Anatomy of the Articulatory System; Evo- semantic (meaning), and pragmatic information. lution of Language 1: Overview; Evolution of Lan- Because one function of the subcortical basal ganglia guage 2: Cognitive Preadaptation; Evolution of in humans, other mammals, and species ancestral to all Language 4: Social Preadaptations Evolution of Language 4: Social Preadaptations ‘Selfish gene’ Darwinism differs from earlier versions cryptic communication gives way to costly and elab- of evolutionary theory in its focus on one key ques- orate display aimed at overcoming the skeptical tion: Why cooperate? The faculty of speech is an resistance of receivers. On this basis, human speech— aspect of human social competence. By inference, it which is low-cost, conventional communication—can evolved in the context of uniquely human strategies of only have emerged under very different social condi- social cooperation. tions, in which listeners could afford to vest trust in Noam Chomsky redirected linguists’ attention away one another’s signaling intentions. from social issues, defining language as an innate Communication begins not when an individual competence of individuals. To study the origins of lan- makes a sign, but when someone else interprets that guage, however, we must strike out in new directions, action as significant. Even a purely instrumental reconnecting linguistics with Darwinism in its mod- action, after all, may be socially understood as a sig- ern, socially aware form. Signal evolution theory nal. Where this has evolutionary significance, the states that where social conflict exists, low-cost behavior may then undergo modification in the service 317 EVOLUTION OF LANGUAGE 4: SOCIAL PREADAPTATIONS of novel, socially conferred, signaling functions. If this tionary precursor for human linguistic communica- is accepted, syntactical structure in the case of speech tion, the most convincing candidate is not the vocal- may have evolved by becoming progressively elicited izations of primates but their imaginative and often and then consolidated by generations of comprehend- humorous social play. ing listeners. First, conceptual complexity is ‘read Conversational speech including humor in the into’ signaling by the attentive mind reader; subse- human case extends and develops the communicative quently, the signaler—given such encouragement— potential of immature primate play. But if this is may succeed in externalizing aspects of that accepted, we must ask a new question. During the complexity explicitly in the signal itself. course of human evolution, how did the conditions for In keeping with this scenario, one speculative such creativity come to be extended from infancy into hypothesis is that speech evolved thanks to novel lev- adulthood? The key restriction on animals’ freedom to els of care, solicitude, and understanding shown by play is reproductive competition and conflict. In many human caregivers toward their offspring. Chimpanzee species, the onset of sexual maturity brings with it the infants may use a characteristic ‘nursing poke’ as a Darwinian imperative to engage in potentially lethal request for the mother’s nipple. To reach the nipple, sexual competition. In the primate case, this impinges the infant at first simply pushes aside its mother’s arm. upon life concurrently with sexual maturity, setting up As mother and infant cooperatively interact, the infant anxieties, conflicts, and resort to defensive alliances saves time and effort by relying on shorthand. The that effectively constitute adult sociality. As competi- abbreviated version now bears some resemblance to a tive stresses intensify, the tendency is for play fights to linguistic sign. A distance has appeared between the give way to real ones—whereupon the play stops. gesture’s underlying meaning and its surface form— Among humans, the transition to adulthood takes a suggesting the rudiments of a ‘secret code’. different form. Youngsters go through an extended Among apes, however, no such code survives into period of childhood, during which they can rely on adulthood. To understand why ape codes fail to sur- social as opposed to ‘fend-for-yourself’ provisioning. vive, we must examine the social context. Whereas the At a certain point, young hunter–gatherer adolescents human infant may anticipate long-term kin-based become coercively incorporated into ritual coalitions. solicitude, benefiting from social provisioning beyond Rites of initiation—central to intergenerational cultur- infancy, a chimp youngster must fend for itself once it al transmission—may be viewed as spectacular ‘pre- has been weaned. Deprived of the prospect of caring tend-play’ performances, drawing on hallucinatory support, it abandons not only the now irrelevant nurs- techniques such as trance, dance, rhythm, body-paint- ing poke but also most other subtle indications of ing, and so forth. Whether or not genital mutilation is need. Given the competitive exigencies of impending involved, the declared aim is to curb individualistic adulthood, the best preparatory training for the ape pursuit of sexual advantage. Bonds of coalitionary sol- youngster may in fact be to avoid excessive reliance idarity, typically modeled on sibling solidarity, are on cooperative understanding from others. From this accorded primacy over sexual bonds. perspective, elaboration of symbolic potential as The outcome of such solidarity is a domain of insti- young apes mature appears constrained less by cogni- tutional reality—of objective facts whose existence tive deficits than by a decisive social one—the obvious depends, paradoxically, on subjective belief. Within absence, in the wild, of any unconditionally supportive human kinship systems, for example, a woman is ‘our or caring audience. Why bother to elucidate one’s aims sister’ (or a man ‘our brother’) because the collectivi- or interests to others who may at best show indiffer- ty asserts it to be so. When in America today, a piece ence—or at worst exploit such intelligence for their of green paper counts as a dollar bill, that too is an own ends? institutional fact, as is the belief that ‘bachelor’ in When delivered in sequence, ape vocal calls English means ‘unmarried man’. Children engaged in emphasize a single holistic meaning. If the calls vary, games of ‘let’s pretend’ grasp the basic principle of they may yield at most a blended compromise between institutional reality when they assert, ‘this rag is meanings. By contrast, the playful bodily gestures of mummy’ or ‘that stick is a horse’. ape youngsters may be rich in cognitive information During a critical transition in human evolution, the and combinatorial complexity. A gesture indicating institutional realities of cultural kinship acquired ‘Let’s play!’, for example, may systematically reverse supremacy over purely biological reproduction. the significance of subsequent ‘chases’ or ‘bites’, in Collective regulation of reproductive relationships ways reminiscent of a grammatical marker in lan- substantially reduced former levels of internal, sexual, guage. Young primates frequently engage in make- and other conflicts. The effect was to open up a new believe play, whose inventiveness may almost suggest social space in which the institutional facts of lan- human verbal creativity. If we are seeking an evolu- guage could cumulatively evolve for the first time. 318  EWE AND GBE LANGUAGES References Power, C., and L.C. Aiello. 1997. Female proto-symbolic strate- gies. Women in human evolution, ed. by L.D. Hager, Dawkins, R. 1989. The selfish gene, 2nd edition. Oxford: 153–171. New York and London: Routledge. Oxford University Press. Searle, J.R. 1996. The construction of social reality. London: Deacon, T. 1997. The symbolic species: the co-evolution of lan- Penguin. guage and the human brain. London: Penguin. Velichkovsky, B.M., and D.M. Rumbaugh (eds.) 1996. Dunbar, R., C. Knight, and C. Power (eds.) 1999. The evolution Communicating meaning. The evolution and development of culture. Edinburgh: Edinburgh University Press. of language. Mahwah, NJ: Erlbaum. Hurford, J.R., M. Studdert-Kennedy, and C. Knight (eds.) 1998. Zahavi, A., and A. Zahavi. 1997. The handicap principle. a Approaches to the evolution of language: social and cogni- missing piece in Darwin’s puzzle. New York and Oxford: tive bases. Cambridge: Cambridge University Press. Oxford University Press. Knight, C. 1991. Blood relations. Menstruation and the origins of culture. New Haven and London: Yale University Press. CHRIS KNIGHT Knight, C., M. Studdert-Kennedy, and J.R. Hurford (eds.) 2000. The evolutionary emergence of language. Social function See also Evolution of Language 1: Overview; Evo- and the origins of linguistic form. Cambridge: Cambridge lution of Language 2: Cognitive Preadaptations; University Press. Evolution of Language 3: Physical Preadaptations Ewe and Gbe languages The Gbe languages are spoken in an area that begins in The Ewe dialects have been grouped into Coastal Lower Volta in Ghana and stretches through Togo and Ewe, which includes Anlo and Tonu, Central Ewe, Benin to Western Nigeria and Lower Weme. These lan- which includes Ho and Kpedze, and Northern Ewe, guages belong to the Kwa family, and have been treated which includes Anfoe and Kpando. The Central and variously as a dialect cluster and a language cluster. Northern dialects are referred to as Ewedomegbe. Earlier, Anglophones used to refer to the whole cluster While considerable variation exists within the Ewe as Ewe while Francophones preferred Aja, Adja, or cluster, the neighboring Gen cluster spoken in Togo Adja-Tado, which are names of major dialects/lan- and Benin is said to lack noticeable variations. guages within the cluster. The term ‘Gbe’ was intro- duced by Hounkpati Capo to refer to the whole cluster because the majority of speakers designate their speech History forms by adding -gbe to their ethnic names. Capo Oral traditions suggest that the Gbe people once lived divides Gbe into five major dialect clusters comprising in Ketu, a Yoruba town in present-day Republic of from West to East Ewe, Gen, Aja, Phla Phera, and Fon. Benin. From there, a group moved south, and some of According to Capo, ‘there is mutual intelligibility them settled at Tado near the Mono river. This settle- between dialects that are contiguous, e.g. Ewe and Gen, ment is associated with the Aja cluster. Some of those Gen and Aja, Aja and Fon, etc.; but the degree of mutu- who moved south founded a settlement at Notsie (ŋɔ al intelligibility is related to geographical distance’. tsie), which is associated with Ewe. The Ewes in Each of Capo’s dialects constitutes a dialect cluster Ghana say their forefathers fled the tyranny of a chief by itself, some of which are provided in Table 1. called Agokoli and moved to settle in Ghana. A dif- ferent group went from Ketu to settle in Adele in Togo TABLE 1 Gbe Dialect Cluster for a while but later left to join the settlement in Ewe Gen Aja Phla-Phera Fon Notsie. Later, some of those who settled at Tado moved to found the Alada kingdom whose political Adan Kpele Agoi Dogbo Alada Agbome nucleus was Agbome and Xɔgbonu. This settlement is Anfoe Peki Anexo Hwe Ayizo Arohun associated with the Fon cluster. The Gen dialect is due Anlo Tonu Gen Sikpi Kotafon Kpase to later settlement by the Fante-Aŋε from Elmina in Gbin Ve Gliji Tado Saxwe Gun Ghana who settled at Anexo, and the Ga also from Ho Watsyi Tofin Maxi Ghana who settled in the plains between Lake Gbaga Kpando Notsie Xwela Weme and the Mono River. 319 EWE AND GBE LANGUAGES Phonology always a nasal. Also, it only allows two syllable-ini- tial consonant clusters, where the second one has to In all, 42 consonants and 16 vowels have been identi- be a liquid ([l, r,…]) or an approximant ([y, w,…]), fied within the Gbe cluster. These are provided in e.g. klɔ´ ‘wash’, tré ‘seal’ fiá [fja] ‘burn’, and bua Tables 2 and 3. [bwa] ‘pretend’. Not every consonant or vowel is used in every dialect. For example, Ewe is the only dialect cluster that has the voiceless and voiced bilabial fricatives, Morphology written [ƒ] and [(], respectively. Some consonants also Nouns in Gbe can occur with the prefixes [a-], [o-], occur in complementary distribution. For instance, the [e], [ə-], and ε-. While [a-] occurs in all the dialects nasal stops occur with nasalized vowels in Ewe while and is obligatory for the words it occurs with, e.g. nonnasal variants occur with oral vowels, e.g. nyɔ˜´ agba ‘plate’ (instead of just gba), the rest are optional y ˜´ in some dialects). Also, the trill /r/ occurs ‘drip’ (or ˜ɔ dialectal variants (optionality indicated by parenthe- only after dentals, alveolars, and palatals while the lat- ses). Thus, ‘breast’ is (ε)no ‘breast’ in Ho and (ə)no in eral occurs everywhere else, including syllable-initial Anlo. The Gbe dialects also have a productive redupli- position. With respect to vowels, Gεn and Aja have cation process that derives different types of lexical merged [ε] and [e] into [e], while Fon and Phla Phera items, including nominals, from a verbal base. have retained the distinction. Within Ewe, the Coastal Reduplication involves the copy of part of the verb dialects have replaced [ε] with [ə]. root to the left. The part that is copied depends on the type of Gbe dialect. In Fon, only the initial consonant Tone is copied, to which /i/ is added. Hence, gba ‘build’ Gbe is a tone language, i.e. intonational pitch is used to becomes gbigba and wlan ‘write’ becomes wiwlan. In distinguish words. All the dialects use three tones name- Ewe, the initial consonant together with the first vowel ly, High, Mid, and Low. These three tones are sub- is copied; hence, va ‘come’ becomes vava. However, grouped into High and Non-High. Consonants play an the second consonant in a syllable-initial cluster is not important role in the realization of these tones. Thus, a copied, e.g. trɔ´ ‘turn’ becomes tɔtrɔ´ . Non-High tone in a noun is realized as Mid if the noun root’s consonant is a sonorant ([m, l,…]) or voiceless Syntax obstruent ([p, t, …]), as in aŋe¯ ‘plastic’. Where the con- sonant is a voiced obstruent ([b, d, g,…]), the tone is Gbe sentences have a basic Subject–Verb–Object word realized as Low. By contrast, a High tone in a noun is order. This order can be changed in Ewe through the realized as High if the noun has a voiceless obstruent focusing of an argument, e.g. Amí kpɔ´ Kofi ‘Ami looked root consonant, and as Mid if it has a voiced obstruent. at Kofi’ can become Kofié Amí kpɔ´, with Kofi being This Mid tone becomes low when the containing sylla- marked by the focus marker -é and moved to the sen- ble occurs before a syllable with a high tone. In addition tence-initial position. The progressive and related con- to these three tones that all the dialects possess, Anlo, a structions in most Gbe dialects use an auxiliary verb, e.g. variant of Ewe, has an Extra High tone, and Adangbe Amí le Kofi kpɔ´ m´ ‘Ami is looking at Kofi’. Note that the has an Extra Low tone, which occurs on the last sylla- complement Kofi occurs in the preverbal position, while ble in questions. This shows that while tone is distinc- the progressive marker -m´ is attached to the verb. In tive and each lexical item possesses a basic tone, it is some variants, a simple high tone serves as the progres- still highly variable in context. As such, it is not repre- sive marker. Some dialects like Aja-Dogbo and Kpesi do sented in the orthographic convention of the Gbe not allow the complement in the preverbal position. dialects. For a few words, the tones are written in Ewe Some verbs in Gbe are frequently mistaken for to avoid confusion with other very similar words. These adjectives because they describe properties that are are the pronouns è ‘you’ vs. e ‘he, she, it’, wò ‘you’ vs. expressed by adjectives in other languages. However, wo [plural marker], nè ‘you’ vs. ne ‘if’, and lé ‘catch vs. they behave like other verbs in Ewe, e.g. lolo ‘become le ‘be located’. The combination of the three tones big’ takes the verbal habitual suffix -na. Gbe also has yields six contour tones namely, Mid–Low, High–Low, obligatory complement verbs (OCVs). These are verbs High–Mid, which are falling tones, and Low–Mid, that cannot occur without complements. Some of the Low–High, and Mid–High, which are falling tones. verbs are fully meaningful by themselves in the sense that they have translational equivalents in languages like English, but they must still occur with comple- Syllable Structure ments of generic meaning, e.g. )u nú ‘lit. eat thing’. The syllable in Gbe is predominantly open and, in the Some verbs with translational equivalents occur with few cases where it is closed, the final consonant is complements that derive from the same root, e.g. fi fi 320  EWE AND GBE LANGUAGES TABLE 2 Consonants in Gbe Bilabial Labiodental Lamino-dental Laminopost-alveolar Apical post-alveolar Voiceless stops p t Voiced stops b d ) Nasal m n Stops Voiceless affricates ts Voiced affricates Voiceless φ f s Fricatives Voiced β v z Fricatives Tap ɾ Oral trills r Nasal trills ˜r Oral laterals l Nasal laterals ˜l Oral approx Nasal approx Alveopalatal Palatal Velar Uvular Labio-palatal Labio-velar Voiceless stops k kp Voiced stops g gb Nasal  ŋ Stops ŋw Voiceless affricates tʃ Voiced affricates d Voiceless ʃ χ Fricatives (χw) Voiced  ʁ Fricatives (ʁw) Tap | Oral trills Nasal trills Oral laterals Nasal laterals Oral approx y γ ɥ w Nasal approx ˜y ɥ ˜ ˜ w TABLE 3 Vowels in Gbe there is the ICV type like da ‘throw’. The complement Front Central Back position of the normal-type DOV is variable: either the Theme (the object transferred) or the Goal (the High i, ˜ u, u˜ entity to which the transfer is effected) can occur Mid-high e, ˜e o, o˜ Mid-low ε, ˜ε ə, ə˜ ɔ, ɔ˜ immediately after the verb. Thus, both fíá ha Kofi and Low a, ˜a fíá Kofi ha mean ‘teach Kofi a song’. The ICV-type DOV only allows the Theme to occur immediately after the verb; hence, da kpé Kofi ‘throw a stone at ‘lit. steal a steal’. A well-known type of OCV is the Kofi’. There is a restriction on objects such that the inherent complement verb (ICV), which is widely dis- Theme cannot be determined while the Goal is not. cussed in the literature on Kwa languages. These are Thus, *Kofi ná ga lá )eví ‘Kofi gave a child the verbs with indeterminate meanings that appear to be money’ is unacceptable. specified by their complements. The combination of an Gbe also has a serial verb construction (SVC), ICV and its complement usually yields concepts that which consists of two or more verb phrases in a clause are expressed by simple verbs in languages like without any syntactic marking of dependency. English. An example is ƒú, which occurs with du ‘race, Additional properties of the clause include having one course’ to yield ‘run’, and tsi ‘water’ to yield ‘swim’. syntactic subject and one tense value. Also, no verb in There are two types of double object verbs (DOV) the clause can be negated independently. An example in Gbe. There is the ‘normal’ type of DOV the only of an SVC is Kofi tsɔ´ ga lá ƒle agbale˜ ‘Kofi bought a examples of which are fíá ‘teach’, and ná ‘give’, and book with the money. 321 EWE AND GBE LANGUAGES The Noun Phrase Japanese where the postpositions are true ‘preposi- tions’ that occur in postnominal position. The noun can occur in a clause without a determiner in which case it refers to a nondefinite entity. Thus, one can say avu le afí má ‘lit. dog located there’ to locate an ani- Ideophones mal belonging to the dog family. Á)é ‘a certain’ is added Finally, Gbe has ideophones — a set of words with when the speaker knows the entity. The definite article, interesting phonological and syntactic properties, e.g. which is (l)á in Ewe, is only used when the entity is bɔhɔbɔhɔ ‘describing the heavy walk of a fat person’, known to the speaker and hearer. Gbe is also known for i.e. to dawdle. It also has utterance particles that signal its logophoric pronoun. This is a pronoun used to refer to the type of utterance or the attitude of the speaker, e.g. a person whose words, wishes, or thoughts are reported. sea ‘you hear?’ The logophoric pronoun, which is ye in Ewe, can occur after every verb provided it occurs in a subordinate References clause introduced by bé ‘say, that’, e.g. Kofi yi be yeava Aboh, Enoch. 2003. The morphology of complement-head syn- kpɔ Ami ‘lit. Kofi went that he [Kofi] see Ami’. tax. New York: Oxford University Press. Ameka, Felix. 1991. Ewe: Its grammatical constructions and illocutionary devices. Ph.D Dissertation. Canberra: Preposition and Postpositions Australian National University. Gbe has prepositions that are derived from verbs and Ansre, Gilbert. 1966. The grammatical units of Ewe: a study of their structure, classes and Systems. Ph.D Dissertation. postpositions that are derived from nouns. The prepo- University of London. sitions in Ewe constitute a closed class of about ten Capo, Hounkpati B.C. 1984. Elements of Ewe-Gen-Aja-Fon elements. They are distinguished from verbs by their dialectology. Peuples du Golfe du Benin (Aja-Ewé), ed. by inability to occur with the habitual suffix –na. For François de Medeiros. Paris: Éditions Karthala. example, tsó is a verb in Kofi tsóna le ga adé me ‘Kofi Capo, Hounkpati B.C. 1991. A comparative phonology of Gbe. Publications African Languages and Linguistics, Vol. 14. (usually) sets off at six o’ clock’, and a preposition in Berlin and New York: Foris. Kofi mlɔna anyi tsó (NOT tsóna) ga ade me ‘Kofi usu- Clements, G.N. 1975. The logophoric pronoun in Ewe: its role ally sleeps from six o’ clock’. Unlike Ewe, the other is discourse. Journal of West African Languages 10. Gbe dialects do not have a habitual suffix and, there- 141–77. fore, it is difficult to determine which elements consti- Duthie, Alan. 1996. Introducing Ewe linguistic patterns. Accra: Ghana Universities Press. tute a preposition without recourse to meaning. Duthie, Alan, and R.K. Vlaardingerbroek. 1981. Bibiography of Postpositions are thus named because they occur after Gbe (Ewe, Gen, Aja, Fon, Xwla, Gun, etc.). Basel: Basler the noun and designate axial parts and regions of Afrika Bibliographien. objects. Although they have mainly evolved from Lefebvre, Claire, and Brousseau Anne-Marie. 2002. A grammar nouns, they are a different class of words. In Ewe, they of Fongbe. Berlin and New York: Mouton de Gruyter. Medeiros (de), François (ed.) 1984. Peuples du Golfe du Benin are distinguished by their inability to take the posses- (Aja-Ewé). Paris: Éditions Karthala. sive marker ƒe. Contrast Kofi ƒé megbé ‘Kofi’s back’, Pazzi, Roberto. 1980. L’homme Ewe, Aja, Gen, Fon et son where megbé is a nominal, with Kofi megbé ‘behind univers. Lome: Nouveau Tirage. Kofi’ where it is a postposition. However, the postpo- Westermann, Diedrich. 1930. A study of the Ewe language. sitions in Gbe and other Kwa languages cannot be said London, New York and Toronto: Oxford University Press. to form a word class with prepositions, unlike e.g. in JAMES ESSEGBEY 322  F Fanakalo Fanakalo is used in interethnic contacts in the eastern its own home turf, as Fanakalo became increasingly parts of South Africa. It is referred to variably as a pid- perceived as a means of oppressing the Black prole- gin with Zulu or Nguni vocabulary, Nguni being the tariat, and barring its members from learning English. Bantu subgroup to which both Zulu and Xhosa belong. Fanakalo continues to be used in South Africa, Although there is no definite consensus (there is a despite explicit efforts on behalf of authorities to tendency to associate the language with the Indian stamp it out. In 1995, for instance, an official reported emigration starting in the 1860s), it would appear that that English ought to be used in the mines instead of the history of Fanakalo starts with the contact between Fanakalo, pretending that the latter was not an ‘ade- the British and speakers of Bantu languages in the quate medium of communication to ensure under- Eastern Cape/Natal area from around 1840. As it ground safety’. Although still known by hundreds of proved to be a useful vehicle of interethnic communi- thousands of Africans, Fanakalo has lost much of its cation, it soon received official encouragement— raison d’être, and it is most likely doomed unless the Fanakalo is rare among pidgins in that it has spread in use of English in South Africa ceases to increase. part through formal education, as the South African More than two thirds of the lexicon of Fanakalo is Chamber of Commerce arranged courses in it in order Bantu in origin—mostly Zulu, but influences from to facilitate comprehension between miners of differ- Xhosa have also been pointed out. Slightly less than a ing ethnic backgrounds. When European-run mining fourth of the vocabulary is derived from English, while operations spread in the second half of the nineteenth Afrikaans has contributed the remainder. In most basic century, so did Fanakalo. In the 1890s, the pidgin vocabulary, not unexpectedly, Zulu is dominant. On began to be used in Rhodesia and in the Copperbelt the 200-item ‘Swadesh list’ of basic vocabulary, only region of Zambia. It even ventured into the Katanga eight words are not of Zulu origin. There are some (Shaba) province of Belgian Congo, and its use has doublets in the lexicon, as in the numeral system, also been attested in Malawi and Namibia. In the where English and Zulu items are in free variation. beginning of the twentieth century, it was suggested In the segmental phonology, the most obvious that Fanakalo might even become the lingua franca of departure from local Bantu languages is the substitu- South Africa, and perhaps even of the entire southern tion of plosives for clicks — at least among native parts of the continent. speakers of nonclick languages. The implosives of The expansion was halted, however, and rapidly Zulu have become explosive in the pidgin, but two lat- turned into a retreat. In Belgian Congo, an official deci- eral fricatives are nevertheless retained. The European sion was taken in 1918 to replace Fanakalo by Swahili component of the lexicon, on the other hand, has been as the working language, and in Zambia, the somewhat largely adapted to the requirements of Bantu syllable streamlined variety of Bemba known as Town Bemba structure. Thus, Afrikaans sterk ‘strong’ surfaces as gradually ousted Fanakalo. Problems gathered even on stelek in the pidgin. 323 FANAKALO While Zulu is a tone language, Fanakalo does not Clauses generally do not use a copula (‘to be’), use pitch differences to distinguish words. Also, vowel although kona ‘to have’ has a secondary use as exis- length and consonant aspiration are irrelevant. tential copula. Intonational stress is on the penultimate syllable, as in Both conjunctions and subjunctions are attested, Zulu. The vowel system includes five phonemes, as in but subordination is normally replaced by coordinat- the local Bantu languages. ing strategies, often simple juxtaposition of sentences. Fanakalo is a strict S(ubject)–V(erb)–O(bject) lan- Several prepositions are also found, although lapa is guage, as are most of those on which it is based. It is semigeneralized in that it covers a wider range of uses noteworthy that adjectives and numerals precede the than most prepositions in the input languages. Also modified nouns, as in English and Afrikaans, but con- common is ga, which — placed between the possessum trary to Zulu and its relatives. and the possessor — marks genitival relationships. Although most meanings are expressed by inde- Few details are known on the variation of Fanakalo. pendent words, Fanakalo does have a few bound mor- Lexical differences have been observed, however, phemes, all of which are inherited from Zulu: past, between the different countries where pidgin is used. passive, and causative are all marked by suffixes (-ile, Not surprisingly, the Zimbabwean variety is reported -wa, and -isa, respectively). to display influences from Shona, the major language More in line with what one would expect from a of the country, in addition to having a larger English pidgin, the intricate Bantu agreement system has left and a smaller Afrikaans component. few traces in Fanakalo. However, different observers disagree on the fate of the Bantu noun class system. It would seem unjustified to describe Fanakalo as a class References language, although the plural prefix ma- has variants Adendorff, Ralph. 1993. Ethnographic evidence of the social that distinguish whether the noun refers to animate or meaning of Fanakalo in South Africa. Journal of Pidgin and inanimate entities. However, Fanakalo only marginal- Creole Languages 8(1). 1–27. ly preserves other features of the original noun class Aubry, Caroline. 1997. Kitchen kaffir, basic Bantu, isilololo, fanagalo... étude ethnolinguistique d’un pidgin sud-africain distinctions. aux multiples visages. M.A. Thesis. Département d’anthro- Different from Zulu, the future morpheme zo is pologie, Université de Montréal. unbound (i.e. not an affix), as is the determiner lo, Cole, D.T. 1953. Fanagalo and the Bantu languages in South which is perceived as shibboleth of the language to the Africa. African Studies 12(1). 1–9. extent that one of its aliases is isilolo — literally ‘the lo- Ferraz, Luís Ivens. 1980. Notes on a pidgin dialect. African Studies 39(2). 209–20. lo language’. The sentence negator is a free morpheme, Heine, Bernd. 1973. Pidgin-Sprachen im Bantu-Bereich. and the bound pronominal markers of Bantu languages Berlin: Dietrich Reimer. correspond to free morphemes in Fanakalo. These pro- MIKAEL PARKVALL nouns demonstrate no case distinctions, and many vari- eties simplify even further through the use of 2sg and See also Pidgins and Creoles; South Africa; Zulu 3sg plus zonke ‘all’ for the corresponding plural forms. and Southern Bantu Languages Farsi Farsi is one of the three major variants of Persian, a The evolution of Persian as the culturally dominant member of the Iranian language family. The other two language of the Near East began with the political dom- variants are Dari, spoken in Afghanistan, and Tajiki, ination of the dynasties that originated in the province spoken in Tajikistan. The Iranian languages belong to Pars. The name of the language, Parsi, is associated with the Indo-Iranian branch of the Indo-European lan- this ancient province. After the Arab invasion in the sixth guage family. Farsi is the official language in Iran, and century, Parsi was utilized as Farsi in official documents is the native tongue of about half the population of this written in Arabic due to the lack of /p/ in this language. country. Other Iranian languages spoken in Iran Farsi is now the term used for the official language, include Gilaki, Kurdi, Lori, and Baluchi. while Parsi is used mostly in elevated literary writings. 324  FARSI Old Persian century, four centuries after the fall of the Sassanian Dynasty, at a time when it had already ceased to exist The oldest version of Farsi is known as Old Persian, as a language in use. the official language of the Achæmenian Dynasty who Although the gap between the last document found had originated in the province Pars. Evidence of this in Old Persian and the first one discovered in Middle ancient language is found in Achæmenian inscriptions, Persian does not exceed four centuries, the morpho- tablets, and coins. There is no document available of logical structure of the latter had radically changed this language following the fall of the Achæmenian compared to the former. The final syllables in nouns Dynasty in the third century BC. and adjectives were lost. Consequently, the rich case Old Persian was structurally similar to Latin and system had disappeared. The number system was Ancient Greek. It was a Subject–Object–Verb lan- reduced to singular and plural, and there was no sign guage with a fairly free word order in the preverbal of the three-way gender system. The plural suffix ân position. The case system was rich, consisting of seven was dominantly used, and the suffix ihâ, the ancestor cases realized as suffixes. Here is a sample: of the generic plural suffix hâ in Modern Farsi, was Nominative: martiya ‘man’ restricted to religious books such as Bondahishn and Accusative: martiyam Minu-ye Kherad. Unlike Old Persian, there were many Vocative: martiyaa functional words present in this language. In spite of Instrumental: martiyaa the loss of the case system, the word order was still Locative: martiyai fairly free in the preverbal position. Also, the copula continued to be absent in most cases, as in to ki (hi) Genitive: martiyahyaa (you who (are)) ‘who are you’. Ablative: martiyaa The language was basically synthetic, and the num- Modern Persian ber of functional words was very limited. Nouns and adjectives were conjugated for mascu- The oldest documents written in Modern Persian can be line, feminine, and neuter, and the number system con- traced to the early tenth century. Persian soon became sisted of singular, dual, and plural. The copula was the language of world class poetry, and has undergone often lacking, as illustrated by the following example: some minor morphological, phonological, and syntac- tic changes in the last 11 centuries. Of the three variants Baq vazrek Ahuramazdaa God great Ahuramazda of this language, Tajiki has preserved the syntactic ‘The great God (is) Ahuramazda.’ properties of Middle Persian more than the other two, while spoken Farsi reveals the greatest changes with respect to word order and other syntactic properties. Middle Persian The oldest document available in Middle Persian, the descendent of Old Persian, belongs to the first century. Properties of Farsi That is, there is a gap of four centuries between the last The phonological system of Farsi consisted of eight documents found in Old Persian (third century BC) vowels in its early stages, six of them short and two and the first one discovered in Middle Persian. long. The long vowels did not survive, and the disap- Middle Persian consisted of several dialects and pearance of the feature [long] in the phonological sys- variants. One of these variants, known as Parti, tem created a number of homophones in this language Pahlavani, or Pahlavi, was the language of the (e.g. sheer and shir, both pronounced as shir now, Ashkanian Dynasty who governed in the northeastern meaning ‘milk’ and ‘lion’). All remaining vowels are part of Persia from the third century BC through the tense. There are 21 consonants and one semivowel /y/ third century AD. Documents related to this variant are in this system. Unlike Middle Persian, Farsi does not limited to a few short inscriptions and coins attributed allow initial consonant cluster, although final conso- to this dynasty, and a few inscriptions related to the era nant clusters are very common in this language. following their fall. Another variant of Middle Persian, The lexicon of Farsi has been highly influenced by called Parsik in some documents and Pahlavi in some borrowings from Arabic. It has been estimated that others, was the official language of the Sassanian 65% of the vocabulary used in modern novels and Dynasty, a family who originated in the province Pars, short stories is of Arabic origin. This number increas- and was the last dynasty before Arab invasion in the es when taking into account the conservative and for- sixth century. Most scholars have this variant in mind mal writings. The influence of modern European when using the term ‘Middle Persian’. The last docu- languages such as French, and most recently English, ments detected in this language belong to the tenth is also evident in the lexicon of this language. 325 FARSI The morphological system consists of verbal pre- specific objects for accusative case has become the fixes and suffixes, the generic plural suffix hâ, the plu- dominant function of râ only after the fifteenth centu- ral suffix ân (only for animate nouns), and derivational ry. Lack of râ still persists in some unofficial dialects, affixes utilized for nouns. Verbs are inflected for per- such as in Jahromi, one of the Southern dialects of son and number. There is no gender distinction, not Persian, as in dâr-am ketâb-et mi-khân-am ‘I am read- even in the pronominal system. ing your book’. It has been argued that some of the One of the interesting characteristics of Farsi is its Middle Persian texts were rewritten in the sixteenth verbal constructions. Middle Persian did not have and seventeenth centuries, and thus the existence of many complex verbs, and the only light verb used in rây as a direct object marker in those texts might be verbal constructions was kardan ‘to do’. Since the thir- the influence of Modern Persian, the native language teenth century, complex verbs have gradually replaced of the authors. In more recent literature, the primary simple verbs. As a result, the number of simple verbs function of râ has been suggested to be marking speci- does not exceed 115 in contemporary Farsi; many of ficity rather than accusative case. The same idea has them are not even used in the everyday application of been proposed for similar cases in regional languages, the language. The tendency of forming complex verbs for example, ko in Urdu and Hindi and possibly -I in has resulted in the existence of two sets of verbs, sim- Turkish. ple and complex, for a number of verbal concepts. In One of the major syntactic differences between many cases, the application of the simple verb is Middle Persian and Farsi is the lack of ergative con- restricted to the written and elevated language (e.g. structions in the latter. In Middle Persian, the transitive geristan, gerye kardan ‘to cry’). The productivity of verb in the past tense agrees with the object, as in man complex verbs is evident by the fact that its nonverbal to did hi (I you saw-2sg) ‘I have seen you.’ Although element is not restricted to native Persian elements, the verb agrees with the subject in Farsi, agreement nor to Arabic infinitives that entered the language cen- between the nonanimate subject and the verb is turies ago, but it also includes recent borrowings from optional, as in barg-hâ sabz shod (and) ‘the leaves European languages (e.g. telefon kardan ‘to call’, tâyp became-sg (pl) green’. However, lack of agreement is kardan ‘to type’). The verbal element of complex possible only in an unmarked word order in these verbs is not restricted to the verb kardan ‘to do’ as it cases. Otherwise, agreement becomes obligatory, as in was in Middle Persian, but it ranges over a number of sabz shod-and / *shod barg-hâ. Another difference simple verbs that are used as ‘light verbs’ in those con- between Middle Persian and Farsi is that the occur- structions (e.g. zadan ‘to hit’, bordan ‘to take’, dâsh- rence of the copula has become obligatory in the lat- tan ‘to have’, etc.). Another characteristic of the ter: compare to ki (hi) (who you are) in Middle Persian complex verb in Farsi is that its nonverbal element with to ki *(hasti) in Farsi. ranges over a number of categories such as nouns Ezafe construction is one of the specific syntactic (shekast dâdan (defeat giving) ‘to defeat’), adjectives properties of Farsi. It is a noun phrase consisting of the (bidâr shodan (awake becoming) ‘to wake up’), head (an element with the feature [N] such as Noun adverbs (bâlâ bordan (up taking) ‘to promote’), and or Adjective), its modifier(s), an optional possesser, even prepositional phrases (az yâd bordan (of memo- and the Ezafe particle e that is structurally utilized as a ry taking) ‘to forget’). These complex verbs receive link between the head and its modifier, (e.g. dokhtar-e either idiomatic reading (e.g. dast andâkhtan (hand bâhush-e dust-e mâ ‘the smart daughter of our friend’. throwing) ‘to mock’) or compositional interpretation The basic word order is S(ubject)–O(bject)–I(ndi- (e.g. fekr kardan (thought doing) ‘to think’). rect) O(bject)–Verb if the direct object is followed by râ Similar to Middle Persian, Farsi lacks a case sys- and receives a specific reading. The nonspecific object tem. The only element arguably associated with case is is semantically closer to the verb, and is syntactically râ, a particle that marks specific direct objects for adjacent to it. Thus, the basic word order is S–IO–O–V accusative case, and has evolved from Old Persian when the object is nonspecific. Farsi is a Null Subject râdi with the meaning ‘for the sake of’. The Middle language, and thus lacks dummy subjects comparable Persian reflex of râdi is rây, which serves several func- to there and it in English. This language has maintained tions such as illustrating purpose or reference, and the fairly free word order observed in Old and Middle marking the indirect object. In late Middle Persian Persian, and therefore is classified as a scrambling lan- texts, rây starts taking on a new function by marking guage. Recent studies on scrambling languages indicate the direct object as in shaw sheerân rây be-ozan ‘Go, that the word order in these languages is not ‘free’, and slay the lions’. However, even in early stages of that the rearrangements of phrasal categories are trig- Modern Persian the usage of this particle following gered by discourse functional features such as Topic specific direct objects is not widespread as in khosro and Focus or grammatical features such as the one rep- Tus be d-u dâd ‘The King gave Tus to him.’ Marking resenting the Extended Projection Principle (EPP). The 326  FEATURE THEORY word-order variations in these languages are subject to Karimi, Simin. 1990. Obliqueness, specificity, and discourse the general principles of Universal Grammar, and may functions. Linguistic Analysis 20. 139–91. ———. 1997. Persian complex verbs: idiomatic or composi- have an effect on the semantic interpretation of the lin- tional. Lexicology 3(2). 273–318. guistic expression. Thus, the term free word order does ———. 2003. On object positions, specificity, and scrambling, not reflect the syntactic facts in Farsi and similar lan- ed. by S. Karimi. Cambridge: Blackwell Publishers. guages. Finally, there is no structural wh-movement in Khanlari, Parviz. 1976. Dastur-e Zab‚n-e Farsi (The grammar this language, although wh-phrases are subject to of Persian language). Tehran: Bonyad-e Farhang. Lazard, Gilbert. 1992. A grammar of contemporary Persian. scrambling similar to other phrasal categories. California: Mazda Publishers. Windfuhr, Gernot. 1979. Persian grammar: history and state of References its study. The Hague: Mouton. Ghomeshi, Jila. 1997. Non-projecting nouns and the Ezafe con- SIMIN KARIMI struction in Persian. Natural Language and Linguistic Theory 15. 729–88. See also Indo-European 3: Indo-Iranian Languages Feature Theory Classical Phonetics felt the need for a classification would be sufficient, using front–back features such as system for speech segments to give some indication of front, central, and back, and high–low features such as how sounds are made and at the same time to identify high, mid, and low. relationships between them. It used the place–manner It was sometimes necessary to add to these basic system for consonants and the high–low/front–back classificatory features to reflect details of the articula- system for vowels. The main purpose here was clearly tion, often depending on how finely the phonetician to enable the phonetician to specify how particular needed to specify articulations. In the case of vowels, sounds were made with respect to their articulation. some additional features were necessary to expand on There was, however, an important spin-off: it became the two-dimensional coordinate systems: vowels could possible to use the features or parameters of the clas- in addition be spread or round, nasal or nasalized, and sification system to label whole sets of sounds or artic- long or short; and consonants could be either tense or ulations. Thus, we might refer to ‘the set of all lax. For example, in English the high front vowel [i] is plosives’ (seven in English: p, t, k, b, d, g, and the glot- said to be long, whereas the low front vowel [æ] is said tal stop), ‘the set of all voiced plosives’ (three in to be short; [t] and [s] are tense sounds, but [b] and [v] English: b, d, g), or ‘the set of all voiced alveolar plo- are not tense. sives’ (one only in English: d), and so on, cutting hor- As a spin-off of being able to label sets of sounds or izontally and vertically around the consonant matrix. articulations in this way, it became possible to describe Similarly, for vowels, ‘the set of all front vowels’ the behavior of various sets—that is, the behavior com- (English: i, e, æ) ‘the set of all rounded vowels’ mon to members of now identifiable sets could be char- (English: u, o), and so on. acterized. Thus, for example, it was possible to say that For the most part, the categories or features used by the set of voiced plosives devoice in word-final posi- Classical Phonetics reflected what phoneticians tion or that all vowels lengthen before voiced plosives regarded as the salient parameters of articulation: in the same syllable, and so on. Hence, rules no longer place features such as bilabial, dental, alveolar, palatal, had to be about the contextual behavior of individual velar, pharyngeal, laryngeal, etc., and manner features sounds, but in terms of how sets or classes of sounds such as plosive, fricative, continuant, etc. This is not behave. Phoneticians now had the ability to capture and an exhaustive list, but the idea was to characterize for express generalization—an important theoretical prin- each consonantal sound where in the vocal tract the ciple in linguistics: generalizations must be expressed sound was made and how it was made. In the case of whenever possible. Capturing generalizations is the vowels, the relative position of the relevant part of the basis of predictive modeling because it means that we tongue on the high–low and front–back axes of a two- can ask the question: What about any other segment dimensional coordinate system within the oral cavity that has these same features?—even though we may 327 FEATURE THEORY not yet have analyzed it. Generalization leads to pre- Halle (1968) had around 45 features, explaining that dictions and hypotheses, and these form a strong basis they found the original set of 14 somewhat inappropri- for scientific advance. ate for characterizing some subtleties in phonology. For all its potential as the basis of a truly scientific Most modern phonologists argue for a binary sys- study of speech, the feature classification system tem of indexing features: a segment either possesses or afforded by Classical Phonetics, however, rested in does not possess any one particular feature. The point fact on very weak theory. For example, a given sound here is that whereas when describing the physical might have its relevant features identified, but phoneti- world of phonetics it may be useful to have a system cians did not yet have a way of saying why there were capable of capturing multivalued features (an n-ary gaps in how features could be combined, nor was it system), in the cognitive or perceptual world of possible to predict why certain sounds could not in phonology the binary system is preferable. For exam- fact ever occur. Why can we not have a bilabial plosive ple, although we might find that there is somewhat that is also a fricative? Why is it not possible to have a more coarticulation between [ nasal] vowels and sur- voiced glottal fricative? We can easily supply answers rounding [nasal] consonants (in words like man or to these questions—but it is hard to construct a theory moon) in some US accents of English than in most around this system that predicts all the constraints and British accents of English, phonologically this obser- at the same time all the possibilities in an explanatory vation is not relevant. There are no nasal vowels in way—which is ultimately what we are after. We want English, although we may want to note that in some to know why things are the way they are, as well as accents they are nasalized in some environments, the simply being able to characterize them as they are. degree of nasalization being of no consequence. It was not until Transformational Generative Clearly, with a binary system of indexing, the maxi- Grammar came along, though, that these generaliza- mum number of features needed to uniquely classify tions became formalized in recent phonological theo- the sounds of a language like English (with around 45 ry. Morris Halle’s Sound pattern of Russian (1959) phonemes) would be six, giving us 26 or 64 segments. was really the first influential textbook in modern More would be needed to uniquely classify the sounds phonological theory (just two years after Noam of all the languages of the world or indeed all possible Chomsky’s Syntactic structures (1957), the first influ- human languages. Larger sets of features were chosen ential textbook in modern syntactic theory). The because it was felt appropriate to sacrifice mathematical Generative Phonologists adopted the theory of distinc- simplicity or elegance in favor of a feature-labeling sys- tive features from the earlier Prague School of tem that appeared to relate these phonological features Linguistics (Nikolai Trubetskoy 1958 [1939])—a with the phonetic set of Classical Phonetics. Thus, the much more formal representational system than that of meaning of the features became more transparent. the classical phoneticians. These ideas are embodied in three principles sur- rounding the distinctive feature set. They should be able to Distinctive Feature Theory in Phonology (1) characterize all contrasting segments in human The use of distinctive features in phonology enables us languages; to capture natural classes, and, by extension, to gener- (2) capture natural classes in a clear fashion; and alize regularly occurring phenomena and to formulate (3) be transparent with regard to phonetic corre- predictions about the behavior of class members. If we lates. wanted to hypothesize about human processing of phonology, we would use this idea to suggest that A claim inherent in the first principle is that the fea- human beings process the patterns of phonology as part ture set might somehow embody the universal human- of speech planning in terms of these classes rather than ness of features rather than any language-specific in terms of individual segments. The regularity of pat- nature. It is predicted that if this set is correctly speci- terning in phonology is part of the evidence for this fied, no other features will be needed even for future claim—but the claim is more solid when based on the languages, as long as human beings do not change evidence that when the users of a language make up how they make and handle language—that is, as long new words they do so by producing utterances that obey as human beings remain human. the rules of the natural classes their sounds fall into. The second principle refers to not just classes, but There have been various sets of distinctive features natural classes. The idea here is that the classes them- proposed as the parameters of segment description and selves reveal something of what is natural in human classification. The original set appeared in Jakobson et language behavior, once again referring to the fact that al. (1952), and consisted of around 14 features; phonological processing is a human activity, and will Jakobson and Halle (1956) had 12. Chomsky and therefore contain elements that are truly universal. 328  FEATURE THEORY The third principle enables us to establish phonetic important to realize, though, that the correlation is not similarity—that is, to group sounds that are phoneti- necessarily linear or one-to-one (see Table 2). cally similar by feature. In the end, there is a very good reason for doing this: it becomes possible to explain some phonological processes in terms of the behavior Example Use of Distinctive Feature Theory: of their phonetic correlates. Redundancy The distinctive feature set most usually found these Redundancy is an important aspect of phonology that days is approximately that of Halle and Clements is captured by the use of distinctive features. Consider, (1983), which is based on the Chomsky and Halle for example, the fact that all segments in English that (1968) set. A detailed description can be found in are [nasal] are also [voice]. We could say that to Gussenhoven and Jacobs (1998). Chomsky and Halle specify [voice] for segments like [m] and [n] is to have a lengthy description of their own set. fail to capture this redundancy. The main distinctive feature here is the nasality—the voicing is secondary Feature Sets and entirely predictable: all nasal consonants are voiced—remember we are discussing abstract phonol- The 12 features found in Jakobson and Halle (1956) ogy, not phonetics. can be related to what it is they characterize articula- One of the principles involved in Distinctive torily and acoustically. These are abstract, phonologi- Feature Theory is to set up a system to capture all the cal features; hence, their relationship with phonetic segmental contrasts in the world’s languages. This can characterizations is one of relative correlation, rather be done, and we can also show where there is no con- than absolute definition. The physical correlates are trast: there is no contrast, nor possibility of contrast, only approximate—there are many variants on these where there is redundancy. If nasals are always voiced, simple descriptions both between languages and then there cannot be a contrast involving voiceless dialects, and between different speakers (see Table 1). nasals. Two things follow from this in the way we use The Chomsky and Halle feature set is more com- features in the theory: prehensive. There are some 27 basic articulatory fea- tures, although each has particular acoustic correlates. ● we need only indicate those feature markings One point about these features is that they can be used that contribute to the contrasts in a particular at an abstract phonological or perceptual level, in language; which case they take on binary values, or they can be ● we can capture the redundancies in a separate used in a fairly limited way at a physical phonetic table of metarules—rules outside the Distinctive level, in which case they can be multivalued. It is Feature specification. TABLE 1 Feature Opposing Feature Articulatory Correlation Acoustic Correlation Vocalic Nonvocalic Vocal cord vibration, relatively unobstructed Periodic vocal cord excitation and vocal tract clear formant structure Consonantal Non- Partial or complete vocal tract constriction Overall energy relatively low consonantal Compact Diffuse Front ‘resonance chamber’ dominates Energy focused toward the center of the spectrum Tense Lax Vocal organs relatively tense or the entire tract High energy, spread throughout the voluntarily greatly distorted spectrum Voiced Voiceless Vocal cord vibration present Periodic (laryngeal) excitation Nasal Oral Nasal cavity brought into play—the velum is Additional nasal formant(s) present lowered Discontinuous Continuous Vocal tract rapidly closing and opening Interruption of the acoustic ‘flow’ Strident Mellow Turbulence created at the place of articulation Temporally unstructured noise at a relatively high frequency Checked Unchecked Glottalized Abrupt energy onset/offset Grave Acute Marginal within the vocal tract (not central) Energy focused on the lower part of the spectrum Flat Plain Constricted aperture (example is lip-rounding) Upper frequencies attenuated Sharp Plain Constriction of the upper oral cavity, but relative Lower frequencies attenuated widening or the lower cavity or pharynx This table is based on Appendix 2 of Clark and Yallop (1990) 329 FEATURE THEORY TABLE 2 Feature Articulatory Correlation Major class features 1. Sonorant Vocal cord vibration possible and usually present 2. Vocalic (or syllabic) Vocal cord vibration possible, but vocal tract constriction restricted to vowel positions 3. Consonantal Significant constriction present in the vocal tract Cavity features 4. Coronal Blade of the tongue raised 5. Anterior Constriction in front of the palatoalveolar place 6. High Raised tongue body 7. Low Lowered tongue body 8. Back Retracted tongue body 9. Rounded Labial rounding 10. Distributed Extended place of articulation 11. Covered Narrow and tense pharynx or raised larynx 12. Glottal constriction Constriction produced by the vocal cords 13. Nasal Velar port open 14. Lateral Tongue sides lowered Manner of articulation features 15. Continuant ([ continuant]stop) Relatively unimpeded airflow 16. Instantaneous release Sudden release (as in plosives) 17. Velar suction Velar closure (used in clicks, for example) 18. Implosive Glottal closure produces inward airflow 19. Velar pressure (Unclear) 20. Ejective Glottal closure and increased pressure 21. Tense ([ tense]lax) Musculature is highly contracted Source features 22. Raised subglottal pressure Increased muscular contraction to raise subglottal pressure 23. Voiced ([ voiced]voiceless) The vocal cords are vibrating 24. Strident Airflow turbulence produced at point of articulation Prosodic features 25. Stressed (Unclear) 26. Pitch (relative, scalar) (Unclear) 27. Length (relative, scalar) (Unclear) This table is based on Appendix 2 of Clark and Yallop (1990) Omitting feature markings where there is redun- A very early original important text on this point was dancy means literally leaving the redundant cells Richard Stanley’s Redundancy rules in phonology, pub- blank in the distinctive feature matrix formed when lished in the journal Language in 1967 (Vol. 43). These segments have their feature specification character- redundancy rules were called segment structure rules to ized. The fact of the redundancy is captured by sepa- contrast them with another type: sequence structure rate rules that take the general form: rules. The latter capture a speaker’s knowledge of redun- dancy in the specification of segments themselves in pat- if X then Y terned sequences. Thus, if we have the sequence at the or, in our specific example, beginning of a syllable in English: CCCV... (with if [nasal] then [voice]. Cconsonant, Vvowel), then the first C must be [s]— a completely redundant situation, since all we need to But why would we want to capture this redundancy, know is that there is a consonant there, followed by two except to show that it is a regularity in the way seg- others. In fact, there are heavy constraints also, of mental features pattern? We want to do this because course, on the remaining two consonants: the second one speakers of a language know about the redundancy. Let must be a plosive and the third must be a liquid ([r, l]) of us look at an example: there are three nasals in some sort or a semivowel ([y, w]). Some phonologists English: the nasal alveolar stop [n], the bilabial nasal have pointed out that the onset of a syllable strings con- stop [m], and the velar nasal []. If we ask an English sonants together with increasing sonority until you get to speaker to invent a new nasal—say, a palatal nasal like the vowel nucleus (the supreme sonorous segment), fol- the one found in the French word angeau—they will lowed by a coda of consonants of decreasing sonority— also automatically make it [voice]. It is as though although there are exceptions to this principle. they know that nasals must be voiced, which is anoth- Stanley’s work, which like much of the earlier work er way of stating the above rule. in modern phonology, adopted an essentially linear 330  FERGUSON, CHARLES ALBERT approach to analysis. Segments were characterized in a to indicate exhaustively just which features are terms of how they behaved with respect to their neigh- present or absent for this particular segment. boring segments in the surface string, which repre- Distinctive Feature Theory has been a significant sented the underlying phonological makeup of step forward in classification from the rather crude utterances. These days it is better to reconstruct such phonetically based ideas of Classical Phonetics. analysis in a nonlinear or hierarchical way, since this Remembering, however, that it is essentially a concept is much more revealing of the underlying constraints in abstract phonology (rather than phonetics), its prin- on the sequencing of segments on the surface. cipal importance lies in how it lends itself to capturing Nevertheless, whatever the approach the focus is on the generality of phonological processes and the structure in phonology and the way such structure is principles underlying the structure of phonological particularly well expressed using sets of constraints on segments. what can and cannot happen in a phonological system. Explanation is the ultimate goal, and for this to be fully effective it must be sourced from outside the References domain of linguistics itself. Thus, for example, we Chomsky, Noam. 1957. Syntactic structures. The Hague: could use Distinctive Feature Theory to express an Mouton. observed regularity that in many languages utterance Chomsky, Noam, and Morris Halle. 1968. The sound pattern of English. Cambridge, MA: MIT Press. final obstruents that are [voice] often become Clark, John, and Colin Yallop. 1990. An introduction to pho- [ voice], and this regularity is explained by reference netics and phonology. Oxford: Blackwell. to the failure of vocal cord vibration as the transglottal Gussenhoven, Carlos, and Haike Jacobs. 1998. Understanding airflow decreases with falling subglottal air pressure— phonology. London: Arnold. that is, the vocal cord vibration tends to stop as we run Halle, Morris. 1959. The sound pattern of Russian. The Hague: Mouton. out of breath toward the end of an utterance. The Halle, Morris, and George N. Clements. 1983. Problem book in explanation for the phonological regularity captured phonology. Cambridge, MA: MIT Press. by regularities expressed through Distinctive Feature Jakobson, Roman, and Morris Halle. 1956. Fundamentals of Theory lies outside phonology—in the domain of language. The Hague: Mouton. aerodynamics. Jakobson, Roman, Gunnar Fant, and Morris Halle. 1952. Preliminaries to speech analysis: the distinctive features and Distinctive Feature Theory makes a distinction then their correlates. Cambridge, MA: MIT Press. between the use of features for characterizing the con- Stanley, Richard. 1967. Redundancy rules in phonology. trastive properties of phonological segments, and Language 43. using them to indicate redundancy. An incomplete dis- Trubetskoy, Nikolai. S. 1958. Grundzüge der Phonologie. tinctive feature matrix uses blanks to indicate redun- Göttingen: Vandenhoeck and Ruprecht (reprinted from Travaux du Cercle Linguistique de Prague, 1939. dancy (and lets you know where cells are the subject MARK TATHAM of redundancy rules), whereas a fully specified distinc- tive feature matrix has all cells filled with either a  or See also Halle, Morris; phonetics; phonology Ferguson, Charles Albert Charles Albert Ferguson is best known for his 1959 It has been reprinted in various collections and trans- article ‘Diglossia’. In it he undertook to describe a par- lated into a variety of languages. ticular ‘language situation’ in which two closely relat- Ferguson’s interests were exceptionally broad, both ed but substantially different varieties of a language geographically and topically. He spent a good deal of (one designated ‘High’, the other ‘Low’) exist side by his career focused on Arabic and with languages of side in a single speech community in a stable and South Asia, but also wrote on speech communities and functionally complementary relationship. This article topics as varied as the phonology of Philadelphia, lit- has spawned approximately 3,000 studies of the topic, eracy in Diyari (an Australian language), loss of agree- with no apparent end in sight. Some view its publica- ment patterns in Swedish, and the Ethiopian tion as the beginning of the field of sociolinguistics. educational system. His interests bridged the ‘applied’ 331 FERGUSON, CHARLES ALBERT and ‘theoretical’ divide. During much of the first half fore part of the conventions of the speech community of his career, he was professionally occupied with (e.g. variation that correlates with an individual’s solving practical language problems, such as develop- social class or gender). However, some language ing language training programs or assisting develop- behavior appears to be idiosyncratic. For Ferguson, an ing nations in matters of language planning and policy. adequate theory of human language must be capable He spent the last half of his career at Stanford of accounting for all of this. University, where he was able to pursue more theoret- Methodologically, Ferguson preferred to study a ical research (e.g. child phonology and language uni- manageable linguistic phenomenon, such as politeness versals), but he remained actively involved in applied expressions, and consider its implications for a more linguistics. general understanding of human language behavior. Over the course of his career, Ferguson pushed for He found comparative case studies particularly reveal- a far more comprehensive theory of language than was ing. He pushed himself and others to aim for high stan- typical of other linguists. Although trained in the dards of accurate observation of actual language use, American structuralist tradition, even as a student he to seek for insights from different perspectives, and to came to see it as too narrow as he attended lectures by include data from overlooked sources. Some of this he Roman Jakobson, a noted European linguist and key accomplished through his own publications—includ- figure in the Prague School. Some time later, he began ing his book reviews, some of which were quite influ- to seriously read the works of the Prague School, ential and have been reprinted. In addition, however, which deeply impressed him. This inclusiveness one should not overlook his considerable influence became a Ferguson hallmark. He never aligned him- from behind the scenes as the chair of influential com- self with any particular school of linguistics nor did he mittees, as a mentor, as the organizer of conferences, found his own. His M.A. and Ph.D. advisor, Zellig and as the architect of a highly influential department Harris (whom Ferguson greatly admired but also con- of linguistics. sidered overly concerned with theory building), felt that Ferguson was much like Edward Sapir. This is Biography evident in Ferguson’s interest in various types of lin- guistic phenomena, particularly those considered mar- Charles Albert Ferguson was born in Philadelphia, ginal by most linguists (slips of the tongue, formulaic Pennsylvania on July 6, 1921. His early curiosity led expressions, rhymes, proverbs, baby talk, etc.). to the study of languages and linguistics at the Rather than espouse or denounce any particular the- University of Pennsylvania, where he received his oretical persuasion, Ferguson worked to bring diverse B.A. (1942) in Philosophy and his M.A. (1943) and parties together, gently pointing out gaps in their theo- Ph.D. (1945) in Oriental Studies; in 1947 he became a ries and opportunities for fruitful research and cooper- founding linguist of the Foreign Service Institute of ation. He seemed to have a gift for identifying new the US Department of State and founder of the Foreign ground that would prove fertile. As a result, he is hon- Service Arabic Field School in Beirut; in 1955–1959 ored as a pioneering figure in fields such as child lan- he taught Arabic and linguistics at Harvard University; guage, sociolinguistics, modern Arabic studies, and and in 1959 he was founding director of the Center for discourse analysis. Ferguson possessed such rare Applied Linguistics, Washington, DC. From 1947 to vision that even much of his early work continues to 1966 he was also actively involved at Georgetown be highly relevant today. University, and in 1967 he was founding chair of the While he ranged broadly, his work is nevertheless Stanford University linguistics program. In 1971, he unified by the continuous pursuit of a better under- was president, Linguistic Society of America; in 1975 standing of the related and fundamental topics of lan- he was elected to the American Academy of Arts and guage universals, conventionalization, and variation. Sciences; and in 1986 he was Professor Emeritus of With his friend and colleague, Joseph Greenberg, he Linguistics. He passed away on September 2, 1998. pressed for studies of a broad representative sample of various languages and language situations that would References illuminate the commonalities in types of language change, structure, and use. Much of language is, how- Belnap, R. Kirk, and Niloofar Haeri (eds.) 1997. Structuralist ever, not universal and Ferguson was fascinated with studies in Arabic linguistics: Charles A. Ferguson’s papers, the question of how the largely arbitrary code that is 1954–1994. Leiden: Brill. the language of a given speech community comes to Dil, Anwar S. (ed.) 1971. Language structure and language use: essays by Charles A. Ferguson. Stanford, CA: Stanford be shared by the individual members of the communi- University Press. ty, who in fact do not all use that language precisely Ferguson, Charles Albert. 1959. Diglossia. Word 15. 325–40. alike. Some of this variation is systematic, and there- ———. 1959. The Arabic koiné. Language 35. 616-30. 332  FIELD METHODS ———. 1971. Absence of copula and the notion of simplicity: tial consonantism in the acquisition of the first fifty words. a study of normal speech, baby talk, foreigner talk and Papers and Reports on Child Language Development 6. 1–60. pidgins. Pidginization and creolization of languages, ed. by Ferguson, Charles A. 1998. Long-term commitment and lucky Dell Hymes, 141–50. New York: Cambridge University events. First person singular III: autobiographies of North Press. American scholars in the language sciences, ed. by E.F. ———. 1972. Verbs of ‘being’ in Bengali, with a note on Konrad Koerner. Amsterdam and Philadelphia, PA: John Amharic. The Verb ‘be’ and its synonyms: philosophical and Benjamins. grammatical studies, Vol. 5, ed. by John W.M. Verhaar, Fishman, Joshua A., Andrée Tabouret-Keller, et al. (eds.) 1986. 74–114. Dordrecht, Holland: D. Reidel. The Fergusonian impact: in honor of Charles A. Ferguson ———. 1976. The collect as a form of discourse. Language in on the occasion of his 65th birthday, 2 vols. Berlin, New religious practice, ed. by William J. Samarin, 101–9. York and Amsterdam: Mouton de Gruyter. Rowley, MA: Newbury House. Huebner, Thom (ed.) 1996. Sociolinguistic perspectives: papers ———. 1976. The structure and use of politeness formulas. on language in society, 1959–1994/Charles A. Ferguson. Language in Society 5. 137–51. New York and Oxford: Oxford University Press. ———. 1995. Variation and drift: loss of agreement in R. KIRK BELNAP Germanic. Towards a social science of language: papers in honor of William Labov, Vol. 1, Variation and change in lan- See also Applied Linguistics: Overview; Arabic; guage and society, ed. by Gregory Guy, John Baugh, Diglossia; Discourse Analysis; Harris, Zellig Sab- Deborah Schiffrin, and Crawford Feagin, 173–98. Amsterdam and Philadelphia, PA: John Benjamins. betai; Jakobson, Roman; Language Planning; Ferguson, Charles Albert, and Carol B. Farwell. 1973. Words and Phonology; Sapir, Edward; Sociolinguistics; Struc- sounds in early language acquisition: the development of ini- turalism Field Methods Language use pervades daily life. Linguistic data are tions may be augmented by elicitation. Word lists aid ubiquitous. Nevertheless, collection of these data in structuring such sessions; the Swadesh list requires careful consideration. Methods for gathering (Swadesh 1952, 1955; Hymes 1960) provides 200 core data are tailored to the questions being investigated. vocabulary words, chosen for universality. Forms may The ‘classic’ field situation in which a linguist starts also be generated using canonical forms for root mor- from scratch to describe a previously unknown lan- phemes and systematically combining these with an guage is increasingly rare. More often, linguists are inventory of the derivational affixes of the language; working within communities exploring particular ele- native speakers can then note which of these potential ments of language use: child language acquisition, words are actually occurring. Linguists interested in gender patterning, and linguistic variation within the syntactic variability may not find all the structures community. their models predict within their corpus of natural texts; again, elicitation may check the acceptability of sentences’ potentials. Insights into the pragmatics of Data Collection language use often flow from ‘mistakes’ in the lin- In the ‘classic’ field situation, linguists often work guist’s production; laughter, awkward silences, and with one or two principal ‘informants’. The linguist correction provide valuable data about appropriate records as much as possible of naturally occurring forms and their felicity conditions. speech, but analyzes this daily for new elements and Sociolinguistic fieldwork requires that the supplements these samples with elicitations designed researcher record the social characteristics of the to test for phonemic status. Once phonemes have been speakers. Since almost any such characteristic may designated, transcription often shifts to a practical or impact speech, most studies restrict the variables noted phonemic orthography, although details that may to some predetermined set, often including age, edu- reflect on register, style, and/or social divisions may cational level, gender, and socioeconomic class. be lost. Classical analysis proceeds to outline the basic Dialectology requires geographic as well as social structures of morphology and syntax, and perhaps information. semantics and pragmatics. In amassing vocabulary, the Random sampling is useful for the study of varia- collection of stories, narratives, and natural conversa- tion within a community. Even when the interview 333 FIELD METHODS schedule is short and minimally intrusive, coopera- although the technology itself can have status implica- tion is often difficult to obtain. Random means of sub- tions for the speakers as well as the researcher. stitution for preselected interviewees should be Recordings, digital or digitized, can be fed directly established. into computers, which provide spectrographic analy- Network interviewing typically meets with less ses, time the phonations and silences, and, in some resistance. An initial contact within the community is cases, provide provisional transcription. Likewise, used to explicate the study and its purpose and to programs can be used to facilitate phonological analy- enlist further contacts. Community members may ses, finding complementary and contrastive distribu- serve as brokers, interviewers, and/or researchers for tions, and free variation. Text processing programs the project. designed for interlinear translation provide consisten- Some projects may be introduced from the top cy in glossing and ease in concordancing and diction- down. This is especially effective in ‘salvage’ situa- ary creation. tions, where the community feels its language is at risk or in need of development, such as bilingual education Transcription Techniques programs. The community leaders may introduce the research and the investigator, as well as provide core- How many and which details are noted in transcrip- searchers, interviewers, and initial contacts. tion correlates with the purpose of the study. In pre- Phone interviews are possible in many countries liminary field analysis, the researcher may be limited and can be effective, where the variable to be investi- to transcribing isolated words. Fine phonetic details gated can be quickly elicited. Snail and e-mail surveys such as degree of aspiration of consonants, degree of may be used to get an ‘ideal’ report of people’s usage, voicing, nasalization, pitch height, loudness, length, if the variable is one that is transparent and non-pre- and speed may be necessary for phonological analy- supposing. Such surveys do not accurately record sis. For those languages with practical orthographies, usage of stigmatized forms. Surveys of educators have these may be used if the problem of study is above the been used in language planning in many countries; phonological level. Researchers involved in conversa- such surveys can provide data on which language of tional analysis find that they need to record the speak- several is used by children in the classroom, by their er turns, intonation units, truncations, overlaps, parents, and the presence of particular forms in the lin- pauses, and laughter. Interactional analyses often guistic repertoire, although underestimation of fluency require kinesic information as well: speaker gaze, in hegemonic languages and, paradoxically, underesti- body orientation, and gestures. Some writers have mation of minority language use are common. used ‘eye-dialect’ annotations to hint at the The ‘rapid and anonymous’ survey is effective for social/regional character of the interlocutors without documenting linguistic variables that serve as social providing actual phonetic transcription. Because ‘eye- markers. Labov’s (1996) survey of department stores dialect’ is nonsystematic and signals ‘different from in New York City used a single question about the standard’ rather than actual production, most location of a store’s product to elicit the answer ‘fourth researchers prefer to provide sociocultural informa- floor’; the presence or absence of postvocalic /r/ in the tion on the speakers in introductory comments, or in response was correlated to socioeconomic status of the comment lines throughout the texts. Such comment store’s staff. lines can also provide background information on the Linguists may analyze any corpus of language use, topics referenced in the recorded text. Some including broadcast and print media, electronic texts, researchers provide key codes to selected forms that documents, and inscribed artifacts. they will want to extract for analysis: e.g. prefixing negative face work with -F, noticings with N, or repairs with R. Linguistic contexts are so rich that no Equipment transcription system can capture the entire event; nar- Early field linguistics was done with few recording rowing is a necessary adjunct of ‘reduction’ to writing. devices beyond the pen(cil) and paper, but modern electronics allows the preservation of more complete Ethics speech events. Videotaping can capture much of the context of the speech act: age and gender of the par- A person’s identity is bound up in her speech patterns. ticipants, part of the physical environment, kinesics. Issues of how people speak are sensitive, especially in Nonetheless, for many field situations the use of a areas where a hegemonic language or dialect is recog- small audio recorder is more practical, as it is often nized as ‘standard’, as such codes are regularly considered less intrusive. Participants tend to forget assumed to be correct, relegating other varieties to the the presence of the equipment after a brief time, status of incorrect. People who are cognizant that their 334  FIGURATIVE SPEECH speech is being monitored may alter that speech to References provide a desired self-image. This may make it diffi- Briggs, Charles L. 1986. Learning how to ask: a sociolinguistic cult for the linguist to elicit or record casual or natural appraisal of the role of the interview in social science research. speech. Nonetheless, clandestine recording and use of Cambridge and New York : Cambridge University Press. language samples without the permission of the speak- Edwards Jane A., and Martin D. Lampert (eds.) 1993. Talking ers is unethical. Where the speakers are aware of the data: transcription and coding in discourse research. Hillsdale, NJ: Lawrence Erlbaum Associates. research and its goals and support these ends, the Hymes, Dell. 1960. Lexicostatistics so far. Current Anthropology influence of overt recording can be mitigated. 1(1).3–44. Familiarity with the recorders/cameras and their Johnstone, Barbara. 2000. Qualitative methods in sociolinguis- increasing unobstrusiveness can yield surprisingly tics. New York : Oxford University Press. candid language samples. Where the language being Kibrik, A.E. 1977. The methodology of field investigations in linguistics: setting up the problem. The Hague: Mouton. recorded is taboo, secret, or designed for use in ritual Labov, William. 1966. The social stratification of English in contexts only, an understanding of how/if the data will New York City. Washington, DC: Center for Applied be published must be reached. When descriptive gram- Linguistics. mars are written, care should be exercised in the exam- Milroy, Lesley. 1987. Observing & analysing natural language. ple sentences used to respect the cultural norms. When Oxford and New York: Basil Blackwell Ltd. Swadesh, Morris. 1952. Lexicostatistic dating of pre-historic possible, publications and preliminary findings should ethnic contacts. Proceedings of the American Philosophical be shared with the host community. Some researchers Society 96. 452–63. have built in evaluation by the speakers as an addi- ———. 1955. Towards greater accuracy in lexicostatistic dat- tional level of data collection, bearing on the attitudi- ing. International Journal of American Linguistics 21. nal base of language use. Language data are all around 121–37. us. Linguistic field methods consist of the selection, JUDITH M. MAXWELL recording, and codification of these data. See also Sociolinguistics Figurative Speech Figurative speech/language is the creative manipula- Inventione, approximately 87 ⌡χE). During the tion of the phonological, syntactic, semantic, pragmat- Renaissance, it occupied a central place in Poetics (i.e. ic structures of texts, or associations of normal the critical study of poetry). language use, producing ‘extra’ patternings to arrive at The books of and on rhetoric all provide lists of vivid expressions and innovative ideas. It is the oppo- terms, with definitions and examples (cf. Leech 1969; site of denotative speech/language. There are many Dixon 1977). However, not all the divisions have different types of figurative speech, ranging from the acquired equal currency/popularity. Puttenham has repetition of a sound (alliteration) to a semantic and classified these devices according to what sense (that pragmatic contradiction in co-occurring items (oxy- of the eye, the ear, or the mind) they appeal to. A use- moron); both of these types are illustrated in the ful distinction is between ‘tropes’ and ‘schemes’. phrase ‘friendly foe’. Tropes play upon the meaning of a word/sentence. The term ‘figure’ or ‘figurative’ comes from the They include: Latin figura (‘shape’, ‘form’, or, more specifically, ‘attitude’, ‘posture’). The concept of figures of speech Metaphor (referring to one thing in terms of is an ancient and powerful interdisciplinary one. It another, e.g. He is a lion in the field. Here, the originated in Sicily in the fifth century ⌡χE as a set of term ‘lion’ is used to refer to the bravery and conventions for speaking within a legal setting. The fierceness of a human being) oratorical functions of rhetoric broadened when it was Metonymy (referring to something in terms of a adopted in Greece. Plato attacked rhetoric; Aristotle conceptually adjacent item, e.g. Crown for defended it, and systematized the conventions in his sovereign) Rhetoric (approximately 330 ⌡χE). The most influen- Synecdoche (using a part for the whole, e.g. hand tial Roman writer on Rhetoric was Cicero (De for worker) 335 FIGURATIVE SPEECH Simile (metaphor by explicit comparison, e.g. He In that sense, every word is a figurative speech. These is like a lion in the field) ideas developed rapidly in the twentieth century. The Personification (envisaging a quality of a person, classical view of figurative speech considered it to be a e.g. Love for the beloved) mere embellishment that deviates from the ordinary Oxymoron (putting together semantically opposite uses of language. This view was dominant until the end terms, e.g. darkness visible) of the nineteenth century. However, the twentieth cen- On the other hand, schemes play upon normal tury accepted the figures to be as old as language itself. sound/word order. Some of them are: The aim of all figurative speech is to create an Alliteration (repetition of consonants/vowels impact, which may be pleasing or shocking, politically, and/or syllable within a line, e.g. the breeze culturally, or socially. These ends are achieved through blew) a striking and novel imagery, through contrasts and Anaphora (repetition of the same word/phrase at comparisons, through juxtaposition, repetition, by the beginning of lines, e.g. the Beatitudes of the evoking echoes of previous linguistic/spatial/tempo- Bible that makes nine statements starting with ral/social context(s). However, once these new con- ‘Blessed are...’) cepts, associations, expressions, and comparisons are Chiasmus (repetition of words or phrases in widely accepted and adopted, the figurative meaning reverse order, e.g. ‘If he is at Waterbath he does slowly dies out or is forgotten; and the word/phrase not care for you. If he cares for you he’s not at becomes a part of the ‘ordinary’, ‘core’ language. Waterbath.’) Figurative speech, not necessarily under that name, Parison (a series of equally constructed clauses, has remained a preoccupation in the twentieth-century e.g. ‘It is the time of war, it is the time of literary criticism, particularly New Criticism, care/It is the time of courage, it is the time of Stylistics, Practical Criticism, with their emphasis on fear’) micropatterns of language. Such paradigm texts as Zeugma (use of one word to apply to two or more Empson’s Seven types of ambiguity (1930) and words, e.g. ‘Stein her honour or her new Brooks’s The well wrought urn (1947) are essentially brocade’) reworkings of texts as structures of figurative lan- guage. Reader-oriented Criticism, like traditional rhet- Another useful distinction is made linguistically, oric, gives focal importance to the addressee. dividing the figures among phonological, orthograph- Linguistic Stylistics, in its early days, was regarded as ic, lexical, and syntactic categories. An example of a a modern ‘descriptive rhetoric’: the descriptive instru- phonological figure can be alliteration, because alliter- ments of linguistics were deployed in the identifica- ation plays with the phonemes or sound units of a lan- tion of rhetorical figures (Leech 1969; Fowler 1986). guage. Orthographical figures are those where the There have been some plausible programs for orthographical symbol, i.e. a letter/s are changed for describing and classifying figurative speech in semi- special effects, such as ‘Amerikkka’ for ‘America’. A otic, pragmatic, and linguistic terms (Plett 1985). The metaphor is an example of a lexical figure, where a most sweeping interdisciplinary treatment of figures lexical item is involved in the figurative speech. A syn- has been that of Jakobson (1956). He discussed tactic figure can be a parison, where the clauses/sen- metaphor and metonymy as fundamental structuring tences create the figurative speech. principles in such diverse fields as poetry, aphasia, and In addition to the examples provided above, other language development. Such tropes as metaphor are devices traditionally used as figures of speech include now widely used in analyzing how people understand anachorism, anachronism, anticlimax, antithesis, and share experience (Lakoff and Johnson 1980). They antonomasia, apostrophe, assonance, bathos, cacophe- have the potential to provide a promising agenda for mism, climax, dysphemism, euphemism, hendiadys, research in Psycholinguistics and Sociolinguistics. hypallage/transferred epithet, hyperbaton, hyperbole, irony, litotes, meiosis/understatement, onomatopoeia, paradox, prolepsis, and pun/paronomasia (for a References detailed discussion, see McArthur 1996). Dixon, P. 1977. Rhetoric. London: Methuen. It is very difficult to draw a line between figurative Fowler, R. 1986. Linguistic Criticism. Oxford: Oxford and literal/plain and conventional speech. The difficul- University Press. ty is enhanced by the fact that language itself is Jakobson, R. 1956. Two aspects of language and two types of aphasic disturbance. Fundamentals of Language, ed. by R. symbolic or figurative — we use some abstract sym- Jakobson and M. Halle. The Hague: Mouton. bols or letters to signify some real objects that we see Lakoff, G., and Johnson, M. 1980. Metaphors we live by. or hear about, or some abstract ideas that we cherish. Chicago, IL: University of Chicago Press. 336  FILLMORE, CHARLES JOHN Leech, G. 1969. A linguistic guide to English poetry. London: Puttenham, G. 1970. The arte of English poesie, ed. by G.D. Longman. Wellock and A. Walker. Cambridge: Cambridge University McArthur, T. (ed.) 1996. The Oxford companion to the English Press. language. Oxford: Oxford University Press. DEEPSHIKHA MAHANTA Plett, H.F. 1985. Rhetoric. Discourse and literature, ed. by T.A. Van Dijk. Amsterdam: Benjamins. See also anaphora; metaphor; metonymy Fillmore, Charles John Charles John Fillmore studied Linguistics at the When he got his degree, he joined William Wang at University of Minnesota, in the late 1940s, followed Ohio State, who was invited to start a linguistics pro- by a brief period of anthropology studies before end- gram there. Fillmore was associated with a funded ing up in the US Army. After one year of training in research project there called Project on Linguistic the United States, he spent two years in Japan, with the Analysis, and that is where he began doing lexical Army Security Agency, where he spent eight hours a work. He taught courses in the transformationalist day listening to short-wave broadcasts from Soviet framework, and he became sort of evangelical about forces in and around Vladivostok. Two years later, he this new movement. He did some work on rule order- was discharged from the Army and attended classes at ing (a booklet on indirect object constructions and a Kyoto University while earning a living teaching few related papers that nobody knows about dealt with English at a Buddhist girls school. When he got back this, and his paper on ‘embedding transformations’, home in 1956, he went to the University of Michigan which seems to have been the first proposal for the in Ann Arbor, at first intending more to become a idea of the transformational cycle, was a part of this). Japanologist than a linguist. The interest in verbs and their complementation pat- As a graduate student there, Charles J. Fillmore terns led up to a group of papers that formulated a pro- started out mainly as a phonologist, having received posal called ‘Case Grammar’, which became quickly most of his phonological training just at the historical popular. This instant popularity made him suspicious. point where descriptive/structuralist phonology was He had received relief from teaching for ten weeks to giving way to generative phonology. He knew Kenneth write something on Deixis, which he somehow devel- Pike’s work on Phonemics very well, as well as all of oped interest in while at Ohio State, but he spent the the papers on phonemic analysis that could be found in time writing ‘The case for case’ instead. the Martin Joos Reader in Linguistics. A major turning He was invited to spend a year at the Center for point in his thinking was a paper by Morris Halle, a Advanced Study in the Behavioral Sciences, in Palo review of work by a Russian linguist, Avanesov. He Alto in 1970, and at that time he wrote the ‘Lectures on had been writing a paper with the phonetician Gordon Deixis’, which was distributed in various informal Peterson (who was his teacher then) when the Halle ways for 30 years and finally appeared in book form in influence came along. The Peterson–Fillmore paper 1997. While in California, he was invited to move to was going to be a set of principles, on the basis of Berkeley, and he accepted. Teaching in Berkeley at first which it would be possible to design an automatic tended to deal with rule ordering arguments in Syntax, program for producing a phonemic analysis of any Case Grammar, and General Linguistics. In later language based on purely phonetic and distributional decades, he focused on Lexical Semantics, Deixis, and facts. Fillmore was not able to influence Peterson to Pragmatics, and all this led to a strong interest in change the direction of the paper and after that he was idiomaticity and fixed expressions in general. This led left on his own. But, aside from a survey paper he did in a sort of natural way to Construction Grammar, his with Professor Peterson and a paper with Bill Wang on strongest interest in the last decade of active teaching. ‘intrinsic cues for consonant perception’, and a Much of this was influenced by the interdisciplinary dissertation on alternative phonemic analysis of connections fostered by the Cognitive Science program Japanese, he never did anything else in Phonetics or that was just getting started on the Berkeley campus. In Phonology. his thinking about Construction Grammar, he has been 337 FILLMORE, CHARLES JOHN influenced most by Arnold Zwicky and Paul Kay. In his Biography lexicographic work, his strongest influence has been Charles John Fillmore, Chuck for everyone around him, the British lexicographer Beryl T. Sue Atkins. And his was born in St. Paul, Minnesota on August 9, 1929. He main interest since his recent retirement from the class- received his B.A. in Linguistics at the University of room has been computational lexicography, through Minnesota in 1951, and his M.A. (1958) and Ph.D. the direction of a project called ‘FrameNet’ (http:// (1961) at the University of Michigan. The title of his dis- www.icsi.berkeley.edu/~framenet). sertation was A system for characterizing phonological With Frame Semantics, Fillmore brings into the pic- theories. The following were his honors: Institute ture the main theoretical tenets of Cognitive Linguistics: Professor, LSA Summer Institute, Salzburg, Austria, the idea that language is an integral part of cognition, 1979; President, Cognitive Science Society, 1980; which reflects the interaction of cultural, psychological, Election to American Academy of Arts and Sciences, communicative, and functional considerations. 1984; President, Linguistic Society of America, 1991; The basic assumption of Frame Semantics […] is that Festschrifts: Masayoshi Shibatani and Sandra A. each word evokes a particular frame and possibly pro- Thompson (eds.), two volumes: Grammatical construc- files some element or aspect of that frame. An ‘evoked’ tions: their form and meaning, Oxford 1996; Essays in frame is the structure of knowledge required for the semantics and pragmatics, Benjamins-Amsterdam understanding of a given lexical or phrasal item; a ‘pro- 1996; and The Berkeley Citation (UCB) 1998. He was filed’ entity is the component of a frame that integrates employed from 1952 to 1955 by the US Army, Army directly into the semantic structure of the surrounding Security Agency; from 1955 to 1957 as English teacher text or sentence. (Fillmore et al. 2000:2) at Kyoto Women’s School, Kyoto, Japan; from 1961 to 1970 as Assistant Professor through Full Professor by The frame notion is the basis for what Fillmore The Ohio State University; 1970–1971, Fellow, Stanford (1985) calls the semantics of understanding (U-seman- Center for Advanced Study in the Behavioral Sciences; tics), which contrasts with what may be called truth 1971–1994, Professor of Linguistics, University of conditional semantics (T-semantics). U-semantics aims California at Berkeley; Department Chairman, to determine what it takes for a hearer to provide an Department of Linguistics, 1979–1983; Acting Director, interpretation of a sentence. This is not only a compo- Institute of Cognitive Studies, 1985–1989; 1994–2001, sitional approach, because its operation relies on a Graduate School Professor at the University of knowledge of words, phrases, and grammatical con- California at Berkeley; Research Associate at the structions for an interpretation to be construed, but also International Computer Science Institute, Berkeley and noncompositional in that the construction process is not director of the ‘FrameNet’ project. He is a member of guided by purely symbolic operations from bottom to many editorial boards. top. In contrast, the goal of T-semantics is to determine under what conditions a sentence may be true, where truth is determined compositionally (Fillmore References 1982:230–2). In fact, for him the connection between Dirven, Rene, and Gunter, Radden (eds.) 1987. Fillmore’s case Lexical Semantics and Construction Grammar goes grammar: a reader. Studies in descriptive linguistics, Vol. 16. beyond the matter of representation. Construction Heidelberg: J. Groos. Grammar views the description of grammatical pat- Fillmore, Charles, and William S.Y. Wang. 1961. Intrinsic cues terns and the semantic and pragmatic purposes they and consonant perception. Journal of Speech and Hearing serve as equally important and necessary. In Research 4(2). 130–6. Fillmore, Charles. 1963. The position of embedding transfor- Construction Grammar, the semantic frame associated mations in a grammar. Word 19. 208–31. with a lexical item provides some of the semantic infor- ———. 1968. The case for case. Universals in linguistic theo- mation needed for the semantic interpretation of a sen- ry, ed. by Emmon Bach, and Robert Harms, 1–89. Austin, tence (Fillmore 1994). Thus, lexical relations as such TX: Holt-Rinehart-Winston. are not as interesting as the semantic properties that ———. 1982. Frame semantics. The Linguistic Society of Korea. Linguistics in the morning calm, 111–37. Seoul: words have by virtue of their participation in frames. Hanshin. His latest book, Language form, meaning and prac- ———. 1985. Frames and the semantics of understanding. tice (Fillmore 2001), traces questions of language Quaderni di Semantica 6(2). 222–54. form, language use, and conventions linking form, Fillmore, Charles, Paul Kay, and M. Catherine O’Connor. 1988. meaning, and practice. But his latest wish is to see Regularity and idiomaticity in grammatical constructions: the case of ‘let alone’. Language 64(3). 501–38. soon a kind of lexicographic practice that recognizes ———. 1992. ‘Corpus linguistics’ vs. ‘Computer-aided arm- the ‘frame’ background of sense descriptions (cf. chair linguistics’. Directions in corpus linguistics ‘FrameNet’). (Proceedings from 1991 Nobel Symposium on Corpus 338  FINITE-STATE MORPHOLOGY Linguistics, Stockholm), 35–60. The Hague: Mouton de Kyratzis, and Jiansheng Go. Mahwah, NJ: Lawrence Gruyter. Erlbaum. Fillmore, Charles, and B.T. Sue Atkins. 1994. Starting where ———. 1997. Lectures on Deixis. Stanford, CA: Center for the the dictionaries stop: the challenge for computational Study of Language and Information, CSLI Lecture Notes, lexicography. Computational approach to the lexicon, ed. by Vol. 65. B.T.S. Atkins, and A. Zampolli, 349–93. Oxford: Oxford Fillmore, Charles J., et al. Construction grammar. Stanford: University Press. CSLI. To appear. Fillmore, Charles, and Paul Kay, first author. 1999. Fillmore, Charles J. 1987. A private history of the concept of Grammatical constructions and linguistic generalizations: case. Concepts of case, ed. by Rene Dirven and Gunter the what’s X doing Y? construction. Language 75(1). 1–33. Radden. Tubingen: Gunter Narr Verlag. ———. (with Charles Wooters, and Collin F. Baker) 2001. Michaelis, Laura A., and Knud Lambrecht. 1996. Towards a Building a large lexical databank which provides deep construction-based theory of language function: the case of semantics. Proceedings of the Pacific Asian Conference on nominal extraposition. Language 72(2). 215–47. Language, Information and Computation, Hong Kong. Petruc, Miriam R.L. 1996. Frame semantics. Handbook of ———. 2001. Language form, meaning and practice. Stanford: pragmatics, ed. by Jef Verschueren and Jan-Ola Östman. CSLI. Amsterdam/Philadelphia: John Benjamins. ———. 1996. The pragmatics of constructions. Social interac- CARMEN BRETONES tion, social context, and language: essays in honor of Susan Ervin-Tripp, ed. by Dan I. Slobin, Julie Gerhardt, Amy See also Deixis; Halle, Morris; Structuralism Finite-State Morphology Finite-state morphology is a computational model of ● Let λ denote the empty sequence consisting of morphology that uses an abstract computing device no symbols. Then called a finite-state automaton or a finite-state machine. ● δ*(s, λ)  s and Such an abstract machine is capable of defining a lan- ● δ*(s, σ1…σn)  δ(δ*(σ2…σn), σ1) when guage and has a number of useful computational prop- n0. erties. Note that the term language is used here to mean The language L(M) defined by M is the set L(M)  a formal language, which is a set of sequences of sym- {σ∈Σ* | δ*(s0,σ)∈F}. bols, and not necessarily a language spoken by humans. Intuitively, one can think of a finite-state automaton The following concise technical definition is included as a diagram where the set of states are circles and the for completeness. Readers who do not require mathe- transition function is a set of arrows connecting states matical rigor may skip the formal definition and pro- as in Figure 1. ceed to the intuitive explanation below. Each state has one arrow pointing out of it for each Formally, a finite-state automaton is a quintuple M  element of the alphabet. States are generally labeled (Σ, S, s0, F, δ), where for identification, and the start state and the set of final states are marked in some way. Such a diagram can be ● Σ is an alphabet, thought of as a machine by imagining a bug sitting on ● S is a finite set of states, the start state. As the symbols of a particular sequence, ● s0 is a designated start state within S, σ, are called out, the bug crawls along the arrow ● F is a subset of S called the set of final states, and labeled with that symbol to the next state. If, after the ● δ is the transition function δ : S Σ→ S. entire sequence has be called, the bug is sitting on a state that is a final state, then the sequence σ is in the Such an abstract machine is capable of defining a language defined by the automaton. In the example language through an extension of the transition func- shown in Figure 1, the language is the set of sequences tion, δ* : S Σ* → S, that extends δ to sequences of of a’s and b’s that contain an odd number of a’s. A members of Σ*, the set of sequences of symbols taken finite state machine used in this way is called a finite- from Σ. We define δ* as follows: state recognizer since it recognizes sequences of a par- ● Let σ1…σn ∈ Σ* be any sequence in Σ*. ticular language (see Figure 1). 339 FINITE-STATE MORPHOLOGY a b 0 1 b start state Figure 1. Odd number of a’s (final a state is shaded). Although every state must have an arrow for each between the surface graphical representation of a word symbol of the alphabet, we generally draw the diagram and its phonological/morphological representations. In with an assumed sink state that is not drawn. The sink Koskenniemi’s model, the multiple machines are not state is one where all arrows simply point back to the usually combined into a single machine, but operate in same state. When an arrow is missing in the diagram, parallel, with each machine representing one morpho- it is assumed to point to a sink state that is not in the logical rule. For example, using a variant of set of final states. Koskenniemi’s original notation, the rule If, on every arrow, we put a pair of symbols, we y/i ⇔ ◊[bcdfghjklmnpqrstvwxyz]~ • [e][s] have what is called a finite-state transducer. Intuitively, we think of the first symbol as the input symbol and would define a rule that requires a ‘y’ that is preceded the second symbol as the output. This provides us with by a consonant and followed by the plural ending ‘s’ a kind of translator. It is finite-state transducers that being replaced by ‘i’ in the surface representation. The are generally used for finite-state morphology. rule is essentially the rule learned in grammar school: Finite-state machines are important because they ‘To form the plural of a word that ends in a ‘y’ pre- are computationally simple to implement and have a ceded by a consonant, then change the ‘y’ to ‘i’ and number of convenient computational properties. For add es’. In the notation here, the symbol ‘⬃’ denotes example, the amount of memory required to determine an instance of the left-hand side of the rule and the if a given sequence is in the defined language is inde- symbol • denotes a morphological boundary. The pendent of the length of the sequence. Another useful square brackets enclose alternative symbols so that property is that both finite-state recognizers and finite- each set of square brackets represents one symbol state transducers can be easily combined to form the from among the alternatives. union or intersection of languages. Hence, specific Koskenniemi originally used his model for Finnish morphological rules can be modeled independently by and hand translated each rule into a finite-state trans- separate finite-state transducers and then combined ducer. Later programs carry out the translation by into a single machine. Although not all languages can computers. The two-level model is very powerful and be defined by finite-state machines, the morphology of can handle the morphologies of at least a large portion many human languages can be modeled quite success- of the world’s human languages. fully with finite-state models. In order to use finite-state transducers for morphol- References ogy, the alphabet is defined as the set of morphemes and the diagram defines the sequences of morphemes Hopcroft, Jjohn E., and Jeffrey D. Ullman. 1979. Introduction that properly define a word for a given human lan- to automata theory, languages, and computation. Reading, MA: Addison-Wesley Publishing Company. guage. The transducer defines a ‘translation’ between Kornai, András (ed.) 1999. Extended finite state models of lan- the morphological rules that are used to construct the guage. Cambridge: Cambridge University Press. word and the surface realization of the word. Roche, Emmanuel, and Yves Schabes (eds.) 1997. Finite-state By far the most prevalent model for finite-state language processing. Cambridge, MA: MIT Press. morphology is two-level morphology as proposed by Sproat, Richard. 1992. Morphology and computation. Cambridge, MA: MIT Press. Kimmo Koskenniemi, who used the model originally ALTON SANDERS to describe Finnish morphology. Koskenniemi uses a set of finite-state transducers to produce a mapping See also Morphology 340  FINNISH AND FINNIC LANGUAGES Finnish and Finnic Languages Finnic (also Fennic, Balto-Finnic) belongs to the Finnic languages. The Saami presumably shifted their Finno-Ugrian languages (also Finno-Ugric), one of language to or were extensively influenced by Proto- the two branches of the Uralic languages (the other Finnic. The genetic origins of the Saami remain a rid- consists of Samoyed languages). In addition to Basque dle. With the expansion of the Finns in Finland, the and the Indo-European languages, it is the only other Saami were pushed northward. language family represented in Europe. The concept The Finnic influences have followed two main Finno-Ugrian is based on the two dominant Finno- lines. There have been west–east contacts between Ugrian languages: Finnish (Finnic) and Hungarian Finnish and Karelian/Ingrian (less with Votic and (Ugric). There are two theories about the original Veps) and north–south contacts between Finnish and Finno-Ugrian homeland. It was situated on the west- Estonian. Estonian, in turn, has been in intensive con- ern side of the Uralic mountains, at River Volga’s great tact with Votic and Livonian. Between the periphery turn south, or is believed to have consisted of a con- languages, for example, Finnish and Veps or Livonian, tinuum of Finno-Ugrian populations from the Uralic there are considerable linguistic differences and mountains to the Baltic sea. The latter view, which has restricted mutual intelligibility. gained increasing support, partly stems from a reeval- The external linguistic contacts have had four main uation of the history and development of the Finnic historical directions: Swedish influenced Finnish, languages. German influenced Estonian, Russian influenced The Finnic languages are spoken around the Gulf of Karelian (as well as Ludian, Ingrian, Votic, and Finland. Finnish is spoken north of the Gulf, in Vepsian), and Latvian influenced Livonian. Reversed Finland; Estonian is spoken south of it; and east of the local influences have also occurred, but to a lesser Gulf, several minor Finnic languages are spoken: Votic extent. (also Votian, Vote), Ingrian (also Izhorian), Karelian, The Finns and Estonians were Christianized by Ludian, and Veps (also Vepsian). In northern Latvia, a Germanic people in the twelfth and thirteenth cen- few people still speak Livonian. Ingrian, Votic, and turies. They remained Catholics until the Reformation. Livonian are likely to lack mother-tongue speakers in The Karelians, Votes, and Ingrians were Christianized approximately one generation. from the south and have remained Eastern Orthodox. There have been continuous contacts in the Baltic The Votes and Ingrians, as well as the Veps, have area among early Indo-European, Baltic, Germanic, belonged to the political spheres of Novgorod, and Finno-Ugrian people since about 500–2000 BCE. Moscow, and, later, Russia. A peace treaty in Estonia has probably been inhabited by Finno-Ugrian Pähkinäsaari in 1323 created the first national border people since 2000 BCE. Three major ancestral (proto) of Finland. This border between Sweden and Finland forms of Finnic were probably established around and Novgorod and Russia has been modified many 300–1000 BCE. The area north of present-day St. times, especially in 1617, when the Swedish conquest Petersburg, the Karelian Isthmus, the area north of of the Baltic-Finnic area in the early seventeenth cen- Lake Ladoga, and that between Ladoga and Lake tury caused an exodus of Orthodox Finnic people into Onega have housed people who later became the Russia. The border has developed into a language Karelians, the Ingrians, and possibly the Veps. By 800 boundary between Finns and speakers of Karelian, CE, the Finnic languages had reached the geographi- Ingrian, and Votic. cal distribution that became the basis of their Finnish is spoken by approximately 5.1 million autonomous linguistic developments. speakers in the Baltic area, including 250,000 speakers The early phases of the Finnic languages are not in Sweden, 10,000 in northern Norway, and 60,000 in well known. The people have been known since the Russia. Approximately 94% of Finland’s population first century CE. The same name has, however, been (5.1 million) are Finnish-speakers. Since 1919, Finnish used to refer to different people. The use of Fenni (by and Swedish have been the national official languages Tacitus in his Germania, 98 CE) was apparently used of the Republic of Finland. Since 2000, Finnish has for the Saami and by others, later, for the Finns. The been an official minority language in Sweden. From use of Aestii probably referred to Baltic people initial- time to time, Finnish has had official status in Karelia. ly and only later became reserved for Estonians. The Migrant communities can be found in the United States Saami languages are related to, but not counted as, and Canada (100,000–150,000 speakers), Germany 341 FINNISH AND FINNIC LANGUAGES (30,000), Estonia (13,000), Norway, Denmark, dow’, pappi ‘priest’, pakana ‘pagan’, and risti ‘cross’. Australia, and elsewhere. From the twelfth century, Scandinavian/Swedish Finnish was codified in the sixteenth century and influence has been quite strong on Finnish: katu recodified during the nineteenth century. Up to 1540, it ‘street’, laki ‘law’, helvetti ‘hell’, lasi ‘glass’, and tuoli was called old Finnish. The period of old written ‘chair’. Swedish has been a mediator of western Finnish continued into the 1820s. The period between European cultural influences to Finnish throughout 1820 and 1870 is called early modern Finnish. During their common history. that period, which coincided with that of the Grand Some of the older core vocabulary in Finnish, such Duchy of Finland (1809–1917) under the Russian as kieli ‘tongue; language’, nuoli ‘arrow’, veri ‘blood’, Empire, a romantic, nationalistically based discussion käly ‘sister-in-law’, silmä ‘eye’, suksi ‘ski’, kala ‘fish’, on the origins, the future direction, and status of päivä ‘day’, koivu ‘birch’, kaksi ‘two’, viisi ‘five’, elää Finnish was largely settled. The birth of modern ‘live’, and kuolla ‘die’, probably stem from the Uralic Finnish is set to around 1870, when Aleksis Kivi’s protolanguage period (approximately 4000–5000 novel Seven brothers was published. During the nine- BCE). Several of the words reflect typical features of teenth century, the older, analytic Swedish-influenced Finno-Ugrian languages. language changed into the modern form of Finnish. Some Finnic and Finno-Ugrian languages have The dialects were used as a source of relexification vowel harmony, for example, Finnish (Estonian lost it (renewal of the vocabulary) and grammar development. around the seventeenth century). Vowel harmony Early Finnish writing was based on southwestern restricts the combination of vowels in a word and its (Turku) dialects. The New Testament (1547) was endings according to the principle of assimilation. translated by the cleric Michael Agricola, who lived in Front vowels (y, ö, ä) combine with each other, and Turku, the political and religious center of Finland back vowels (u, o, a) combine with each other. For (Stockholm was the capital). The southwestern impact example, käly ‘sister-in-law’ is based on front vowels, was weakened in the translation of the Bible (1642). and kuolla ‘die’ is based on back vowels. The vowels Of great importance was the collection and compila- /i/ and /e/ are considered neutral. tion of the Kalevala epic (by Elias Lönnrot; it was In Finnish and the Finno-Ugrian languages, words completed in 1849), which had eastern origins (Savo can be attached to each other in long rows (agglutina- and Karelian). The written standard formed a compro- tion) to express the same content and grammatical mise based on older western and later eastern sources. relations that languages such as English express with A new spoken standard was also developed. The separate words. The Finnish two-word phrase orthography closely matched the spoken standard: one Menisimmekö iltakävelyllemme? corresponds to the sound largely corresponded to one sign/letter, and one English sentence ‘shall we go for our evening walk?’ letter corresponded to one way of pronouncing it. This Finnish and Finnic languages lack definite and ideal relationship does not persist today. indefinite articles. The phrase pimeässä metsässä ‘in Finno-Ugrian studies have contributed important  dark  forest’ may mean ‘in a dark forest’ and also hypotheses about the older cultural development in the ‘in the dark forest’. Finnish has taken steps on the way area. The contacts with Indo-European languages to develop a definite article during the last two cen- were established early, maybe around 2000 BCE. turies, se ‘it’, and an indefinite article, yksi ‘one’. Among the oldest Indo-European loanwords in Finnic languages lack grammatical gender. The Finnish, one finds nimi ‘name’, vesi ‘water’, and kesä third-person singular pronoun hän in Finnish is identi- ‘summer’. Some of the oldest Indo-European vocabu- cal for ‘he’ and ‘she’. lary belongs to the core vocabulary. There has also Finnish has two main dialect groups, the western been a period of close contacts with Aryan/Iranian and eastern dialects, which can be subdivided: Häme languages, the time of which is uncertain: jumala dialects, heirs of early Proto-Finnish; southwestern ‘God’, syntyä ‘be born’, and taivas ‘heaven’ stem from dialects with clear traces of Scandinavian and these contacts. Among the old loanwords are also Estonian; Ostrobothnian dialects; Savo dialects, prob- Baltic ones: hirvi ‘elk’, lohi ‘salmon’, heinä ‘hay’, ably developed from Karelian; north Finnish dialects kirves ‘ax’, seinä ‘wall’, silta ‘bridge’, and heimo developed from trade and contacts with the north from ‘tribe’. The Baltic influence has been interpreted as a Karelia, Häme, southwestern Finland, and, later, Savo result of intimate contacts and interethnic marriages. areas; and southeastern dialects. Meänkieli, the other Another long-term influence was Germanic: juusto official minority language in northern Sweden, origi- ‘cheese’, leipä ‘bread’, ranta ‘shore’, kulta ‘gold’, nates from the ‘mixed’ northern dialects of Finland. rauta ‘iron’, kauppa ‘shop’, raha ‘money’, and The Finnish dialects in Tornedalen in turn became the kuningas ‘king’. Slavonic influence is much more main source for Kven Finnish in northern Norway recent (600–800 CE): lusikka ‘spoon’, ikkuna ‘win- (seventeenth century). For Finnish dialects, few 342  FINNISH AND FINNIC LANGUAGES problems of mutual intelligibility exist, except possi- loanwords. The lexical influence of Russian grew dur- bly between the southwestern and the eastern Savo ing the twentieth century, but its linguistic impact dialects. remains smaller than that of German. Capital Helsinki speech plays a central role in pres- Linguistic differences between Estonian and ent-day Finnish. Helsinki has experienced several lan- Finnish include the central high Estonian vowel /õ/, guage shift periods between Swedish and Finnish võõras, cf. Finnish vieras, ‘stranger; guest’. Estonian since the sixteenth century. It has received many also lost word-final vowels: Finnish silmä vs. Estonian waves of speakers of different dialect backgrounds. silm ‘eye’, Finnish piimä ‘sour milk’ vs. Estonian piim Western dialects, especially the southern Häme ‘milk’, Finnish Raamattu ‘Bible’ vs. Estonian raamat dialects, have dominated. At the end of the 1800s, the ‘book’. The last two pairs also exemplify the many influx of both Finnish-speaking and Swedish-speaking shifts in meaning. Estonian lacks the Finnish diph- labor forces created a basis for extensive language thongs, for example: Finnish (minä) syön corresponds contacts. Finally, standard spoken Finnish, which was to Estonian ma söön, ‘I eat’. established in the young often bilingual bourgeoisie of Estonian has resisted the dominance of both the late 1800s and early 1900s, has contributed to German and Russian. The great social distance Helsinki speech. These different linguistic origins between the German speakers and Estonians may have have created a mix not found elsewhere. For approxi- been one reason, because Estonians experienced a mately 40–50 years, Helsinki speech has been an overt harsh serfdom under Balto-German landlords up to the prestige model for the rest of Finland. nineteenth century. Another reason may be that Peter Estonian has two main dialects: the northern the Great created an administrative and cultural center dialects (with Tallinn as the main city) and the south- east of Estonia when he founded St. Petersburg, his ern dialects (main city Tartu), including the southeast- window to the west at the Baltic Sea. Other Finnic lan- ern Võru dialect. The orthography is based on the guages in that area were submersed by Russian domi- northern dialects, but the southern dialects have had a nance. For example, Votic and Ingrian today rely on writing of their own. The main dialect division may some dozens of speakers. The area around St. reflect the two early Finnic protolanguages south and Petersburg has also been a battlefield during several east of the Gulf. Estonian is the third largest Finno- wars between Sweden/Finland and Russia and Ugrian language, with approximately 1 million speak- between Russia and Germany. Genocide of Finnic ers out of the 1.5 million inhabitants in Estonia. The people (especially Ingrians) during some of the Soviet remainder (0.4 million) mainly speaks Russian. regimes and Russification of all citizens of the Soviet Estonian was formally an official language, together Union have contributed to the language death of the with Russian, during the Soviet rule, but this changed Finnic neighbors of Estonian. after independence in 1991. In 1995, the new Karelian is the third largest of the Finnic languages. Language Law made Estonian the sole official lan- It is believed to have developed more independently guage of the Republic of Estonia. Estonian communi- from the ninth to eleventh century CE. Proto-Karelian ties live in northwestern Russia, Finland (10,000), and made up the basis for Ingrian, but Karelian itself has Sweden (10,000). Migrant communities live in North developed in close contact with Veps. Karelian con- America and Western Europe. As a result of Russian sists of four major dialects, which are partly an effect and Soviet time deportations, Estonians also live in of different language contacts. North Karelian dialects Siberia and the Far East. are close to the eastern Finnish dialects. South The translation of the Catechism in 1535 by Johann Karelian is spoken west and north of Lake Ladoga. Koell (in Wittenberg, Germany) introduced written Aunus is spoken southeast of south Karelian, in and Estonian. During the following centuries, Estonian around the town of Aunus. Ludian is spoken west of books were mainly written by German speakers; Petrozavodsk. Ludian is sometimes described as a hence, German had an extensive influence on written contact language of Karelian and Veps. As a result of Estonian. Beginning in the nineteenth century, the exodus into Russia in the seventeenth century, German influence faced a nationalistic ‘purification’ some Karelian ‘islands’ developed close to Kalinin. similar to that of Swedish in Finland. Estonians were These so-called Tver-Karelians still speak Karelian. inspired by the Finnish attempts, to the extent that a During the 1930s, three attempts were made to cre- folk epic similar to Kalevala was compiled: Kalevpoeg ate a standard Karelian orthography. Karelian is a (in 1862; by Friedrich Reinhold Kreuzwald). The minority language within its own autonomous ‘nationalization’ of Estonian brought about greater lin- Karelian Republic (approximately 10% of the 800,000 guistic distance to Finnish, because the language inhabitants speak Karelian). Their distribution is activists in Estonia and Finland preferred different uneven, however: Karelian has been better maintained solutions when they created new words for Germanic in rural areas. The dialect split has made the choice of 343 FINNISH AND FINNIC LANGUAGES Numerals 1–10 in Finnish, Estonian, Veps, Northern Saami, and Hungarian Finnish Estonian Veps Northern Saami Hungarian 1 yksi üks üks okta egy 2 kaksi kaks kaks guokte ketto″ 3 kolme kolm koume golbma három 4 neljä neli nel’l’ njeallje négy 5 viisi viis viD vihtta öt 6 kuusi kuus kuz guhtta hat 7 seitsemän seitse siiJ men J ieDá hét 8 kahdeksan kaheksa kahcan gávcci nyolc 9 yhdeksän üheksa ühcan ovcci kilenc 10 kymmenen kümme kümnen logi tíz one variety and one writing system difficult. Speakers Karlsson, Fred. 1983. Finnish grammar. Porvoo, Helsinki and of the northern dialects and the notable group of Juva: WSOY. Korhonen, Mikko. 1981. Finno-Ugrian language studies in Finnish speakers in Karelia have preferred standard Finland 1828–1918. Helsinki: Societas Scientiarum Finnish orthography. The fact that all Karelians know Fennica. Russian, and the strong Orthodox faith, are other fac- Laakso, Johanna (ed.) 1991. Uralilaiset kansat. Tietoa suomen tors complicating the introduction of a Latinized sukukielistä ja niiden puhujista (The Uralic people. Facts Karelian writing. about the language relatives of Finnish and their speakers). Porvoo, Helsinki and Juva: WSOY. Veps is often described as an archaic Finnic lan- ———. 1999. Karhunkieli. Pyyhkäisyjä suomalais-ugrilaisten guage. In historical time, it remained in contact with kielten tutkimukseen (The language of the bear. Strokes of the eastern Finno-Ugrian languages. It is spoken south Finno-Ugrian language studies), SKST 729. Helsinki: of Ludian and west and south of Lake Onega and Lake Suomalaisen Kirjallisuuden Seura. Valgjärvi. The number of speakers of Veps is estimat- Laanest, Arvo. 1982. Einführung in die ostseefinnischen Sprachen. Hamburg: Buske. ed at 10,000 to 12,000. Three of four main dialects Lallukka, Seppo. 1990. The East Finnic minorities in the Soviet have survived: North, Central, and South Veps. The Union. Suomalaisen tiedeakatemian toimituksia 252. number of speakers of Veps and its area have dimin- Helsinki: Suomalainen tideakatemia. ished continuously. The post-Soviet society has made Lehiste, Ilse. 1988. Lectures on language contact. Cambridge, it possible to support and study Veps (also for western MA and London, England: MIT Press. Niemi, Jussi, Terence Odlin, and Janne Heikkinen, (eds.) 1998. Finno-Ugrists). In-depth studies of Veps would cast Language contact, variation and change. Studies in lan- additional light on the contacts among the Indo- guage, Vol. 32. Joensuu: Faculty of Humanities, University European, Baltic, Finnic, and other Finno-Ugrian of Joensuu. people. They could also add new aspects to the study Pyöli, Raija. 1996. Venäläistyvä aunuksenkarjala. Kielenulkoiset of the potential development of a Baltic Sprachbund. ja –sisäiset indikaattorit kielenvaihtotilanteessa (The Russification of Aunuksenkarjala. Language external and The study of and support to Veps are urgent tasks internal indicators in a language shift situation), University of because of its ongoing language shift. Joensuu Publications in the Humanities, Vol. 18. Joensuu: University of Joensuu. Sinor, Denis (ed.) 1988. The Uralic languages: description, his- References tory and foreign influences. Handbook of Uralic studies, Abondolo, Daniel (ed.) 1998. The Uralic languages. London: Vol. 1. Leiden: Brill. Routledge. Stipa, Günter J. 1990. Finnisch-ugrische Sprachforschung: Von Anttila, Raimo. 1988. Historical and comparative linguistics. der Renaissance bis zum Neopositivismus, Mémoires de la Amsterdam: John Benjamins. Société Finno-Ougrienne, Vol. 206. Helsinki: Suomalais- Fogelberg, Paul (ed.) 1999. Pohjan poluilla. Suomalaisten ugrilainen seura. juuret nykytutkimuksen mukaan (Tracing the North. The Uibopuu, Valev. 1988. Finnougrierna och deras språk: kapitel roots of the Finns in the light of recent research), Bidrag till om de finskugriska folkens förflutna och nutid (The Finno- kännedom om Finlands natur och folk, 153. Helsinki: Ugrians and their languages: chapters about the past and Finska Vetenskapsocieteten. present of the Finno-Ugrian people). Lund: Studentlitteratur. Häkkinen, Kaisa. 1996. Suomalaisten esihistoria kielitieteen Virtaranta, Pertti, and Seppo Suhonen. 1978. Lähisukukielet: valossa (The pre-history of the Finns as described by lin- Näytteitä uralilaisista kielistä, III, (The closely related lan- guistics), Tietolipas 147. Helsinki: Suomalaisen guages: examples of the Uralic languages, Vol. III) Suomi Kirjallisuuden Seura. 119.3. Helsinki: Suomalaisen Kirjallisuuden Seura [Texts Heinsoo, Heinike. 2000. On the formation of the Finnic lin- and cassettes of the Finnic languages]. guistic situation. The roots of peoples and languages of JARMO LAINIO Northern Eurasia, vols. II and III, ed. by Ago Künnap, Fenno-Ugristica 23. Tartu: Division of Uralic Languages, See also Case; Hungarian and Ugric Languages; University of Tartu. Language: Contact - Overview; Vowel Harmony 344  FIRTH, JOHN RUPERT Firth, John Rupert The British linguist John Rupert Firth (1890–1960) has matism in the works of G.H. Mead, J. Dewey, and G. traditionally been associated with two quite different de Laguna. approaches in linguistics. The first approach relates to Descriptive linguistics identifies the social compo- a sociological linguistic theory called the Theory of the nent of language as a critical aspect of language, and context of situation, influenced by the works of the Firth introduced a linguistic category encompassing anthropologist Bronislaw Malinowski. The second is a the elements that usually occur in a communicative sit- technique of phonological analysis developed by the uation. He labeled this category Context of situation, so-called London School. However, a detailed study of and its elements are: Firth’s work and teaching (mainly from 1930 to 1957) (1) The participants: persons, personalities, and forces us to describe his influence in linguistics in relevant features of these. somewhat different terms. Throughout his works, Firth developed a linguistic theory of his own that places (a) The verbal action of the participants. special theoretical emphasis on two levels of analysis: (b) The nonverbal action of the participants. the social level (equivalent to the pragmatic level, in (2) The relevant objects and nonverbal and nonper- modern terms) and the phonological one. sonal events. Firth explained his theory on language in Speech (3) The effect of the verbal action. (1930) and The tongues of men (1937). Both works present functionalist and relativist thinking in the Firth used a recurring metaphor—the dispersion of study of language, which Firth called ‘descriptive lin- light into a spectrum—to explain another important guistics’. Firth developed a theory for language in part of his theory: levels of analysis. He described lin- action, not a universal theory of human linguistic abil- guistics as the prism through which light (the lan- ity. These books are a presentation of his thinking sim- guage) passes. The result of this division—or ilar to Sapir’s Language (1921) and Bloomfield’s dispersion—are the different modes of meaning that Language (1933), but Firth did not expound a method we can find in the linguistic analysis of an utterance. of linguistic analysis. Rather, he wrote for the general These modes of meaning must be described by their public and showed a multidisciplinary theoretical own linguistic categories, which establish different approach, which he maintained throughout his works. levels of analysis (e.g. phonetics vs. phonology vs. Thus, these books develop a study of language from morphology, etc.). The different meanings must be the biological, historical, sociological, and linguistic revealed and analyzed by the linguist, keeping in mind point of view. The rest of Firth’s work—its greatest a very important premise in firthian theory: that part—is published as highly specialized theoretical Meaning is Function (cf. Wittgenstein’s Philosophical articles or collections of articles. Firth’s style has been investigations, 1954). described as obscure, mainly because he rarely Thus, descriptive linguistics studies different mean- focused on one topic throughout any given article. His ings on the different levels of analysis, but not neces- essays are usually an amalgam of theoretical knowl- sarily in a hierarchical way (from phonetics to edge, readings, and opinions. semantics or vice versa). Apart from the context of sit- Firth understood language as a system of signs that uation level, Firth posits the lexical, grammatical (syn- we use to do things; that is to say, language works in tactical and morphological level), phonological, and different ways in different contexts. For Firth, lan- phonetic levels, but he is committed to a specific num- guage is somewhat magical, and this linguistic power ber of levels. is what we learn from the day we are born. All the Two theoretical concepts developed by Firth now events in a speaker’s life (society, culture, roles, per- form part of the technical vocabulary of linguistics: sonalities, etc.) dictate the form of this person’s lan- collocation and colligation. The first is used at the lex- guage. Some critics detect here a behaviorist attitude ical level and refers to a word that acquires part of its in Firth’s writing, inherited from Malinowski’s works meaning by the company of the words it usually keeps. and teaching. However, one could point out that this The second term is colligation, very similar to agree- may be just one of the many functionalist aspects of ment, which is used at the grammatical level to refer to his thinking, and that there is a relationship between the syntactical relationship that an element maintains Malinowski (and later Firth) and the American prag- within a construction with others. 345 FIRTH, JOHN RUPERT The theoretical points of view that Firth maintained 1960s by authors such as E. Henderson, J. Carnochan, in his works are very peculiar. Since the publication of K. Sprigg, F.R. Palmer, and R.H. Robins. Although De Saussure’s Cours de linguistique générale (1916), some bibliographies claim that FPA failed, it is still language has been divided into systems and structures being developed (with some theoretical changes due to with their respective paradigmatic and syntagmatic recent phonological studies) in some parts of the relationships. One concept, however, separates Firth United Kingdom. (and later the London School) from the rest of the lin- Thus, Firth developed a linguistic theory that found guistic paradigms: polisystemicity. Polisystemic different applications on different levels of analysis. means ‘several systems’, and this concept is applied to Firth’s influence, although important, is not well different aspects of linguistic analysis. For example, a known. While his name is mainly associated with the language can contain several phonetic systems: a London School and Firthian prosodic analysis, native system with its own phonetic patterns, and sys- systemic-functional grammar, which is associated tems of foreign phonetic patterns borrowed from other with the work of M.A.K. Halliday, he is also known as languages. At the same time, a language can contain Neo-Firthian. More complete works on Firth and FPA several subsystems within its structures: a conso- are currently becoming available, which, in some nant–vowel–consonant pattern (CVC) is a structure sense, is a vindication of his theoretical contribution to that contains a system for initial consonants C-, a sys- linguistics. tem for medial vowels -V-, and a system for final con- sonants -C. This polisystemic approach is opposed to Biography the monosystemic approach (typical of structuralism), where only one system is posited for the whole lan- John Rupert Firth was born in Leeds, England, Great guage (a good example is the hypothesis of a ‘phono- Britain, on June 17, 1890. He did his B.A. (1911) and logical system’ of a language containing a closed set M.A. (1913), both in history. He was a lecturer in of phonemes). History at the City of Leeds Training College in 1913, Firth criticized this monosystemic and paradigmat- and that year joined the Indian Education Service. He ic approach because it provides only an incomplete was Professor of English at the University of the analysis of a language. His polisystemic approach, in Punjab, Lahore (now Pakistan) (1920–1928), and contrast, studies not only the paradigmatic relation- Senior Lecturer in the Department of Phonetics at ships but also the syntagmatic ones. These are the the- University College, London (1928–1938), holding oretical underpinnings of the main categories that meanwhile a number of part-time appointments. He Firth established at the phonological level: phonemat- had an Assistantship in Sociology of Languages at the ic unit and prosody. These two categories form the London School of Economics (1928), where he conceptual basis of Firth’s phonology, and later of the worked with Bronislaw Malinowsky, was Senior Prosodic Analysis developed by the London School. Lecturer in the School of Oriental and African Studies Phonematic unit is a phonological category, applied in 1938, and Reader in Linguistics and Indian at the phonological level. Sound structures are repre- Phonetics in 1940. sented using the phonematic units ‘C’ for consonants Firth held the first chair of General Linguistics in and ‘V’ for vowels. Each one of these units is, at the Great Britain (1944) at the University College, same time, the representation of a system given at that London. He was also Head of the Department of place in the structure where some potential elements Phonetics and Linguistics since 1941. Although he appear. Phonematic unit is thus the expression of a retired in 1955, he continued attending the paradigmatic or vertical relationship. On the other Philological Society meetings, of which he was a hand, prosody, far from classical definitions (related to member since 1933, President from 1954 to 1957, and poetry, metrics, or suprasegmentals), is a syntagmatic Vice-President in 1959. He died on December 14, or horizontal relationship given in a structure, i.e. 1960. something that happens at the same time in the struc- ture as a whole and characterizes it. In fact, a prosody is usually the guide for setting up a structure. These References concepts are implicit in Firth’s early works but are Anderson, S.R. 1985. Phonology in the twentieth century. most clearly developed in his 1948 article ‘Sounds and Chicago: University of Chicago Press. prosodies’. This article is known as the foundation of Bazell, C.E., J.C. Catford, M.A.K. Halliday, and R.H. Robins. the Prosodic analysis (later Firthian prosodic analy- (eds.) 1966. In memory of J.R. Firth. London: Longmans. De Beaugrande, R. 1991. Linguistic theory. The discourse of sis, FPA), a technique of phonological analysis almost fundamental works. London: Longmans. solely used by the London School. The most important Firth, John Rupert. 1948a. Sounds and prosodies. Transactions work of this school was done during the 1950s and of the Philological Society. 127–52. 346  FISHMAN, JOSHUA A. ———. 1950. Personality and language in society. Kelly, J., and J.K. Local. 1989. Doing phonology. Manchester: Sociological Review 42(2). 37–52. Manchester University Press. ———. 1957. Ethnographic analysis and language with refer- Langendoen, D.T. 1968. The London School of Linguistics. ence to Malinowski’s views. Man and culture: an evaluation A study of the linguistic views of B. Malinowski and J.R. of the work of Bronislaw Malinowski, ed. by Raymond Firth. Cambridge, MA: MIT Press. Firth, 93–118. London: Routledge & Kegan Paul. Palmer, F.R. (ed.) 1968. Selected papers of J.R. Firth ———. 1957. Papers in linguistics, 1934–1951. London: (1952–1959). London: Longmans. Oxford University Press. ———. 1970. Prosodic analysis. London: Longmans. ———. 1957. A synopsis of linguistic theory, 1930–1955. Robins, R.H. 1963. General linguistics in Great Britain Studies in linguistic analysis, Special volume of the 1930–1960. Trends in modern linguistics, ed. by Philological Society, 1–32. Oxford: Basil Blackwell. Mohrmann, Norman and Sommerfelt. Antwerp: Spectrum. ———. 1964. The tongues of men and speech. London: Oxford ELENA BATTANER-MORO University Press, Reprinted 1966 and 1970. Harris, Roy (ed.) 1988. Linguistic thought in England 1914–1945. London: Duckworth. See also Phonology; Pragmatics; Prosody Fishman, Joshua A. With his prolific research work spanning almost four Broad perspectives on the issues involved in lan- decades, Joshua Fishman has been recognized as a guage planning typify Fishman’s work—he has leading sociolinguist/sociologist of language. He has probed into interrelated topics such as the relationship been influential in scholarly studies of the relationship of language, ethnicity, nationality, and state (see between language and society—at the macro level (i.e. Fishman 1972); language maintenance and shift the broad societal level) as against micro (i.e. the indi- (1964); bilingualism and diglossia (1967); languages vidual level). His major theoretical contribution is the of wider communication (1989); the stages involved in concept of ‘domains’ (i.e. a larger unit than situation, adequate language planning (1974); and issues of cor- e.g. that of family, work) of language use and the pus planning such as the development of vocabulary in extension of the concept of ‘diglossia’ (i.e. the exis- Israel (Fellman and Fishman 1977). His edited book, tence of two or more varieties of the same dialect/lan- Handbook of language and ethnic identity, is a collec- guage) to include bilingualism. The fact that three tion of 28 unpublished papers that provide a kaleido- volumes of scholarly articles were brought out on his scopic vision of a variety of social issues—including 65th birthday indicates the extent of his academic Economics, History, Political Science, Ethnography, influence. Sociology, Sociolinguistics, Psychology, and Fishman’s first scholarly article appeared in Yidishe Religion. Shprakh in 1947. It was an outcome of his ardent inter- A macrosociological and quantitative (i.e. statistical est in the study of his mother tongue, Yiddish. In 1949, data collection) approach is the distinguishing feature he received a prize from the Yiddish Scientific Institute of Fishman’s work. His work is characterized by for an unpublished monograph on bilingualism. An meticulous analysis of large bodies of data collected in extended study of minority languages and bilingual- major surveys using methods of sociology and more ism started in response to a routine questionnaire cir- recently, in collaboration with Gella Fishman, it has culated by the US Census Bureau for the national also incorporated the exhaustive elucidation and inter- census of 1960, for which he suggested a revision of pretation of archival material. the language questions. This initiated his campaign to Fishman’s concept of ‘domain’ has been particular- establish the importance of collecting data on the sta- ly influential in promoting research into bilingualism tus of non-English languages in the United States. and diglossia and related issues. For example, Parasher Eventually, it led to the publication of the book (1980) has applied it to the Indian bilingual cum Language loyalty in the United States (1966), the diglossic situation. Similarly, Rubin (1968) applied it major reference work in the preparation of the to the national bilingual situation in Paraguay. Bilingual Education Act (Senate Bill 428). It also Fishman has been criticized for having a European established a new area of research in Sociolinguistics: bias, and for threatening the fundamental assumptions language maintenance and shift as a field of study. of nation building. There is an internally consistent 347 FISHMAN, JOSHUA A. response in Fishman’s work to these criticisms: the Journal of the Sociology of Language since its incep- preservation and promotion of the minority groups and tion in 1973. their languages do not necessarily constitute a chal- lenge to the larger societal groups. In fact, with plan- References ning awareness and appreciation of the potential Fellman, J., and Joshua A. Fishman. 1977. Language planning contribution of all the elements, a rich and ultimately in Israel: solving terminological problems. Language plan- stronger social organization can be achieved. ning processes, ed. by J. Rubin et al. New York and Berlin: Mouton de Gruyter. Fishman, G.S. 1991. Bibliographical inventory. The influence Biography of language on culture and thought: essays in honor of Joshua A. Fishman’s sixty-fifth birthday, ed. by R.L. Cooper Joshua A. Fishman was born in Philadelphia, PA, on and B. Spolsky. Berlin and New York: Mouton. July 18, 1926. He received B.S. and M.S. degrees from Fishman, Joshua A. 1964. Language maintenance and language shift as a field of inquiry. Linguistics 9. 32–70. the University of Pennsylvania (1944–1948). He stud- ———. 1966. Language loyalty in the United States. ied Yiddish, his mother tongue, at UCLA. He did his ———. 1967. Bilingualism with or without diglossia, diglossia Ph.D. in social psychology from Colombia University with or without bilingualism. Journal of Social Issues 23. 2. (1953), and served as Educational Psychologist for the 29–38. Jewish Education Committee of New York in ———. 1972. Language and nationalism: two integrative essays. Rowley, Mass: Newbury House Publishers (1973, 1951–1954. He directed research for the College ©1972). Entrance Examination Board, while teaching sociolo- ———. 1989. Language and ethnicity in minority sociolin- gy of language at City College (CUNY), in guistic perspective. Clevedon, Avon, England; Philadelphia: 1955–1958. He was Associate Professor of Human Multilingual Matters Ltd., ©1989. Relations and Psychology at the University of ———. 1968. Readings in the sociology of language. The Hague: Mouton. Pennsylvania (1958–1960); Professor of psychology ———. 1985. The rise and fall of the ethnic revival: perspec- and sociology at Yeshiva University, New York (1960); tives on language and ethnicity. Berlin, New York and Dean of the Ferkauf Graduate School of Social Amsterdam: Mouton. Sciences (1966); served as academic vice president ———. 1991. Yiddish: turning to life. Amsterdam: John (1973–1975); and became professor emeritus in 1988. Benjamins. ———. (ed.) 1999. Handbook of language and ethnic identity. He held visiting appointments at ten major universities New York: Oxford University Press. in the United States of America, Israel, and the Parasher, S.V. 1980. Mother-tongue-English diglossia: a case- Philippines, and fellowships at the Center for study of educated Indian bilinguals’ language use. Advanced Study in the Behavioural Sciences Anthropological Linguistics 22(4). (Stanford, CA), the East–West Center (Honlulu), the Rubin, J. 1968. National bilingualism in Paraguay. The Hague: Mouton. Institute for Advanced Study (Princeton, NJ), the SHREE VALLABH PARASHER Netherlands Institute for Advanced Study (Wassenar), and the Israel Institute for Advanced Study See also Diglossia; Language Planning; Sociolin- (Jerusalem). He has been editor of the International guistics Focality The terms focus, focal, and focality are used in two pitch that is often used at the beginning of a sentence: in different domains: phonetics and syntax/semantics. the following examples, the latter is written in italics, the In phonetics, focus is understood as a specific stress former in capital letters (with examples of typical pre- or pitch; in a narrow sense, focal stress means the (typi- ceding co-text segments closed in parentheses): cally falling, or rising–falling) sentence stress, placed in the unmarked (normal) case at the end of the sentence, (1) (Mary visited us last week.) Paul was here and is distinguished from contrastive (typically rising) YESTERDAY. 348  FOCALITY (2) (As a Christmas present, Jim got a picture from groups contained in the focus are not connected with me.) For Martin his mother has brought a new such a presupposition (a presupposition is an asser- BOOK. tion, the truth of which follows not only from the sen- (3) (Last Saturday I went to an exhibition with tence triggering it but also from its negative Jane.) On Monday we visited an old CASTLE. counterpart), cf.: (4) (Jerry and Marilyn were there too.) Him she (11) Mike’s sister has visited that exhibition (pre- SAW (, but she did not recognize Marilyn). supposition: Mike has a sister). The contrastive pitch has to be distinguished from (12) That exhibition was visited by Mike’s sister the presence of a focus in a marked position, which is (no such presupposition, the existence of the case in (1): Mike’s sister follows only from the positive sentence, but not from its negative counter- (1) (Who was visiting you yesterday?) PAUL was part). here yesterday. It is more exact not to work only with the dichoto- In the analysis of the sentence and its information my of topic and focus, but to describe the information structure, focus is understood as that part of the sen- structure of the sentence on the basis of an opposition tence referring to ‘new’ information, rather than to the of contextually bound and nonbound items. This ‘given’ information (corresponding to the topic of the allows for a more subtle classification, cf. the follow- sentence). Phonetic focus is carried by a part of the ing examples, in which the indices b and n are used for informational focus; cf. the following examples, in contextually bound and nonbound items, respectively. which normal intonation is assumed (with the stress at the end of the sentence) and the sign ‘/’ marks the (13) (A young pair was present there.) Ib recog- boundary between topic and focus): nizedn onlyn himn. (14) (I prefer fiction written by a not much known (5) (When did your friends visit you?) Some of lady from Edinburgh.) Thisb author’sb booksb them were present/at our party last Friday. haven a specificn flavourn for meb. (6) Many men/read few books. —Few books are read/by many men. In the unmarked case, nonbound items belong to (7) Mary was writing her dissertation/on the week- the focus of the sentence and bound ones are included ends. —On the weekends, Mary was writing in its topic; exceptions concern bound items embedded her dissertation. in the focus (such as for me, if understood as a modi- fier of flavour in (14)) and nonbound items in the As the last two examples show, the articulation of the topic. sentence in its topic and focus is semantically relevant. Note that the opposition of contextual boundness is The passivization may be understood as triggering a understood as a pair of grammatical values, which cor- word-order variation, for which otherwise English responds to the cognitive notions of ‘given’ and ‘new’ allows only in specific syntactic contexts, such as: information in the prototypical case, but not always; (8) They moved/from a village to a large industri- e.g. Him in (13) as a stressed pronoun belongs to the al center. —They moved to a large industrial focus of the sentence and is nonbound, although its center/from a village. pronominal (anaphoric) character is connected with (9) We went/by car to a lake. —We went to a the fact that it refers to an entity known to the speaker lake/by car. (at least in some sense) and that is assumed to be known also to the hearer. In many other languages (Latin, German, Czech, and most other Slavic languages), the word order is References much freer than in English, so that e.g. an object belonging to the topic can precede the subject: Bosch, P., and R. van der Sandt (eds.) 1998. Focus. Linguistic, cognitive and computational perspectives. Cambridge: (10) Das Buch hat mir mein Vater gegeben (lit.: the Cambridge University Press. book has me my father given—‘My father HajiJ ová, E., B.H. Partee, and P. Sgall. 1998. Topic-focus artic- ulation, tripartite structures and semantic content. gave me the book’). Dordrecht: Kluwer Academic Publishers. While a definite noun group included in the topic Steedman, Mark. 2000. Information structure and the syntax–phonology interface. Linguistic Inquiry 31. 649–89. prototypically triggers a presupposition of the exis- EVA HAJICOVA tence of an entity that it refers to (more precisely, to the communicative availability of the referent), noun See also Context; Definiteness; Prosody 349 FORENSIC LINGUISTICS Forensic Linguistics Forensic linguistics is the scientific study of language although studies in forensic phonetics often use both as applied to forensic purposes and contexts—the auditory and acoustic methods of analysis. The primary application of linguistic knowledge to legal problems. areas of auditory research in forensic phonetics are It is a new area of linguistics and a rapidly growing speaker discrimination and identification by victims and area of modern applied linguistics. witnesses; voice perception, discrimination, imitation, Recent research demonstrates increased interest and disguise; and identification of group characteristics and research in forensic linguistics. Overviews of lan- of speakers, including first-language interference, guage, law, and the legal process have been elaborated regional or social accent and dialect, and speaker age. by many scholars, e.g. Levi (1994b). Work has contin- An introduction to the use of auditory (and acoustic) ued on the language of the courtroom and on analysis phonetics in forensic examinations is French (1994). of discourse in the language used in legal settings Acoustic phonetics is the study of the physical (Shuy 1993, 1998). Many specific studies and collec- characteristics of speech sounds as they leave the tions relating to linguistic applications to the law have speaker, move into the air, and gradually dissipate. The been written or edited, one of the first and most sig- acoustic analysis of speech sounds requires laboratory nificant being that of Levi and Graffam Walker (1990). observation with instruments and specialized comput- Documentation of advances in forensic linguistics er analysis. appears in the research of individual forensic linguists Vowels and consonants are presented as waveforms and in the bibliographic work of Levi (1994a) and of having three features: amplitude, corresponding to the University of Birmingham. loudness; frequency of complete repetitions, corre- Recent milestones in forensic linguistics include sponding to high/low pitch; and complexity, corre- the following: Judith Levi and Anne Graffam Walker sponding to the periodic waves of a simple sound, or organized and coordinated the 1985 Georgetown to a mixture of waves from a complex sound, like a University conference on Language in the Judicial vowel sound. The waveforms are presented visually as Process. In 1995, Bethany Dumas started Language in a spectrogram. See Hollien (1990) for the forensic the judicial process, an electronic newsletter aimed at application of acoustic phonetics. disseminating information on bibliography, organiza- The primary area of acoustic analysis in forensic tions, courses and programs, and legal cases. During phonetics is speaker identification, but many studies the 1990s, university courses on language and law have also been carried out to identify group character- were developed and presented around the world. The istics of speakers, including their physical height and most important advances in the study of forensic lin- weight; regional, social, or language group; voice and guistics were the University of Birmingham’s initia- accent disguise; effect of intoxication on speech; and tion of Forensic Linguistics: The International Journal technical aspects of speech samples and recordings. of Speech, Language and the Law in 1994 and their Semantics is the study of meaning as expressed in founding of the International Association of Forensic words, phrases, sentences, or texts. The focus of Linguistics, which have since provided serious venues semantic analysis in forensic contexts is on the com- for the presentation of research that are more regular, prehensibility and interpretation of language that is unified, and formal than ever before. difficult to understand. Some studies combine seman- The classification of areas in forensic linguistics is tic and pragmatic approaches to meaning interpreta- evolving as the field develops, usually following those tion. A thorough introduction to this area can be found of the structure and function of language. The research in Solan (1999). The point of view that expert linguists and casework of forensic linguists presently define the have no role to play in helping judges interpret statutes field in a taxonomy that includes the following areas: is presented by Murphy (1998). auditory and acoustic phonetics, semantics, discourse Primary areas of research in forensic semantics are: and pragmatics, stylistics, language of the law and of interpretation of words, phrases, sentences, and texts; the courtroom, and interpretation and translation. ambiguity in texts and laws; and interpretation of Auditory phonetics is the study of speech sounds meaning in spoken discourse, such as reading of rights based on what is heard and interpreted by the human lis- and police warnings, police interviews, and jury tener (the aural–perceptual characteristics of language), instructions. 350  FORENSIC LINGUISTICS Analysis of discourse is the study of units of become very common. In criminal and civil cases with language that are larger than the sentence, such as questioned writings, there is often a need to determine narratives and conversations. Discourse in spoken and whether one author wrote all the writings in a ques- written language can take many forms, especially in tioned set, whether one of a number of possible conversations tied to specific social contexts. Analysis authors authored the questioned writing, or whether a of a speaker’s intended meaning in actual language use single suspect-author can be eliminated or identified is the study of pragmatics. Pragmatics is important for as the writer. forensic purposes because speakers and writers do not Linguistic stylistics uses two approaches to author- always directly match the words they use with the ship identification: qualitative and quantitative. The meaning they intend to convey. This leaves the work is qualitative when features of writing are iden- speaker’s intended meaning more open to interpreta- tified and then described as being characteristic of an tion by the listener, sometimes resulting in miscom- author. The work is quantitative when certain indica- munication. tors are identified and then measured in some way, The linguist most closely associated with develop- e.g. their relative frequency of occurrence in a given ing forensic discourse analysis for a broad range of set of writings. Certain quantitative methods are cases is Roger Shuy. Some of his cases are document- referred to as stylometry. Qualitative and quantitative ed in Language crimes: the use and abuse of language methods complement one another and are often used evidence in the courtroom (Shuy 1993) and in The lan- together to identify, describe, and measure the pres- guage of confessions, interrogation and deception ence or absence of style markers in questioned and (Shuy 1998). known writings. Primary areas of research in forensic discourse and Examples of descriptive (qualitative) style markers pragmatics include: analysis of spoken and written lan- from two separate cases appear in Figure 1. guage; study of the discourse of specific contexts, such Another recent development in forensic authorship as dictation, conversations, and hearings; the language identification is the application of stylistic analysis to of the courtroom, i.e. of lawyers, clients, questioning, computer programming. Researchers (Krsul 1994) and and jury instructions; and the language of specific security analysts, working on proprietary issues and speech acts, such as threats, promises, and warnings. virus sources, have identified various style indicators The focus of forensic stylistics is author identifica- as indicative of authorship in programming code, e.g. tion of questioned writings. While the methods were variable names, layout, upper/lowercase letters, place- developed mainly for literary purposes, applications of ment of comments, debugging symbols, line length, stylistic analysis to forensic authorship problems have and ratio of white lines to code lines. Linguistic Unit In QUESTIONED In KNOWN writing Writing Same-Author Case A: confidentiality Same-Author Case A: I’ll see you. Different-Author Case B: Spanish hice Different-Author Case B: Spanish estaba Figure 1. Style-marker examples for spelling and word formation. 351 FORENSIC LINGUISTICS Primary areas of research in forensic stylistics for physical traces that assist in uncovering the history relate to the descriptive and quantitative methods of a document. The document examiner observes fea- of authorship identification, dialect evidenced in writ- tures of handwriting (letter size, formation, and rela- ten language, the questioned time and occasion of tive proportions; letter slant, spacing, pressure, line writing, stylometry, statistical methods of analysis, use quality, connecting strokes, etc.) and of typing (type- of corpus linguistics in forensic analysis, and comput- face or font style, spacing between letters and lines, er programs for the analysis of style. A relatively com- association of a document with a particular machine or plete survey of forensic stylistics can be found in type of typewriter, printer, or copier). McMenamin (2002). Previous studies indicate that examination for Perhaps the most important development in the lan- authorship determination includes numerous charac- guage of the law was that begun by David Melinkoff in teristics of writing style (punctuation, spelling, abbre- his 1963 book, Language of the law. Melinkoff pressed viations, forms of dates, etc.), but newer studies are for clarity and brevity in the law, which extended to a drawing a sharper line between forensic document later movement to simplify the language of laws, insur- examination and the elements of writing style studied ance policies, and consumer literature. Systematic in linguistic stylistics. analysis of legal language is carried on today with, for Psycholinguistics is the academic discipline that example, Tiersma’s (1999) emphasis on plain legal lan- integrates the study of psychology, linguistics, and guage, language rights, and the pragmatics of written cognitive science. The acquisition, comprehension, legal language as well as of courtroom language. and production of language are studied in various The courtroom personae who speak are witnesses, related ways: cognitively, neurologically, and concep- lawyers, and judges. Research related especially to the tually. However, this is not what has been referred to discourse and pragmatic use of courtroom language as ‘psycholinguistics’ in forensic contexts. The late has made this one area of forensic linguistics that can Murray S. Miron developed the so-called ‘psycholin- significantly affect case outcomes. Studies of court- guistic approach’, which examines written or spoken room language are analyses of the language of all the language to profile the origins, background, and players: the language of witnesses (witness examina- psychology of the originator, especially in the context tion, victims, children, men vs. women); the language of threats. It is difficult for a nonpsychologist (e.g. a of lawyers (trial language, legal debate, closing argu- judge or jury) to understand and evaluate the nexus ments); and the language of judges (trial participation, made between a written-language threat and the jury instructions). diagnostic profile of the writer, and the practice of Interpreting is a complex skill and especially diffi- ‘psycholinguistics’ can be risky and may lead to mis- cult in forensic contexts. Interpretation studies such as takes. This kind of ‘psycholinguistics’ is not common- those in Cooke et al. (1999) focus on interpretation ly regarded as a subdiscipline of forensic linguistics. tasks specific to questions and answers in testimony, the perceived role of the interpreter, interpreter educa- tion, the right to interpretation, etc. Forensic and aca- References demic scholars are giving more and more attention to Bryant, M. 1930, 1962. English in the law courts: the part that the theory and practice of interpretation and transla- articles, prepositions and conjunctions play in legal deci- tion, especially in these areas, and in others such as sions. New York: Frederick Ungar. pretrial interpreting, courtroom interpretation, inter- Cooke, M., D. Eades, and S. Hale. 1999. ‘Introduction,’ to spe- pretation with cultural and dialect differences, ques- cial issue on legal interpreting, with eight other articles. tioning in interpreted testimony, and the absence of Forensic Linguistics 6(1). 1–5ff. Danet, B. 1980. Language in the legal process. Law & Society interpretation. Review 14(3). 445–564. Translating in the legal context requires much more French, P. 1994. An overview of forensic phonetics with partic- than a literal, word-for-word matchup of two lan- ular reference to speaker identification. Forensic Linguistics guages. Good translations are constrained by the 1(2). 169–81. intended meaning of the writer, the new text created by Hollien, H. 1990. The acoustics of crime: the new science of forensic phonetics. New York: Plenum. the translator, and the meaning given the translated Krsul, I. 1994. Authorship analysis: identifying the author of a text by the reader. program. Technical Report CSD-TR-94-030, Department of Areas associated with forensic linguistics include Computer Science, Purdue University. document examination and aspects of psycholinguis- Levi, J.N. 1982. Linguistics, language, and law: a topical bibli- tics. The examination of questioned documents relies ography. Bloomington: Indiana University Linguistics Club. ———. 1994a. Language and law: a bibliographic guide on the scientific study of the physical evidence of a to social science research in the USA. Chicago: University document. Forensic document examiners look to the of Chicago Press and Washington, DC: American Bar writing instrument, the writing surface, and the writer Association, Teaching Resource Bulletin No. 4. 352  FRANCE ———. 1994b. Languages as evidence: the linguist as expert wit- Solan, L.M. 1999. Can the legal system use experts on mean- ness in North American courts. Forensic Linguistics 1. 1–26. ing? Tennessee Law Review 66 (summer) 1167. Levi, J.N., and A. Graffam Walker (eds.) 1990. Language in the Tiersma, P.M. 1999. Legal language. Chicago: University of judicial process. New York: Plenum Press. Chicago Press. McMenamin, G.R. 2002. Forensic linguistics. Boca Raton: CRC Wetter, J.G. 1960. The styles of appellate judicial opinions. Press. Leyden: A.W. Sythoff. Melinkoff, D. 1963. The language of the law. Boston: Little Electronic Resources Brown. Forensic Linguistics: The International Journal of Speech, Murphy, H.F. 1998. Linguistics and law: an overview of foren- Language and the Law: http://www.bham.ac.uk/forensic sic linguistics. Journal of Law, Intellectual Property, and linguistics/ Technology 1. 1–20. International Association for Forensic Phonetics: http://www. O’Barr, W. 1982. Linguistic evidence, language, power and iafp.net/ strategy in the courtroom, New York: Academic Press. Language in the Judicial Process: http://www.outreach.utk. Shuy, R.W. 1984. Linguistics in other professions. Annual edu/ljp/ Review of Anthropology 13. 419–45. Plain Language—USA http://www.plainlanguage.gov/ ———. 1986. Language and the law. Annual Review of Plain Language—UK http://www.plainenglish.co.uk/ Applied Linguistics 7. 50–63. GERALD MCMENAMIN ———. 1993. Language crimes: the use and abuse of language evidence in the courtroom. Oxford: Blackwell. ———. 1998. The language of confessions, interrogation and See also Discourse Analysis; Psycholinguistics; deception. Los Angeles: Sage Publications. Semantics France This article deals with the linguistic situation of Strasbourg Oaths (842 AD). The dialects of the north, France, and with the French language in its interna- the langues d’oïl, showed more Germanic influence, tional context. Since 1992 the French constitution has as against the Romance langues d’oc in the south. stated that the language of the republic is French. As a (This division is still reflected today in the existence of major western European state (population 59 million) three main linguistic zones, the north being divided and former imperial power, France has a long history from the south by a line running east from Poitiers, of standardizing and promoting French, although there with an intermediate franco-provençal triangle fan- are other languages within its borders. ning out from the region of the Alps.) Among the major dialects of the north with a written form, it was the language of the Ile de France (around Paris) which The French Language was set to become the high prestige variety. So-called French is classified as a Gallo-Romance language. The ‘Old French’ of the twelfth and thirteenth centuries Romans who colonized Gaul in the first century BC had a wide range of vocalic sounds and had reduced to had a profound and lasting influence on the institu- two the Latin system of cases. During the fourteenth tions, culture and language of the country that we century case endings disappear, with a corresponding know as France. Of the period prior to Roman domi- reliance on determiners, the vocalic system is simpli- nation, only a few traces remain of the Gaullish fied and we find that the final consonants have been (Celtic) substratum—everyday words e.g. for trees if dropped. Through the Edict of Villers-Cotterêts of ‘yew tree’, or place names, e.g. the dun (‘fort’) ending 1539, French replaced Latin as the official language found in Verdun. The Romans spread their language for administrative and legal documents. Ten years later (or languages, from classical written Latin, to a later the writer Joachim du Bellay in his Défense et illus- more popular and largely spoken form) so successful- tration de la langue française advocated the use of ly that when Germanic tribes invaded in the fifth cen- French in literature; du Bellay and other writers con- tury AD, their languages influenced the Latinate tributed greatly to the enrichment of the French lan- language in use (e.g. in vocabulary items such as guage, although the Renaissance also brought in its guerre from werra) but did not supplant it. Evidence wake an influx of Italian loan words. The seventeenth for the emergence of a recognizably distinct French century saw the codification of the ‘standard’ lan- language (or set of dialects) is found in the bilingual guage, with the Académie Française being set up in 353 FRANCE 1635. Nevertheless, there were still significant num- in the mid twentieth century. Their modest revival was bers of speakers of other languages by the mid-eigh- eventually made possible by the place won for them teenth century when the revolutionaires tried to ensure within the French education system, and by a shift in the development of a unified state. A report by the government thinking initiated by the socialist presi- Abbé Grégoire estimated that no more than three mil- dent Mitterand’s recognition of the ‘right to differ- lion (about a fifth of the population) mastered standard ence’ in 1981. The loi Falloux of 1851 had stipulated French. From the late nineteenth century onwards, that French was the only language to be used in standard (Parisian) French was further promoted by schools, but 100 years later the loi Deixonne allowed free, universal schooling (from 1880 on by Jules schools to give a few hours optional teaching a week Ferry), conscription in World War 1 and the mobility in those regional languages that were not the language of civil servants. Today, facilitated also by the media, of a foreign power (Corsican, considered a dialect of standard French dominates across the whole of Italian, was only added in 1974). Since 1981 govern- France, sometimes alongside a regional form of ment policy has become gradually less hostile to the French and sometimes alongside a regional language regional languages, which have been boosted by the or dialect. Even apart from dialectal variation, there is écoles associatives in which they can be the medium a good deal of sociostylistic (register) variation. While of instruction. Today, the following are recognized as the efforts of bodies such as the Académie to conserve ‘regional languages’: three Romance languages, the language have applied mainly to a formal, usually Occitan, Corsican (close to Italian) and Catalan (an written, register, the spoken language demonstrates official language of Catalonia in Spain); two Germanic certain syntactic differences (e.g. deviation from languages (Alsatian in Alsace and part of Lorraine) canonic Subject–Word–Object word-order, dropping and Flemish for a small number of speakers near the of pre-verbal negative particle ne) and characteristic Belgian border; and one Celtic language (Breton) and lexical features (a rich range of colloquial vocabulary the non-Indo-European language Basque (also spoken for everyday items (bouquin-livre ‘book’) with many on the Spanish side of the border). It is difficult to get abbrieviated forms (apéro-apéritif ‘before-dinner an objective estimate of the numbers of speakers, or drink’). So-called français jeune refers to the French even of the health of these languages (but see Judge spoken by some young people, characterized by a mix 2000, Sanders 1993 and the publications of the of colloquial French, ‘slang’ vocabulary such as le ver- European Union on the Lesser Used Languages of lan based on the inversion of syllables with subsequent Europe EBLUL). modifications (thus laisse béton—laisse tomber ‘leave Basque (approximately 80,000 speakers) has been it’, meuf —femme ‘woman’) and borrowings from var- boosted by the ikastolak (Basque language schools), ious languages. Borrowings from English are plentiful which have received state funding for primary and in all registers of contemporary French, hence recent secondary levels since 1994. The first school opened attempts to legislate in favor of the protection of its doors in 1969 to five pupils and by 1997 there French, from the Bas-Lauriol law of 1975 to the con- were 1630 pupils receiving nursery, primary or sec- troversial loi Toubon passed in 1994. A debate that had ondary education through the medium of Basque. rumbled on for years about orthographic reform was There are four dialects of Breton and probably about finally settled in 1991 with a compromise solution, 500,000 speakers. In 1997, 1,751 pupils sat for the with a small number of spelling rationalizations being French school-leaving exam (baccaulauréat) having recommended but not imposed. Apart from the followed a complete cursus through the diwan Académie which still pronounces (usually prescrip- (Breton schools) created 20 years earlier, and had the tively) on language matters, there are currently various possibility of taking some courses in Breton e.g. at commissions devoted to protecting French, such as the the University of Rennes. There are an estimated Conseil supérieur de la langue française which pro- 100,000 speakers of Catalan in the Languedoc- motes ‘good usage’, and the Délégation générale à la Roussillon area of France, who are at once encour- langue française et aux langues de France part of aged and overshadowed by the support given to the whose job is to liaise between the terminological com- language by the provincial government of Catalonia missions set up by each government ministry to stem on the Spanish side of the Pyrenees. Catalan schools the influx of English words, as well as to promote started in 1976 and it is now possible to go up to doc- French world-wide. toral level at the University of Perpignan. The future of Corsican (approximately 200,000 speakers) is linked to on-going discussions about the status of the The Other Languages of France island, including the possibility of a regional assem- The regional languages that the revolutionaries and bly and of the increased use of Corsican in schools, educationists had tried to suppress began to fight back many inhabitants being bilingual in French and 354  FRANCE Corsican. In a 1982 survey by the French agency The largest groups currently originate from North INSEE, 96% of Corsicans on the island understood Africa: thus, there are estimated to be approximately the language and about 86% claimed to speak. It is 1.5 million speakers of Arabic and 0.5 million Berber possible to study Corsican up to postgraduate level speakers in France. There have been some attempts to (University of Corti), and there has been a strong lit- provide instruction for Arabic heritage pupils, but so erary revival. In some respects the Germanic lan- far these have partly foundered on such issues as guages of north-east France fare less well in the which dialect to teach (classical, Algerian, etc.). To education system (and in the media) than other sum up, levels of educational provision vary from lan- regional languages. As Flemish and Alsatian were guage to language, as do levels of competence. As for classified as official languages of a neighbouring most languages with no official backing, the regional state, no support was given under the terms of the loi languages have suffered from dialect fragmentation, Deixonne. A current estimate of 80,000 speakers of and the norm adopted has not pleased all activists. Flemish may be optimistic. Alsatian is treated as a Those now being educated in the écoles associatives spoken dialect for which the written form is German, may be in a position of having parents who do not and the high number of those represented as studying speak the language and grand-parents who speak a dif- this ‘regional language’ is misleading, as it includes ferent dialect of it; few will have the chance to speak pupils taking German, and for Alsatian itself there are the language in the work-place. Within the European no schools, only classes associatives. Occitan should Union, French is one of the two major working lan- be the giant of the regional languages, covering as it guages, although constantly under pressure from does 31 départments in the southern half of France. English. After some delay France became a signatory However, various factors such as migration to and to the European Union’s Charter for Regional or from the area, and dialectal differences between Minority Languages but did not ratify it, on the Gascon and Northern and Southern Occitan (the lat- grounds that it is incompatible with the constitution ter dividing further into Provençal and Langdocien), which, holding all citizens to be equal, does not rec- mean that the survival of this language group, with its ognize minorities. However, the title of the Délégation rich medieval literary heritage, cannot be taken for générale à la langue française has been amended to granted. The orthography proposed by the nineteenth include the ‘languages of France’ which may be a century poet Mistral and his association, the hopeful sign for the regional languages. Félibrige, was not necessarily suitable for Occitan in the twentieth century and a compromise form has The French-Speaking World now been reached. A flowering of theatre and song in (la Francophonie) the second half of the twentieth century was accom- panied by a degree of political activism. There are France’s first wave of colonial expansion in the now over 30 calendretas (schools), and 12,532 chil- sixteenth century took the French language to the dren were taught Occitan in the school year 1997–8. Americas, and its second during the nineteenth and the Overall, there may be up to 10 million people who twentieth centuries to North and sub-Saharan Africa, understand the langues d’oc, and two million who the Pacific and parts of Asia. Today a reminder of this practise them. The langues d’oïl, the dialects of imperial past remains in the use of French in France’s northern French, were treated as ‘patois’ and given former territories (Québec) as well as in her overseas little consideration for many years, but a number of departments (such as the French Antilles) or territories studies have recently demonstrated the tenacity of (e.g. French Polynesia). Depending on the country or picard or the chtimi dialect of the Lille area. area, French may be the only official language (as in Beside the indigenous languages of metropolitan Québec), one of two or more official languages (cf. France, there are other languages that are worthy of French and English in New Brunswick or the same mention. A report by Bernard Cerquiglini in 1999 two languages in Cameroun), one of several official identified 75 languages in ‘greater’ France. Firstly, the languages of a multi-lingual state (Belgium or other languages of the overseas departments of France Switzerland), or it may be a major language of educa- (e.g. the Polynesian and Melanesian languages of the tion and business without being an official language French Pacific territories) are technically ‘languages (e.g. Morocco or Madagascar). It is sometimes of France’. Secondly, so are the non-territorial lan- claimed that the future of French lies with its use out- guages (Romany languages, Yiddish) and the immi- side France and that will be, for example, more speak- grant languages of metropolitan France. The latter are ers of French in Africa than in France. However, this the result of successive waves of migration—Polish, depends on the definition of a ‘French-speaker’, not all Italian, Spanish, Portuguese, Chinese, and speakers of of those who use French as a vehicular language being sub-Saharan African languages, to mention only a few. fluent speakers, and such varieties as le français 355 FRANCE populaire d’Abijan differing substantially from stan- Caribbean, both French and French-based Creole are dard metropolitan French. What follows is an attempt spoken in independent Haiti and in the French depart- to estimate the numbers of French-speakers in illustra- ments of Guadeloupe, Martinique and Guyane, but tive situations from around the world. only French is recognized as the official language in Europe To the 59 million citizens of metropolitan the latter. France must be added about 6.5 million speakers of Africa French is widely used as a vehicular lan- French in Belgium, Switzerland, Monaco, guage in the former French colonies in sub-Saharan Luxembourg, Andorra and small pockets such as the Africa. The situation is fluid, but currently the follow- Val d’Aosta in Italy. 4.5 million of these are in ing are among the countries in which French is an offi- Belgium where French is one of three official lan- cial language but co-exists in daily use with African guages, the other two being Flemish (i.e. Dutch) and languages: Burkina Faso, Congo, Guinea, Ivory Coast, German. The original superior status accorded to Mali, Niger, Senegal. In North Africa (Tunisia, French when Belgium became a nation in 1830 gave Algeria, Morocco), there has been a policy of ‘ara- rise to a succession of protests, which eventually led bization’, but French is still widely used. In the Indian to the establishment of a trilingual state, with parallel Ocean, French is used in education in Mauritius (offi- organizations for the French and (now majority) cial language: English) and Madagascar (official lan- Flemish speaking populations. Thus, apart from the guage: Malagasy) and Réunion (a department of small German-speaking community near the German France, where Creole co-exists with the official lan- border, the country is divided into three linguistic guage French). entities: the French-speaking community (which Since some speakers have French as a second or includes Belgian French and the Romance wallon third language, estimating the number of speakers dialect), the Flemish-speaking community, and worldwide is not easy, but the estimate of 80–100 mil- Brussels, accorded special bilingual status. Educated lion with French as a mother-tongue is sometimes Belgian and Parisian French are very close, with some given. A great strength of French is that it is present on particular features characterizing Belgian French such every continent, and another is that educated speakers as the use of the numerals septante, octante etc. for from around the world speak a mutually comprehensi- standard French soixante-dix (‘sixty’) and quatre- ble form of French which approximates to the Parisian vingts (eighty). norm. An international ‘francophone summit’ meets The Americas French and English are the official every two years to promote cooperation among languages of federal Canada, though each province French-speaking countries. In addition to France, cer- determines its linguistic status. The majority of tain other countries now play an important role in pro- French-speakers are in Québec (population 6.5 mil- moting French worldwide. In certain areas e.g. lion) where French is the only official language, but terminology banks, or the ‘feminizing’ of the language there are also French speakers elsewhere, notably in (particularly for job titles), Francophone countries the Maritime provinces, Ontario, Manitoba and such as Canada have acted more effectively than has Saskatchewa. Québécois French retains traces of the France herself. French is indubitably the vehicle for a sixteenth-century regional speech of the original vibrant francophone culture, in literature, cinema and migrants, as in the characteristic pronunciation of media in many countries. /mwe/ for moi (‘me/I’). The linguistic range extends from educated québécois French, which may show References certain differences, largely in pronunciation and lexis, from metropolitan French, to a variety of rural and Ayres-Bennett, W., and J. Carruthers, with R. Temple. 2001. Studies in the Modern French language: Problems and per- urban dialects—in the latter case most notably the spectives. London: Longman. urban working-class joual of Montreal. In the last half Ball, R. 1997. The French-speaking world. London and New century, French in Quebec has benefited from strenu- York: Routledge. ous provincial government efforts to promote its use Judge, A. 2000. France: ‘one state, one nation, one language?’ through linguistic legislation and the setting up of bod- Language and nationalism in Europe, ed. by S. Barbour and C. Carmichael. Oxford: Oxford University Press. ies such as the Office de la langue française. Lodge, A., N. Armstrong, Y. Ellis and J. Shelton. 1997. French-speaking emigrants deported by the Exploring the French language. London: Arnold. English from the Arcadian area of the Maritimes took Sanders, C. (ed.) 1993. French today. Cambridge: Cambridge their language to New England and to Louisiana University Press. where French was further bolstered by French Walter, H. 1994. French inside out. London and New York: Routledge. Canadian economic migrants (in New England) and CAROL SANDERS by French and Creole-speaking migrants from the Caribbean (in the case of Louisiana). In the See also Indo-European 4: Romance 356  FRENCH LANGUAGE French Language French is one of the major languages of Europe, spo- French (Occitan area), for example, mute (written) e’s ken as a first language not only in France (where it tend to be sounded, while they are silent in Standard originated) but also in parts of Belgium, Switzerland, French. Alsatian French shares many features of pro- Luxembourg, and a small region of Italy (the Aosta nunciation with the Germanic dialect spoken in the Valley). In addition, as a result of colonization, French region. is spoken in many other parts of the world. In Africa, The same is true of French spoken outside France. it is the official language of 18 countries, which occu- The French spoken in the rest of Europe tends to dif- py a good half of the land area of Africa, from fer from the French of France most notably in pronun- Mauritania in the west to the Central African Republic ciation and somewhat in vocabulary. The numbers 70, to Chad, Mali, and Senegal in the east to Congo, Zaire, 80, and 90, for example, have different names in Rwanda, and Burundi in the south. It is also spoken Belgium and Switzerland than in France. But it is the and understood by a considerable section of the popu- French outside of Europe that shows the most diversi- lation in the three Arab countries of North Africa that ty—e.g. québécois in Canada, Cajun in Louisiana, the constitute what is called the Maghreb (Morocco, various types of French in former Indo-China, the Algeria, and Tunisia). Across the Atlantic, in North Indian Ocean, French Polynesia, and Africa all have America, it is one of the two official languages of developed their own specific traits of the language. As Canada and is the first language for many Canadians one example, in Canadian French, stress placement in Québec and other provinces of Eastern Canada. (word accent), intonation (inflection of the voice), and There are vestiges of French in the United States as the vowel system are not the same as in Standard well, particularly in Louisiana, Missouri, and parts of French, nor is the vocabulary. As for French creoles New England, and in the Caribbean islands of (found in Louisiana, French Guyana, the Caribbean, Martinique and Guadeloupe (which are still dependen- and the Indian Ocean), they are not intelligible to cies of France), as well as the Republic of Haiti. On speakers of French. Their vocabulary is based on that the continent of Asia, it was spoken by many in former of French, but they are entirely new languages with French Indo-China and still retains its prestige in Laos their own particular morphology and syntax. and Kampuchea (former Cambodia), but is much less French has a variety of roles in the linguistic reper- used in Vietnam than formerly. There are many islands toire of its users. It can be a native language (for exam- in the Indian Ocean and Oceania where French is still ple, in France, Belgium, Canada, Luxembourg, and spoken: Madagascar and smaller islands nearby, and Switzerland). In many countries, it is an (or even the) New Caledonia, French Polynesia (including Tahiti), official language used for governmental and other pur- New Hebrides, and various smaller islands. In all, poses (in the nations just mentioned; also in Haiti, the there are about 88 million francophones (speakers of Comoros, and in many former French and Belgian French) in the world. After English, it is the second colonies in sub-Saharan Africa). In still others, it func- international language, used by institutions such as the tions as a lingua franca (a language of communication United Nations and UNESCO (United Nations by those whose own languages are very different from Educational, Scientific, and Cultural Organization). each other), for example, in sub-Saharan Africa and in Although French is the official language of France New Caledonia. For many people, it is their second or and, since the beginning of the twentieth century the third language in contexts where there may be many media have been accessible everywhere, not everyone other languages around; in such cases, it is highly in France speaks French. First of all, there are still prestigious to speak French and thus only the well- other languages spoken in France: e.g. Basque in the educated elite do (in French Polynesia and in Creole- southeast (on the border with Spain), Breton in the speaking areas such as Haiti, French Guyana, west, Flemish in the north (on the border with Mauritius, some of the Caribbean Islands, and the Belgium), and Alsatian and Lorrain in the east (on the Seychelles). It is also, for some, a language of culture border with Germany). In addition, there is Occitan (in Lebanon and the Maghreb). In many countries out- (including Provençal) in the south, Franco-Provençal side Francophonie (those countries where French is (spoken in France, Switzerland, and Italy), as well as recognized as a means of communication), French many small local languages spoken all through France. remains—after English—a preferred foreign language All these have their effect on French. In Southern learned at school. 357 FRENCH LANGUAGE History of French: Old and Middle French language. In 1635, the French Academy was founded to give ‘firm rules’ to the language, to make it ‘pure and French is a descendant of the Vulgar Latin (common eloquent’, and to compose a grammar and a dictionary. spoken language) brought by Roman settlers to the Claude Vaugelas, in his Remarks about the French lan- area occupied by modern-day France in the last centu- guage (1647), established that the linguistic norm ry BC, after it was conquered by Julius Caesar in the should be based on ‘good usage’, i.e. the French spo- Gallic Wars. It became the spoken language of Gaul, as ken at the Court and used by the best writers. During the region was then known, and remained so even after the Classical period, French acquired great prestige the fifth-century invasions by Germanic tribes. But this abroad. It was the international language used in Gallo-Romance way of speaking diverged increasing- European courts and became the medium of diploma- ly from Latin, as recognized by the Council of Tours cy. Such was the fame of French that competitions (813), which recommended not using Latin in preach- were held to extol its virtues. In 1784, Antoine Rivarol ing because the common people could not understand won first prize from the Academy of Berlin for his it. Moreover, by the end of the ninth century, the north- ‘Discourse on the universality of the French language’. ern dialects were most influenced by the Germanic Great writers also contributed to the preeminence of speech of the dominant Franks, and the southern French and the development of the literary language. dialects remained closer to Latin. This resulted in the emergence of two separate languages: langue d’oïl in Modern French the north and langue d’oc in the south (oïl and oc were the words used for ‘yes’ in the two languages). Modern By the time of the French Revolution in 1789, French French is a descendant of langue d’oïl. The oldest doc- had spread to the bourgeoisie (middle class) in urban ument in langue d’oïl is the Strasbourg oaths (843), areas; but outside the Paris region, the patois were still which marks the beginning of the Old French period very much the vernacular of the people. In 1790, Abbé (ninth to fourteenth centuries). Latin being virtually Grégoire conducted the first survey about the use of the only written medium of the early middle ages, doc- French and found that 45% of the population of France uments in Old French are rare. Literary texts appear were not able to speak French. The Revolutionaries held first in langue d’oc, the language of the troubadours, at that linguistic uniformity was essential to the cohesion the end of the eleventh century, a hundred years before of the nation and equated the patois with backwardness their counterparts in Old French. and antirepublican sentiments. They wanted, therefore, In the tenth century, French kings made their court to exterminate the patois and to have French taught at in Paris and, as the royal domain increased, so too did school to every citizen. These ideas were not imple- the prestige of the dialect they spoke (called francien). mented until almost a century later, during the Third In the twelfth century, Parisian speech was already a Republic. Through laws devised by Jules Ferry and model to be emulated; by the end of the Old French passed in 1881–1886, primary school education was period, the King’s French was in the process of made free, secular, and compulsory. The teaching of becoming the written norm in France. French was central to the curriculum, and the use of any In the Middle French period (fourteenth to six- other language at school, inside or outside of the class- teenth centuries), the use of French continued to room, was banned and punished. In addition, railways, expand. In 1539, a royal edict banned Latin from court roads and cities, military service, mass media (newspa- proceedings and legal deeds and replaced it with the pers, radio, television)—all contributed to impose French spoken at the time. French was also promoted French as the only language of the nation in the twenti- as a vehicle of learning and culture at the expense of eth century. In 1992, an amendment to the Constitution other dialects, which were relegated to the status of officially proclaimed French to be ‘the language of the patois (local vernaculars). Republic’. At the same time, there has been a more tol- In 1549, Joachim Du Bellay’s Defense and illustra- erant attitude toward regional languages and regional tion of the French language maintained that French differences. In 1951, the educational policy was relaxed was not inferior to the classical languages (Latin and to allow some regional languages, e.g. Basque, Occitan, Greek) and should be used for literary works. During and Breton, to be taught in French schools (the this period, the King’s dialect began to be codified Deixonne law). By then, however, there were hardly and, with the advent of printing, attempts were made any school-age native speakers of patois left and the to simplify and regularize the orthography. demise of these languages is all but irreversible. Classical French Protecting the French Language The Classical French period (seventeenth to eighteenth With the rise of Anglo-American influence in the years centuries) witnessed further efforts to standardize the following World War II in international commerce, 358  FRENCH LANGUAGE science, technology, and especially popular culture, a typically comes after the noun it modifies, and it strong view has emerged in some French government agrees with that noun in gender and number (singular and intellectual circles that French is under serious or plural). The noun is often used with a definite arti- threat from English. While the proportion of English cle (le, la, les in French) and an indefinite article (un, borrowings in everyday speech is in fact very low, une, des in French), both of which are placed before concern for maintaining the purity of French has in the noun. The informality/formality of situations is recent times led to the creation of the General also characteristically reflected by the choice of pro- Delegation for the French Language (1989) and vari- nouns of address, tu/vous ‘you’ in French. ous ministerial committees charged with developing During the course of time, French has lost or French vocabulary in economic and technical acquired features that did not exist in Latin and many domains. The Bas-Lauriol (1975) and Toubon (1994) of which are not found in the other Romance lan- laws were passed to limit the use of foreign languages guages. Some of its most characteristic sounds are in the spheres of trade, employment, and services. innovations: pharyngeal (guttural) R, nasal vowels as Similar actions were taken in Québec where Anglo- in vin ‘wine’ and bon ‘good’, and rounded front vow- American influence is overwhelming. The Charter of els as in rue ‘street’ or peu ‘little’. It is true that adjec- the French Language (1977) makes French the only tives agree with the nouns they modify in gender and official language in Québec and gives precedence to number, but the gender of the noun is often not overt- French in all areas of public life (education, work, ly given and simply has to be memorized, and for both economy). adjectives and nouns the plural found in writing (with ‘s’) is typically not sounded in the spoken language. Unlike Latin, French still has a rich verb morphology, Francophonie but it has been greatly restructured since Latin times. In the 1960s, the notion of a francophone identity for The imperfect and simple past of Latin were aug- users of French around the world became prevalent. mented by a conditional and a new future tense, as African leaders suggested a new type of association, well as a series of compound tenses (present perfect, culturally and linguistically based, called pluperfect, future perfect, etc.) and other tenses with Francophonie, which would replace the old colonial some sort of auxiliary verb (e.g. future with aller ties. The francophone movement was enthusiastically ‘go’). French also has a subjunctive, although its use supported by France, Québec, and French-speaking is often predictable: it is most often used in sub- Belgium. In France, the Supreme Council for ordinate clauses after certain kinds of verbs in the Francophonie (founded in 1984) has contributed main clause or after certain kinds of conjunctions. actively to the promotion of French outside of France. Modern French syntax is characterized by a fixed sub- And since 1986, Francophonie summits have been ject–verb–object word order; adjectives typically are organized every other year with about 40 participating placed after the noun they modify, as are relative countries. clauses. Questions are asked by placing the subject after the verb, especially in formal usage, or by using the expression est-ce que ‘is it that’. In the spoken lan- Romance Languages guage, questions are also often asked through a rising French is a Romance language because it has descend- inflection (intonation). ed from the language spoken by Roman settlers as they Where French differs most is in how the verb is went out to conquer neighboring lands and founded formed: the other Romance languages typically have the Roman empire. The other Romance languages suffixes that give the person (1st, 2nd, or 3rd) and include Spanish, Catalan, Portuguese, Occitan, Italian, number (singular, plural) of the subject of the verb, but Rhaeto-Romance (spoken mostly in Switzerland), French has lost most (but not all) of its verb suffixes. Dalmatian (spoken on the Dalmatian coast, in former As a result, it has had to rely on nouns or pronouns Yugoslavia), Rumanian, and Sardinian. As descen- (that come before the verb) to give that information; dants from Latin, the Romance languages have much and typically, if no noun is present, then a pronoun has in common, in particular their vocabulary. They also to be used. In most of the other Romance languages, share other traits that they inherited from Vulgar Latin. the use of a subject pronoun is optional. Compare Noun morphology was simplified in most of the French je chante with Latin/Spanish/Italian canto Romance languages, so that there are two genders ‘I sing’ (where cant- means sing and -o means ‘I’, and (masculine and feminine) instead of three (neuter), the use of the pronoun ego/yo/io is optional). and the case system (by which grammatical relations The vocabulary of French has also evolved over in the sentence are indicated by various endings) also time. The lexical stock is essentially of Latin origin, disappeared. They also share the fact that the adjective but many learnèd words from Greek were added in the 359 FRENCH LANGUAGE Renaissance period, as were words borrowed from inherited from Latin, but on, which used to mean Latin. There are borrowings from numerous other lan- ‘one’. The negative marker ne is also just about gone. guages (Germanic, Spanish, Arabic, Italian, and In other words, twenty-first century French is evolving English), which reflect the history of contacts that the further and further away from its origins in Vulgar French people have had with other cultures. Latin and from the other Romance languages. Current Changes in Spoken French References Social background, sex, and age affect the way in Ager, Dennis. 1990. Sociolinguistics and contemporary French. Cambridge: Cambridge University Press. which French is spoken. Young people from the outer ———. 1996. ‘Francophonie’ in the 1990s: problems and suburbs of large French cities (many of whom are poor opportunities. Clevedon: Multilingual Matters. and/or from immigrant families) use more nonstan- Ball, Rodney. 1997. The French speaking world: a practical dard features of the language, as well as verlan, a introduction to sociolinguistic issues. London: Routledge. back-to-front slang, and borrowings from English and Harris, Martin. 1988. French. The Romance Languages, ed. by M. Harris, and N. Vincent. London: Routledge. Arabic. The forms of language that are used are also Lodge, Anthony. 1993. French: from dialect to standard. influenced by the kind of linguistic situations in which London: Routledge. speakers are involved. In the written language, the Offord, Malcolm. 1990. Varieties of contemporary French. simple past is a marker of formal registers and in other London: Macmillan. usages is replaced by the present perfect form. Posner, Rebecca. 1996. The romance languages. Cambridge: Cambridge University Press. But what is perhaps most notable right now is the ———. 1997. Linguistic change in French. Oxford: Clarendon fact that the ordinary spoken language used by educat- Press. ed speakers is moving further and further away from Rickard, Peter. 1989. A history of the French language, 2nd edi- the written language and the standard French spoken tion. London: Hutchinson. only 50 to 10 years ago. For example, subject pronouns Walter, Henriette. 1994. French inside out, transl. by P. Fawcett. London: Routledge. like je, tu, and il are becoming prefixes, attached to the front of the verb, along with object pronouns like me, LINDA R. WAUGH te, and le. The pronoun for ‘we’ is no longer nous, See also Algeria; Indo-European 4: Romance Fromkin, Victoria Alexandra Victoria Fromkin described herself as a ‘professional linguistic competence and linguistic performance and radical revolutionary’ in her youth (quoted in Cheng his conception of linguistic theory as an attempt to and Sybesma, ‘Interview with Victoria Fromkin’, Glot understand the nature of language and the human mind International 2:5.1(1996)), and the qualities of intel- were considered revolutionary at the time, and that, of lectual independence that sustained her in that work course, appealed to Fromkin. later came to characterize her career in linguistics. Fromkin’s doctoral dissertation was written under She began the study of linguistics as a graduate stu- the direction of the linguist and phonetician Peter dent in 1961, when the American linguistic world was Ladefoged, and for some years thereafter she worked enmeshed in challenges posed by Noam Chomsky and on the physical nature of speech sounds and the prop- generative grammar to the uniquely American form of erties of a universal phonetic theory. Some linguists at structuralism developed in the 1940s and 1950s by fol- the time rejected phonetics as part of the discipline, lowers of Leonard Bloomfield. Her first teachers at the arguing that it dealt merely with physical phenomena, University of California, Los Angeles (UCLA) were not with the structure of language. Fromkin’s view of not part of the generative movement, but at an institute linguistics was more generous: on machine translation in Venice, Italy, she met The question of the boundaries of phonetics and lin- Yehoshua Bar-Hillel, a philosopher and colleague of guistics, or whether such boundaries should be drawn, Chomsky. Talking with Bar-Hillel, Fromkin’s under- is an important one. . . . When I first came into the field standing of linguistics and the goals of linguistic theo- I was interested in electromyographic registrations of ry were clarified. Chomsky’s distinction between linguistic units, and there were people who said: ‘That is 360  FROMKIN, VICTORIA ALEXANDRA not linguistics’, and I said: ‘But linguistics is whatever the linguistic capacities of the human brain and the tells us more about the nature of human language and modularity of language. how language is realized in speech and in perception’. Modularity—the theory that there is a unique lan- (Phonetica 37. 22–3, 1980) guage ‘organ’ distinct from other cognitive systems It was just this ability to see beyond the constraints and abilities—was central to much of Fromkin’s work, that some would impose upon the discipline that and her support for the theory of a genetically deter- enabled Fromkin to expand her own work, and that of mined language faculty placed her within the general the field, in ways that led eventually to the cross-disci- scope of Chomskyan generative linguistics. However, plinary specialization of neurolinguistics. For her work was often on the bridge between linguistics Fromkin, the research began with slips of the tongue. and other disciplines (aphasia studies, neurobiology, Her interest lay not in merely describing, categorizing, psychology), and so, except for debates over modular- or even explaining slips and other speech errors, but ity, she was not a core figure in controversies of theo- rather in determining what such errors might imply retical linguistics. about issues of language structure and language pro- Partly because of this, and partly because of her cessing that were relevant to linguistic theory. In her strong administrative skills, she played a significant landmark paper on speech errors published in 1971, leadership role in national and international linguistic she demonstrated that slips of the tongue involve dis- organizations, serving as president of the Linguistic crete units of a language (for example, sounds, sylla- Society of America (1985), as secretary and then chair bles, words) in accord with the rules of that language. of Linguistics Section Z of the American Association That is, speech errors are not random. They are struc- for the Advancement of Science (1993–1996, tured in the same way that language is structured, and 1997–1998), and as an executive board member of the studying such errors may provide insights into the way Permanent International Committee of Linguists. Her in which language is organized in the brain. accomplishments were recognized by her election to In her studies of lexical substitutions, for example the National Academy of Sciences. (saying ‘Take the milk out of the oven’ when a speak- In both her academic and her administrative work, er intended ‘Take the milk out of the refrigerator’), Fromkin was always concerned with building bridges Fromkin argued that such substitutions suggest that in between linguistics and other disciplines: ‘You can’t do our mental dictionary, words are specified with their work on language without linguistics’ (quoted by semantic features. Other errors point to specification of Cheng and Sybesma, ‘Interview with Victoria phonetic, phonological, morphological, and syntactic Fromkin’, Glot International 2:5.23 (1996)). She features, as well as orthographic (spelling) properties. encouraged linguists to contribute their understanding The vocabulary in our brain is not a mere alphabetical of language to other fields, and she searched those listing, but a complex set of independent and intercon- fields for evidence that might support claims that nected subsystems of linguistic information. linguists made about language and its organization in Fromkin’s work on speech errors and their implica- the brain. tions for linguistic theory was highly regarded and Fromkin also took seriously the responsibility of very influential. She edited two books of contributions bringing linguistics and its findings to the public and by scholars from around the world, the first in to the undergraduate population of colleges and uni- 1973 (Speech errors as linguistic evidence), the sec- versities. At UCLA, she taught the Introduction to ond the result of a symposium she led at the 12th Language class for more than 30 years, and it was International Congress of Linguists in Vienna, Austria from that class that she developed, with co-author in 1977 (Errors in linguistic performance: slips of the Robert Rodman, her book Introduction to language, tongue, ear, pen, and hand, 1980). In 1988, she con- without doubt the most widely read of all introductory tributed the state-of-the-art article on speech errors to linguistics textbooks in the second half of the twenti- Linguistics: the Cambridge survey. As her friends and eth century. colleagues were well aware, she always carried a note- book in which to record any speech errors she heard. Biography The compilation eventually reached more than 12,000 examples. Victoria Fromkin was born in Passaic, New Jersey, on In her later work, Fromkin investigated language May 16, 1923. She did her B.A. in economics from the use and loss in people with aphasia, dyslexia, and University of California, Berkeley (1944). She married Parkinson’s disease, and she promoted the study of in 1948 and had one child. She returned to school in sign languages and of language acquisition in children 1961, University of California, Los Angeles; and who had been denied language input by illness or neg- received her M. A. (1963) and Ph.D. with dissertation lect. In these studies, she explored evidence revealing on phonetics guided by Peter Ladefoged (1965). She 361 FROMKIN, VICTORIA ALEXANDRA was Assistant Professor, English, California State Fodor, Jerry A. 1983. The modularity of mind. Cambridge, MA: University, Los Angeles, spring 1965; joined the MIT Press. Fromkin, Victoria. 1971. The non-anomalous nature of anom- University of California, Los Angeles faculty in 1966 alous utterances. Language 47. 27–62 in Speech; in 1967 transferred to Linguistics and ——— (ed.) 1973. Speech errors as linguistic evidence. The remained affiliated there throughout her career; in 1966 Hague: Mouton. she was Assistant Professor, 1969 Associate Professor, ———. (ed.) 1980. Errors in linguistic performance: slips of 1972 Professor, Chair of Department of Linguistics the tongue, ear, pen, and hand. New York: Academic Press. ———. 1980. Discussion [following ‘The goal of phonetics, its 1970–1971 and 1973–1977, Dean of the Graduate unification and application’, by Björn Lindblom]. Phonetica Division 1979–1989, Vice-Chancellor of Graduate 37. Programs 1980–1989, and she retired in 1991. She ———. 1987. The lexicon: evidence from acquired dyslexia. served on the faculty of the Linguistic Institutes of the Language 63. 1–22 Linguistic Society of America in 1966, 1976, 1977, and ———. 1988. Grammatical aspects of speech errors. Linguistics: the Cambridge survey, Vol. 2, ed. by Frederick 1983; taught in the summer program in linguistics at the J. Newmeyer 115–27. Cambridge: Cambridge University University of California, Santa Cruz 1971, 1972, 1973; Press. and was visiting professor at University of Stockholm ———. 1988. Sign languages—evidence for language univer- 1977, Cambridge University 1977, Wolfson College at sals and the linguistic capacity of the human brain. Sign Oxford University 1983 and 1987. She was President of Language Studies 59. 115–27. ———. 1991. Brain and language: redefining the goals and the Linguistic Society of America 1985; President of methodology of linguistics. The Chomskyan turn, ed. by Association of Graduate Schools 1988; Chair, Board Asa Kasher. Oxford, UK and Cambridge, MA: Basil of Governors of Academy of Aphasia 1991–1993; Blackwell. 78–103. secretary 1993–1996 and then chair 1997–1998 ———. 1993. Speech production. Psycholinguistics, ed. by of Linguistics Section Z of the American Association Jean Berko Gleason and Nan Bernstein Ratner. Fort Worth, TX: Harcourt Brace Jovanovich. for the Advancement of Science; fellow of the ———. 1998. Introduction to language, co-authored with Acoustical Society of America, the New York Academy Robert Rodman, 6th edition. Fort Worth, Tx: Harcourt Brace of Sciences, the American Association for the (previous editions 1974, 1978, 1983, 1987, 1992). Advancement of Science, and the American ———. (ed.) 2000. Linguistics: an introduction to linguistic Psychological Society; executive board member of theory. Oxford, UK and Malden, MA: Blackwell Publishers. Hyman, Larry M., and Charles N. Li (eds.) 1988. Language, the Permanent International Committee of Linguists; speech and mind: studies in honour of Victoria A. Fromkin. elected member National Academy of Sciences, London and New York: Routledge. American Academy of Arts and Sciences, Aphasia Jackendoff, Ray. 1994. Patterns in the mind: language and Research Group of the World Federation of Neurology. human nature. New York: Basic Books. She died in Los Angeles on January 19, 2000. Pinker, Steven. 1994. The language instinct: how the mind cre- ates language. New York: William Morrow. ———. 1999. Words and rules: the ingredients of language. References New York: Basic Books. Zwicky, Arnold M. 1980. Mistakes. Reynoldsburg, OH: Cheng, Lisa, and Rint Sybesma. 1996. Interview with Victoria Advocate Publishing. Fromkin. Glot International 2 (May). 1–22 Emmorey, Karen D., and Victoria A. Fromkin. 1988. The men- JULIA S. FALK tal lexicon. Linguistics: the Cambridge survey, Vol. 3, ed. by Frederick J. Newmeyer, 137–8. Cambridge: Cambridge See also Bar-Hillel, Yehoshua; Chomsky, Noam; University Press. Neurolinguistics Function Words Words are divided into two basic classes: (1) lexical, freely add new words. Function words, on the other or open class words and (2) function, or closed class hand, as seen in these English examples, include deter- words. Lexical words refer to nouns (e.g. dog, house), miners, such as the and a(n); auxiliary verbs (or sim- verbs (e.g. to go, to see), and adjectives (e.g. green, ply ‘auxiliaries’), such as might, have, and be; nice), and supply the bulk of the meaning in a sen- conjunctions, such as and, that, and whether; and tence. This class is called ‘open’ because speakers can degree adverbs, such as very and too. These words are 362  FUNCTION WORDS called ‘functional’ or ‘grammatical’ because they carry tain a determiner. ‘Functional projections’ were thus little meaning (have no synonyms) and typically ‘help’ assigned a structure and importance similar to ‘lexical another word. For instance, determiners add grammat- phrases’. ical information about specificity and definiteness (the Research in the late 1980s and 1990s revolved dog vs. a dog), but do not essentially alter the meaning around the question of exactly which functional pro- of the phrase. They are also called ‘closed class words’, jections a sentence may contain. Each function word since speakers do not easily add new words to the set. that had a grammatical function was soon regarded as Lexical words typically carry intonational emphasis the main structural building block of a sentence. The or stress, while function words are generally increasing importance of function words in linguistic unstressed. Therefore, function words are prone to theories went hand in hand with an increasingly contraction—for example, the auxiliary have in I’ve abstract description of sentence structure. This shift seen it. provided many empirical and theoretical advantages. The distinction between function and lexical words First, sentence structure could now be divided into has been very fruitful for linguistic description. So- three function domains: (1) a lexical domain around called analytical languages, such as Chinese, use a the verb, which establishes semantic relations between large number of function words. In contrast, function the main sentence elements; (2) a grammatical domain words are typically lacking in the speech of young chil- around the auxiliary, which establishes grammatical dren, certain kinds of aphasia, and telegraphic speech. relations such as agreement (the auxiliary agrees in It is also widely confirmed that languages rarely bor- number and person with the subject: I am/She is/They row function words from other languages or invent new are leaving); and (3) a discourse domain around the ones (hence their status as a ‘closed’ class). Most of the complementizer that, which links an embedded clause recent innovations in the English vocabulary, such as to a main clause (I know that this is true or I wonder pizza, angst, fax, E-mail, phat, AIDS, website, browser, whether this is true). screenager, to surf, Nethead, and technobabble, are Second, differences between languages could be lexical rather than functional in nature. explained by how the function words, and the domains Function words primarily add grammatical infor- they define, were used. For instance, so-called verb- mation, which means that they are defined above all by second languages, such as German, Middle English, their syntactic behavior. Most traditional grammarians Dutch, and Swedish, move the verb to the comple- assumed (and their descendants continue to assume) mentizer domain, whereas languages such as English that the structure of sentences and phrases is influ- refrain from doing so. For example, the word order in enced mainly by lexical words. Function words were the English sentence, ‘Yesterday the rabbit saw the regarded as mere additions to lexical phrases. Thus, fox’ would be, in German, ‘Yesterday saw the rabbit the sentence, ‘The rabbit will see the fox’ was ana- the fox’. Differences even between unrelated lan- lyzed as a noun phrase the rabbit, followed by a verb guages were thus reduced to very basic principles. phrase will see the fox. The determiner the was thus Function words and lexical words are not sharply simply an addition to the noun phrase, and the auxil- distinct categories, but rather form a continuum. iary verb will was added to the verb phrase. Certain classes of words can thus share features with A shift in this thinking came in the 1980s within the both prototypical lexical words and prototypical func- framework of generative grammar. From then on, aux- tion words. The English preposition is a case in point: iliaries were seen as providing their own, independent, some prepositions have lexical meaning, such as loca- contribution to sentence structure. Nevertheless, func- tion (behind) and direction (toward), while others have tion words still did not play a part in determining the little meaning (of or to). Many prepositions are used to category in which a phrase was placed—for example, introduce sentences (after, for, like) and are therefore a phrase such as the rabbit continued to be regarded as similar to prototypical function words, namely, com- a noun phrase that happened to contain a determiner. plementizers. This view changed radically by the mid-1980s, Grammatical meaning can be expressed in different when function words were increasingly treated as the ways. English uses independent auxiliaries to express key factor when deciding which sentence elements present or past tense (I am thinking vs. I was think- should be placed in which category. To use the techni- ing), but also inflects the verb for the same purpose cal terminology, function words were ‘projecting to a (I think vs. I thought). Languages exhibit great varia- phrase’ or ‘heading a phrase’. Determiners, for exam- tion along these lines: some languages express all ple, now came to be regarded as the head of determin- grammatical meaning via independent function words, er phrases—that is, the rabbit was now interpreted as and such languages are called ‘analytic’. In contrast, a determiner phrase the . . . containing the noun phrase so-called ‘synthetic’ languages use inflection and rabbit, instead of a noun phrase that happens to con- other markings on lexical words throughout. 363 FUNCTION WORDS This distinction between analytic and synthetic lan- References guages also represents a continuum, and languages Abney, Steven. 1987. The English noun phrase in its sentential change in this respect over time. Old English made aspect. Ph.D. Dissertation. Massachusetts Institute of extensive use of grammatical markings on lexical Technology. words. Modern English has lost much of this capabil- Chomsky, Noam. 1995. The minimalist program. Cambridge, ity and instead uses auxiliaries to fill the gap. In fact, MA: MIT Press. Fukui, Naoki. 1986. A theory of category projection and its the auxiliary verb will used to be a lexical verb in Old applications. Ph.D. Dissertation. Massachusetts Institute of English, but its meaning (‘to want’) changed when it Technology. was recruited to express future tense. In modern theo- Gelderen, Elly van. 1993. The rise of functional categories. retical approaches, such as the Minimalist Program, Amsterdam and Philadelphia: Benjamins. which tend to focus more on underlying differences Lightfoot, David. 1979. Principles of diachronic syntax. Cambridge and New York: Cambridge University Press. than on surface variation, the distinction between ana- Radford, Andrew. 1990. Syntactic theory and the acquisition of lytic and synthetic languages becomes negligible. English syntax. Oxford and Cambridge, MA: Blackwell. Function words have little lexical meaning and usu- ———. 1997. Syntactic theory and the structure of English. ally carry no stress. In traditional grammars, they do Cambridge and New York: Cambridge University Press. not carry the function of governing an independent Traugott, Elizabeth, and Berndt Heine (eds.) 1991. Approaches to grammaticalization, 2 vols. Amsterdam and Philadelphia: projection or phrase, whereas in some modern Benjamin. approaches they do. Function words are quite similar ELLY VAN GELDEREN (and are historically related) to grammatical markers on lexical words. See also Auxiliaries; Grammar, Traditional Functional Approaches Functional linguistics appeared as a reaction to formal Today, there are two main approaches that claim to approaches to grammar, especially generative and be ‘functional’. These are M.A.K. Halliday’s systemic transformational approaches. One of its basic assump- functional linguistics, and Simon Dik’s functional tions is that language is a symbolic system with a cer- grammar, both discussed below. tain purpose or purposes, mainly communication, Michael Halliday’s systemic grammar follows the although there are other possibilities too, such as the tradition of the London school and, more concretely, use of language as an instrument of thought. the work of John Rupert Firth, who developed his own Apart from structures and form, any linguistic sys- theory studying language as part of a social system. tem also has functions. Functional approaches to lan- According to Firth, language was used with a specific guage assume that there is a correspondence between purpose in situational contexts, which, in turn, reflect- form and function, and this correspondence is always ed cultural contexts. Other minor influences on motivated. However, a very subtle theoretical matter is Halliday’s work are the linguistic theories of the what kind of function we are referring to. Apart from Prague school around Nikolay Trubetskoy, Louis the most general functions of communication and Hjelmslev’s glossematics, and the ideas of the organization of thought, it is possible to refer to func- American linguist Benjamin Lee Whorf. tions at more atomic levels (i.e. functions of linguistic In 1961, Halliday, one of Firth’s disciples, devel- elements or functions of linguistic constituents, oped from Firth’s ideas a theory that dealt with differ- considered in different levels of linguistic analysis ent scales and categories. He used the categories of such as phonetics, morphology, semantics, syntax, structure, system, unit, and class, as well as three etc.). At the most general level, however, there are scales (rank, delicacy, and exponency) that connected several typologies with different functions that try to the categories with one another and with the data. This provide different functional alternatives, like the ones grammar received criticism, especially concerning the proposed by authors such as Karl Bühler (cognitive, categories of class and structure, the scale of rank, and expressive and conative/instrumental), M.A.K. the taxonomic nature of the theory, as mentioned by Halliday (ideational, interpersonal and textual), or Christopher Butler (1985:29–38; 1995:529). Later, Roman Jakobson. this grammar evolved toward what was first called 364  FUNCTIONAL APPROACHES systemic grammar (Halliday 1967/1968), and some model receives ideas already put forward by William years later, systemic functional grammar (Halliday Foley and Robert Van Valin in 1980 in their Role and 1973, 1985), which is the current form of the theory. reference grammar, but their proposal is original and Systemic grammar is based on a distinction between different. three main meta-functions of language: (1) the ideation- Dik (1989) proposes adequacy standards for his al function, which deals with the expression of content grammar, such as psychological adequacy, typological and with the experience of the speaker within the real adequacy, and pragmatic adequacy, without which it world; here, it is possible to distinguish two subfunc- would not be possible to account for language as an tions: experiential and logical; (2) the interpersonal instrument of social interaction, and therefore the gram- function, which is used for establishing and maintaining mar would no longer be a functional grammar. However, social relations; and (3) the textual function, which the proposed model presents a high degree of formal- deals with the creation of texts and the relations that are ization that is characterized by having an internal archi- established within them. These meta-functions occur tecture with different levels that appear included one simultaneously in language. This simultaneity can be within another. The resulting embedding could be repre- applied to two axes for the organization of the theory: sented as: [Level 4 [Level 3 [Level 2 [Level 1 Nuclear Predicate]]]]. the systemic (paradigmatic) and the structural (syntag- Any clause has this basic configuration, to which the matic) axes. Each of these meta-functions involves dif- rules of expression can be applied in order to produce ferent systems that can be organized according to the the concrete form of the sentence in English. In a poste- different units of the scale of rank (clause, phrase, rior elaboration of the model by Kees Hengeveld (2004), group, word, informative unit). In fact, these three the mentioned scheme constitutes the representational descriptive dimensions of the theory eventually become level, but this level is produced from information in the even more complicated, because Halliday also talks of a interpersonal level, and it is determined as much by the stratification in phonology, lexicogrammar, semantics, cognitive context as by a cognitive component. and context. The use of many axes for the organization In this new formulation of Dik’s functional gram- of this theory makes systemic grammar a complicated mar, the roles of the interpersonal factor and the com- one that tries to gather the complexity of language into municative context appear more clearly, which many dimensions. In this respect, it does not have characterize this model as even more functional. among its priorities the criteria of parsimony and ele- gance that appear in other theories. Additionally, sys- References temic grammar is characterized by its being a theory Butler, Christopher S. 1985. Systemic linguistics: theory and that is sociologically oriented: that is, it classifies dif- applications. London: Batsford. ferent contextual parameters of a social nature into ———. 1995. Systemic functional grammar. In Verschueren, dialectal and diatypical (based on different registers), Östman, and Blommaert. and the latter into: field, which is the type of social Dik, Simon C. 1978. Functional grammar. Amsterdam: North- activity in which language is inserted; tenor, which con- Holland. ———. 1989. The theory of functional grammar, part I: the cerns the role relationships of roles among those who structure of the clause. Dordrecht: Foris. interact; and mode, which refers to the medium of com- Halliday, M.A.K. 1961. Categories of the theory of grammar. munication. These types of parameters are related to the Word 17. 241–92. meta-functions of language: field to the ideational func- ———. 1967/1968. Notes on transitivity and theme in English. tion, tenor to the interpersonal function, and mode to the Parts 1, 2 and 3. Journal of Linguistics 3(1). 27–81, 3(2). 199–244, 4(2). 179–215. textual function. Halliday’s theory encompasses all ———. 1973. Explorations in the functions of language. usage domains of language, and in that respect it can be London: Edward Arnold. rightly considered a functional theory. ———. 1985. An introduction to functional grammar. London: Simon Dik’s Functional grammar presents a differ- Edward Arnold, 2nd edition, 1994. ent approach. It claims to be a functional theory, since Hengeveld, Kees. 2004. The architecture of a functional dis- course grammar. A new architecture for functional grammar, language is conceived mainly as an instrument of ed. by J.L. Mackenzie and Gómez González M.A. Berlin/ social interaction, and this characteristic is incorporat- New York: Mouton de Gruyter. ed into the model. Despite this claim, the results, both Nuyts, Jan. 1995. Functionalism vs. formalism. ed. by In in this first version of the theory (1978) and in the Verschueren, Ostman and Blommaert. second (1989), are not as satisfactory as Halliday’s Verschueren, J., J.-O. Ostman, and J. Blommaert (eds.) 1995. Handbook of pragmatics. Amsterdam and Philadelphia: proposal. Dik’s grammar was influenced by the work John Benjamins. of many previous authors, like Joseph Greenberg, CARLOS INCHAURRALDE James Fillmore, David Perlmutter and Paul Postal, Emmon Bach, H. Paul Grice, Herbert Clark, and See also Firth, John Rupert; Halliday, Michael Susan Haviland, among others. The architecture of its Alexander Kirkwood 365  G Gender and Language One meaning of the term gender, in the phrase ‘Gender be ‘generic’ or ‘sex-indefinite’, i.e. to have the capaci- and language’, is gender as a characteristic of lan- ty to include both males and females. Examples guage as an abstract system. In this sense, gender include man (and its compounds) and he (as well as refers to a grammatical category. A second meaning of some animal names such as dog and lion). Thus, in gender refers broadly to the ‘human’ and necessarily principle, it is grammatically correct to say to a class social concepts of femininity and masculinity. of male and female students: ‘Everyone will get his homework back tomorrow.’ But since his here is intended to include females, this illustrates that gender Gender as Grammatical Category in English is in fact in part grammatical. Traditionally, languages have been described as hav- The genericity of those masculine forms in English ing either ‘natural’ or ‘grammatical’ gender. A lan- that can in principle additionally refer to ‘humans’ has guage with natural gender requires that the gender of however been seriously challenged—one reason being an animate noun or pronoun corresponds to the bio- that phrases like ‘Man breastfeeds his young’ sound logical sex of the person or animal to which that noun odd; another that what may be intended as generic refers. In this way, woman and girl in present-day may not be so interpreted (by either sex), and accord- English are feminine nouns and she and her are femi- ingly that women and girls may be effectively exclud- nine pronouns. Similarly, bull is a masculine noun and ed or rendered relatively invisible by the use of these vixen is a feminine noun. Nouns like computer and ‘generics’. The ‘generics’ man and he have thus been marmalade are neither feminine nor masculine, but seen by feminists as one form of ‘sexist language’ rather ‘neuter’. Despite a very few odd but well-known (other forms being those that define, stereotype, trivi- exceptions (like a ship being sometimes referred to as alize, and/or degrade women, e.g. Miss/Mrs (vis à vis she), the English language is usually seen as having Mr), usherette, air hostess, ‘she’s a blonde’, dumb natural gender. blonde). (see Lakoff (1975) for an influential— Grammatical gender, in contrast, is ‘formal’. In lan- although, in retrospect, problematic—pioneer account guages with grammatical gender, all nouns have a gen- of such language.) der, although whether this is masculine, feminine, or The masculine ‘generics’ as well as other ‘sexist neuter is unlikely to be evident from the noun itself. In language’ items now have a somewhat old-fashioned French, for example, chaise (chair) is feminine and ring to them, and are often substituted for by alterna- pain (bread) is masculine. The determiner (in these tives (although not replacements) such as people, he or cases, the definite article) indicates the gender (la she, s/he, ‘singular they’ (especially in spoken English, chaise, le pain). e.g. ‘Everyone will get their homework back tomor- Some masculine nouns and pronouns in English row’), flight attendant, and Ms. These alternatives are and other languages supposedly have the potential to now included in grammars and dictionaries, and 367 GENDER AND LANGUAGE ‘inclusive language’ is required by many journals and the overriding research question was whether and how institutional Codes of Practice. Other nonsexist alter- women and men used language differently. natives have been adopted for languages other than This project was characterized by two different English (Pauwels 1998). approaches. The first was the ‘dominance’ paradigm, The vociferous and in many ways effective cam- i.e. a concern with mixed-sex talk and whether men paign against sexist language (see Lakoff (1975) for an actually dominated women linguistically, making their influential—although, in retrospect, problematic— use of language a form of ‘doing power’ (Kramarae pioneer account of such language) was loosely based 1981; Thorne et al. 1983). A British pioneer was Dale on a ‘structuralist’ view of language—one form, one Spender, who in the 1980s popularized what were then meaning. Several limitations of this have now been new and important findings about male dominance in identified, in particular that (a) word meanings classrooms, in terms of boys being the recipients of change, (b) the meaning of a word will vary with con- most teacher talk, and the producers of most class- text, (c) people will interpret a given word in a whole room talk (Spender 1982). range of ways, and, perhaps most importantly, (d) sex- The other, slightly later approach was the more lib- ist discourse can very easily occur without a single eral ‘(cultural) difference’ paradigm, exemplified in its ‘sexist language item’ (for example, people can be most extreme form by Deborah Tannen (1990), in a used in a sexist way if it actually refers only to males; less extreme form by Jennifer Coates (1998). Coates’ a text can still be degrading to women without using work concerns particularly single-sex (women’s) talk. any ‘sexist language’). Partly because of the evident In this approach, in which dominance was much less limitations of ‘nonsexist language’, feminist critique of an issue, men and women were seen as speaking has shifted to discourse, in the sense of ‘ways of struc- differently as a result of having grown up in different turing knowledge and social practice’ (Fairclough linguistic subcultures. 1992), or of ‘ways of seeing the world’. Gendered dis- However, these apparently different approaches courses represent men and women in particular ways, also had a great deal in common. Focusing variously e.g. in contemporary childcare texts, ‘Father as baby on women’s disadvantages (‘dominance’) and entertainer’ (see Sunderland 2004). Other contempo- women’s strengths (‘cultural difference’) in their con- rary sexist discourses can be seen as (a) ‘universal het- cerns with identification of gender differences in talk, erosexuality’, i.e. the apparent assumption that every they played down similarities, and underemphasized adult is either with a partner of the opposite sex or differences ‘within’ men as a group and women as a seeking one, (b) ‘vive la diffrénce’, i.e. that apparent- group. With exceptions, they both tended to generalize ly essential differences between women and men are (women do this, men do that), to underplay the impor- enjoyed by both and should/need not be problema- tance of context and local meanings, and to stress the tized, (c) ‘the battle of the sexes’, i.e. that women’s importance of gender over other identities (see below). gains represent losses for men, (d) ‘how to get your More profoundly, they implicitly shared a model of man and keep him’ (Hollway 1984), a discourse sine gender as broadly ‘fixed’, and somehow shaping or qua non of many women’s magazines, and (e) ‘male even determining talk, and accordingly linguistic gen- sex drive’ (Hollway 1984) (or ‘men can’t help their der differences—rather than of gender as itself being sexually aggressive behavior’). Masculinity and femi- shaped by language use. ninity, as well as individual women and men, and A combination of impatience with difference/domi- indeed gender relations, can be represented in particu- nance, and developments in cultural studies and femi- lar ways in discourse in a whole range of written and nist theory, led feminist linguists to turn the previous spoken texts (Mills 1995). But representation can only conception of the language/gender relationship on its ever be part of the story—it says nothing about how a head, i.e. from ‘How does gender shape language?’ (or text will be responded to by a reader. ‘How is language a reflection of gender’?) to ‘How does language (or discourse) shape gender?’ (especially over the last 20 years) Accordingly, one of the major Gender as Social Concept ways in which the language and gender field (particu- The second, social use of the term gender in the con- larly that with a sociolinguistic orientation) has changed text of gender and language study has witnessed con- (especially over the last 10 or 15 years) is the growing siderable changes in meaning. Feminist linguistics in recognition of the problematic nature of a model of gen- the 1970s largely bought into the traditional, sociolin- der as fixed (established at the age of 16 or 18 or what- guistic, variationist paradigm, in which gender was ever), and the newer understanding of gender as broadly mapped onto the biological category of sex. changing all the time. Although the feminist view of gender was always that it Corresponding to this is a recognition that if gender was primarily social or cultural rather than biological, is not fixed, the binary, ‘gender differences’ approach 368  GENDER AND LANGUAGE to language use (which underpinned both the ‘domi- which gay men speak when they are not interested in nance’ and the ‘(cultural) difference’ paradigms) is signaling their sexual identity. also problematic (see Cameron (1992) for a useful cri- Thus, it is now possible to see the previous ‘two tique). Gender and language study now acknowledges prongs’ of gender and language research as having the further complexity of gender since it can be seen as dovetailed into discourse, and discourse as social coexisting with considerations of, inter alia, class, eth- practice as a shaper of identity. The relationship nicity, and sexuality. between language/discourse and gender is now One response to the question of how language explored across a range of genres and settings (see e.g. shapes gender is ‘through discourse, which is both Sunderland 2000, 2004). However, one particular shaped by and shapes identity’. To see gender (mas- recent and influential context-related concept for fem- culinity, femininity) as identity excludes neither shap- inist linguistics is that of ‘Community of Practice’ ing by social forces nor an individual’s own agency (Holmes and Meyerhoff 1999). A ‘Community of (Bucholtz et al. 1999; Norton 2000). Neither does it Practice’ can be defined both by its membership and entail fixity. Johnson for example sees masculinity and its practice. The ‘practice’, both nonlinguistic and lin- femininity as ‘on-going social processes dependent guistic, will include heterogeneity as well as homo- upon systematic restatement’ (1997:22). geneity in the ways things are done, and the way they If identity is further conceptualized as multiple, are talked about. It can be, and normally is, both gen- gender can be seen as one of several identities (e.g. dered and gendering. masculinity/femininity, ethnicity, sexual identity)— To conclude, gender and language study is now although one identity may be foregrounded at any characterized by considerations of (i) gender as identi- given time. Hence, while, say, ethnic identity might be ty, and this identity as both multiple and fluid experienced as most salient for some individuals in (‘becoming’ rather than ‘being’), (ii) representation, in some contexts (for themselves and/or for others), this a range of written and spoken texts and text types, (iii) cannot be seen as separate from gender. Each identity the ongoing social and linguistic construction of gen- will mediate and be mediated by the others. der, (iv) individual agency, (v) the possibility of seeing Masculinity, already referred to in passing, is one gender as ‘performance’, (vi) discursive practices and recent concern of gender and language study (see discourse(s) as social practice (Fairclough 1992), and Johnson and Meinhof 1997; Coates 2003). Masculinity (vii) gendering through a ‘Community of Practice’ can be manifested and constructed linguistically in a (see Bergvall et al. 1995; Hall and Bucholtz 1995; range of ways, as can femininity. Many of these are Bucholtz et al. 1999). likely to be forms of what has been referred to as hege- monic (crucially, heterosexual) masculinity. The study References of gender identities, both masculinities and feminini- Bergvall, Victoria, Janet Bing, and Alice Freed (eds.) 1996. ties, from gay as well as the previous heterosexual per- Rethinking language and gender research. London: spectives has thus paved the way for ‘Queer Theory’, Longman. which rather than advocating inclusion and equal Bucholtz, Mary, A.C. Liang, and Laurel Sutton (eds.) 1999. opportunities, entails problematizing all sexual identi- Reinventing identities—the gendered self in discourse. New York: Oxford University Press. ties, and thus advocates continual enquiry as regards Butler, Judith. 1999. Gender trouble: feminism and the subver- heterosexual and other ‘hegemonic’ identities as well sion of identity, 2nd edition. New York: Routledge. as more marginalized ones (Butler 1999; Nelson Cameron, Deborah. 1992. Feminism and linguistic theory, 2nd 1999). edition. London: Macmillan. Gender and language study does not however con- Coates, Jennifer. 1998. Women talk. Oxford: Blackwell. Coates, Jennifer. 2003. Men talk: Stories in the making of mas- ceptualize gender only as identity. It can also be seen culinities. Oxford: Blackwell. as performance—of doing rather than of being (Butler Fairclough, Norman. 1992. Discourse and social change. 1999). Butler does not deny identity—rather, she sees London: Polity Press. gender identity as coming from, rather than preexist- Hall, Kira, and Mary Bucholz (eds.) 1995. Gender articulated: ing, performance. For Butler, people perform an iden- language and the socially constructed self. New York: Routledge. tity—linguistically, in discourse, and in other ways. To Hollway, Wendy. 1984. Gender differences and the production illustrate this, looking at the talk of gay males, it is of the subject. Changing the subject, ed. by J. Henriques et possible to distinguish between ‘Gayspeak’, the iden- al. London: Methuen. tifiable talk of gay men (a form of talk whose exis- Holmes, Janet, and Miriam Meyerhoff (eds.) 1999. The com- tence has been challenged)—and the idea of gay men munity of practice: Theories and methodologies in language and gender research. Language in Society 28. 173–83. deliberately speaking in a stereotypically camp way in Johnson, Sally, and Ulrike Meinhof (eds.) 1997. Language and order to signal their sexual identity. This ‘perform- masculinity. Oxford: Blackwell (includes ‘Introduction’ by ance’ may however bear no relation to the way in Sally Johnson). 369 GENDER AND LANGUAGE Kramarae, Cheris. 1981. Women and men speaking. Rowley, Sunderland, Jane. 2000. Review article: issues of language and MA: Newbury House. gender in second and foreign language education. Language Lakoff, Robin. 1975. Language and woman’s place. New York: Teaching 33(4). 203–23. Harper and Row. Sunderland, Jane. 2004. Gendered discourses. London: Mills, Sara. 1995. Feminist stylistics. London: Routledge. Palgrave Macmillan. Nelson, Cynthia. 1999. Sexual identities in ESL: Queer theory Talbot, Mary. 1997. Language and gender: an introduction. and classroom enquiry. TESOL Quarterly 33(3). 371–91. Cambridge: Polity Press. Norton, Bonny. 2000. Identity and language learning: gender, Tannen, Deborah. 1990. You just don’t understand. New York: ethnicity and educational change. Harlow, Essex: William Morrow. Longman/Pearson Education. Thorne, Barrie, Cheris Kramarae and Nancy Henley (eds.) 1983. Pauwels, Anne. 1998. Women changing language. London: Language, gender and society. Rowley, MA: Newbury House. Longman. JANE SUNDERLAND Spender, Dale. 1982. Invisible women: the schooling scandal. London: The Women’s Press. See also Gender: Class Marking Gender: Class Marking Nouns can be marked according to number, but there languages, gender is marked by special endings and is also the possibility of marking nouns according to the required agreement with certain forms of the arti- their belonging to certain classes. These classes were cle and demonstratives. Gender is also present as a originally based on semantic criteria, which may still grammatical distinction that requires different forms be present in many nouns, but it is customary to find for personal pronouns. In English, there are no dis- nouns that do not match the criteria in their own class. tinctive endings for nouns according to gender In English and Indo-European languages in general, (although nouns sometimes have special forms that we normally refer to class marking as ‘gender’, and reflect gender, e.g. heroine is a feminine noun and the original semantic basis for the marking is sex. hero is a masculine noun). In addition to this, there are Since sex applies to most animate beings, it is easy and no gender endings for articles and demonstratives. The straightforward to distinguish between male and only word class that should be clearly marked at all female, which are marked grammatically as masculine times according to gender is the class of personal pro- and feminine, respectively. Inanimate beings, on the nouns, following these criteria: male humans (or ani- other hand, cannot be either male or female, and that mate beings in some situations) are masculine (John is why in some Indo-European languages there is still came, but he left early), female humans are feminine a third class, the neuter, which marks grammatically (Mary didn’t come. She stayed at home), animals and those nouns denoting entities that have no sex. The plants in general are neuter (I saw a cat yesterday. It picture becomes complicated by the fact that many was black), and inanimate beings are neuter (I like this nouns are classified within the different gender class- bicycle because it does not weigh much). There are es regardless of semantic considerations. Some exam- some special cases that do not follow these rules (e.g. ples follow. Isn’t she a beautiful BMW? refers to a car, which is an In German, there are three genders, which reflect inanimate being). the semantic distinction in terms of sex in nouns like Class marking can be more complex than what gen- Mann, ‘man’, which is masculine, or Frau, ‘woman’, der alone suggests. In Bemba, a Bantu language, we which is feminine. However, there are also cases in find 16 different classes, which are marked morpho- which there is no correlation with the semantic notion logically in combination with number, corresponding of sex, as can be seen in nouns like Mädchen, ‘girl’, to the original semantic values of humans, plants, which is neuter. In Spanish, there is no neuter, but fruits, mass/liquid, small objects or inanimates, ani- there is a neat distinction between masculine and fem- mates, inanimate elongated objects, and paired body inine, which closely follows the sex notion in nouns parts. The marking is carried out by means of prefixes like hombre, ‘man’, and mujer, ‘woman’, but which is (e.g. in umu-ana, ‘child’, the morpheme marking class- arbitrarily assigned in many other cases. In these two inclusion is umu). Swahili, another Bantu language, 370  GENDER: CLASS MARKING has 11 classes, with the following prefixes for nouns, Russian, it is even more difficult to establish corre- adjectives (which agree with the noun they accompa- spondences between suffixes and gender, since the ny), and pronouns: endings also reflect grammatical case. The forms that can be used in German for the nouns Tisch ‘table’, Class Nominal Adjectival Pronominal Prefix Prefix Prefix Körper ‘body’, Mensch ‘man’ (masculine), and Jahr ‘year’, System ‘system’ (neuter) in the singular are the 1 m- m- a-/m-; yu-; following: w-; ye- 2 wa- wa- wa- 3 m- m- u- Masculine Neuter 4 mi- mi- i- 5 zero or ji- zero or ji- li- Nominative -/-/- (Tisch, -/- (Jahr, 6 ma- ma- ya- Körper, System) 7 ki- ki- ki- Mensch) 8 vi- vi- vi- Accusative -/-/-en (Tisch, -/- (Jahr, 9 zero or n- zero or n- i- Körper, System) 10 zero or n- zero or n- zi- Menschen) 11 u- m- u- Genitive -es/-s/-en (Tisches, -es/-s (Jahres, Körpers, Systems) Menschen) Swahili’s classes can be grouped according to Dative -/-/-en (Tisch, -e/- (Jahre, semantic criteria (similar to those of Bemba), but there Körper, System) can be many inconsistencies in the groupings. A clas- Menschen) sification proposal that takes into account all the nouns that do not fit into the system divides the noun classes The feminine and plural forms are simpler than into two subsets: a ‘derived’ set of classes, in which these, since they are in most cases invariable; but it there is a fit between meaning and form (this is also may be impossible to find out gender simply by look- called ‘fixed’ gender), and an ‘inherent’ set of classes, ing at the ending (e.g. Jahre is the nominative plural with arbitrary membership (‘free’ gender). There is and the dative singular of Jahr, ‘year’, which is neuter, overlap between these two sets formally, but the dis- whereas Frage, ‘question’, is the nominative singular tinction helps in making the system consistent, of a feminine word; all these forms end in -e). because we can tell which nouns are classified accord- Articles and adjectives normally agree in gender ing to semantic criteria and which are not. with nouns in Indo-European languages (English is As for the linguistic form used, Swahili and Bemba one of the exceptions), which means that they have mark class membership by means of prefixes; but, as full-ending paradigms for masculine, feminine, and, in we have seen, there are other means. In Indo- some cases, neuter. Pronouns also use different ending European languages, such as Spanish, German, or paradigms according to gender, so that they can agree Russian, class marking is in the ending. That is, class- with the gender and number of the noun that corre- es are marked by means of information conveyed by sponds to the entity to which they refer. suffixes. In Spanish, -a is normally the ending of fem- inine singular nouns, whereas -o is normally the end- ing of masculine singular nouns. However, this is not References always so, because there are some nouns that end in - Corbett, Greville. 1991. Gender. Cambridge: Cambridge a which are masculine (e.g. poeta, the Spanish word University Press. for ‘poet’). In this language, gender information is Craig, Colette (ed.) 1986. Noun classes and categorization. Amsterdam and Philadelphia: John Benjamins. combined with number, so that all plurals add -s to the Givón, Talmy. 1983. Syntax. A functional–typological intro- singular form, and we can find in general the follow- duction. Amsterdam and Philadelphia: John Benjamins. ing pattern: Heine, Bernd. 1982. African noun class systems. Apprehension. Das sprachliche Erfassen von Gegenstanden, Vol. 1, ed. by H. Seiler and C. Lehmann. Tübingen: Narr. Masculine Feminine Heine, Bernd, Ulrike Claudi, and Friederike Huennemeyer. Singular -o -a 1991. Grammaticalization: a conceptual framework. Plural -os -as Chicago: University of Chicago Press. Matthews, Peter H. 1991. Morphology. Cambridge: Cambridge University Press. However, there are numerous exceptions to this CARLOS INCHAURRALDE rule, and there are also other endings, which are diffi- cult to assign to one of the two classes. In German or See also Gender and Language 371 GENERATION Generation Generation is the process by which thought is rendered of grammar that linguists have developed has been into language. Within computational linguistics, it is applied to NLG including some that are relatively referred to as ‘natural language generation’ (NLG) to unknown in the wider community such as Melcuk’s help distinguish it from Chomsky’s generative gram- Meaning-Text Theory and Halliday’s Systemic mar and to contrast with natural language understand- Functional Grammar (SFL). Interest in these theories ing. It is the study of how actual speakers, people or of grammar stems from the need to reason about the computers, construct utterances in actual contexts— alternative choices that are available, the functions the situations that motivate them to speak. As such, they perform, and their consequences for other choic- NLG is part of the larger field of Cognitive Science, es later on. For example, saying ‘the house is red’ vs. where it is also referred to as ‘production’, particular- ‘the red house’ will express the same fact about the ly by psycholinguists. house, but with differences in emphasis and in what The first NLG systems were developed in the 1950s parts of the sentence they leave open for other infor- as part of machine translation systems. The field mation to fill. Surface realization is also the only com- gained maturity in the 1980s, with its own series of ponent within the architecture of NLG systems that is workshops and conferences and its own unique prob- sufficiently mature for ‘plug and play’ reusable com- lems. As gauged by membership in the special interest ponents to have emerged, notably KPML (Bateman group SIGGEN, there are several hundred people 1997) and FUF/SURGE (Elhadad and Robin 2001) today actively pursuing research on generation. both of which use SFL. NLG takes its methodology from Artificial The state of the art in NLG is measured by a com- Intelligence. To study a cognitive capability, you bination of the fluency of the texts it is possible to pro- design and implement computer programs that attempt duce and comparative difficulty of adapting to new to replicate it: in this instance to produce fluent utter- subjects or genres. Consider this example, an automat- ances for a purpose. To do this, the program (‘genera- ically generated recipe for butter bean soup (Dale tor’) starts with a body of data or information. This 1992: 14). might be the daily movement of a stock market index (Kukich 1988) or numerical data about temperature Soak, drain and rinse the butter beans. Peel and chop the onion. Peel and chop the potato. Scrape and chop and wind patterns (Goldberg et al. 1994). In cases like the carrots. Slice the celery. Melt the butter. Add the veg- these, the first thing the generator must do is analyze etables. Sauté them. Add the butter beans, the stock and the data to determine both what information it contains the milk. Simmer. Liquidize the soup. Stir in the cream. (e.g. what is particularly salient: have winds increased Add the seasonings. Reheat. or diminished) and what concepts—ultimately what words or phrases—could be used to communicate that Notice how this text is tailored to its genre. All the information. sentences are imperatives; objects can be omitted when There is a consensus among NLG researchers that they are obvious (‘Reheat __’), and the sentences are generation involves three broad-brush components: (1) simple and short. One of the problems in microplan- determining and organizing the information content to ning is how to formulate this ‘tactical’ knowledge be expressed; (2) ‘microplanning’, where the senten- about a genre’s preferred constructions in such a way tial and referential structures are determined; and (3) that it can be deployed by other NLG systems produc- ‘surface realization’, where the text plan is processed ing texts about a different subject, especially when they by a grammar to construct the sequence of syntactical- use different processing methodologies. We know how ly and morphologically appropriate word forms, to do this with grammars, but not with the knowledge which are then rendered on some output medium, typ- of how to balance the consequences of alternatives ically formatted text displays or web pages. (For when a text is composed. details of alternative NLG architectures, see This class of problems, unique to generation, is fur- McDonald (2000) or Reiter and Dale (2000).) ther illustrated with the example below, which was pro- Surface realization is the most advanced of these duced by Robin’s STREAK system (1993). It is an three components since it draws on the well-estab- example of the best that can be done today as it is indis- lished knowledge of grammar in linguistics and com- tinguishable from what a human sports journalist would putational linguistics as a whole. Virtually every kind produce as a capsule summary of a basketball game. 372  GENERATIVE GRAMMAR Dallas, TX—Charles Barkley matched his season record References with 42 points Friday night as the Phoenix Suns routed Bateman, John. 1997. Enabling technology for multilingual the Dallas Mavericks 123–97. natural language generation. Natural Language Engineering STREAK uses an architecture based on revisions to 3. 15–35. Dale, Robert. 1992. Generating referring expressions. an initial draft, where it continually looks for opportu- Cambridge, MA: MIT Press. nities to incorporate historical knowledge into a skele- Elhadad, Michael, and Jacques Robin. 2001. SURGE: a com- ton of reported facts. In this instance, for example, prehensive plug-in syntactic realization component for text there is the fact that the Mavericks are on a long los- generation. Computational Linguistics, in press. ing streak. It is not possible to add the number of loss- Goldberg, Eli, Norbert Driedger, and Richard Kittredge. 1994. Using natural language processing to produce weather fore- es to the sentence, given its present structure; however, casts. IEEE Expert 9(2). this generator has extensive tactical knowledge about Kukich, Karen. 1988. Fluency in natural language reports. the choices available to it and knows that if it uses an Natural language generation systems, ed. by David alternative way of phrasing the fact that the Suns lost, McDonald and Leonard Bolc. New York: Springer-Verlag. one that reifies the loss as a noun, it can then incorpo- Levelt, Willem J.M. 1989. Speaking: from intention to articula- tion. Cambridge, MA: MIT Press. rate the number of losses by modifying the noun with McDonald, David. 2000. Natural language generation. Handbook the count. Underlines indicate the text that has of natural language processing, ed. by Robert Dale, Herman changed: Moisi and Harold Somers. New York: Marcel Decker. Reiter, Ehud, and Robert Dale. 2000. Building natural language . . . the Phoenix Suns handed the Dallas Mavericks their generation systems. Cambridge, UK: Cambridge University 27th defeat in a row at home 123–97. Press. Robin, Jacques. 1993. A revision-based generation architecture The focus of ongoing research is in text planning for reporting facts in their historical context. New concepts problems such as this illustrates; in the extension of in natural language generation: planning, realization and established capabilities to larger texts (as this is systems, ed. by Helmut Horacek and Michael Zock. written, the limit is multiple page, individually tai- London: Pinter. SIGGEN see www.aclweb.org lored instruction, or advice pamphlets), and in inte- DAVID MCDONALD gration with other modalities such as real-time graphics and the production of speech with appro- See also Computational Linguistics; Machine priate prosodics. Translation Generative Grammar Generative grammar is a conceptual model whose cen- ogists, attributes knowledge of language to general- tral tenet is that language is a property for which purpose learning mechanisms, and thus assumes this human beings are biologically prewired. While all ability to be no different from any other cognitive models of language assume some role for both biolo- capability; the generative approach, in contrast, sub- gy and environment, they differ with respect to the scribed to by many linguists, views linguistic knowl- emphasis that they place on each: empiricist models edge as being unique and specialized, and stemming attribute a greater role to the environment, and focus from brain structures devoted specifically to the pro- on differences across speakers’ grammars (for cessing of language. instance, how the particular input that children receive The generative framework had its origins in the influences the development of their grammar). 1950s with the publication of Noam Chomsky’s 1957 Nativist models, in contrast, attribute a greater role to book Syntactic structures, which built on the work of the biological component, and focus on commonalities his teacher Zellig Harris. Chomsky’s approach was a across speakers’ grammars. While all nativist models reaction to the behaviorist theory of language preva- assume language to be biologically determined, they lent at the time, championed by the psychologist differ with respect to the nature of such knowledge: Skinner. Under a behaviorist model, the brain is consid- The emergentist approach, favored largely by psychol- ered a blank slate with regard to linguistic knowledge; 373 GENERATIVE GRAMMAR children must thus be explicitly taught their language book,’ each further embedding resulting in a new sen- by the adults around them in a stimulus–response tence. Another central feature of such a grammar is manner, their behavior being rewarded when they imi- that it is highly constrained. For instance, all syntac- tate the adults’ language correctly. Chomsky instead tic rules make reference to the internal structure of advocated a view subscribed to in the previous centu- the sentence (known as structure dependency). Thus, ry (to which behaviorism had been a reaction) that yes–no questions in English are formed by moving some brain activities are unconscious and reflexive, the auxiliary to the front of the sentence (‘Will the just as is the case for many physical processes. Much student read the book?’), such a rule being framed in of human beings’ linguistic knowledge, Chomsky terms of an internal grammatical unit (the auxiliary argued, is abstract and unconscious, but can be ‘will’). No language has rules that are structure inde- brought to conscious awareness by examining speak- pendent, such as moving the third word to the front ers’ usage of such linguistic knowledge (known as the of the sentence to form a question, which refer competence/performance dichotomy). instead to surface properties such as linear position. Under a generative approach, human beings are That the latter formulation will not work can be seen assumed to be prewired for language, beginning life when the subject is replaced with a pronoun: while not with a blank slate but rather with a linguistic tem- the structure-dependent rule will generate the correct plate or blueprint that they flesh out upon exposure to question since it always moves the auxiliary (‘Will he specific linguistic data. Instead of learning language read the book?’), the structure-independent formula- by imitating those around them, children create their tion will not, as the auxiliary is no longer the third own grammars. One reason for assuming this is that word (‘Read he will the book?’). Constraints such as they make errors that adults do not (e.g. ‘I hurted structure dependency support the idea of speakers myself’) and generate novel forms, neither of which possessing unconscious, abstract linguistic knowl- should occur if they learned solely by imitating. In edge, as there is no overt evidence of the syntactic addition, most children are not given explicit instruc- groupings of words in the input that they hear. And, tion in their language or corrected on errors, necessary the fact that speakers were never taught such con- in an imitative model, yet all manage to acquire lan- straints, let alone being aware that they exist, coupled guage. Furthermore, although the environments in with the fact that these hold across all languages, which children acquire their language vary, they all go supports the idea of there being a wired-in universal through similar stages in acquiring a language and component to language. during the same general time frame (for instance, While the idea of an innate, prewired blueprint has forming sentences with content words such as nouns remained constant in generative grammar, the concep- and verbs around 18 to 24 months, and function words tual details have varied across the decades. In the like ‘will’ and ‘my’ between 24 and 30 months). 1960s, the emphasis was on the distinction between a Moreover, although children do not hear examples of deep structure, which conveyed the semantic proper- every possible structural pattern, they nonetheless ties of a sentence, and a surface structure, which sup- attain a grammar capable of generating all the possible plied its pronunciation. Thus, the passive sentence sentences in their language (known as the poverty of ‘The book was read by the student’ was assumed to the stimulus argument). And, although each is exposed come from the same deep or underlying structure as its to different data and in a different order, they all end active counterpart, ‘The student read the book’, as up with the same basic grammar for their language, both have the same meaning. In the 1970s, the empha- which would be unexpected under an imitative sis shifted to finding the set of transformations used to account. derive the various syntactic patterns of each language. The name ‘generative grammar’ is used to refer to The list included transformations for passives, yes–no this model since speakers are assumed to possess a questions, and wh-questions (‘Which book did the stu- grammar capable of generating all the possible sen- dent read?’). While the number of sentences in a lan- tences in their language (while excluding all the guage is potentially infinite, it was assumed that the impossible ones). The grammar consists of a finite number of transformations could be reduced to a finite number of rules, yet is capable of generating an infi- set. However, it soon became clear that there were nite number of sentences from such rules due to their many more transformations than it was possible to ability to refer back to each other repeatedly (known enumerate. Emphasis then shifted to narrowing down as recursion). For instance, one can continue to the transformations by type. Two general types were embed sentences within one another as in the exam- established: noun phrase (NP) movement, which moved ple ‘John thought that Mary said that Fred believed a phrase within a sentence, as in passives, and wh- that Cindy suspected that the student had read the movement, which moved a phrase outside a sentence to 374  GENERATIVE GRAMMAR a presentential landing site, as in wh-questions. assumed to be attached directly in the lexicon; the syn- Eventually, these two transformations were collapsed tactic component then checked to see that features on into one general transformation, move alpha, which the words matched. If so, the derivation was said to allowed movement of any constituent anywhere, sub- converge, otherwise, it crashed. Thus, the sentence ject to certain constraints. ‘The student enjoys the book’ would be acceptable In the 1980s, the model was flipped on its head. since ‘student’ and ‘enjoys’ are both third-person sin- Rather than looking for the possible structures in a lan- gular, whereas ‘The student enjoy the book’ would be guage, the emphasis instead shifted to determining the ungrammatical as the number agreement on the noun impossible structures. This represented a significant and verb do not match. The formation of sentences evolution in the conceptual model: whereas the num- was now assumed to occur by means of a few basic ber of patterns possible in a language is potentially operations such as merge, used to generate basic infinite, the number of constraints is thought to be very declarative sentences, and move, used to derive pat- small. In addition, it also made it possible to shift the terns such as passives and questions. Another concep- emphasis to universal aspects of language, rather than tual model that developed during this decade was simply to those properties that an individual language optimality theory, which attributed variation among possessed. The grammar was now taken to consist, not languages to their different rankings of a set of univer- of a set of rules, but rather of a set of autonomous sal constraints. modules that interacted with each other; one concep- Since any model proposed for a grammar must be tion of such a model became known as government- one that is learnable by the child, language acquisition binding theory, named after two of the modules, while research has helped to shape development of the gen- alternative models were also proposed such as lexical- erative framework, and much first- and second-lan- functional grammar. The overall conceptual model guage research today is solidly grounded in such a became known as the principles-and-parameters framework. First-language researchers are interested in model since it considered language to consist of a set questions such as whether the principles of language of wired-in principles that all languages shared, along are all present at birth or instead come online gradual- with a set of parameters that they also shared, but ly as the child develops cognitively (the continuity/ whose values varied cross-linguistically and needed to maturation debate). Second-language researchers are be set upon exposure to language-particular data (such interested in determining whether second-language innate knowledge being referred to as Universal learners have full access to the language faculty as first- Grammar). An example of a principle would be a language learners do, partial access only (properties movement constraint known as subjacency, which pro- that are the same transferring, but new ones not being hibits movement of a phrase out of more than one acquirable), or no access at all. And, a central question clause or noun phrase in a single step (the name ‘sub- for both first- and second-language acquisition is jacency’ referring to the fact that movement can occur whether there is a critical time period during which lan- to an adjacent clause, but not a subadjacent one). Thus, guage must be acquired, as is true for other biological- one cannot say ‘Which book do you know the student ly determined properties. who read?’ since ‘which book’ has been moved from While the conceptual details of the generative within the relative clause (‘the student who read which grammar model have changed greatly over half a cen- book’) and the upper sentence (‘you know the student tury, the basic underlying tenet, that language is a who read which book’). While such a constraint is species-specific property for which human beings thought to be universal, the constituents out of which come prewired, has remained constant. Future the element may move (the bounding nodes) vary research will undoubtedly yield new insights into the cross-linguistically. English is freer in its movement specific shape of the grammar, while remaining true to allowances than Russian, but less so than Italian or the model’s belief in a wired-in blueprint. Swedish; a language like Japanese, in contrast, allows no overt syntactic movement. Thus, a parametric dif- ference linked to this principle would dictate what the References bounding nodes for a given language are. Baker, Mark. 2001. The atoms of language: the mind’s hidden In the 1990s, the emphasis turned to making the rules of grammar. New York: Basic Books; Oxford: Oxford model even simpler conceptually. The new approach, University Press, 2002. called minimalism, assumed a much more limited role Chomsky, Noam. 1957. Syntactic structures. Mouton: The Hague. for the syntactic component. It now was seen as a Chomsky, Noam, Adriana Belletti, and Luigi Rizzi (eds.) 2002. computational device that simply checked that sen- On nature and language. Cambridge and New York: tences were formed correctly. All morphology was Cambridge University Press. 375 GENERATIVE GRAMMAR Radford, Andrew. 1997. Syntactic theory and the structure of Smith, Neil, and Deirdre Wilson. 1979. Modern linguistics: the English: a minimalist approach. Cambridge and New York: results of Chomsky’s revolution. Brighton: Harvester, and Cambridge University Press. Indiana: Indiana University Press. Radford, Andrew. 1988. Transformational grammar: a first NAOMI BOLOTIN course. Cambridge and New York: Cambridge University Press. See also Chomsky, Noam; Harris, Zellig Sabbetai Genericity Much of our knowledge about the world is expressed An alternative approach denies that there is any quan- in sentences such as tification involved. Rather, generics are evaluated with respect to rules and regulations. For example, A raven (1) A raven is black. is black is true not because of the properties of indi- (2) Birds lay eggs. vidual ravens, but because there is a (biological) rule (3) The tiger has stripes. stating that ravens are black. (4) Mary smokes. Perhaps one of the difficulties in determining These are examples of generics. whether generics are quantificational is that generics Generics can occur in many syntactic forms: (1) are lawlike; they cannot express accidental generaliza- involves the indefinite singular noun phrase a raven, (2) tions. For example, even if it turned out that all contains the bare plural noun birds, and the subject of Supreme Court judges had an even Social Security (3) is the definite singular noun phrase the tiger, where- number, the generic Supreme Court judges have an as that of (4) is the name Mary. All these forms can have even Social Security number would be odd. Contrast nongeneric uses as well. Indeed, there does not seem to this with Supreme Court judges are appointed by the be any sentence form that is unique to generics; what President, which is fine. Intuitively, this is because characterizes them is their meaning. Intuitively, the there is a law stating that Supreme Court judges must above sentences seem to express some sort of general- be appointed by the President, but no law governs their ization, about ravens, birds, tigers, and Mary. Yet, it is Social Security number. far from clear exactly what generics mean. What does it To account for lawlikeness, some researchers treat it mean to say that some generalization holds? as a form of necessity. According to standard accounts Possibly the first hypothesis that comes to mind is of necessity, a necessary statement is not merely true in that generics express some sort of quantification. A the actual world, but in all possible worlds. That is to raven is black, for example, could be taken to mean that say, there are any number of ways the world could have a certain number of ravens are black. But how many? been different from the way it actually is; but in all such What is the generic quantifier? It cannot be every, hypothetical situations, a necessary statement would because A raven is black is true despite the existence of still be true. The idea is, then, that a lawlike statement albino ravens. It cannot be most, since Birds lay eggs is is true in ‘sufficiently many’ possible worlds. true despite the fact that fewer than half of all birds lay Indeed, it seems that the truth or falsity of a gener- eggs (adult females only). The problem is even harder ic does not depend on specific events in the actual with Mary smokes, because it is quite unclear how often world. In the sentence A computer computes the daily Mary has to smoke in order for the sentence to be true. weather forecast, for example, the daily weather fore- Faced with the diversity of interpretations of gener- cast cannot be taken to refer to the current weather ics, researchers have taken two general approaches. forecast. For example, if we know that today’s forecast The first attempts to define the quantifier in such a way predicts a blizzard and is the main news item, we still that its nature, possibly in conjunction with context, cannot change the above sentence to A computer com- intonation, and world knowledge, may account for all putes the main news item. Yet, if the daily weather the interpretations of generics. For example, one could forecast did refer to the current weather forecast, this still maintain that the generic quantifier is most, and last statement should be true. propose that to determine the truth or falsity of Birds However, although generics do not seem to depend lay eggs, only female birds need to be considered. on specific events in the actual world, their interpreta- Since most female birds lay eggs, the sentence is true. tion nevertheless depends on real-world circumstances. 376  GENETIC RELATIONSHIP Suppose the weather report is Mary’s favorite newspa- References per column. Although the world could well have been Carlson, Greg. 1977. Reference to kind in English, Ph.D. such that Mary would have no interest in the weather Dissertation. University of Massachusetts; New York: forecast, we may still truthfully say that A computer Garland, 1980. computes Mary’s favorite newspaper column. Carlson, Greg. 1989. On the semantic composition of English Thus, the truth or falsity of the generic statement A generic sentences. Properties, types and meaning, ed. by computer computes the daily weather forecast depends Gennaro Chierchia, Barbara H. Partee and Raymond Turner, 167–92. Dordrecht: Kluwer. on the actual world, rather than what may be the case Carlson, Greg, and Francis J. Pelletier. 1995. The generic book. in hypothetical circumstances, but the interpretation Chicago: University of Chicago Press. apparently cannot be tied to a specific point in time. Cohen, Ariel. 1996. Think generic: the meaning and use of To account for this aspect of genericity, some theo- generic sentences. Ph.D. Dissertation. Carnegie Mellon ries propose to restrict the possible worlds applicable University; Stanford: Center for the Study of Language and Information, 1999. to interpretation only to those that are normal, or those Cohen, Ariel. 1999. Generics, frequency adverbs, and probabil- that are close to the actual world in terms of its essen- ity. Linguistics and Philosophy 22. 221–53. tial properties. Such theories have to face the problem Dahl, Östen. 1975. On generics. Formal semantics of natural of defining normality or essence in such a way that a language, ed. by Edward L. Keenan, 99–111. Cambridge: world in which, say, Mary is not interested in the Cambridge University Press. Delgrande, James P. 1987. A first-order conditional logic for weather is not included in this definition. prototypical properties. Artificial Intelligence 33. 105–30. Alternatively, it has been proposed that generics Heyer, Gerhard. 1990. Semantics and knowledge representation are evaluated with respect to a variety of possible in the analysis of generic descriptions. Journal of Semantics futures (e.g. that tomorrow the weather forecast may 7. 93–110 not be the main news item) but not with respect to Pelletier, Francis J., and Nicholas Asher. 1997. Generics and defaults. Handbook of logic and language, ed. by Johan van alternatives to the world as it actually is (e.g. that Benthem and Alice ter Meulen, 1125–77. Amsterdam: Mary might not ever have been interested in the Elsevier. weather forecast). Schubert, Lenhart K., and Francis J. Pelletier. 1987. Problems It is probably evident from this discussion that, in the representation of the logical form of generics, plurals, despite much progress, generics remain a puzzling and and mass nouns. New directions in semantics, ed. by Ernest LePore, 385–451. London: Academic Press. deeply contested phenomenon, with many questions ———. 1989. Generically speaking, or using discourse repre- and few widely accepted answers. In fact, one of the sentation theory to interpret generics. Properties, types and interesting and, so far, unanswered questions is this: if meaning, ed. by Gennaro Chierchia, Barbara H. Partee and it is so difficult to define what generics actually mean, Raymond Turner, 193–268. Dordrecht: Kluwer. why do we use them so often? ARIEL COHEN Genetic Relationship If we look at the word for ‘brother’ in a handful of a period of common development from which their European languages, we can see some striking resem- similarities arise. This is the case, for example, with blances: for example, between English brother, German the major modern Romance languages (i.e. Italian, Bruder, Polish brat, Welsh brawd, Irish brathair, Spanish, Portuguese, French, and Romanian), which Swedish broder, Dutch broer, and Czech bratr. Clearly, all have their origins in the Latin language spoken there is some kind of relationship between these lan- throughout the ancient Roman empire. As time pro- guages, but how exactly do we account for these simi- gresses, however, divergence occurs, mostly as a result larities? And why is it that differences also exist? of population migrations or through loss of a central The most salient way in which languages come to unifying standard: certain groups of speakers of the be related is through a genetic relationship. This protolanguage are split off from one another and, sub- means that, at some point in their history, the several sequently, their now separated language varieties pur- languages in question were actually the same language sue their own courses of development, from which the (normally known as the protolanguage) and underwent differences between them arise. 377 GENETIC RELATIONSHIP This model of language relationship is usually words, or grammatical features) is marked by equally known as the family tree model, and was proposed by individual borders (known as isoglosses). These August Schleicher in 1871. However, in its purest isoglosses may well form relatively strong borders form, the family tree model gives rise to a number of between languages when they bundle closely togeth- problems. Firstly, it assumes that there are clear divid- er, but they are by no means necessarily located in the ing lines between languages (both temporally and geo- same places. The wave theory proposes that individ- graphically), so that there is a point at which we can ual features spread from one place to another by make a distinction between, say, late vulgar Latin as a means of diffusion. In other words, the extent of a fea- protolanguage and early Spanish as a separate devel- ture’s isogloss spreads as it gradually moves from one opment from it; however, it is very difficult to draw location to the next, rather like passing a message these lines clearly. Secondly, the pure family tree along a chain of people; and, as more features spread model assumes that no further interaction takes place out, the more similar the language spoken in the dif- between the separated languages and the protolan- ferent places along the chain becomes. Because it guage after a split has occurred; however, this is again does not assume that shared features are always the often not so, as Romance demonstrates—the protolan- result of a single period of shared historical develop- guage (Latin) continued to be used quite extensively ment, this model is able to accommodate later con- alongside the modern Romance languages, for exam- vergence between languages. Similarly, it also allows ple, by the church and the universities. Similarly, the for the influence of non-genetically-related languages model also does not allow for a later convergence in these contexts, for example, the influence of the between languages that resulted from an earlier split. Semitic Arabic language on the Indo-European Urdu For example, it is estimated that roughly a third of the language. English vocabulary is of French origin. However, the The two other causes of language similarity are (1) ancestors of English and French split off from one chance similarity and (2) the operation of universal lin- another at a very early date, and this shared vocabulary guistic processes, which lead to similar developments does not stem from that more distant genetic relation- after languages have split off from one another. As an ship between English and French; rather, it is a result example of the latter, we can consider the development of large-scale later borrowing into English, especially of the definite articles (= ‘the’) in two of the Romance in the years following the Norman conquest of languages—French and Italian. These are descended in England in 1066 CE. This example is salutary, because both languages from the Latin pronoun ille and its vari- it shows that a large degree of similarity (e.g. in vocab- ants. In both French and Italian, the feminine singular ulary) does not necessarily result directly from a forms are the same (la) and both also lose their /a/ genetic relationship. vowel before another vowel (French l’allumette; Italian Clearly, therefore, although genetic relationship is l’amica). However, this does not mean that the word la often a very important determining factor in language came either directly from Latin or was borrowed from similarity, as with the Romance languages, the purest Italian into French (or vice versa); rather, universal form of the family tree model must be considered an phonetic principles (loss of an unstressed vowel and the idealization of reality. merging of adjacent vowel sounds) have operated on There are, in fact, broadly three other reasons the original Latin word (illa) with the same end results. (apart from genetic heritage) that can lead to a close interrelationship between languages. Probably the next most important, which has already been alluded References to, is borrowing from one language into another. Aitchison, Jean. 1991. Language change: progress or decay? Although small numbers of loans are very common, Cambridge: Cambridge University Press. large-scale borrowing that leads to a substantial over- Anttila, Raimo. 1972. An introduction to historical and com- lap in vocabulary and/or grammar normally arises parative linguistics. New York: Macmillan. Hock, Hans Henrich. 1986. Principles of historical linguistics. either in a context where many speakers are bilingual Berlin: Mouton de Gruyter. or in a situation where two languages are in close geo- Hock, Hans Henrich, and Brian D. Joseph. 1996. Language his- graphical proximity to one another. To account for tory, language change, and language relationship: an intro- this, another model of language relation was proposed duction to historical and comparative linguistics. Berlin: by Johannes Schmidt in 1872. This is known as the Mouton de Gruyter. Jeffers, Robert J., and Ilse Lehiste. 1979. Principles and meth- wave model and it is based on the principles of dialec- ods for historical linguistics. Cambridge, MA: MIT Press. tology. The wave model rejects the notion that precise Lass, Roger. 1997. Historical linguistics and language change. geographical borders exist between different lan- Cambridge: Cambridge University Press. guages. Instead, it assumes that the spatial extent of ANDREW WILSON individual linguistic features (be they sound patterns, See also Indo-European 4: Romance 378  GENRE Genre ‘Genre’ stems from Latin genus and French genre, More recently, however, language and communica- words meaning ‘kind’. Hence, the popular meaning of tion are increasingly viewed as multimodal and sys- genre in Do you have some shoes in the same genre? temic. Hence, genres are often perceived as the is close to its etymological roots. Historically, howev- interactive interplay of form, content, and use. er, and largely as an inheritance from Aristotle, ‘genre’ Consequently, discourse and hence genres interrelate was defined normatively as a given category of litera- three major aspects of the lifeworld: ‘uttering selves’, ture, with the choices originally restricted to ‘dramas’ ‘referred worlds’, and ‘acting societies’. This triadic and ‘novels’. With Romanticism, ‘poetry’ came to be perspective considers genre as a recognizable balance included as well. Hence, at the end of the nineteenth of aesthetics (form), epistemology (content), and century, under the influence of literary criticism and ethics (use). However, this multidimensional view mass-produced literature, subgenres and mixtures of raises new questions: Should genres be defined by the three classic forms were accepted as genres and a their degree of stability? Can they be understood and comprehensive system of defined kinds of literary analyzed one by one in isolation? Where are they real- writing was established. ly located? Can genres be defined properly? This normative understanding prevailed until the To address the questions in order, genres are cer- late 1960s, when it was challenged from within literary tainly not just given or static. Studying forms of com- theory and by other disciplines, such as anthropology, munication historically and sociologically makes it pragmatics, and text theory. ‘Genre’ as an exclusive lit- clear that genres come into being, live, change, and erary concept was rejected. Genres were simply text fade out. Genres have to mix cultural stability and types and were approached descriptively. However, dif- change. Hence, genres are now more often seen as rel- ferent forms of opera, film, video, music, dance, pro- atively stable and flexible, some more stable than oth- fessional and everyday practice—in brief, any kind of ers, the variation motivated by social needs for cultural communication—were soon considered genres openness and closeness. Further, it is clear that genres as well. Genre had returned to its origin, ‘kind’. are inherited and create system-like connections. Genre had traditionally often been defined one-sid- Narratives, from short jokes to ‘endless’ television edly by proceeding from one of the three main aspects series, for instance, share some aspects but differ in of text: form, content, or use. For instance, formal (or others. Genres are thus systemic, i.e. no aspect of a syntactic) orientations would argue that genres such as particular genre can be analyzed without defining its limericks and death notices could be recognized from relationships with others. certain textual structures. Contentual (or semantic) Whether a verbal genre is in a text is disputed. The views would focus on typicality in content, seeing understanding of language as a describably closed sys- patents and psalms, for instance, as particular kinds of tem is often combined with the view that genres are descriptions. Finally, focusing on use (or function), repetitive, fixed macropatterns of communication locat- genres were primarily seen as (social) action, such as ed in texts. Approaches seeing genres as flexible and commands and recipes. open phenomena, however, perceive them more as a set As long as linguistics focused mainly on pronuncia- of potential forms, meanings, and uses. Metaphorically, tion, grammar, and meaning at the sentence level or genres can be seen as nine tenths of an iceberg, where below, higher levels, such as text and context, were of the visible part (the utterance) only symptomatically less interest. However, once linguistic pragmatics start- hints or directs communicators to a potential intended ed focusing on the use of language, perceptions of gen- meaning framed by a possible genre. Hence, even a ver- res went through a paradigmatic change. Genres were bal genre is a growing, immanent potential of mind and now seen as function-oriented tools for social action. body, not just a kind of text; and, by the same token, a This created a tension between a functionalist/pragmat- text is not in a verbal genre. ic and a semantic/syntactic approach. In the 1980s, The relation between the two planes is regulated by there was successful advocacy for a syntactic–semantic a mechanism called, in text theory, theme/rheme, approach to genre within the field of film/media, with which balances the given (or the expected in relation both form and content defining genre. In other words, to the genre, the context, or the situation) and the new: the relationship between the main communicational Once upon a time there was (theme/given in the genre aspects of genre was unresolved. ‘fairy tale’) a big troll (rheme/new). The relationship 379 GENRE between utterance/text and genre/context is dialogical Coe, Richard, et al. (eds.) 2001. The rhetoric and ideology of in the sense that the genre influences the utterance. genre: strategies for stability and change. Hampton Press: Cresskill, New Jersey, USA. However, once an utterance is uttered, it contributes in Cope, Bill, and Mary Kalantzis (eds.) 1993. The powers of lit- principle to changing the genre. Over time, genres are eracy: a genre approach to teaching writing. London: The changed by utterances. Falmer Press. The unpredictable and growing hidden and imma- Dubrow, Heather. 1982. Genre. London and New York: nent potential makes it hard, if not impossible, to Methuen. Fowler, Alastair. 1982. Kind of literature: an introduction to the describe the ‘true’ nature of genres. Hence, a dilemma theory of genres and modes. Cambridge, MS: Harvard for applied linguistics is that focusing on stability may University Press. yield a more trustworthy and precise scientific disci- Freadman, Anne. 1994. Anyone for tennis? In Freedman and pline, but its value for application and practical under- Medway. standing will be reduced. Some academic disciplines Freedman, Aviva, and Peter Medway (eds.) 1994a. Learning and teaching genre. Portsmouth: Heinemann Boynton/Cook. favor precise definitions to obtain valid operationaliza- Freedman, Aviva, and Peter Medway (eds.) 1994b. Genre and tion in research. This has led either to a rejection of the the new rhetoric. Critical perspectives on literacy and edu- concept of genre or to a predisposition for searching cation. London: Taylor and Francis. only for fixed aspects and hence to a self-fulfilling Habermas, Jürgen. 1984. The theory of communicative action, hypothesis of genre as fixed. Vol. 1, translated by T. McCarthy. Boston: Beacon. Hauptmeier, Helmuth. 1987. Sketches of theories of genre. Thus, when the concept of genre has at the same Poetics 16. 397–430. time become more general in daily life as well as in Hirsch, Edward D. 1967. Validy in interpretation. New Haven: the perception of several academic professions, a trust- Yale University Press. worthy and overall usable definition of the concept is Martin, Jim. 1997. Analysing genre: functional parameters. endangered. However, while the acceptance of too Genre and institutions. Social processes in the workplace and school, ed. by Frances Christie and Jim Martin. London much openness may lead to an increased understand- and Washington: Cassell. ing of the dynamics of text and context, it may reduce Miller, Carolyn. 1994. Genre as social action. In Freedman and its scientific validity. Thus, genre urges analysts to Medway. decide whether the purpose of definitions and expla- Ongstad, Sigmund. 2001. Genres—from static, closed, nation is proof or understanding, precision or likeli- extrinsic, verbal dyads to dynamic, open, intrinsic semiotic triads. In Coe et al. hood, description or application. Paltridge, Brian. 1997. Genre, frames and writing in research settings. Amsterdam: John Benjamins. References Swales, John. 1990. Genre analysis. English in academic and research settings. Cambridge, UK: Cambridge University Altman, Rick. 1986. A semantic/syntactic approach to film Press. genre. Film genre reader, ed. by Barry Keith Grant. Austin: Threadgold, T. 1989. Talking about genre ideologies and University of Texas Press. incompatible discourses. Cultural Studies 3(1). 101—27. Altman, Rick. 1999. Film/Genre. London: British Film Institute Todorov, Tzvetan. 1990/1978. Genres in discourse. Cambridge, Publishing. UK: Cambridge University Press. Bakhtin, Mikhail. 1986. Speech genres and other late essays. Voloshinov, V.N. 1973. Marxism and the philosophy of lan- Austin: University of Texas Press. guage. New York: Seminar Press. Ben-Amos, D. (ed.) 1976. Folklore genres. Austin: University of Texas Press. SIGMUND ONGSTAD Georgian and Caucasian Languages The Caucasus is a range of mountains that run from Georgian the northwest near the Sea of Azov at the top of the Among the best known of the native languages is Black Sea in southern Russia in a nearly straight line Georgian. Georgian belongs to the Kartvelian or South southwestward to the Caspian Sea coast where they Caucasian language family, to which Mingrelian and end in the Apsheron Peninsula near Baku in Laz also belong, often grouped together as Zan or Azerbaijan. Known as the Mountain of Tongues, the Chan, and the distantly related Svan in the mountains Caucasus Mountains harbor roughly 50 languages, at themselves. The family is located, as one of its names least 37 of which are indigenous to the region. 380  GEORGIAN AND CAUCASIAN LANGUAGES suggests, on the southern slopes of the Caucasus mas- which is CCSCCSCVC. In fact Georgian is typologi- sif, extending down toward the lowlands along the cally highly unusual in allowing such strings, but also eastern Black Sea coast and well into eastern Turkey. admitting strings of vowels, as in /aairebadi/ ‘transient, Georgia is a republic and has roughly 3.5 million volatile, volatility,’ or /da-a-aprak’-d-a/ direction- inhabitants who speak Georgian or a related language. active-warp-past-3, “It warped it,” with each /a/ given The 1989 Turkish census listed 8.6 million people of a slight pulse of energy. Georgian has the canonical ‘Georgian’ ancestry, most in the eastern regions. A few pure vowels /a, e, i, o, u/, with only Svan showing immigrant communities exist in North America and vowel length. Georgian has a light percussive intona- Europe, as well as in Russia. tional stress, which is highly unusual in that it is vari- Georgian itself is a literary language that dates back able, falling sometimes on the penultimate and at other to the fifth century CE, and has an older (khutsuri or times on the antepenultimate syllable, for example clerical) script and a later modern one (mkhedruli or /tbili´si/, /tbi´lisi/. Despite its consonantism Georgian is knightly). Old Georgian exists as a bible and in other mellifluous. older church writing. The modern language is a vehi- Georgian has very complex word forms and word cle for a whole culture and has an extensive literature, formation processes. Derivational processes changing which may be dated to the 12th century. the class of a word, e.g. from verb to noun, generally Georgian is a language of bewildering complexity, proceeds by means of suffixation: not merely for reasons of its grammatical richness, but a. /tavisupal/ ‘free’ vs. /tavisupl-eba/ also because of a high degree of irregularity exhibited ‘freedom’ (with syncope) by that grammar. The sound system has many features typical of a Caucasian language generally, with the b. /bawšw/ ‘child’ vs. /bawšw-oba/ addition of two features that make it typically ‘childhood.’ Kartvelian. First, the system shows a three way con- Circumfixes, i.e. combinations of prefixal and suf- trast in the stop consonant series and a two-way con- fixal elements that are used simultaneously, are also trast in the fricative series, taking the dental stops and widely used: affricates as examples: stops /t, d, t’/, affricates /c, , a. /or/ ‘two’ vs. /me-or-e/ ‘second’ c’/, fricatives /s, z/. It also has a series of ejective con- b. /xšir/ ‘frequent’ vs. /si-xšir-e/ sonants. Georgian stands apart from other Caucasian ‘frequency’ languages, as do other members of its family, in lack- c. /kartwel/ ‘Georgian’ vs. /sa-kartwel-o/ ing lateral fricatives and affricates, and in lacking pha- ‘Georgia’ ryngeal fricatives. Georgian also exhibits the peculiar d. /γrma/ ‘deep’ vs. /u-γrm-es-i/ “harmonic clusters” of Kartvelian. These are two ‘very/the most old’ series coronal stops and affricates that are made with simultaneous velar or uvular closure. Thus Georgian Georgian nouns and pronouns do not distinguish has the coronals in (1). gender, but are marked for seven grammatical cases. These encode grammatical relations (ergative, absolu- Velarized and uvularized coronals tive, dative), address (vocative), possession (genitive), a. /tkeši/ pouring rain /txa/, /tqa/ goat means (instrumental) and manner (adverbial). b. /dgoma/ to stand /dγe/ day Most Georgian morphology takes place in the com- c. /t’k’bili/ sweet /t’q’e/ forest plex verb, which can consist of up to ten morphemes, d. /ckera/ to look /cxeni/, /cqeni/ horse with many irregularities. Georgian verbs can express e. /gera/ beat, pulse /γola/ to lead highly complex meanings for which other languages f. /c’k’riala/ clear, cheerful /c’q’ali/ water need whole sentences, as the following examples show g. / J kmeta/ to pinch / J xiri/, / J qiri/ rod, stick – each given with a rough literal translation and a pos- h. /Egup’i/ group / Eγabna/ to scribble sible free translation: i. / J ’k’ua/ mind, intellect / J ’q’ap’i/ slush a. ga-v- c’itl-d-e-t If the harmonic (rounded) clusters are treated as surface- I- blush- start- maybe- plural unitary complex segments, then the syllable onsets of ‘…(that) we might blush’ Georgian become relatively straightforward: C(onso- b. ga-mo-gw-e-c E’ri-a surface- nant)V(owel), CCV, CSV, and SCV, where S stands toward.speaker-us-medial-wound-was for any sonorant (consonants with acoustic properties ‘We wounded him.’ similar to vowels, e.g. r, l, y, w, m, n). One can have c. mo-m-k’lav-en great strings of consonants, such as /wprckwnd-i/ = toward speaker me kill they [fp‘rc‘ Gk‘ Gndi] 1-peel-past-1.past, ‘I peeled it,’ (coll.), ‘They will kill me.’ 381 GEORGIAN AND CAUCASIAN LANGUAGES Eob-s d. s- dara across the word. The Richa dialect of Aghul is the only him watch he(subject) language known to contrast pharyngealized uvulars, ‘He was watching him’ true pharyngeals, “adytals” (sounds made with the epiglottis covering the adytus, the opening of the lar- Georgian shows ergative – absolutive case mark- ynx), and laryngeals. ings for subject and direct object with verbs of com- pleted action, absolutive – dative markings for verbs of Richa Aghul contrasts incomplete or future action, and dative – absolutive for a. uvular /xaE/ ‘house’ /γad/ ‘hammer’ –a w/ ‘nut’ /γ –a b/ ‘stack’ perfect tense verbs. Determining how this rich system b. pharyngealized uvular /x E E of case marking interacts with more conventional c. pharyngeal /haw/ ‘udder’ /ʕan/ ‘belly’ aspects of syntax is a central problem in the study of E/ ‘apple’ /ʕak˚/ ‘light’ d. adytal /h.ac Georgian. The vowel systems are also rich. While the Lezghian and Lakk-Dargwa languages lack /o/, the others can have any number of vocalic contrasts, including front Nakh-Daghestanian Languages rounded and high back unrounded vowels. The vowels To the east of the Kartvelian family lies the Nakh- can also come nasalized, long, and adytalized, the last Daghestian or Northeast Caucasian family. This con- having a harsh quality. Dargwa shows an opposition sists of roughly thirty-two languages. See the table, between truly pharyngealized vowels and adytalized where * denotes literary languages, all written in ones, the only language on earth to show this in its modified Cyrillic and, with the exception of scant vowels. Godoberi (an Andic language) has “stiff” material, all dating from the Soviet period. They vowels, made with great articulatory force. Vocalic extend from the center of the northern Caucasus down sequences are prohibited. to Azerbaijan, where they trail off into isolated vil- The morphology of these languages is character- lages. Hinukh may be the smallest speech community ized by a grammatical class system, ranging from the on earth. familiar gender system of male, female, neuter, and Nakh-Daghestanian languages (approximate popula- plural, to systems (in Vai Nakh) with six classifica- tion) tions, both in singular and in plural, or with fifteen in Godoberi, with masculine, feminine, and neuter singu- 1. Vai Nakh *Chechen (792,000), *Ingush lars, and with human and neuter plurals, each with (197,000), Batsbi (Kisti) (3,000) three distinct sub-types. The Lezghian languages tend 2. Avaro-Andi-Tsez to lack any grammatical class. A few nouns will exhib- a. *Avar (7 dialects) (501,000) it these markers, for example Avar /w-ac:/ mascu- b. Andic – Andi (9,000), Karata (5,000), line(masc)-sibling, ‘brother’, /y-ac:/ feminine(fem)- Chamalal (4,000), Bagwalal (4,000), sibling, ‘sister’, /b-ac:/ animal-sibling, ‘littermate of Akhwakh (5,000), Botlikh (3,000), Godoberi an animal’. Generally, however, such markers occur as (2,500), Tindi (5,000), Udi (formerly 6,000), prefixes or suffixes on pronouns, adjectives, associat- c. Tsezic - Tsez (Dido) (7,000), Hinukh (200), ed adverbs, and verbs, which reflect the class of the Hunzib (400), Khwarshi (1,000), Bezhta noun in the absolutive, as for example with Avar /há- (Kapuchis) (3,000), Inkhokwari (unknown) n-iw do-w ha-w-ize/ here-adverb masc 3rd-masc d. Lakk-Dargwa - *Lakk (92,000), *Dargwa be.born-masc-infinitve, ‘(for) him to be born here’. (282,000), Kubachi (3,000), Khaidaq The languages tend to have a rigidly S(ubject)- (28,000), O(bject)-V(erb) word order, with adjectives and e. Lezghian - *Lezgi (367,000), *Aghul relative clauses preceding the word they modify. (14,000), *Tabasaran (78,000), *Rutul Only in Vai Nakh do OVS and VSO orders occur in (15,000). Archi (1,000), *Tsakhur (19,000), certain discourse environments. Embedded clauses Budukh (1,000), Khinalug (2,000) generally precede the main clause. Some examples 3. Kryz (6,000) from Avar: These languages contrast plain stops with “intense” a. /w-as w-ekér-ula/ ones (marked with full colon). The numerous lateral masc-child masc-run-he.now affricates are velar in their articulation. Contrasts occur ‘The boy runs.’ among velar, uvular, pharyngeal, and laryngeal conso- E´i t’ax b. /c Eá-l r-ós-ul-e-w w-úgo/ nants. Some, such as Archi and Aghul, add pharyn- man book-plural plural-buy-now-masc masc-is gealized uvulars as well. In many languages the ‘The man is buying the books.’ presence of a pharyngeal or pharyngealized sound ele- c. /h.´i nc . . . E’.-a- u-l mac’ b-úgo du-r, k’aa-l-é- u-l/ ment in a word triggers the spread of pharyngealization bird-its-animal voice animal-is you-of, 382  GEORGIAN AND CAUCASIAN LANGUAGES speak-now-of ‘When you speak you sound like as the syntactic function of the complex (subordinated, a bird.’ Literally, ‘Bird’s voice is yours, while adverbial, gerundive), any deictic setting for the (you are) speaking.’ action, benefactives and adversatives, other adjuncts, all the arguments of a sentence and their number, capa- bility or inability, the geometry of the scene of action, Northwest Caucasian Languages details of motion to or from or around that scene, any incorporated nouns that modify the sense of the verb, The Northwest Caucasian family lies to the west of causative agent(s), valence of the root, reiteration of a Kartvelian, where Abkhaz adjoins Mingrelian, and state or action, adverbs, tense, mood, number of the extends up along the Black Sea coast to the Sea of absolutive noun, adverbial and modal suffixes, and Azov, thence eastward toward the center of the North subordinating complementizers at the very end. It is Caucasus. Perhaps half a million speakers still live in more a phrase than a simple lexical unit. Discourse this homeland, but ten times that number reside in a often consists of strings of verbs with only the occa- diaspora that has communities in Turkey, the Middle sional noun: East, Europe, and North America. The languages are Circassian, in western (Adyghean) and eastern Bzhedukh West Circassian verbal morphology (Kabardian) forms, with the moribund Ubykh transi- a. /sa s -q -z-a-t-y-a-y- -γa-w c´G -Dy -Dy -γa/ tional to Abkhaz and Abaza (on the northern side of I me-change.of.state-self-dative-surface- the massif from Abkhaz). Apart from some tales direction-to-3-past-let-stop-again-finally-past recorded in the 19th century, and perhaps some ‘He let me stop again at last.’ ancient monuments in cuneiform or in runes, the lit- b. /ø-q -z-f -w-s´’G -z-a-tn-y-a-w c´G -s´G -ah.e-γa- erary languages come from the Soviet period. All are gGara-r/ written in modified Cyrillic except for Ubykh texts, 3-change.of.state-what-for-you-despite-self- which are written in Latin script. These languages are dative-surface-direction-to-stop-able- famed for the complexity of their consonantal sys- around-past-definite.gerund-absolutive tems and the simplicity of their vocalic one. Ubykh, ‘why he was able to stop all around despite for example, has 81 consonants and only two vowels. you/your efforts’ These languages can show aspirated and ejective ’y ’y fricatives, as well as pharyngealized labials and uvu- c. /sa a-r ø-q -s-ʔa+cE -w-fa-s-w c E -na/ I 3-absolutive 3-interest-my-arm+instrumental- lars, palatalized uvulars, and labialized dentals, [pt], you-for-I-kill-immediate.future etc. Consontatal clusters abound, but vocalic ‘I shall kill him with my own hands for your sequences are absent. sake.’ Derivational morphology proceeds by suffixation and compounding. What is striking is the limited num- Syntax is ergative with SOV order. Modifiers follow ber of basic roots, only several hundred. Most of the and embeddings precede heads. vocabulary is made of compounds, often for terms that would seem basic. Some forms are extremely elabo- rate, as with Bzhedukh West Circassian /s´ha-a-n -γG References y -pcEa-y-a-sE -a-a-’-a/ head-/a/-purpose-door-direc- Aronson, Howard I. Georgian, a Reading Grammar. Columbus, tion-to-lead-in-/a/-near-to, ‘shutter for a window.’ In Ohio: Slavica, 1982. some cases the semantic machinery at work is clearly Colarusso, John. A grammar of the Kabardian Language. visible. Calgary, Alberta: The University of Calgary Press, 1992. Ebeling, Carl L. “The Grammar of Literary Avar (A Review Progressively abstract compounds (Bzhedukh West Article),” Studia Caucasica, 2: 58–100, 1966. Circassian) Haspelmath, Martin. A Grammar of Lezgian. Berlin, New York: a. /pq / ‘bone, frame, strut’ Mouton de Gruyter, 1993. b. /s´h.a-pq/ head-bone, ‘skull’ Kibrik, Alexandr E. (editor). Godoberi. Munich: Lincom- Europa, 1996. c. /ʔa-pq+a-pq/ arm-frame+leg-frame, ‘body’ (as Kibrik, Alexandr E. and Sergo V. Kodzasov. Sopostavitel’noe a structure) izuchenie dagestanskikh jazykov, imja. fonetika. Moscow: d. /w na-pq/ house-frame, ‘framework of a house’ Moscow University Press, 1990. Schultze, Wolfgang. Tsakhur. Munich: Lincom-Europa, 1997. e. /warad -pq/ song-frame, ‘melody’ Tschenkeli, Kita. Einführung in die Georgische Sprache, band 1, theoretischer Teil. Zurich: Amirani Verlag, 1958. It is the verb in these languages that is one of the won- Tuite, Kevin. Svan. Munich: Lincom-Europa, 1997. ders of morphology. It can bear morphemes expressing Van den Berg, Helma. A Grammar of Hunzib. Munich: Lincom- (roughly from start to end) the absolutive noun, the Europa, 1995. speaker’s interest in the action or incredulity, as well JOHN COLARUSSO 383 GERMAN German German is one of the main cultural languages of the German is spoken, runs eastward from Aachen, south Western world, spoken by approximately 100 million of Düsseldorf, Kassel, Magdeburg, and Berlin, to people around the world. It is the supraregional stan- Frankfurt-an-der-Oder. Although dialectal differences dard of a connected language area in Central Europe within both the High German and Low German and the national language of both Germany (77 mil- regions remain, a trend toward uniformity in the direc- lion) and Austria (8 million), and it is one of the four tion of the written standard is expected partly as a official languages of Switzerland, spoken by three result of widespread broadcasting, diminishing isola- fourths of the population (4.5 million speakers). tion, and increased socioeconomic mobility. Additionally, it is spoken in Alsace-Lorraine in eastern High German is the standard written language, even France, in the region of Alto Adige in Italy, and also in in the regions where Low German is more commonly eastern Belgium, Luxembourg, and Liechtenstein. spoken; High German is also the standard used in the German forms a continuum from Switzerland north to schools of Austria and Germany, taught to foreigners, the sea; a local dialect can be understood by speakers and used in the worlds of business and communica- of nearby dialects, but not necessarily by speakers of tions. High German is divided into Upper German, in far-away dialects. Switzerland, Austria, Liechtenstein, and southern Reliable statistics are not available concerning the Germany, and Middle German, across Luxembourg number of German-speaking persons who inhabit and the middle of Germany. those regions of Eastern Europe from which the Historically speaking, German falls into three main Germans were expelled at the end of World War II: periods: Old German (750 CE–1050 CE), Middle regions of the former Soviet Union (especially German (1050–1500), and Modern German (1500 up Ukraine), Hungary, the Czech Republic, Slovakia, to the present). The earliest existing records in German Poland, and Romania (all of them former regions of date back to about 750 CE. At that time, the language the Austro-Hungarian Empire). Outside Europe, the was developing in the Kingdom of Franks, especially largest number of people using German as their moth- in its eastern parts where the population was formed er tongue live in the United States (1.5 million). An mainly of Germanic tribes. Dialects of this region important group of German-speaking people in the formed a single language around the eighth century. United States speak the so-called Pennsylvania Dutch. Old German is characterized by the use of local German is also spoken in Australia, Canada, dialects in writing and by the absence of a standard Argentina, Brazil, Chile, and in former German language. During the Middle German period, a rela- colonies (Namibia, Congo, Cameroon). German is tively uniform written language developed in the gov- important as a cultural and commercial second lan- ernment: during the reign of Louis IV (1314–1347), guage for millions of people in Central, Northern, and the Holy Roman emperor, German was adopted as the Eastern Europe and in North and South America. language of official court documents, instead of the Latin that until then was dominating official writings. Middle German had additional syllable vowels and a History of the German Language more analytic syntax. Besides, it was the first common German is a member of the Western Germanic family. language of the people that was the language of chival- A descendant of the Old High German language, ric poetry. Between 1480 and 1500, it was introduced German is a mixture of dialects that, following the so- for official use in many municipalities and in the called ‘High German Consonant Shift’ (sixth to courts of Saxony and Meissen and was also adopted eleventh centuries) that affected the southern areas, by the universities of Leipzig and Wittenberg. By were divided into Low German (Niederdeutsch, 1500, German had generally been accepted as the offi- Plattdeutsch), spoken in the lowlands of the North of cial language in all parts of Saxony and Thuringia. It Germany, and High German (Hochdeutsch), spoken in was this German that Martin Luther, often referred to the highlands of the South of Germany and Austria. as ‘the creator of New High German’, adapted for his The term ‘Low German’ is essentially a geographic translation of the Bible in 1523. His translation had one, referring to the coastal, or lowland, area of the such a wide distribution and acceptance that the East German region, as opposed to the High German area. Middle German dialect in which it had been written The usually cited dividing line, south of which High came to serve as the authoritative exemplification of 384  GERMAN modern High German. In addition, the publication of (High German hoffen, Low German hopen, ‘to hope’). books in German developed in the East Middle Another consequence is the transformation of the German towns of Wittenberg, Erfurt, and Leipzig, as Germanic t in z (Pflanze) or ss (High German essen, well as in such western and southwestern cities as Low German eten, ‘to eat’). After vowels, k became ch Mainz, Strassburg, Basel, Nuremberg, and Augsburg. (High German machen, Low German maken, ‘to These developments helped reduce regional differ- make’); in all other cases, k remained unchanged ences and standardize the literary language. except in the extreme south of Germany, where it first Standard written German emerged during the first became kch, and later ch. A later change, also found in quarter of the sixteenth century in the eastern midland Low German, is that of the Germanic th to d (High area of Erfurt, Meissen, Dresden, and Leipzig, where German das, Low German dat, ‘that’). the inhabitants, originally from regions farther west and southwest, spoke a dialect based on the Middle Characteristics of German and Upper German dialects of High German. Despite the progress made in the sixteenth and seventeenth Another characteristic of German, as well as of all the centuries to develop and maintain a Nationalsprache Germanic languages, is that the principal accent falls (national language), French still continued to be a con- regularly upon the first syllable of a word; in verbal siderable influence on the German language because it combinations, however, the root syllable, not the pre- was the language of the German courts. During the fix, is stressed (see examples below). reign of France’s ‘sun king’, Louis XIV (1643–1715), The German language makes extensive use of inflec- it seemed that French was very much in fashion with tional endings, both for nouns and verbs. However, Old the upper and middle classes. Not surprisingly, never- High German was much more inflected than Modern theless, several societies began to develop a tendency German. Before the tenth century, its speakers used against the influence of foreign languages on German, numerous endings: the noun could vary in gender (three more particularly dedicated to the preservation and genders), case (five cases), and number. Over time, this purification of the German language. complex system of inflections was simplified. By the eighteenth century, German-speaking lands Modern German nouns are inflected for case and had generally adopted a written language, even though number, but not for gender. Of 43 nominal inflections common spoken language continued to be marked by that existed in Old High German, only nine survived in local and regional variations. the modern language. Still, the German language As for Low German, its history started in the thir- remains much more inflective than its relatives teenth century as the heir of Old Saxon (Old Low English or Dutch. German). Formerly a flourishing language, spoken in German nouns are classified according to three gen- the rich cities of Hansa, it later suffered the strength- ders (masculine, feminine, and neuter). The gender is ening influence of High German. Nowadays, it is used most easily identified by the noun’s definite article in mostly in rural areas, although some revival has been the nominative case: der (masculine), die (feminine and taking place for the last 50 years in the universities of plural), and das (neuter). Sometimes, the grammatical Northern Germany. The Low German vocabulary has gender of a noun has little to do with biological gender: generally remained much more archaic. In the last two die Sonne ‘the sun’ is feminine, but der Mond ‘the centuries, there have been considerable loanwords of moon’ is masculine. There is a gender distinction for High German origin, especially in the field of culture German nouns in the singular, but there is no gender and industry. distinction in the plural. All plural nouns (in the nomi- The history of the German language was affected native and accusative) take the definite article die. by several systematic shifts of certain consonants. The Nouns can appear in four cases: nominative (indi- so-called Germanic consonant shift distinguished the cating the subject of the sentence), accusative (the ancient Proto-Germanic language from other Indo- object), dative (the indirect object), and genitive (the European members. In this shift, described by possessor of any other noun). The last three cases are Grimm’s Law, Indo-European p, t, k changed to also used after prepositions, according to very specif- Germanic f, th, h, respectively; Indo-European b, d, g ic German grammar rules. This case system allows for to Germanic p, t, k; and similarly Indo-European bh, a freer word order, so that subject, object, and indirect dh, gh to Germanic b, d, g. After Western Germanic object can be placed in any order in the sentence. This had developed its own distinctive traits, the High flexibility of word order also allows for word play and German sound shift occurred (500–700 CE): the greater poetic license. Germanic p became pf under certain conditions (High The function of a noun in a German sentence deter- German Pflanze, Low German Plante, ‘plant’); when mines its case, which, together with its gender, deter- used medially or finally after vowels it became ff or f mines the article, and the endings on modifiers and 385 GERMAN adjectives preceding the noun (or on the noun itself): The only tenses the German verb can form by itself e.g. Der junge Mann sitzt hier. ‘The young man is sit- are Präsens (present tense), Präterium (past tense), ting here.’ (nominative); Kennen Sie den jungen Konjunktiv I (subjunctive I), and Konjunktiv II (sub- Mann? ‘Do you know the young man?’ (accusative); junctive II). In all other tenses, one of the verbs haben Ich gebe es dem jungen Mann. ‘I give it to the young (to have), sein (to be), werden (to become, to get) is man.’ (dative); Wie ist der Name des jungen Mannes? used as an auxiliary verb, in conjunction with the ‘What is the young man’s name?’ (genitive). Partizip II (past participle) or the Infinitiv (infinitive) There are seven types of pronouns in German: per- form of the principal verb. Generally, German uses the sonal pronouns (e.g. ich ‘I’, wir ‘we’, etc.), reflexive verb werden ‘to become’ to form the future tense, the pronouns (e.g. Sie wäscht sich. ‘She washes herself’), verbs haben ‘to have’, and sein ‘to be’ to form the per- demonstrative pronouns (with the same form as the fect tenses. The verb haben is generally used with definite articles), indefinite pronouns (e.g. viele transitive verbs and sein with certain intransitive ‘many’, etwas ‘something’, etc.), relative pronouns verbs. Nearly all of the verbs conjugated with sein are and interrogative pronouns (e.g. wer ‘who’, wen strong verbs or irregular. ‘whom’, wessen ‘whose’, etc.), and intensive pro- There are six modal auxiliary verbs in German: dür- nouns. fen ‘to be permitted, may’, müssen ‘to have to, must’, The proper German personal pronoun for the second können ‘to be able, can’, mögen ‘to like to’, wollen ‘to person depends upon the closeness to the person one is want to’, and sollen ‘to suppose to, ought to’. addressing: i.e. Sie and its related forms are used when German has a large number of phrasal verbs made addressing one or several persons whom you would up of a verb (e.g. kommen ‘come’) and one or two par- normally call by the last name, while du and its related ticles or prefixes (e.g. an ‘at’). These particles and forms are used when addressing one person (ihr more their verbs have the force of single-word verbs (e.g. than one person) whom you would normally call by the ankommen ‘arrive’). Many of these particles attached first name: e.g. Herr Schmidt, kennen Sie meine Frau? to verbs are also ordinary prepositions: ab ‘away ‘Mr. Schmidt, do you know my wife?’, Robert, hast du from’, ein ‘into’, vor ‘before’, etc. They provide addi- meinen Hut? ‘Robert, do you have my hat?’, Kinder, tional (or sometimes changed) meaning to the verb. ihr seid zu laut! ‘Children, you are too loud!’ Occasionally, some of these prefixes become separat- Depending upon its context, the personal pronoun sie ed from the verb in clauses. In a simple sentence, these can have various meanings. When it is capitalized and prefixes are separated from the verb at the end of the does not begin a sentence, it always means ‘you’ (sin- clause: Sie kommen nun an. ‘They are arriving now.’ gular or plural in a formal way): e.g. Was meinen Sie They join the verb in the past tense: Wir sind angekom- dazu? ‘What do you think of that? When it is not capi- men. ‘We have arrived.’ When an infinitive is used or talized and appears as the subject of a singular verb, it if the verb appears at the end of a subordinate clause, means ‘she’ or ‘it’ if it refers to an object of feminine the prefix remains attached: Wir werden ankommen. grammatical gender: e.g. Was meint sie dazu? ‘What ‘We will arrive.’ does she think of that?’ When it is not capitalized and Certain verbs have inseparable prefixes that are not appears as the subject of a plural verb, it means ‘they’. used alone, have no fixed meaning alone, and are When it is used as the direct object of a verb, it can never separated from the verbs they are accompany- mean ‘her’ or ‘it’ (if it refers to an object of feminine ing. An example would be be in bekommen ‘get.’ grammatical gender), or ‘them’: e.g. Ich fahre sie durch Separable prefixes have intonational stress, as in die Stadt. ‘I drive her (or them) through the town.’ ankommen, but prefixes are never stressed, as in There were several changes during the history of bekommen. the German verbal system. During the Old High Adjectives and adverbs have the same form in German period, the verb could take many different German. The context in which they are used indicates forms. Nowadays, the number of main forms of the how to interpret them: Dieser Käsekuchen ist sehr gut. verb is reduced to three: helfen - half - geholfen ‘help ‘This cheesecake is very good’ vs. Sie bäckt den - helped - helped’ instead of Middle High German Kuchen sehr gut. ‘She bakes the cake very well.’ helfen -half - hulfen -geholfen. The major word classes are interconnected by a There are four types of verb in German: Hilfsverben complex network of word-formation rules. A distinc- (auxiliary verbs), Modalverben (modal auxiliary tive feature of German is its extensive use of lengthy verbs), regelmäßige Verben (regular verbs), and compound words. For example, the German counter- unregelmäßige Verben (irregular verbs). There are 173 part of the English ‘history of antiquity’ is irregular verbs in Modern German, but this class is Altertumswissenschaft. decreasing in number. However, it still contains many Traditionally, German was written in a Gothic or verbs belonging to the basic vocabulary. Black Letter style of the Roman alphabet, known in 386  GERMANY German as Fraktur (meaning ‘fractured’), which dates typical folk literature and its speakers are all bilingual, from the twelfth century. In the period following as the official language is English. World War II, however, Fraktur was largely supersed- ed by the Roman characters exclusively used through- References out the rest of Western Europe. The Roman script contains the symbol ß or double s, which is used only Beedham, Christopher. 1995. German linguistics: an introduction. Munchen: Iudicium. in the lower case and represents a voiceless s (as in Engel, U. 1988. Deutsche Grammatik. Heidelberg, Julius English mouse), and the umlauted vowels ä, ö, and ü. Groos. German is the only language in which all nouns are Hammond, Robin. 1981. A German reference grammar. capitalized, common as well as proper (as they were in Oxford: Oxford University Press. English several hundred years ago). This makes the Keller, R.E. 1978. The German language. London: Faber. Lippi-Green, Rosina L., and Joseph C. Salmons (eds.) nouns easy to spot when reading a sentence. Verbs and 1996. Germanic linguistics. Amsterdam, Philadelphia: adjectives functioning as nouns are also capitalized. John Benjamins Publishing Company. Russ, C.V.J. (ed.) The dialects of modern German. A linguistic survey. London: Routledge. Pennsylvania German Russ, Charles C.V. 1994. The German language today: a linguistic introduction. London, New York : Routledge. Pennsylvania German, also known as ‘Pennsylvania Stevenson, Patrick (ed.) 1995. The German language and Dutch’ or simply ‘Dutch’, is spoken by 300,000 native the real world: sociolinguistic, cultural, and pragmat- speakers, descendants of German colonists, who left ic perspectives on contemporary German. Oxford: the Palatinate region of Germany during the late sev- Clarendon Press. enteenth and eighteenth centuries and settled in the Waterman, J.T. 1966. The history of the German language. Prospect Heights; IL: Waveland Press, Inc. southeastern part of Pennsylvania in the United States. LAURA DANILIUC They speak the Rhine-Franconian dialect with rela- tively few admixtures of English. It is the vehicle of a See also Germany Germany Germany, or more accurately the Federal Republic of substantially different regional spoken language Germany (Bundesrepublik Deutschland), consists of 16 forms. Such regional variation was a result of the High states (Bundesländer), five of which constituted the German Sound Shift (also termed the Second Sound German Democratic Republic (East Germany) until Shift), which had taken place to differing degrees in reunification in 1990. Germany is situated in Central different parts of the German-speaking community Europe, with the Baltic Sea, the North Sea, and from the sixth to eighth centuries CE. In this sound Denmark to the north, Poland and the Czech Republic to shift, the voiceless stops /p/, /t/, /k/ changed to voice- the east, Austria and Switzerland to the south, and less fricatives /f/, /s/, /x/ and affricates /pf/, /ts/, /kx/, France, Luxembourg, Belgium, and the Netherlands to and the voiced stops /b/, /d/, /g/ to voiceless stops /p/, the west. It has an area of 357,020 square kilometers and /t/, /k/. The dialects in the south of Germany, termed a total population of 82,536,700 (as of December 2002). Upper German (Oberdeutsch) in keeping with the The official language of the country is German topography of the area, were affected first and most Standard German (as opposed to e.g. Austrian Standard completely. Gradually, the shift moved northward. It German). The development of this supraregional writ- became weaker as it did so, however, and the dialects ten norm is a relatively recent phenomenon and a good in the center of Germany (Central German, example of language planning, i.e. the use of deliber- Mitteldeutsch) were only partially affected, while ate, often official, measures relating to the status and/or those in the north (Low German, Niederdeutsch) form of a language. At the end of the Middle High remained unaffected. German period (c. 1050–c. 1350), the written standard During the Early New High German (Early was still Latin, and the orthography, grammar, and Modern) period (c. 1350–c. 1750), however, the func- vocabulary of the German texts of that time were not tional motivation for a standard increased, because of standardized but rather were strongly influenced by the development of towns and cities, the increase in 387 GERMANY secular education, the invention of the printing press, German is generally less than that of middle-class and, in particular, because of Martin Luther’s transla- children. tion of the Bible into a variety of German intended to Adherence to the ‘stem principle’ is one of the rules be understood by all speakers of German introduced with the recent reform. Unlike in the old (1522–1534). Aside from these influences, a conscious spelling, all consonants are to be maintained in the drive was made by grammarians and lexicographers in spelling of compound words, even in cases where the sixteenth century toward standardization, so that three identical consonants may come together. The this process is thought to have been largely complete word for ‘ballet dancer’, a compound of Ballett ‘bal- by the mid-eighteenth century. The variety of German let’ and Tänzer ‘dancer’, thus becomes Balletttänzer that ultimately served as the basis for the development of with three ts rather than the former form, Ballettänzer, written Standard German (High German, Hochdeutsch) with two ts. Also, the decision as to whether to spell was based on East Central German. Without doubt, the words apart or together was regularized. Rather than adoption of the standard was gradual, so that Charles writing radfahren ‘to ride a bicycle’ but Auto fahren Ferguson’s concept of ‘diglossia’, i.e. a situation in ‘to drive a car’, for example, both are now written which two varieties have mutually exclusive functions apart, as Rad fahren and Auto fahren. within a particular speech community, was relevant in The reform process, which began in the 1950s, was the period following the introduction of the standard: a lengthy and complex one, clearly exemplifying the speakers at that time would have used their local difficulties of language planning. Not only was inter- dialect (the Low variety) in the private domain of state and international consensus concerning the home and friends and the standard variety (the High orthographic norm difficult to reach, but public sup- variety) in the public domain. port was also weak. It was not until several attempts at The strict regulation of German orthography dates reform had been rejected that a list of recommenda- back to the mid-nineteenth century. Prior to this, there tions (drawn up by a group of representatives from existed no unified spelling. Indeed, until a decree in Germany, Austria, and Switzerland as well as 1862 ruled otherwise, various spellings could even be Belgium, France (Alsace), Hungary, Italy (South taught in a single school. The first significant attempts Tyrol), Liechtenstein, and Romania) was accepted in to standardize German orthography took place after the July 1996 at an orthography conference in Vienna. establishment of the German Reich (1871) at the These new spelling rules have represented the official Verhandlungen der zur Herstellung größerer Einigung norm since August 1, 1998. Their implementation and in der deutschen Rechtschreibung berufenen Konferenz review is the task of the Institut für deutsche Sprache (Conference on the Establishment of Greater in Mannheim. The initial grace period during which Standardization of German Orthography) in Berlin in both the old and new spelling may be used ends on 1876. The proposals of this conference, however, were July 31, 2005. never enforced. The process nevertheless continued, In public opinion, the spelling reform met with gen- and at the second national orthography conference, eral rejection. Indeed, the state of Schleswig-Holstein, held in Berlin in 1901–1902, a set of orthographic in a referendum on September 27, 1998, voted not to rules was developed on a trial basis for use in accept the reform. The reasons for the lack of public Germany. These were also adopted by Austria and support were varied, although many involved emotion, Switzerland. The seventh edition of Konrad Duden’s laziness, or ignorance. Many individuals were reluc- orthographic dictionary (Orthographisches Wörterbuch tant to change since they identified with the old der deutschen Sprache, Orthographic dictionary of the spelling, were not prepared to learn the new rules, or German language), produced in 1902, incorporated mistakenly equated language with spelling, believing these rules, and in December 1955, the education and that such planning would disturb the natural progres- arts ministers of the states of the Federal Republic of sion of a natural language. Other reasons involved the Germany awarded the Duden legal status, declaring it belief that the costs of the reform were not justified the orthographic norm for German. The rules of the since the reform did not go far enough, representing a second national orthography conference served as the mere tidying-up process. orthographic norm for German until August 1998, As for the standardization of German pronuncia- when a spelling reform was instituted. The central tion, it was not until 1898 that the norm, codified by motive for the reform process, which aimed at stan- Theodor Siebs and based on the pronunciation of dardization and simplification, was educational, since German Standard German in the north of Germany, it was argued that the time devoted to teaching spelling was established (Stage German, Bühnendeutsch). to children could be better spent otherwise and also However, although radio announcers from different that the complex spelling rules discriminated against parts of Germany may sound the same even today, working-class children whose exposure to written regional pronunciations of the standard variety based 388  GERMANY on the accents of the local nonstandard German standard written variety. Colloquial German itself can dialects reign, since Bühnendeutsch was never com- be situated near or far from the standard, depending on pletely adopted. Indeed, it is not only in pronunciation the region and the specific situational constellations. that German Standard German presently shows In the north of the country, for example, colloquial regional variation; German Standard German is col- speech is closer to the standard than in the south. ored with regional variation on every level of lan- Germany is a multilingual country, and although guage, except the strictly regulated orthographic level. German is the principal language, many other lan- On the level of vocabulary, for example, the word guages are also spoken in the Federal Republic. While Samstag is commonly used in the south and west of many of these, such as Frisian, Sorbian, and Danish, the country for ‘Saturday’, whereas Sonnabend is gen- are long-established minority languages, others, such erally used in the north and east. Indeed, in recent as Arabic, Greek, Turkish, Hindi, Italian, and Spanish, years, the trend toward regionalism has become were only introduced in the late 1950s via the immi- increasingly evident. gration of guest workers (Gastarbeiter). In more Despite standardization, the dialects of German are recent years, there has also been immigration of indi- many and reflect the uneven effect of the aforemen- viduals of German descent (Aussiedler) from Eastern tioned Second Sound Shift in various parts of Germany. countries. In 1999, 7.2% of the population were not of The ‘Benrath line’ is an isogloss (i.e. a line drawn on a German nationality. linguistic map to locate the outer limits of some charac- Frisian is the West Germanic language most close- teristic feature) that runs from Aachen in the west ly related to English. Today, two mutually unintelligi- through Benrath, a suburb of Düsseldorf, to Frankfurt ble main varieties are spoken in Germany: North an der Oder in the east, dividing the country into two Frisian and East Frisian. East Frisian, or more specifi- regions. To the north are found Low German dialects cally Saterland Frisian, is still spoken in the Saterland, not affected by the sound shift, and to the south, High west of Oldenburg in East Frisia in the north of German dialects affected by the shift. Dialects from the Germany (fewer than 1,000 speakers). It is being dis- very north and very south of this divide are virtually placed by German. The same is also true of North mutually incomprehensible. In the transition zone in the Frisian (approx. 8,000–9,000 speakers), which itself middle (Central German), partial traces of the shift are includes nine very different, partially mutually incom- found. Within the divide into Low, Central, and Upper prehensible, dialects spoken on the west coast of German, further divisions are made based on particular Schleswig-Holstein north of Husum and on the off- linguistic criteria. The accompanying map provides a shore islands of Föhr, Amrum, Sylt, Nordstrand, sketch of some of the principal differentiations. It must Pellworm, the ten islands of the Halligen group, and be pointed out, however, that none of these dialect areas Heligoland. Reasons for the dwindling numbers of are as homogeneous or as strictly defined as they appear Frisian speakers are improved communication, the (see Figure 1). centralization of the education system, the large num- Standardization has led to an overall decline in the ber of dialects of the language, and the fact that no use of local dialects, and not all are equally healthy, standard written or spoken form exists. A form of col- despite a general trend toward regionalism. In the loquial language close to German Standard German is south, in Bavaria, Baden-Württemberg, and to a lesser thus favored. Efforts to save the Frisian language from degree in Hesse and parts of the Rhineland, dialects extinction have generally been rather insignificant. function as markers of regional identity to a substan- While the people of the tourist island of Sylt manage tial extent and are therefore relatively healthy, being to nurture their identity and speak Frisian, speakers in used fairly far up the social ladder. In the north, how- other areas on the mainland have a more difficult task ever, their use is more socially stigmatized. Not sur- since they do not form an isolated group. In addition, prisingly, therefore, German Standard German has many speakers wish to discard the ‘socially backward already replaced many of the Lower German dialects and poor’ image commonly associated with speakers of the north. Diglossia is rare in Germany, except in of Frisian. Since the late 1980s, political engagement the case of elderly people living in rural parts of the for the Frisian language has increased; money has Upper German dialect region and in remote rural areas been made available to teach Frisian on a voluntary in the Low German and Central German regions. basis in schools, for example. While such efforts may Instead, most speakers today commonly shift their serve to delay the extinction of Frisian, it is doubtful choice of variety along a continuum between the poles whether they will prevent it. of dialect and standard, with most everyday spoken Standard Danish is spoken by a recognized minori- German being of a colloquial nature (Umgangssprache), ty in southern Schleswig. Despite the existence of a relatively new level of language that has developed schools there in which Danish is spoken, the future for in urban areas due primarily to the presence of a the language in this area is also rather pessimistic. 389 GERMANY Hamburg East Low German West Low German Berlin Franconian Low East Cologne Central German Dresden West Central German East Franconian Frankfurt (North Upper German) Nuremburg ) an Eas n nic erm t Up p a G (Ba er Ger em r Munich vari m Al ppe an) an Basle t U es (W Austria Switzerland Italy Slovania Figure 1. German Dialects. Since the seventh century, the Slavic language example, a foundation was established for the Sorbian Sorbian (approx. 67,000 speakers) has been spoken in people (Stiftung für das sorbische Volk) with the task Lusatia (Lausitz), a relatively vaguely defined area in of promoting the language and culture. However, former East Germany bordering on Poland and the despite such efforts, the survival of the Sorbian lan- Czech Republic and belonging administratively to the guage, although the least endangered of the estab- German states of Brandenburg and Saxony. Two vari- lished minority languages in Germany, is uncertain eties, Lower Sorbian and Upper Sorbian, are differen- since all speakers are bilingual in German, the tight tiated. Lower Sorbian is spoken in the area around bonds that existed between Sorbians in former times Cottbus (Chos´ebuz) in Lower Lusatia and Upper have dissolved in the meantime, and many young Sorbian in the area around Bautzen (Budyšin) in Upper Sorbians living in German-speaking areas do not iden- Lusatia. The Sorbians themselves never formed their tify with their language. own state but have remained under German rule. After As for foreign language teaching in German a period of repression during the Nazi era, Sorbian was schools, not surprisingly, English, given its interna- promoted in the German Democratic Republic. Its use tional status, is primary. in local government was authorized and children had Gastarbeiter ‘guest workers’, also and more appro- the right to learn it in schools, although in reality it priately termed migrants, immigrants, work immi- was often not possible to practice such rights. German grants, and foreign workers, came to the Federal reunification brought generous support for the lan- Republic primarily from countries such as Greece, guage and culture of the Sorbian minority. In 1991, for Italy, Turkey, and the former Yugoslavia from the late 390  GERMANY 1950s onward, following a policy decision designed to in Germany, and the subjective social and psychologi- ameliorate the acute labor shortage in West Germany cal distance between the individual and native speak- at the time. They settled mostly in urban areas such as ers of German. Munich, Stuttgart, and Frankfurt am Main, often in Guest Worker German is also of interest to concentrated ethnic groups (e.g. Turks in West Berlin researchers of pidgins, i.e. languages without native (Kreuzberg), Greeks in Munich, Yugoslavs in speakers that arise out of a basic need to find a lingua Stuttgart). The early 1970s witnessed an end to this franca (a common system of communication), since it policy, and immigration laws became stricter. With shares many features of pidgins, even though it arises changing economic circumstances, these ‘guest’ work- as a result of communication between migrant workers ers were encouraged to return to their home countries and German native speakers rather than between from the early 1980s onward. But many have migrant workers of various cultures, the usual occa- remained, settled down, and brought relatives to sion for the development of a pidgin. Germany. Unfortunately, this group leads a marginal Finally, individuals of German descent (Aussiedler) existence in Germany, being denied dual citizenship, from Eastern Europe, primarily from Romania, for example, on the basis of policy assumptions that Poland, and the former Soviet Union, have been immi- this workforce was in Germany temporarily and that grating to Germany since the 1950s. However, such Germany is not an immigrant country. immigration, particularly from the former Soviet Most first-generation foreign workers are not profi- Union, increased dramatically between the fall of the cient in German despite having lived in the country for Iron Curtain and 1995; since then, numbers have been a number of decades. The majority of them had not steadily decreasing. While some of those from the for- enjoyed any instruction in German prior to their arrival mer Soviet Union and Romania speak a variety of and, since integration was not foreseen, few have ever German, others from Russia speak an archaic dialect been taught German systematically. With the excep- of German and can understand but not speak German tion of much of the Turkish population, though, many Standard German. Those from Poland, on the other second-generation immigrants are bilingual, with hand, have little or no competence in German. The German as the dominant language. German of these speakers shares many similarities The linguistic study of the German spoken by the with Guest Worker German. migrant workers has provided a number of insights In sum, the study of the linguistic situation in the into the process of uninstructed second-language Federal Republic of Germany is a rewarding endeavor. acquisition. It has been demonstrated, for instance, The study of the standardization of German Standard that although speakers do transfer features from their German sheds light on the difficult process of lan- native language to their second language, the German guage planning, and the study of minority languages of speakers of languages as different as Spanish and highlights the possible consequences of language con- Italian (Romance languages) and Croatian and tact, the issue of language shift, and, in the case of the Slovenian (Slavic languages) is characterized by simi- immigrant languages, aspects of uninstructed second- lar simplifying measures, such as deletion and reduc- language acquisition. tion. A typical example of the former is the utterance Ich fahre zwei Wochen Irland ‘I’m going two weeks Ireland’, in which two prepositions have been delet- References ed—the utterance should read Ich fahre in/ für zwei Ammon, Ulrich. 1995. Die deutsche Sprache in Deutschland, Wochen nach Irland ‘I’m going to Ireland in/for two Österreich und der Schweiz. Das Problem der nationalen weeks’. An example of reduction is the use of the Varietäten. Berlin and New York: de Gruyter. Barbour, Stephen, and Patrick Stevenson. 1990. Variation in informal ‘du’ whatever the formality of the situation. German. A critical approach to German sociolinguistics. Given such similarities between speakers of different Cambridge: Cambridge University Press. native languages, it is possible to speak of the exis- Clahsen, Harald, Jürgen M. Meisel, and Manfred Pienemann. tence of Guest Worker German (Gastarbeiterdeutsch). 1983. Deutsch als Zweitsprache. Der Spracherwerb Considerable variation is also present, though. An ausländischer Arbeiter. Tübingen: Narr. Clyne, Michael G. 1995. The German language in a changing order of acquisition of grammatical structures has also Europe. Cambridge and New York: Cambridge University been established based on the analysis of Guest Press. Worker German. How far a particular individual pro- Duden, Konrad. 1880. Vollständiges Orthographisches gresses along the path of acquisition has been found to Wörterbuch der deutschen Sprache (= Duden), 1st edition. depend on a number of factors, the most important Leipzig: Bibliographisches Institut. Duden, Konrad. 1902. Orthographisches Wörterbuch der being the degree and intensity of contact with deutschen Sprache (=Duden), 7th edition. Leipzig: Germans, age on arrival, level of education acquired in Bibliographisches Institut. home country, years of schooling, period of residence Ferguson, Charles A. 1959. Diglossia. Word 15. 391 GERMANY Heller, Klaus. Rechtschreibreform—Eine Zusammenfassung Sauer, Wolfgang W., and Helmut Glück. 1995. Norms and [HTML document], IDS (Institut für deutsche Sprache) reforms: fixing the form of the language. The German lan- Sprachreport Extraausgabe (July 1996), Retrieved guage and the real world: sociolinguistic, cultural, and prag- [4.05.2001] from the World Wide Web: http://www.ids- matic perspectives on contemporary German, ed. by Patrick mannheim.de/pub/sprachreport/reform/. Stevenson. Oxford and New York: Clarendon Press. Langer, Nils. The Rechtschreibreform—a lesson in linguistic Stevenson, Patrick. 2002. Language and German disunity. A purism [HTML document], GFL: German as a foreign lan- sociolinguistic history of East and West in Germany guage [online] 3 (2000), Retrieved [4.05.2000] from the 1945–2000. Oxford: Oxford University Press. World Wide Web: http://www.gfl-journal.com/. ANNE BARRON Russ, Charles V.J. (ed.) 1990. The dialects of modern German: a linguistic survey. London: Routledge. See also Acquisition Theories; Diglossia; German; Russ, Charles V.J. 1994. The German language today: a lin- Language: Contact—Overview; Language Death; guistic introduction. London and New York: Routledge. Language Planning; Standard Language Givón, Talmy Talmy Givón’s outstanding contribution to the study of Considering antimodularity in the first place, it must language in a wide scientific setting cannot be under- be borne in mind that generative grammars, including stood without the influence of two leading figures of formal approaches like Chomskian Transformational North-American linguistics: Dwight Bolinger and Grammar and functional approaches like Simon Dik Joseph Greenberg. Like his teacher Bolinger, Givón has and Kees Hengeveld´s Functional Grammar, consider followed the central claim of the functional tradition of grammar as a set of descriptive devices in such a way linguistics, which ultimately can be traced back to the that several components or modules turn out different Prague School of Linguistics and European functional- aspects of the final linguistic expression. The output of ism, namely that language cannot be studied independ- each module constitutes the input to the next module: ently of its context. Like Greenberg, Givón has been typically, the lexicon turns out lexical items, the com- concerned with describing cross-linguistic phenomena ponent of predication-formation fills in the semantic- and finding functional explanations for them. syntactic slots of the clause with lexical items, thus It is significant that among the many influences that turning out a sentence, and a morphophonological com- Givón acknowledges in his works, which include fel- ponent gives the sentence its final (pronounced) form. low functionalists like Wallace Chafe, Bernard Cognitive plausibility implies that it is not enough Comrie, Robert Dixon, John Haiman, Bernd Heine, to propose descriptive devices such as rules or opera- Paul Hopper, and Sandra Thompson as well as Noam tions, or even explanatory devices like specific or gen- Chomsky, two teachers of Givón´s are mentioned with eral principles. Much more than that, it is necessary to particular warmth and respect. The first is Bolinger, to consider the plausibility of grammatical devices in the whom Givón attributes an insistence that the form of light of human abilities of memory and processing. language cannot be studied in isolation from its mean- For example, sentence elements or constituents bear- ing, that grammar is a set of meaningful devices, and ing contrastive information, which typically occupy a that the aim of the forms of language is to express clause-initial position in English, can be inserted into thought. The other—rather unexpected—is a high the correct slot by a rule that places contrastive focus school teacher of humanities who eventually became a constituents in the clause-initial position; or as a result reputed Egyptologist, Sarah Groll. Since Givón con- of the application of a functional principle that stipu- siders Groll´s teachings a formative experience that lates that more important information is more likely to has lasted him a lifetime, it does not seem out of place be placed earlier in the clause. For Givón, the cognitive to put down the interest in cultural and linguistic basis of this functional principle is that the earlier a diversity shown by this author to the influence of chunk of information is placed within its relevant unit, Groll. the more attention it attracts. Moreover, the informa- Along with these influences, three main principles tion that attracts more attention is memorized, stored, could be said to characterize Givón´s work: the princi- and retrieved more efficiently. Another point of interest ple of antimodularity, the principle of cognitive plau- related to cognitive plausibility is Givón´s epistemolo- sibility, and the principle of linguistic relativity. gy: for him, Western Philosophy cannot be reduced to 392  GOTHIC Plato and Aristotle. Kant, Peirce, and Wittgenstein also the UCLA Linguistics Department (1968–1981) and shed light on the social and communicative dimensions the University of Oregon Linguistics Department of language and, more importantly, on language cate- (1981–2002). He was Chair, Linguistics Department, gorization, which is not discrete but continuous, that is, University of Oregon (1981–1985) and Director, based on the notion of prototype of a category. Cognitive Science Program, University of Oregon Givón follows in Franz Boas and Edward Sapir´s (1986–1987). He held visiting and consulting positions footsteps in the question of linguistic relativity, advo- in Papua-New Guinea, New Zealand, Germany, Perú, cating the study of each language on its own terms. Spain, and Japan. He was also Founding Editor of The ideological basis of linguistic relativity is even Studies in African Linguistics and Typological Studies more obvious in Givón than in the work of his prede- in Language and Distinguished Professor (Emeritus) cessors: in order to avoid Eurocentrism (the scientific of Linguistics and Cognitive Science, University of hegemony of Indo-European languages over lan- Oregon. guages belonging to other families), the conclusions reached after the study of English cannot be automat- References ically generalized to all languages independently of Bates, Elizabeth, and Brian MacWhinney. 1982. Functionalist their typological characteristics, and English cannot be approaches to grammar. Language acquisition: the state of the recurrent term of comparison for any linguistic the art, ed. by Eric Wanner and Lila Gleitman. Cambridge: topic. For this reason, Givón prefers a formally uncon- Cambridge University Press. strained grammar, which consists of functionally Butler, Christopher. 1991. Standards of adequacy in functional oriented principles, rather than a formally con- grammar. Journal of Linguistics 27. Butler, Christopher, Ricardo Mairal, Javier Martín Arista, and strained grammar involving language-specific rules Francisco Ruiz de Mendoza. 1999. Nuevas Perspectivas en whose typological validity is, at least, questionable. Gramática Funcional. Barcelona: Ariel. Functionally oriented principles are based on explana- Givón, Talmy. 1979. On understanding grammar. tory devices like iconicity (or resemblance between ———. 1984. Syntax: a functional-typological introduction, the linguistic code and the extra-linguistic reality) and Vol. I. ———. 1989. Mind, code and context: essays in pragmatics. markedness (or asymmetry between two members of a ———. 1990. Syntax: a functional-typological introduction, pair or one of the members of a cluster and the rest of Vol. II. the members of such a cluster). ———. 1993. English grammar: a function-based introduction. Givón has been a source of refreshing and thought- ———. 1995. Functionalism and grammar. provoking inspiration; without him, some recent Martín Arista, Javier. 1997. La adecuación psicológica de la gramática funcional tipológica. Cuadernos de Filología developments in functionalism would not be easy to Inglesa 6(2). explain. Newmeyer, Frederick J. 2001. The Prague School and North American functionalist approaches to syntax. Journal of Linguistics 37. Biography Nichols, Johanna. 1984. Functional theories of grammar. Annual Review of Anthropology 13. Talmy Givón was born in Afula, Palestine on June 22, 1936. He received his B.S. in Agriculture in 1959; JAVIER MARTÍN ARISTA M.S. in Horticulture in 1962; M.A. in Linguistics in See also Functional Approaches; Generative Gram- 1966; and Ph.D. in Linguistics in 1969. He taught at mar; Greenberg, Joseph Harold Gothic Gothic, the only documented member of the Eastern moved from Scandinavia, through the Baltic islands, Germanic branch of the Indo-European family, is an and to the Black Sea, where they divided into two extinct language once spoken by the Goths. The major groups around AD 270. Gothic historian Jordanes (AD 551) says that his peo- The West Goths—the Visigoths—invaded the terri- ple came originally from Scandinavia; place names, tory of the Roman Empire during the reign of Marcus including the island of Gotland, show their presence Aurelius (AD 270–275) and settled in the province of there around the first millennium BC. The Goths then Dacia (modern-day Romania) near the lower Danube 393 GOTHIC and then migrated to Gaul and Spain, where they and silver on purple velum. Discovered in the six- established a powerful kingdom that lasted for over teenth century, it is now in the care of Uppsala three to four centuries. Around AD 490 in University in Sweden. Constantinople, the East Goths—the Ostrogoths— Other Gothic texts are limited, but include the under the leadership of Theordoric established an Skeireins (eight unconnected pages of a commentary Ostrogothic kingdom in Italy that was subsequently on the Gospel of St. John), some deeds of sale, a frag- destroyed by the Byzantines in the middle of the sixth ment of a calendar of martyrs, the five Codices century. By the eighth century, both groups were Ambrosiani, and bilingual manuscripts like the assimilated into the larger populations who surround- Gothic-Latin Codex Carolinus. However, the Codex ed them. Argenteus is the earliest literary remains in any Proto-Gothic, the direct predecessor of Gothic and Germanic language; the only earlier records are limit- one of the descendants of Proto-Germanic, can only be ed to a few Norse runic inscriptions and some loan- studied through probable reconstructions of Proto- words preserved in a few non-Germanic languages. Gothic words. However, early Gothic literary records Historians generally agree that Wulfila created the that provide language samples for linguistic study are Gothic alphabet (which should not be confused with available and date from the fourth century, predating the Gothic script also known as Black Letter that is records of any of the other Germanic languages. used by printers) with the sole purpose of translating Gothic is the cornerstone of Germanic linguistics the Bible. The Gothic alphabet has 27 letters; the because it provides the oldest extensive collection of majority of the letters were derived from Greek, but written samples in the Germanic group of languages, five or six of the letters came from Latin, and two were except for a few brief Norse runic inscriptions. borrowed from runic script. In general, though, the Comparable specimens of other Germanic language, Gothic alphabet resembles Latin and Greek scripts like English, Frisian, High German, and Norse, do not with some differences in the order and sounds of the appear until four to nine centuries later. Very little is letters. Gothic is written from left to right with only known about the other Eastern Germanic languages of very rare spacing between letters. What we know Burgundian, Gepidic, Rugian, and Vandalic. about the pronunciation of Gothic is based on how the Gothic had two main dialects: Ostrogothic was the Greek and Latin foreign words and names in the language of the Ostrogoths in eastern Europe and Italy, Gothic Bible were transcribed. and Visigothic was the language of the Visigoths of Gothic, separated from other Germanic languages east central Europe, Gaul, and Spain. Some scholars early on, does not exhibit many of the linguistic devel- believe that the Ostrogothic dialect became extinct opments that appear in other Germanic languages in after the fall of the Ostrogoth kingdom in Italy in the later periods. Therefore, the phenomena that are sixth century; however, Ogier Ghislain de Busbecq, a unique to Gothic are considered important to all Flemish ambassador who served in Constantinople Germanic linguistics. between 1560 and 1562, found traces of Gothic in the One way in which Gothic can be distinguished words and phrases of the area. Even though his schol- from other members of the Germanic family is by its arship and transcription were less than ideal, this vari- use of the long [ê] sound (as in mênoþ ‘month’), a ety is thought to be a dialect of Ostrogothic and is now sound that is assumed to have been in Proto-Germanic known as Krimgotisch (Crimean Gothic). The and that disappeared in all other Germanic languages. Visigothic dialect is believed to have become extinct Another Gothic singularity is how it preserved the around the time of the Arab conquest of Spain in 711. original Proto-Germanic [z] sound (as in huzd ‘hoard’) The earliest Gothic literary work and the corner- that changed to [r] in all other Germanic languages. stone to all Germanic linguistics is the translation of Gothic nouns and pronouns had five Germanic the Bible by Wulfila (311–383), a Visigothic bishop cases: nominative, vocative, genitive, dative, and accu- from Moesia (modern Bulgaria). Wulfila’s original sative; however, some of the pronouns and adjectives translation of the Bible from Greek has not survived; also retained the instrumental case that was lost in however, several reproductions of his work from the other Germanic languages. Gothic nouns, pronouns, fifth and sixth centuries still exist. The Codex and adjectives had three genders (masculine, feminine, Argenteus, written in the first half of the sixth century and neuter) and three numbers (singular, dual, and plu- by Ostrogothic scribes in northern Italy, consists of ral). The dual number (a plural form expressing two translations of the gospels and epistles, providing persons or things) is an archaic feature that does not about three-quarters of the New Testament. A bound appear in later varieties of other Germanic languages. volume of manuscripts, the Codex Argenteus, the most Gothic also had a reflexive pronoun (seina, sis, sik studied of the Wulfila reproductions, is written in gold ‘self’) that can be traced back to Proto-Indo-European. 394  GRAMMAR, TRADITIONAL Some verb features specific to Gothic occur seldom, guþa (‘then-then-in-month-sixth-sent-was-angel- if at all, in the other Germanic languages that appear in Gabriel-from-God’), yielding the English translation later written samples. Germanic verbs, in general, can be ‘Then in the sixth month was sent the angel Gabriel divided into strong (regular) and weak (irregular) verbs, from God.’ and each of these can also be divided into subclasses. The seventh class of Gothic strong verbs is distinguished from other Germanic languages in that it uses the References process of reduplication (repetition of syllables) to form Bennett, William H. 1972. An introduction to the Gothic lan- the verbs. For example, the past tense of the verb háitan guage. New York: MLA. ‘to call’ is formed by repeating the first consonant of the Braune, Wilhelm, and Ernst A. Ebbinghaus. 1981. Gotische verb stem and attaching it after the vowel ái, resulting in Grammatik, 18th edition. Tübingen: Niemeyer. Krause, Wolfgang. 1968. Handbuch des Gotischen, 3rd edition. a form that attaches as a prefix to the basic stem to form Munich: Beck. haíháit ‘called’. Another verb feature unique to Gothic Mierow, Charles C. 1915. The Gothic history of Jordanes. are the retained passive inflectional suffixes for verbs; all Princeton, NJ: Princeton University Press. other Germanic languages use the periphrastic construc- Penzl, Herbert. 1972. Methoden der germanischen Linguistik. tion of auxiliary plus verb plus past participle (as in ‘is Tübingen: Niemeyer. Robinson, Orrin W. 2002. Old English and its closest relatives: called’ in English). Gothic uses passive inflectional suf- a survey of the earliest Germanic languages. Stanford, CA: fixes, as seen in the use of -ada to form the passive hái- Stanford University Press. tada ‘is called’ from the verb háitan ‘to call’. ALLISON SMITH A sample Gothic sentence reads as Þanuh þan in menoþ saihstin insandiþs was aggilus Gabriel fram See also Indo-European 2: Germanic Languages Grammar, Traditional Traditional grammar refers to the type of grammar The analytical study of language began around 500 study done prior to the beginnings of modern linguis- BC in Greece and India. The work of Greek scholar tics. Grammar, in this traditional sense, is the study of Dionysius Thrax is the model for all grammars of the structure and formation of words and sentences, European languages that follow. His He¯ grammátike¯ usually without much reference to sound and meaning. tékhne¯ (c. 100 BC; The Art of Letters) was the first In the more modern linguistic sense, grammar is the widely recognized text to provide a curriculum for study of the entire interrelated system of structures— learning proper Greek. His lessons included an intro- sounds, words, meanings, sentences—within a lan- duction to the alphabet, lessons on how to join sylla- guage. bles together properly, and instruction in the Traditional grammar can be traced back over 2,000 appreciation of word arrangement (syntax). To Thrax, years and includes grammars from the classical period grammar was the technical knowledge necessary to of Greece, India, and Rome; the Middle Ages; the produce the prestige language of poets, orators, and Renaissance; the eighteenth and nineteenth century; writers. and more modern times. The grammars created in this Around the same time, the Roman scholar Marcus tradition reflect the prescriptive view that one dialect Terentius Varro produced the 25 volumes of his De lin- or variety of a language is to be valued more highly gua latina (c. 100, About the Latin Language). Varro than others and should be the norm for all speakers of contrasted Latin with Greek, changed Greek grammat- the language. Traditional grammars include prescrip- ical terms into Latin, and formed his grammar of Latin tive rules that are to be followed and proscriptive rules by adapting Greek rules. of usage to be avoided. ‘When describing an emotion, Other Latin grammars, influenced by the works of use of an English word descended from Latin is pre- Thrax and Varro, were produced in the Middle Ages. ferred over an Anglo-Saxon word’ is an example of a Aelius Donatus published Ars Grammatica (c. fourth prescriptive rule, and ‘Never split an infinitive’ is an century, Art of Letters), and Donat Priscianus example of a proscriptive rule. Caesariensis (Priscian) wrote Institutiones grammaticae 395 GRAMMAR, TRADITIONAL (c. sixth century, Grammatical Foundations), which is a mixture of English dialect varieties, not just the pre- the only complete surviving Latin grammar. scribed, and sometimes inconsistent, prestige forms. As printing became more widely available in the Linguists or teachers using a descriptive approach say Renaissance, European grammarians began the mass that it allows students to investigate language on a production of grammars of their languages by mirror- deeper level, enabling students to see the system at ing the Latin grammars of Varro, Donatus, and work, instead of teaching them isolated prescriptive Priscian. These traditional grammarians presumed that and proscriptive rules based on Latin, a dead language the grammatical descriptions of Latin could be rou- no longer in flux as English constantly is. tinely applied to their own languages; this perception, Linguists also believe that the rules of traditional however, was not accurate and resulted in many artifi- grammar are inadequate because many of the rules are cial prescriptive and proscriptive rules. Many of these oversimplified, inconsistent, or not consistently con- false assumptions still carry over to attitudes about formed to. The grammars of classical Greece and English today. Rome were based on the best orators or poets of the Continuing with this tradition, grammarians in the day. However, the best poets or speakers of our day are eighteenth century studied English, along with many lauded for their poetic use of language that breaks pre- other European languages, by using the prescriptive scriptive rules. For example, a traditional grammar approach in traditional grammar; during this time rule of modern English, often found in usage guides alone, over 270 grammars of English were published. and student handbooks, forbids the use of fragment During most of the eighteenth and nineteenth cen- sentences like ‘The train running up the hill.’ However, turies, grammar was viewed as the art or science of e.e. cummings or Maya Angelou could use this sen- correct language in both speech and writing. By point- tence for poetic effect without question. ing out common mistakes in usage, these early gram- Many teachers themselves want to be trained in tra- marians created grammars and dictionaries to help ditional grammar, even though its inconsistencies settle usage arguments and to encourage the improve- may not help them when they have to explain gram- ment of English. matical points to their students. The National Council One of the most influential grammars of the eigh- for Accreditation of Teacher Education (NCATE) teenth century was Lindley Murray’s English gram- requires that teachers in training take linguistics or mar (1794), which was updated in new editions for language courses to teach them to examine the differ- decades. Murray’s rules were taught for many years ences between traditional grammar and more modern throughout school systems in England and the United grammars. However, many English teachers view tra- States and helped to create modern attitudes about the ditional grammar as necessary and newer grammars existence of a correct or standard variety of English. as little help to them. And even though more modern Murray’s grammar represents a practice that continued types of grammatical analysis exist, many students, to develop throughout the nineteenth century and was future teachers, and the general public still believe still dominant in the 1960s when linguistics began to grammar means the traditional Latin-based grammar focus more on generative and transformational gram- of old. mar due to Noam Chomsky’s groundbreaking and influential ideas. Even though linguists today view traditional gram- References mar as an unscientific way to study language and Allen, Robert. 1972. English grammars and English grammar. grammar, many of the basic Latin-based notions of New York: Scribner. grammar can still be found in all levels of the class- Battistella, Edwin. 1999. The persistence of traditional gram- mar. Language Alive in the classroom, ed. by Rebecca S. room and in textbooks and usage guides available to Wheeler. Westport, CT: Praeger. educators and the public. Traditional grammar books Chomsky, Noam. 1965. Aspects of the theory of syntax. usually provide lists of grammatical terms, definitions Cambridge, MA: MIT Press. of those terms, and advice on using so-called ‘stan- Haussamen, Brock. 1993. Revising the rules: traditional gram- dard’ grammar, including suggested correct usage of mar and modern linguistics. Dubuque, IA: Kendall/Hunt. Leitner, Gerhard (ed.) 1991. English traditional grammars: an punctuation, spelling, and word choice. This advice is international perspective. Philadelphia: John Benjamins. usually based on the prescriptive rules of prestige Noguchi, Rei. 1991. Grammar and the teaching of writing. varieties of English, varieties often only able to be Urbana, IL: National Council of Teachers of English. used by those in power either economically or politi- Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan cally. Svartvik. 1985. A comprehensive grammar of the English language. London: Longman. Linguists, along with many English faculty, would ALLISON SMITH rather have students study language with a descriptive approach that includes the analysis of real samples of See also Chomsky, Noam; Generative Grammar 396  GRAMMAR, THEORIES Grammar, Theories Grammar theories have several overlapping main pur- which we have called ART, is defined by rule (4) and is poses in linguistics: to either the word ‘a’ or the word ‘the’ (the vertical bar is read as a logical ‘or’). The third rule requires a bit more ● provide a model that clarifies our understanding explanation. The constituent ADJ-SEQ is defined, in of a particular language, part, in terms of itself. Such a definition is called a ● model language in general, recursive definition. Using ‘The big black dog’ as an ● provide a convenient way of working with lan- example, the ADJ-SEQ would be defined as an ADJ guage, particularly with computers, (the adjective ‘big’) followed by another ADJ-SEQ that ● describe the fundamental structures of language, consists only of an ADJ, which is the adjective ‘black’. and The process of matching of the actual words and phras- ● model human communication. es to the constituents defined by the grammar is called Noam Chomsky’s transformational grammar has, in parsing, and the result of that matching is called a recent decades, been by far the most prevalent theory. parse. Since, when one diagrams the matching, the result is a tree-like hierarchical structure, the result is Phrase-Structure Grammar often referred to as a parse tree. Certain constituents, such as ART, ADJ, and Phrase-structure grammar was introduced by Noam NOUN, clearly indicate particular word classes. Chomsky as a grammatical production system that cat- Generally, we assume a lexicon of the words of the egorized grammars and languages into a hierarchy of language so that the grammar rules would assume a four types. A very similar notational scheme to specific set of word classes: describe one of those types (context free) was devel- oped independently by John Backus for the purpose of (1) S → NP VP describing programming languages. Phrase-structure (2) NP → article ADJ-SEQ noun grammar describes a language literally in terms of the (3) VP → verb | verb NP structure of its phrases. The basic idea is that a sen- (4) ART → a | the tence of a language is defined in terms of its phrase (5) ADJ-SEQ → adjective | adjective ADJ-SEQ structure. The definition typically includes con- In such a characterization, an ‘article’ is any entry stituents that have their own phrase structure and in the lexicon that is identified as belonging to the hence require their own definitions. For example, in word class ‘article’. Nouns, adjectives, verbs, and the sentence ‘The big black dog chased the car’, the other classes are similarly defined. phrase ‘The big black dog’ serves as the entity that is Although the example is greatly oversimplified, it doing the chasing while the phrase ‘the car’ is the enti- will be adequate to illustrate many of the major issues ty being chased. We could define the structure of a addressed by various grammatical models. Several simple sentence as follows: points should be noted by this first simple example: (1) Each rule defines a constituent, regardless of the (1) S → NP VP context of that constituent, and is therefore called a con- (2) NP → ART ADJ-SEQ NOUN text-free grammar. There are various ways to extend (3) VP → Verb | Verb NP context-free grammars to account for context and some (4) ART → a | the of those extensions are included in the grammatical (5) ADJ-SEQ → ADJ | ADJ ADJ-SEQ models discussed below. Note also that phrase-structure (6) ADJ → big | little | black | white | good | bad grammars are not necessarily context free. In (7) NOUN → boy | girl | dog | cat | car | truck Chomsky’s hierarchy, context-free grammars are in the (8) Verb → chased | followed | saw second lowest (in terms of expressive power) level of Each rule defines the phrase structure of one type of the hierarchy. Generally, context-free grammars are constituent (the arrow is read as ‘is defined as’). The highly desirable because of the relative ease in reading first rule defines a sentence (called an S) as a sequence and understanding them. Greater expressive power may of an NP followed by a VP. An NP is defined in rule (2) be achieved by allowing the left-hand side of a rule to as being a sequence of three components, each of contain more than one symbol, but such models are sel- which requires further definitions. The first component, dom used. Generally, other kinds of extensions are used. 397 GRAMMAR, THEORIES (2) We omit discussions here of whether morpholog- the lexicon. Although categorial grammars generally ical rules are accounted for by the lexicon or the gram- have more than one rule, the number of rules is always mar. Hence, the word ‘chased’ may be in the lexicon small (often two) with the structure primarily being explicitly, or there may be a morphological component handled by the lexicon. The underlying principle of that identifies ‘chased’ as being the past tense of the lex- categorial grammars is that each word is a functor, an icon entry ‘chase’. Although that is not relevant for our object that operates as a function that maps one or example, it is certainly relevant in general when evalu- more adjacent words into a new object, generally a ating grammatical theories. Some languages (e.g. new function. Turkish, Hungarian, Finnish, or Mohawk) have mor- phological complexities that would make the positing of Relational Grammar a lexicon that holds all morphological forms untenable. (3) The discussion so far has addressed only the Relational grammar is a direct descendant of transfor- syntax of the language. It has not addressed semantic, mational grammars that was developed primarily by pragmatic, or other issues. David Perlmutter and Paul Postal in the early 1970s. (4) The example deals only with defining a sen- One important characteristic of relational grammars is tence, but generally any model that attempts to deal their cross-language generality. Early transformational with meaning must address larger units of language in grammars dealt almost entirely with English, but rela- order to account for context. tional grammars were developed using a large variety of languages. Relational grammars view syntactic Categorial Grammars structures as a hierarchy of relations that are taken to be uninstantiated primitives. For example, the subject The term categorial grammar was introduced by is level 1, the direct object is level 2, and the indirect Yehoshua Bar-Hillel and combined some of Bar- object is level 3. Clausal structure is represented by a Hillel’s own work with earlier work of some Polish hierarchical diagram, called a stratal diagram, of the logicians, most notably Kazimierz Adjukiewicz. clause’s terms. Categorial grammars in their original form are equiva- One of the main motivations for the creation of rela- lent in expressive power to context-free phrase-structure tional grammars was to correct a perceived weakness of grammars, but are based much more on words and their transformational grammars. Perlmutter and Postal influence on adjacent or nearby words than are phrase- argued that logical sentence components such as the structure grammars. Each word class is a ‘category’. subject are defined directly in terms of the sentence Categories are either basic categories or derived cate- structure, which did not provide for a general character- gories. Derived categories have one or more operators ization of relationships such as that between active and in them that allow them to be combined with adjacent passive voice. As a simple example, ‘The dog chased words in a manner much like multiplication. In the fol- the boy’ and ‘The boy was chased by the dog’ have dif- lowing example, the only basic category is C. Hence, a ferent syntactic subjects, but the same logical ‘subject’. phrase ‘the boy’ would have the categories (NP/C)C. That is, the dog does the chasing in both sentences. Treating the slash much as one would division, we can Hence, a model that defines the term ‘subject’ strictly in ‘multiply’ (NP/C) by C, arriving at NP. Hence, the terms of constituent structure will have difficultly phrase ‘The boy’ will be an NP. Similarly, the phrase accounting for the similarity of the two sentences. ‘The big boy’ would have the category sequence (NP/C)(C/C)C, which, when ‘multiplied’ out, gives NP. A categorial grammar equivalent to the phrase- Case Grammar structure grammar given above would have only one Another descendant of transformational grammar is case rule and a lexicon as follows: grammar as introduced by Charles Fillmore in the early Grammar Rule Lexicon 1970s. Case grammar is based, according to Fillmore, on (Ci/Cj)Cj → Ci a:(NP/C) two principles: (1) the centrality of syntax and (2) covert big: (C/C) categories. The former implies that the sentence’s syn- black(C/C) tactic structure is primary (as opposed to, say, morphol- boy:C ogy) and the latter refers to the dominance of the cat:C dog:C meaning underlying the use of words and constituents. girl:C For example, in the sentence ‘The boy hit the dog with a little(C/C) stick’, syntactically, ‘the boy’ is the subject, ‘the dog’ is the: (NP/C) the direct object, and ‘a stick’ is the object of a preposi- white(C/C) tion. A model consistent with case grammar would view The example of categorial grammar has only one the sentence as requiring the covert categories ‘actor’, grammar rule. The rest of the language structure is in ‘affected’, and ‘instrument’. In the example, ‘the boy’ 398  GRAMMAR, THEORIES would be the actor, ‘the dog’ would be the affected, and Lexical-Functional Grammar ‘a stick’ would be the instrument. Lexical-functional grammar was developed in the One place where we can see a more illustrative early 1970s by Joan Bresnan and Ron Kaplan to pro- example is in the analysis of coordination. For exam- vide a model that could serve as the basis of both a ple, given that Tom is cooking and Susan is cooking, it computationally feasible and psychologically realistic is reasonable to assert that Tom and Susan are cooking. model of human language. Like relational grammar, it However, given that Tom is cooking and the omelet is deviated from the transformation model by including cooking, we would not say that Tom and the omelet are nonstructural factors in the syntactic component of cooking (except, perhaps, to be humorous.) The point language. Also, like relational grammar, it gives a dis- here is that although the sentence clauses ‘Susan is tinguished status to the notion of ‘subject’ and cooking’ and ‘the omelet is cooking’ appear to be syn- ‘object’. The ‘lexical’ in the title is due to its emphasis tactically similar sentences, the role that ‘Susan’ fills in on the role of the lexicon. the sentence is quite different from the role filled by In a lexical-functional description of a sentence, ‘the omelet’. In this example, the respective roles there are two components: a constituent structure and would be difficult to distinguish via purely syntactic a functional structure. The constituent structure is means. Of course, sometimes the roles cannot be deter- strictly structural and is based on a context-free mined by syntactic means but depend on semantics and phrase-structure grammar. The functional structure is the context of the situation. For example, in a cartoon, by additional computations, expressed as equations, Susan and a chicken could be cooking in the sense that associated with the context-free rules. The computa- both are in the act of preparing a meal, or both could be tional aspects of lexical-functional are quite similar to in the other role in a context involving cannibals. a different formalism, developed in the late 1960s by Donald Knuth, called an attribute grammar. The basic Dependency Grammar idea is that each constituent of the language has a set An alternative to describing the syntax of a language of attributes. For example, the ‘number’ attribute of the via its phrase structure is to describe the dependencies NP ‘the big black dog’ could be ultimately assigned as that exist among the words, phrases, and clauses. The ‘singular’ since the words ‘the’ and ‘dog’ are both idea is a very old one, but the modern characterization marked as singular by the lexicon. Note that if their is based largely on the work of Lucien Tesnière in the number attributes had been different, the noun phrase late 1950s. The dependency approach to syntactic would not have been well formed. structure was overwhelmed by Noam Chomsky’s trans- Lexical–functional grammar is similar to case gram- formational grammar approach and was thus largely mar in that the attributes constitute a set of named slots ignored by the modern linguistic community. In recent that are filled by the computational rules. Note also that years, more interest has been shown in the dependency attributes and their associated computational rules need approach as well as other alternatives to phrase-struc- not be confined to syntactic properties. Semantic fea- ture representations for syntactic structures. tures with complex values, including values that are Dependency grammars take the head and modifier actually active computational agents such as automata approach to constituent structure. For example, in a or computer programs, fit naturally into the model. phrase-structure approach, one might analyze the sen- tence ‘cars burn gas’ by focusing on the verb, ‘burn’. Generalized Phrase-Structure Grammar Cars are the things doing the burning and gas is the stuff being burned. An alternative approach would be Unlike lexical–functional grammar, generalized phrase- to consider cars as the head of the sentence and the structure grammar (GPS) does not posit a separate, dis- burning of gas as an attribute of cars. Basically, the tinct functional component. The model presented here ‘head’ of a sentence, clause, or phrase is where we was introduced by Gerald Gazdar, in the late 1970s, but focus our attention. The dependencies are generally the formal mechanisms are very similar to W-grammar determined via the semantics of the sentence. An easy as developed by A. Van Vijngaarden in the late 1960s. thumb rule is to imagine that you are recording the The basic idea is to use phrase-structure grammar, but to information of the sentence in, say, some kind of ency- allow more powerful metarules that define additional clopedia. The head of the phrase is then the entry rules rather than syntactic constituents. This seemingly under which the information is stored. In the case of small change in the formalism changes the expressive ‘cars burn gas’ it is likely that storing the information power of the grammar substantially and allows the spec- under ‘car’ rather than ‘burn’ would ultimately prove ification of complex syntactic rules such as matching a to be more useful. Dependency grammars have often sentence in the active voice to its corresponding passive been used in information retrieval systems, particular- sentence. In particular, it allows GPS to dispense with ly early systems. the transformations of transformational grammar. 399 GRAMMAR, THEORIES In Gazdar’s model, grammatical constituents Definite Clause Grammar included components called derived categories, which Definite clause grammar was first proposed by Alain were denoted using notation similar to the derived Colmerauer in the late 1970s and developed further by categories in categorial grammar. However, in GPS Fernando Pereira and David Warren. Definite clause the derived category X/Y (where X and Y stand for grammar is primarily a computational model that is grammatical constituents) represents a constituent X based directly on logical formulas. For example, the that is missing the constituent Y, that would normally phrase structure rule of the first example, S → NP VP, be included as a component of X. This ability to could be written as the logical formula: define constituents with ‘holes’ allows a relatively For all sequences of words x, y, and z, if x is an NP straightforward treatment of sentences such as and y is a VP and the concatenation of x and y is z, ‘Which book is that article on Gazdar in?’ In this then z is an S. example, the verb phrase of the sentence is missing a All of the rules of any Phrase-structure grammar noun phrase (‘which book’), and hence requires a could be written according to such a logical formula. noun phrase elsewhere in the sentence that can ‘fill in The purpose of stating a rule in that fashion is that it the hole’. allows for direct implementation of a parser on a com- puter. Definite clause grammar was originally created Cognitive Grammar as a natural way to write parsers in the programming language Prolog. Like GPS, definite clause grammar Cognitive grammar was developed in the mid-1970s does not posit a separate distinct functional compo- by Ronald Langacker. It was originally called space nent, but simply adds additional components to the grammar and was considered by its creator to be fun- logical formula. For example: damentally at odds with current linguistic trends of the For all sequences of words x, y, and z, if the con- time in that it claimed the inseparability of syntax and catenation of x and y is z and either (1) x is a singular semantics, rejected the primacy of formal logic as the NP and y is a singular VP, or (2) x is a plural NP and foundation of grammar, and argued for integration y is a plural VP, then z is an S. rather than specialization of theories for separate lin- This is a logical formula that would include number guistic components. agreement between the noun phrase and the verb Cognitive grammar, as its name implies, considers phrase. The logical formulas can be translated directly language as an integral part of human cognition and into computer code that provides a parser. therefore knowledge of language cannot be wholly independent of knowledge of human cognition. Langacker argues that a separate processing unit that References deals with reason and cognition has not been firmly Blake, Barry J. 1990. Relational grammar. London: Routledge. established and may not exist. His model is intended van Brenthem, Johan, and Alice ter Meulen (eds.) 1997. to be valid whether such a separate module exists or Handbook of logic and language. Cambridge, MA: MIT Press. not. Walter J. Cook, S.J., 1989. Case grammar theory. Washington, There are exactly three kinds of grammatical DC: Georgetown University Press. structures in cognitive grammar: semantic, phono- Gazdar, Gerald, Ewan Klein, Geoffrey Pullum, and Ivan Sag. logical, and symbolic. Symbolic structures are not 1985. Generalized phrase structure grammar. Cambridge, distinct from the other two, but combine them. MA: Harvard University Press. Grishman, Ralph. 1986. Computational linguistics, an introduc- Cognitive grammar attempts to characterize the map- tion. Cambridge: Cambridge University Press. ping and providing the correspondence of the sounds Koerner, E.F.K., and R.E. Asher. 1995. Concise history of the we hear to the thoughts we associate with those language sciences from the Sumerians to the Cognitivists. sounds. Cambridge: Cambridge University Press. Cognitive grammar is based on several underlying Langacker, Ronald W. 1987. Foundations of cognitive gram- mar, Vol. I: theoretical prerequisites. Stanford, CA: Stanford principles. First, formal logic is not suitable for mod- University Press. ˆ eling the meaning of a sentence, since meaning is not Mel’cuk, Igor A. 1988. Dependency syntax: theory and prac- based on truth values, but rather individual experi- tice. Albany, NY: State University of New York Press. ences. Second, the meaning of a particular word is Oehrle, Richard T., Emmon Bach, and Deirdre Wheeler. 1988. defined by the universe of experience of the speaker Categorial grammars and natural language structures. Boston. MA: D. Reidel Publishing Company. (or listener) and not by a constrained formula such as Sells, Peter. 1987. Lectures on contemporary syntactic theories: is implied by a dictionary entry. Finally, like words, an introduction to government-binding theory, generalized linguistic categories are not simple structures, but gen- phrase structure grammar, and lexical–functional grammar. erally complex networks of linked structures. Menlo Park, CA: CSLI. 400  GRAMMATICAL FUNCTION Winograd, Terry. 1983. Language as a cognitive process: syn- See also Bar-Hillel, Yehoshua; Chomsky, Noam; tax. Reading, MA: Addison-Wesley Publishing Company. Fillmore, Charles; Morphology; Morphological ALTON SANDERS Typology Grammatical Function The concept of ‘grammatical function’ or ‘grammatical is marked with a final -s indicating third-person singu- relation’ comes from traditional grammars, which differ- lar. Thus, the verb is said to ‘agree’ with the subject he, entiate four grammatical functions. For example, in a which is a third-person singular pronoun. Moreover, in sentence like John gave this book to Peter at school, it so-called accusative languages, case-marking on nouns can be said that John functions as a ‘subject’ of the verb and pronouns can serve as an indicator of grammatical to give; this book is the verb’s ‘direct object’ and to Peter functions. Thus, in English we observe different plays the role of its ‘indirect object’; finally, the locative pronominal forms for subject and object, as in He saw at school is ‘oblique’. These labels underlie a network of him, where he is the form for nominative case (subject) relations between the verb and its arguments. and him is accusative (object). Nevertheless, the corre- In early studies of natural language, a sentence was spondence between case-marking and grammatical identified with the logical proposition, which has two functions is different in so-called ergative languages essential components: the substance (subject) and its (including Inuit, Georgian, Basque, Hindi, and many property (predicate). The subject is supposed to be others). In these languages, the subject of an intransi- known to both speaker and hearer, while the predicate tive verb (I run) has the same marking as the object of adds something new about the subject. Therefore, sub- a transitive verb—that is, the equivalent of the English jects usually have a definite reference, and precede the sentence The teacher has seen me would be something predicative part of an utterance (the verb with its like The teacher I has seen in Burushaski (a language objects). To put it differently, grammatical subjects isolate spoken in northern Pakistan). ensure the coherence of a discourse that leads from old Finally, some more complex grammatical properties to new information. This perspective was developed also differentiate the subject from the object. For exam- by Prague School linguists of the 1920s and is fol- ple, subjects (but not objects) can ‘control’ a missing ele- lowed by modern functional approaches to grammar. ment in a subordinate clause: thus, in Paul helps Mary From a semantic point of view, grammatical func- without asking for money, it is Paul (subject), not Mary tions are usually associated with different situational (object), who does not ask for money, i.e. the complete roles, known as ‘thematic roles’. For example, the sub- sentence would be Paul helps Mary without Paul asking ject is usually associated with the role of Agent (or for money. The subject of the subordinate clause can Actor) (John writes …), while such roles as Theme (to only be left out if it is identical to the subject of the main give a book) and Goal (to give it to John) are related clause. Notice that in Paul helps Mary without Mary ask- to direct and indirect objects, respectively. Oblique ing for money the subject of the subordinate clause can- arguments can express Instrumental and Locative the- not be left out. Moreover, in some languages (e.g. Latin, matic roles (to be loved by Mary, to live in London). Russian, Norwegian, Icelandic) there are reflexive pro- However, with psychological verbs the direct corre- nouns whose antecedant must be the subject (so-called spondence between thematic roles and grammatical ‘subject-oriented reflexives’). Accordingly, in the functions is problematic. For example, the verb to fear Icelandic sentence Pall gaf barninu bók sín (literally assigns the Experiencer role to the subject and the Paul gave child book self’s), the book is construed as Theme role to the object (John (subject) fears this belonging to Paul (subject), not to the child (object). story (object)), while the synonymous verb to frighten In languages with rich noun inflection, a noun may assigns the same roles, but in inverse order (This story behave syntactically as a subject but be marked as an (subject) frightens John (object)). object (so-called ‘grammatical function-splitting’). With respect to relevant word forms, the subject, Icelandic, for instance, has such constructions as Mér rather than the object, induces verbal agreement in a likuði konan sín (literally Me likes wife self’s), I like my finite clause. In He knows them, for example, the verb wife. Here, mér does not have the word form normally 401 GRAMMATICAL FUNCTION associated with the subject, since it is dative (not nom- means of expression in a given language. For instance, inative) and does not trigger verbal agreement. the English lexicon has the verb to invite, whose Agent However, mér is the subject from the syntactic point of role is mapped to the subject, while its Patient role is view, since it can refer to the same entity as a subject- mapped to the object, as in John (subject-Agent) invit- oriented possessive sín. In the literature, such cases are ed the president (object-Patient). The passive counter- known as the ‘quirky subject’ phenomenon. part of this verb is formed by remapping Patient to the Many linguists (following Noam Chomsky) main- subject and Agent to the oblique, as in John (subject- tain that objects are more closely related to the verb, Patient) is invited by the president (oblique-Agent). and consequently more likely to form idioms than sub- This is the view elaborated in the framework of ‘lexi- jects (John is pulling my leg, this house catches the cal-functional grammar.’ eye, etc.). This asymmetry is said to be due to the fact Summing up, the traditional notion of grammatical that verbs and objects form the predicative unit, while function is controversial in modern linguistics. The the subject is a more independent entity. From this core grammatical functions, subject and object, cannot point of view, the notions of ‘subject’ and ‘object’ can be defined absolutely, since they are differentiated on be defined on structural grounds. the basis of several possibly independent properties. However, nonconfigurational languages, in which From the semantic point of view, they represent the word order is free (e.g. the Australian language two main participants in the event denoted by the verb. Warlpiri), have been used to argue against a purely Grammatical functions may be useful descriptively, structural interpretation of grammatical functions. since they make it possible to abstract away from lan- Since the mid-1970s, some new theoretical approach- guage-specific properties and to generalize over cer- es have been created on the basis of grammatical func- tain syntactic phenomena. tions being autonomous categories. For example, grammatical functions may be regarded as syntactic primitives forming the following hierarchy: subject > References direct object > indirect object > oblique. This is the Bhatt, Darbhe N. S. 1991. Grammatical relations: the evidence stance taken by the framework of ‘relational gram- against their necessity and universality. London and New mar’. It is argued that certain sentence structures may York: Routledge. Bresnan, Joan. 2001. Lexical—functional syntax. Oxford and be described in terms of ‘grammatical function-chang- Malden, MA: Blackwell. ing processes’ . For example, the formation of pas- Davies, William D., and Stanely Dubinsky (eds.) 2001. Objects sives, such as John has broken the glass—The glass and other subjects: grammatical functions, functional cate- has been broken (by John), is characterized as follows: gories, and configurationality. Dordrecht and Boston: Kluwer. the object is ‘promoted’ (in hierarchy) to the subject, Perlmutter, David M. (ed.) 1983. Studies in relational grammar, Vol. 1. Chicago: The University of Chicago Press. while the subject is ‘demoted’ to the oblique (the pres- Ura, Hiroyuki. 2000. Checking theory and grammatical func- ence of the latter is optional). Dative shift in English tions in universal grammar. New York and Oxford: Oxford (e.g. I gave my favorite book to Mary and I gave Mary University Press. my favorite book) also represents ‘grammatical func- Van Valin. Jr. Robert D. and Randy J. LaPolla. 1997. Syntax: tion-changing’, here the promotion of the indirect structure, meaning and function, Cambridge: Cambridge University Press. object to the direct object. EGOR TSEDRYK Another possibility is to conceive of grammatical functions as categories that relate thematic roles See also Case; Configurationality; Predication; (Agent, Patient, Experiencer, etc.) to the concrete Thematic Structure; Word Order Grammaticalization Grammaticalization is defined as the development constructions to which they belong, the study of from words with full meaning to forms with only gram- grammaticalization is also concerned with construc- matical function, and from grammatical to even more tions, and with even larger discourse segments. The grammatical forms. Since the development of gram- primary goal of grammaticalization theory is to matical forms is not independent of the grammatical describe how grammatical forms and constructions 402  GRAMMATICALIZATION arise and develop through space and time, and to transfer, leading from the domain of concrete objects explain why they are structured the way they are. to that of space, from space to time, from (‘real- In the history of grammaticalization studies, three world’) space to discourse space, and so on. main phases can be distinguished. The first phase is There is wide agreement that grammaticalization associated with the work of eighteenth-century French involves three interrelated mechanisms: (1) ‘deseman- and British philosophers who argued that grammatical ticization’ (or ‘semantic bleaching’), loss of meaning complexity and abstract vocabulary derive historically content; (2) ‘decategorialization’, loss of morphosyn- from concrete words. In the nineteenth century, it was tactic properties (e.g. inflection) characteristic of lexi- mostly German linguists, especially Franz Bopp and cal or other less grammaticalized forms; and (3) Georg von der Gabelentz, who used findings of gram- ‘erosion’ (or ‘phonetic reduction’), loss of phonetic maticalization to understand grammatical change and substance. Desemanticization results from the use of language structure. The third phase started in the forms for concrete meanings that are reinterpreted in 1970s and was shaped by scholars using functionalist specific contexts as more abstract, grammatical mean- frameworks of linguistic analysis, most of all Talmy ings. Having acquired grammatical meanings, these Givón. His slogan ‘Today’s morphology is yesterday’s forms tend to become increasingly divergent from syntax’ opened a new perspective for understanding their old uses: they lose categorial properties charac- grammar, and he proposed the following unidirection- teristic of their old uses, hence undergo decategorial- al cycle (where the endpoint marks the beginning of a ization, and they tend to be used more frequently, new cycle again leading from Discourse to Zero; see become more predictable in their occurrence and, con- Givón 1971, 1979): sequently, tend to lose sounds. While all three mecha- nisms involve a loss of properties, there are also gains. Discourse > Syntax > Morphology > In the same way as linguistic items undergoing gram- Morphophonemics > Zero. maticalization lose in semantic, morphosyntactic, and While earlier work dealt mostly with the study of phonetic substance, they also gain properties charac- individual words and grammatical items, modern work teristic of their uses in new contexts. is more concerned with the use of linguistic forms in Grammaticalization requires specific contexts to take constructions and larger discourse units. place, and it can be, and has been, described as a prod- The following are the main assumptions underlying uct of context-induced reinterpretation or inferencing. grammaticalization theory. First, it is argued that lan- Grammaticalization thus begins with concrete, lex- guage is a historical product and has to be accounted ical forms and constructions, and ideally ends in zero, for, first of all, with reference to the historical forces that is, grammatical forms increasingly lose semantic that are responsible for its present structure. and phonetic content and, in the end, they may be Accordingly, it is claimed that this theory offers more replaced by altogether new forms; grammaticalization powerful explanations of language structure than any has therefore been described as a cyclical process. approach that is confined to the analysis of a language Cyclicity is a frequent but not a necessary property of at a particular point in time. grammaticalization processes. Furthermore, grammat- Second, the development of grammatical forms is icalization has been described by some as a process unidirectional, being the result of a process whereby that involves the reanalysis of grammatical forms forms for concrete meanings are used to express more and/or constructions, while other authors argue that abstract (grammatical) meanings as well, e.g. when a there is no necessary relationship between grammati- form used for a visible object (e.g. the body part back) calization and reanalysis. is also used to refer to a nonvisible item (the spatial notion behind), or a form used for an action (go to) is used to also refer to a grammatical notion (FUTURE References TENSE). But equally common are processes whereby Bybee, Joan L., Revere D. Perkins, and William Pagliuca. 1994. items that are already part of the inventory of gram- The evolution of grammar: tense, aspect and modality in the matical forms give rise to more strongly grammatical- languages of the world. Chicago: University of Chicago Press. ized and more abstract items. A number of examples Givón, Talmy. 1971. Historical syntax and synchronic mor- contradicting the unidirectionality principle have been phology: an archaeologist’s field trip. Chicago Linguistic found, but as acknowledged by most scholars who have Society 7. 394–415. identified exceptional cases, such examples are few _________. 1979. On understanding grammar. New York, San compared to the large number of cases that conform to Francisco, London: Academic Press. Haspelmath, Martin. 1998. Does grammaticalization need the principle (see especially Newmeyer 1998). reanalysis? Studies in Language 22(2). 315–51. On account of its specific directionality, grammati- Heine, Bernd. 1997. Cognitive foundations of grammar. calization has been described in terms of metaphorical Oxford, New York: Oxford University Press. 403 GRAMMATICALIZATION Heine, Bernd, Ulrike Claudi, and Friederike Hünnemeyer. Lehmann, Christian. 1985. Grammaticalization: synchronic 1991. Grammaticalization: a conceptual framework. variation and diachronic change. Lingua e Stile 20(3). Chicago: University of Chicago Press. 303–18. Heine, Bernd, and Tania Kuteva. 2002. World lexicon of gram- Newmeyer, Frederick J. 1998. Language form and language maticalization. Cambridge: Cambridge University Press. function. Cambridge, MA: MIT Press. Heine, Bernd, and Mechthild Reh. 1984. Grammaticalization Traugott, Elizabeth C., and Bernd Heine (eds.) 1991. and reanalysis in African languages. Hamburg: Buske. Approaches to grammaticalization, Vol. 1. Amsterdam, Hopper, Paul J. 1991. On some principles of grammaticization. Philadelphia: Benjamins. In Traugott and Heine, pp. 17–35. BERND HEINE Hopper, Paul J, and Elizabeth C. Traugott. 1993. Grammaticalization. Cambridge: Cambridge University Press. See also Bopp, Franz; Givón, Talmy Great Britain Great Britain, or more properly the United Kingdom England of Great Britain and Northern Ireland, includes England, which gets its name from the Old English England, Scotland, and Wales and Northern Ireland. Englaland (the land of the Angles), is in the southern The total area of the current United Kingdom is part of the island of Britain and is the original home of 244,820 square kilometers with a total population of English, a Germanic language that came with the 59,511,464 (2000 est.). Angles and Saxons to England from modern-day Wales was politically integrated with England in northern Germany, starting in the fifth century. The 1536, Scotland officially joined England and Wales in Celts, the earlier inhabitants, intermarried with the 1707 with the Act of Union, and Ireland was added in Anglo-Saxons or migrated to the outer parts of the 1801 to create the political union of the United British Isles. Even though the Anglo-Saxons lost their Kingdom of Great Britain and Ireland. The union with power in 1066 with the Norman Conquest, English Ireland was in effect until the Irish Free State was and England remained strong, and by the sixteenth formed in 1921. The six counties of Northern Ireland, century, English was the official language of England. with a predominantly Protestant population, remained Various minority groups live in England, using lan- in the union, and the official name was changed to the guages other than English as their only language or as United Kingdom of Great Britain and Northern Ireland a bilingual partner to English. The Irish are the largest in 1927. Now known as the Republic of Ireland (Eire), minority group, and there are also groups of the Irish Free State became a separate republic in 1949. Europeans (Italians, Greeks, Turkish Cypriots, and The approximate historical periods of English with- Poles), Asians, Africans, Arabs, West Indians, in Great Britain are recognized as Old English from Pakistanis, and Indians. In addition, Welsh and 450 to 1100, Middle English from 1100 to 1500, early Cornish are also still spoken. Cornish, a Celtic lan- modern English from 1500 to 1700, and modern guage, has been revived and is now taught in some English from 1700 to the present. Cornwall schools, and some speakers of Welsh in Many languages have existed in Great Britain; how- places like London maintain their language. ever, English has been the dominant language since the sixteenth century, with various other languages and dialects existing throughout the area. Languages that English were once in Great Britain but are now extinct include English is the official language and is sometimes Anglo-Danish, Anglo-Latin, Anglo-Saxon (Old referred to as Anglo-English, England English, English), Brythonic, Cornish, Manx Gaelic, Norman English English, and British English. The term French, Norn, and Pictish. ‘British English’ is often used when trying to compare Indigenous languages currently found in Great the dialects of England with dialects around the world. Britain include English, Irish Gaelic, Scots, Scottish Many social and regional English dialects exist. Gaelic, and Welsh. Various other immigrant languages Even though some people from different dialect (including Caribbean English Creole, Hindi, Urdu, groups may sometimes have difficulty understanding Punjabi, and Cantonese) are now spoken in cities like each other, the media has helped spread standardized Birmingham, Cardiff, Glasgow, Leeds, and London. words and pronunciations. 404  GREAT BRITAIN Standard English and Received Pronunciation Birmingham area, East Midlands English around the Standard English (SE) and Received Pronunciation East Anglian region, North-West English around (RP) are prestige social forms of English. SE, in its Liverpool and Manchester, and North-East English linguistics sense, relates to mostly the standardized around Leeds, York, and the Yorkshire countryside. grammar and vocabulary of English, and RP to the pronunciation of the educated upper-middle and upper Wales class. However, the terms are often used interchange- ably by those outside of linguistics. Even though it is Wales, in the southwest corner of the island of Britain, assumed that SE includes RP, various accents are part gets its name from the Old English word Wealas (for- of SE use. Most people write SE with slight variations eigners). Celts from northwestern Europe originally in grammar and spelling; many also speak SE with settled in this area and were subsequently invaded by some variation in regional vocabulary. Some speakers Romans. When Anglo-Saxons took control of England use two dialects—a social or regional one for friends in the fifth century, Wales was the only Celtic region and family and a version of SE for formal use, and left in southern Britain, and the Welsh fought off the most urban and rural areas have a variety of spoken English, Irish, Scandinavians, and Normans from the English that has been influenced in some way by fifth to eleventh centuries. In the thirteenth century, the social or regional factors. last native Welsh prince was killed during an English invasion, and from 1301 on, the Prince of Wales has been British. In 1536, Wales was officially united with Cockney England, and English became the official language. Another social dialect at the opposite end of the social spectrum from RP is Cockney, the speech of about 7 million working-class Londoners. Although often stig- Welsh English matized, Cockney has a long history and makes up the Close to 100% of the Wales population speaks most prominent and widely spoken urban dialect in English. Welsh English represents a spectrum of vari- Great Britain. eties and includes the social dialects of bilingual Welsh/English speakers, the English of the working class, SE speakers with a Welsh accent, and SE speak- Regional Dialects ers with an RP accent. Welsh has a strong influence Different regional dialects of English can be traced on the northern, mid, and southern regional dialects back in England to the eighth century when the of Wales, with a very strong influence in the north- Germanic tribes, who had settled in different areas of ern counties, where English/Welsh bilingualism is England in the preceding three centuries, had Englisc common. as their common language. The Angles settled in the There is little or no stigma attached to the use of Midlands and along the eastern coast, and their Welsh English in Wales; however, many believe that dialects included Mercian and Northumbrian. The education should be bilingual and hence all Welsh Jutes settled around Kent and spoke the Kentish people have access to Welsh as the national language. dialect. The Saxons settled in Essex, Middlesex, Others, especially non-Welsh speakers, do not believe Sussex, and Wessex. Under the leadership of Alfred that speaking Welsh is a necessary part of being Welsh. the Great, the West Saxon dialect of Wessex became Tensions are high, however, concerning the use of the the dominant dialect and served as the prestige and lit- Welsh language, since many believe that education erary dialect of the Middle Ages. should be bilingual so all Welsh people have access to Eleventh-century dialects included Northern, West Welsh as their national language. Midland, East Midland, Southern, and Kentish. After the Norman Conquest, the East Midland dialect from the London and East Anglian region became the new Welsh standard with the help of William Caxton, the first Welsh, the Celtic language of Wales, is known as major British printer, who used East Midland as the Cymraeg to its speakers and was the main language printing standard. Regional speech came to be consid- of Wales for many centuries. However, Welsh use has ered as substandard but was used well into the eigh- been diminishing since the fifteenth century. An teenth century by the middle and upper classes. eighteenth- and nineteenth-century revival led to Today, regional varieties of English dialects are still Welsh being taught in all schools; in some schools, it strong, especially in large cities and rural areas. is even the language of instruction. Currently, about Regional dialects include South-East English around 26% of the population is bilingual (with Welsh and the London area, South-West English around the English); however, only a few monolingual Welsh Cornwall area, West Midlands English around the speakers exist. Even so, Welsh is relatively stable; a 405 GREAT BRITAIN Welsh-only television channel and bilingual road was formerly called Erse or Irish, but speakers today signs show that its steady decline might be slowing. often refer to it as the Gaelic. Welsh has several dialects that differ mostly in pro- Until the eighteenth century, Gaelic was the princi- nunciation; the vocabulary has mostly been standard- pal language of Ireland; however, with the increased ized by education and the media. influence and pressure of Northern English, the use and prestige of Gaelic have declined since the twelfth century. Since the late twentieth century, Gaelic has Scotland been a first language for very few people (usually in Scotland, a part of Great Britain since 1707, gets its the Hebrides) but has regained some of its prestige name from the Old English ‘land of the Scots’ and through efforts to encourage bilingual policies, and includes the mainland, which is divided into three has made Gaelic a part of the educational system. main regions (the Highlands, the Lowlands, and the According to the 1981 census, approximately 80,000 Southern Uplands), and the outlying islands, which people speak Gaelic, but most of these are bilingual include the Western Isles (also known as the Hebrides) English/Gaelic speakers. and the Northern Isles (the islands of Orkney and Scottish Gaelic has a strong literary tradition, and Shetland). there is a resurging twentieth-century Gaelic literature When the Romans invaded Britain in the first cen- encouraged by An Comunn Gaidhealach (The tury, they called the area now known as modern Highland Association) and Comunn na Gàidhlig (The Scotland by the name Caledonia and called its inhab- Gaelic Association), two organizations that support itants the Picts (from Latin picius ‘painted’) because and promote the Gaelic culture and language. of the Pict tradition of painting their bodies. In the fifth Scottish Gaelic has influenced Scots English and century, after the Romans withdrew from England, other varieties of Scottish English in pronunciation, syn- and Germanic tribes began their invasions into tax, and vocabulary, and this influence runs both ways. England, Christian Celtic immigrants from Ireland began to settle the area of modern Scotland, converted the Picts to Christianity during the sixth century, and Scots added the Pictish kingdom to the Scots kingdom by Some language scholars consider Scots (or Scots the ninth century. English) a language and others consider it a dialect of By the tenth century, this area was known as English. Whatever way it is defined, since the begin- Scotland. After the Norman Conquest, many Anglo- ning of the eighteenth century, scholars have special- Saxons settled in the Lowlands of Scotland, bringing ized in studying it in its different periods: Old English English customs and the English language. England (to 1100), Early Scots (1100–1450), Middle Scots and Scotland were joined formally together in the Act (1450–1700), and Modern Scots (1700 onward). of Union of 1707 in which both parliaments were Scots is first recorded in the seventh century in merged, thus creating the Kingdom of Great Britain. southern Scotland as a variant of the Old English of Scotland’s language history is complex. In the Northern England. By the fourteenth century, this vari- Middle Ages, six languages were in regular use: ety known as Inglis had replaced Gaelic as the lan- Cumbric, Gaelic, Inglis, Norn, Norman French, and guage of the Lowlands. By the mid-sixteenth century, Latin. Modern Scotland has two or three (depending Scots began to be heavily influenced by the English of on who does the analysis) languages: Scottish Gaelic, southern England. In the eighteenth century, Scots was English (Scottish and RP), and Scots. Up to the fif- considered substandard, but a revival of vernacular lit- teenth century, Scots was the term applied only to erature in Scots in the nineteenth century encouraged Gaelic and its speakers; now this variety is most com- the middle and upper classes to become more interest- monly referred to as Gaelic, Scots Gaelic, or Scottish ed in the Scots language. In the twentieth century, there Gaelic. From the late fifteenth century, the term Scots have been more attempts to revive Scots as the nation- has referred to Scots English, the language of the al language; even so, English and Scottish English are Lowlands. Scottish English refers to the other varieties still the dialects of power and prestige, and there has of English used in Scotland and, depending on the been a steady decline of native Scots speakers. analysis and the discussion, can include or exclude The dialects of Scots fall into four main regional Scots English. dialect areas: the Northern Isles Scots of Orkney and Shetland, the Northern Scots of northern Scotland, the Southern Scots near the English border, and the Scottish Gaelic Central Scots of middle Scotland, including the work- Gaelic is the English name for the Celtic language of ing-class dialects of Edinburgh and Glasgow, which Ireland, Scotland, and the Isle of Man. In Scotland, it are often referred to pejoratively as Gutter Scots. 406  GREAT BRITAIN Scottish English eties: Ulster Scots, Anglo-Irish, Hiberno-English, and If Scots is a language, then Scottish English includes Belfast English. all other varieties of English in Scotland, including the Highland English of the Highlands and Islands, the Ulster Scots Scottish standard English of the past three centuries, Ulster Scots, also known as Scotch-Irish, was brought and the RP dialect of the upper middle and upper class. to Ireland in the seventeenth century by about 200,000 Scottish English has many similarities with the English migrating Lowland Scots and is mainly found in parts of England but also has many features of Scots, and of Antrim, Derry, and Down counties in Northern many modern speakers of Scottish English acquired Island. Ulster Scots is usually identified as the dialect Scots as their first language in their childhood. of the working class and is found throughout Northern Ireland and the Irish Republic. Northern Ireland Anglo-Irish Northern Ireland, a part of the United Kingdom, includes Anglo-Irish, or Ulster English, is a variety of English six of the nine counties that originally made up the early spoken over most of Northern Ireland and the Irish Celtic kingdom of Ulster: Antrim, Armagh, Derry or Republic, except for the most northern counties. Like Londonderry, Down, Fermanagh, and Tyrone. The Celt Ulster Scots, the beginnings of Anglo-Irish came to leaders, who were Druids, lost their power in the fifth Ireland in the seventeenth century, but with English set- century after St. Patrick introduced Christianity. After tlers and not Scottish. There are regional and social vari- many centuries of Celtic challenges to both Christianity eties that are heavily influenced by the education of the and the power of England, the Celt leaders fled Northern speaker. RP is usually part of the Anglo-Irish of the mid- Ireland in the early seventeenth century, and about two dle class, but less educated speakers usually have other thirds of the current Northern Ireland population are distinct pronunciations. As time goes by, a standardized descendants of Protestant English and Scottish settlers variety of Anglo-Irish is becoming more common. who came to Ulster in the same century. The other third are Irish in origin and Catholic. By the mid-nineteenth Hiberno-English century, English was the language of power and prestige. Hiberno-English is another variety of English in In modern Northern Ireland, an omnipresent tension Ireland, used mainly by less educated speakers. exists between Protestants and Catholics, and even Mostly found in rural counties (Armagh, Fermanagh, though the right to free language choice is part of this and Tyrone) in Northern Ireland, it is heavily influ- tension, both sides of the argument are likely to use some enced by Irish Gaelic. variety of English now. Belfast English Irish Gaelic Belfast English, another variety of English in Northern Irish Gaelic, or Irish, is a Celtic language that remained Ireland, varies with the level of education of the speak- the main language of Ireland until the eighteenth centu- er. It is normally described as rapid informal speech ry when the British passed and enforced laws that made that has the vocabulary of Ulster Scots with some non- Irish the language of the powerless. The school system standard grammatical features of Anglo-Irish. was restructured with new National Schools, in which English was the only language of instruction. Gaelic was also negatively impacted by the Irish famine of the References mid- nineteenth century, which resulted in the mass Baugh, Albert C., and Thomas Cable. 1992. A history of the emigration of Gaelic first-language speakers. English language. Englewood Cliffs, NJ: Prentice-Hall. By the end of the nineteenth century, there was a Crystal, David. 2000. Language death. Cambridge: Cambridge University Press. revival of interest in the Irish language, literature, and Hughes, Arthur, and Peter Trudgill. 1998. English accents and history of Northern Ireland and the Republic of Ireland. dialects: an introduction to social and regional varieties of This new national attitude helped slow the steady decline English in the British Isles. London: Edward Arnold. of Irish Gaelic but only after English had already gained McArthur, Tom. 1998. The English languages. Cambridge: enormous power and Irish had died out as a spoken lan- Cambridge University Press. Price, Glanville. 2000. Languages in Britain and Ireland. guage except in isolated rural areas of Northern Ireland. Oxford and Malden, MA: Blackwell Publishers. Wells, John C. 1982. Accents on English: an introduction. Cambridge: Cambridge University Press. Northern Irish English ALLISON SMITH Northern Irish English refers to the English used in Northern Ireland; however, there are four distinct vari- See also Celtic Languages 407 GREAT VOWEL SHIFT Great Vowel Shift English is unique among the European languages supposed to be the result and the trigger of a chain- using the Roman alphabet in that it has a special set of like shift of all ME pure long vowels. In most cases, ‘names’ for the vowel letters <i, e, a, o>. The names the phonemic contrasts between the original entities reflect the pronounciation of these letters when they were preserved, although their phonetic realizations represent long or tense vowels, as in mine, cede, fame, were changed. This position is well represented by grove. Until the fourteenth century the value of these Jespersen’s summary statement (1909:231): ‘The letters was close enough to their Latin pronounciation great vowel-shift consists in a general raising of all so that Chaucer (d. 1400) could rhyme A with omnia: long vowels with the exception of the two high vow- els /i:/ and /u:/, which could not be raised further On which was first i-write a crowned A, without becoming consonants and which were diph- And after, Amor vincit omnia. Prol. 161 thongized into …[ai, au].’ The presumed structural connectedness among the changes is represented in The long vowels acquired their English names the chart below: subsequent to a set of changes known collectively as The Great Vowel Shift (GVS). The GVS is one of the main reasons for the apparent anomalies of English spelling today. The changes covered by this label are universally acknowledged as the most dramatic readjustment of the system of long vowels in the recorded history of English. The phe- nomena subsumed under the general designation of GVS have been the focus of extensive scholarly inquiry during the last 150 years, and the debate is ongoing. The all-encompassing, chain-shift view of the The research history of the GVS encapsulates the GVS dominated the scholarship throughout the history of phonological research in general, and the early parts of the twentieth century. Hypotheses history of English historical language studies: from addressing the initiation, causation, and propagation atomistic description, to structuralist and generative of the putative massive chain shift have been pro- generalizations, and, more recently, increased atten- posed in terms of numerous phonological theories, tion to surface phonetic facts in the context of including structuralism, generative phonology, lexi- regional and social variation. Nineteenth-century cal phonology, dependency phonology, particle philologists described the changes in great detail, phonology, lexical diffusion, and optimality theory. focusing on the results in southern standard British With rare exceptions, a representation of the English, or, more specifically, on Received changes from Middle to Modern English, with each Pronounciation (RP). The latter is the supraregional long vowel linked to an arrow pointing upward or variety of British English based on nineteenth-cen- outward, is still repeated in textbook accounts of the tury southern pronounciations, and fostered until at history of English phonology. The characteristics least the 1960s as the standard in the media and the associated with the position that the GVS was a ‘correct’ pronounciation in (British) English comprehensive and coherent chain shift of the ME instruction around the world. The vowel correspon- dences reconstructed for the changes between late c. 1350-1400 RP Examples Middle English (ME), c. 1350–1400, and twentieth- bind, design, wife century RP are as follows (see Figure 1). fiend, see, tree meal, sea, steak The last century has witnessed both the ‘making’ bake, save, strange and the ‘undoing’ of the GVS. Starting with Karl cow, crown, how do, moon,scoop Luick (1898) and Otto Jespersen (1909), the boat, nose, stone changes shown above have been treated as a unified set of phonological events, where each new value is Figure 1 408  GREAT VOWEL SHIFT monophthongal long vowels, and the reasons why Another problem with such restrictive dating this position has been challenged, are as follows: comes from the post-1700 instances of continuing shifts in non-RP varieties of English. Figure 2 (1) Scope. The GVS affected all and only long shows some examples of post-GVS changes of the vowels. high vowels that are not covered by the traditional It is true that all long vowels underwent some conceptualization of the shift as a single event kind of change during the early Modern English bounded by cut-off time points. The first column period (c. 1450–1750). Nevertheless, any GVS rep- represents the ME input, the second column shows resentation ignoring the other functionally long the reflexes of these high vowels in RP, the third one vocalic units in ME is too restrictive. Diphthongal shows the developments in London English, also variants of the high vowel /i:/, /iy/, existed or arose Australian and New Zealand English, and the fourth in ME from earlier native and borrowed sequences column represents current realizations of these vow- of high vowels and following consonantal /j/ and /h/, els in some southern US varieties of English. thus ME /stiy/ ‘sty’, /sliy/ ‘sly’ from earlier <stig> The changes of the long vowels outside RP can- /stij/, <sligh> /slij(h)/. The high back long vowel /u:/ not be fitted within the conventional GVS picture. had noncontrastive variants /uw/, as in ME /fuwl/ Regarding the GVS as a one-time coherent event, ‘fowl’, /buw(h)/ ‘bough’ from earlier native and bor- which began and ended at well-defined points of rowed <fugol> [fu0(ə)l], <boh> [buw(h)]. In both time, causes us to lose sight of the rich phonetic cases, the diphthongal and the steady-state variants variation that precedes and follows it. Reference to merged toward a diphthongal realization. These ini- the dating of the GVS must be accompanied by the tial diphthongs were further optimized by differenti- realization that it is an artificially isolated portion of ation of their endpoints: /ai/ for bind, design, wife, a continuum, representing the evolution of the entire sty, and /aυ/ for cow, crown, how, fowl. Thus, /ai/ and phonological system of the language through time, /aυ/ are the attested RP results of the gradual change with no starting- and no endpoints. of both ‘pure’ long vowels and minimal diphthongs (3) Mechanism. The GVS started with the high of high vowels followed by a homorganic glide; it and the high mid vowels. was the diphthongal realizations that were formative in terms of the later history of these vocalic units. (a) The high vowels /i:/ and /u:/ were diph- Similarly, the history of the ME mid long vowels /ε:/ thongized first. The raising of the high (as in steak), /a:/ (as in bake), and /ɔ:/ (as in boat) midvowels /e:/ and /o:/ was motivated by cannot be separated from the history of preexisting the existence of vacant slots for /i:/ and diphthongs as in day, play, weigh, blow, stow, dough /u:/. This is the ‘drag chain’ theory that from earlier /dæj, plej, wejh/, /blow, stow, dowh/. originates with Jespersen. Therefore, any description of the reorganization of (b) The impulse for the diphthongization of the vowel system of ME should take into considera- the high vowels /i:/ and /u:/ came from tion both the history of the long vowels proper and the raising of the high mid vowels /e:/ the diphthongal entities with which the long vowels and /o:/. This is the ‘push chain’ theory merged in the course of the GVS. that originates with Luick. (2) Dating. The GVS began by c. 1400, and was The ‘drag chain’ position is now mostly of histor- completed by c. 1700. ical interest. Luick’s observation that there was no diphthongization of /u:/ in those dialects of northern This chronological span can accommodate only ME where /o:/ was fronted to /ø:/ c. 1300, has been some of the changes that shaped the long vowel sys- accepted as a good argument in favor of the causal tem of Modern English. The more dramatic changes link between the initial raising of the mid-vowels associated with the GVS, the full diphthongization and the diphthongization of /i:/ and /u:/ in the south- of the high vowels /i:/ and /u:/ and the raising of the ern dialects. In that sense, the chain linking the high high mid vowels /e:/ and /o:/ in RP, were indeed and the mid high vowels is a ‘push chain’. arguably accomplished within this time-frame. However, the mergers of the ‘pure’ high vowels with preexisting diphthongs that determined the future ME RP Lon, Aus, NZ Am. South of these vowel types were already under way in bind, wife early ME, if not earlier. High-vowel diphthongiza- cow, crown tion and mid-vowel raising can be traced in the spelling starting from c. 1350. Figure 2 409 GREAT VOWEL SHIFT (4) Stages: The GVS proceeded in two distinct ent way, and only a minichain shift can be recon- unrelated stages. structed reliably, the survival of the term Great Vowel Shift as a synonym for a comprehensive chain shift in Separating the developments of the long vowels the history of all English long vowels is in doubt. into two distinct stages is a relatively recent addition to the GVS accounts. Earlier, GVS scholarship had assumed that all long vowels were proper partici- References pants in a single ‘great’ chain shift, yet research dur- Jespersen, Otto. 1909. A modern English grammar on ing the last three decades has shown that only the top historical principles, Vol. 1. Heidelberg: Carl Winter’s two heights of the ME long vowel system were Universitätsbuchhandlung. involved in chain shifting. The low mid vowels of Lass, Roger. 1992. What, if anything, was the Great Vowel Shift? History of Englishes: new methods and interpreta- southern ME, /ε:, ɔ:/ and the low vowel /a:/ began to tions in historical linguistics, ed. by Matti Rissanen, Ossi raise and merge with preexisting /e:~ey/, /o:~ow/ Ihalainen et al. Berlin: Mouton de Gruyter. only in the late sixteenth century. The merger of the ———. 1999. Phonology and morphology. The Cambridge vowels in e.g. see (ME /e:/> /i:/ c. 1550) with the history of the English language, Vol. III, 1476 to 1776, ed. vowel of sea (ME /ε:/>/e:/ c. 1650) did not begin by Roger Lass. Cambridge: Cambridge University Press. Luick, Karl. 1898. Untersuchungen zur englischen Lautgeschichte. until the end of the seventeenth century, and contin- Strassburg: Trübner. ued into the eighteenth century. These later develop- ———. 1921–1940. Historische Grammatik der englischen ments are no longer claimed to be part of the chain Sprache. Leipzig: Tauchnitz. shift that can properly be reconstructed for the four Stockwell, Robert P. 1972. Problems in the interpretation of the top vowels in southern standard British English. great English vowel shift. Studies in linguistics: papers in honor of George L. Trager, ed. by M. Estellie Smith, In summary, the term GVS is not well defined. It 330–53. The Hague: Mouton. has been applied as a cover term for a variety of often ———. 1978. Perseverance in the English vowel shift. Recent unrelated changes of the southern English long vow- developments in historical phonology, ed. by Jacek Fisiak. els and diphthongs that started in ME and can be The Hague: Mouton. shown to continue to this day. In that sense, it Stockwell, Robert P., and Donka Minkova. 1988. The English vowel shift: problems of coherence and explanation. Luick includes raisings, diphthongizations, mergers, and revisited, ed. by Dieter Kastovsky and Gero Bauer. Gunter minichain shifts within the long vowel system of one Narr Verlag: Tübingen. language variety. A redefinition of the term GVS, Stockwell, Robert P., and Donka Minkova. 1997. On drifts and suggested by Lass (1992:153), restricts it to the shifts. Studia Anglica Posnaniensia 31. 283–303. changes of the original high and high mid vowels Wolfe, Patricia. 1972. Linguistic change and the Great Vowel Shift in English, Berkeley: University of California Press. [i:, u:, e:, o:] occurring in southern English in the fif- DONKA MINKOVA teenth to sixteenth centuries. Since the remaining vowels do not participate in chain shifting in a coher- See also English Greek, Modern Modern Greek is the direct descendant of Koine cial language. Modern Greek is also one of the two Greek, the language of the New Testament, which in official languages of the Republic of Cyprus (Turkish turn was based on the Attic dialect of Ancient Greek. is the other one). Greek speakers are located mostly in It is thus part of the Greek or Hellenic branch of Indo- the Republic of Greece itself, with about 10,000,000 European. The language is called eliniká (Hellenic) by living there, but large numbers can also be found in its modern speakers. Up until about the end of the last Cyprus (c. 600,000) and parts of the diaspora (e.g. century, the spoken language was also referred to as about 1,000,000 in Australia, chiefly in Melbourne roméika (Romaic) by the common people who spoke and Sydney). Historically, Greek speakers have settled Greek as citizens of the Ottoman Empire before 1821 all over the eastern Mediterranean, in Southern Italy, and then after the successful war of independence in along the Black Sea coasts, in Egypt, the Levant, 1821–1832 of the Kingdom of Greece. The modern Cyprus, and much of Asia Minor. This geographical Republic of Greece has Modern Greek as its sole offi- spread continued throughout the Hellenistic period 410  GREEK, MODERN and on through the Byzantine and Medieval periods, was the Greek of Athens before the 1821 War of and is valid to some extent even into the Modern era, Independence, and is still found elsewhere in Greece although most of the Greek inhabitants of Asia Minor due to various resettlements. (present-day Turkey) were removed to Greece (and A key aspect of the development of Modern Greek many Greek-speaking Moslems from Greece to pertains to its external history, namely the fact that Turkey) after the population exchanges of the early throughout post-Classical Greek, the language and its 1920s in the aftermath of World War I. Within Greece, speakers were never able to escape the important cul- the greatest concentration of speakers, some 3,500,000 tural influence of the Classical Greek language and or more, lives in the greater Athens area alone, most of Classical Greece itself. The importance of Classical them speakers of the current standard language. Greece—in the Mediterranean, the Balkans, parts of Faced with the difficult problem of distinguishing the Middle East, and even Western and Central between dialects of a language as opposed to separate Europe—meant that Classical Greek was taken as the languages, the highly divergent modern form of Greek prescriptive norm against which speakers of later known as Tsakonian, still spoken in the eastern stages of Greek generally measured themselves. This Peloponnesus (in Greece), could well be considered situation led to a ‘two-track system’ for the language, now as a separate language from the rest of Modern in which a high-style consciously archaizing variety Greek, and the Pontic dialects once spoken along the that speakers and writers modeled on Classical Greek Black Sea coast of Asia Minor but now spoken in was set against a vernacular innovative variety. After many parts of Greece due to the 1923 population the War of Independence from the Ottoman Empire in exchanges are divergent enough to warrant considera- 1821 and the founding of a new nation-state of Greece, tion now as a separate language from the rest of Greek. this distinction crystallized into a significant register Similarly, modern Cypriot shows significant differ- and stylistic difference between what has come to be ences on all levels (phonological, morphological, and known as Katharevusa (‘Puristic’, literally ‘(the) puri- syntactic) that invite classification as a separate lan- fying (language)’) as the high-style variety associated guage, although this judgment is perhaps a more diffi- with official functions, i.e. those pertaining to govern- cult one than in the case of Tsakonian or Pontic. ment, education, religion, and such, and Dimotiki Still, it is customary to treat Modern Greek as a uni- (‘Demotic’, literally ‘(the) popular (language)’) as the fied language with a range of dialects, much as was the language of the people in ordinary, day-to-day, mun- case with Ancient Greek. While the dialect complexi- dane affairs. This sociolinguistic state of affairs was ty of Ancient Greek was largely leveled out in the basis for the formulation of the notion of diglossia Hellenistic times with the emergence of the relatively (Ferguson 1959), and struggles between advocates of unified variety of Greek known as the Koine (see the each type of Greek, carrying with them certain social chapter on Ancient Greek), the natural forces of lan- attitudes and political positions, continued throughout guage change led to new dialect diversity in the most of the twentieth century. After a number of gov- Byzantine period, with the modern regional dialects ernmental acts and actions in 1976, Dimotiki became emerging after about the tenth to twelfth centuries AD. the official language, and the diglossic situation was The main exception to this characterization is resolved, at least from an official standpoint. Tsakonian (as mentioned above), which derives more Significant for understanding variation in Greek is the or less directly from the ancient Doric dialect, fact that all throughout, both the official and unofficial although with an admixture of standard Modern Greek periods of diglossia, speakers’ usage was actually in recent years; in addition, the Greek of Southern somewhat mixed, with borrowing between the two Italy, still spoken, for instance, in some villages in varieties, especially with Puristic forms incorporated Apulia and Calabria, seems to have ancient Doric into Demotic. The present state of Demotic, what has roots. The Pontic dialects (mentioned above) may emerged as ‘Standard Modern Greek,’ hereafter SG, derive more directly from the Hellenistic Koine. based on the everyday Greek of the largest city and Stemming from the later Byzantine form of the capital of Greece (Athens), reflects a number of such Koine, the major modern regional dialects are (follow- borrowings from Katharevusa, involving both gram- ing Newton 1972) Peloponnesian-Ionian, Northern, mar (morphology and syntax) and pronunciation, as Cretan, Old Athenian, and Southeastern (including the well as the lexicon, as discussed below. Greek of the Dodecanese islands and, traditionally at Also relevant along with these stylistic/register dif- least, Cypriot Greek as well). Peloponnesian-Ionian ferences is the effect of orthography. There is a long has formed the basis historically for what has become tradition of written forms of Greek, with the familiar the contemporary Standard language, and is the basis Greek alphabet being the most enduring writing system for the Greek of modern Athens, as by far the leading for the language; as is so often the case, written forms population center in Greece; the Old Athenian dialect tend toward the conservative, especially with respect to 411 GREEK, MODERN the representation of pronunciation. There is thus with- replacement by ‘native’ Greek elements. These efforts in Greek, especially regarding phonology, a basis for have met with varying degrees of success, but in any influence from the written language, and the potential case there are still large numbers of Turkish words in for variation therefrom. Moreover, spelling reforms of the language today, especially at the most colloquial the late 1970s and early 1980s, leading to the so-called and everyday levels of usage. monotoniko (‘monotonic’) system, changed certain Thus, for a number of historical reasons having to aspects of Greek orthography, in particular doing away do in large part with the geographic distribution of with several phonetically irrelevant accent marks and Greek speakers and with the particular circumstances diacritics that reflected Ancient Greek orthography; of the relationship of later Greek speakers with their still, the old orthography can be encountered in books cultural past and heritage, Modern Greek today shows published before 1981 and in private use (e.g. personal considerable variety in its realizations. Regional letters), so that there is variation to be found in the form differences cut across social differences, and all this of written Greek even today. has come despite the existence in most periods of What the long-term diglossia and associated influ- various strong centralized standard forms of the ence from a written language have meant for Greek is language (e.g. archaizing varieties in Medieval and the emergence of dialect differences that are not just early modern times, the demotic standard of today, regional (geographic) in nature. Rather, there are etc.) that have provided norms for prescriptive usage. important socially based distinctions that have been As far as the linguistic structure of SG is concerned, fed by diglossia and by associations between conser- it would perhaps be useful to draw a quick comparison vative social and political attitudes and conservative between the most widely studied Classical Greek (see linguistic usage on the one hand, and progressive atti- entry) and its evolution into Modern Greek. In terms of tudes and innovative linguistic usage on the other. its phonology, Ancient Greek voiceless aspirated stops Thus, within Greek one has to reckon with mixing of /th/ (θ), /ph/ (ϕ), /kh/ (χ) and voiced unaspirated stops /d/ varieties and borrowing among them of both a region- (δ), /b/ (β), /g/ (γ) have changed to voiceless fricatives al and stylistic/social nature. /T/ (T), ,/f/ (ϕ), /x/ (χ) and voiced fricatives /∆/ (δ), /v/ Other types of socially based variation can be found (β), and /⊗/ (γ), respectively, making SG a language too, although beyond the omnipresent one based on the with a very rich fricative inventory. Vowel length is lost Katharevusa vs. Dimotiki distinction. From a functional and all of the Ancient Greek diphthongs have been standpoint, mention should be made of the existence of monophthongized in the modern language (e.g. AG /ei/, certain institutionalized trade jargons, e.g. that of cop- /oi/ becoming SG /i/, AG /ai/ becoming SG /e/, etc.). persmiths, and several varieties of ‘disguised languages’ Furthermore, many AG vowels and diphthongs (e.g. /ν/, (e.g. one involving switching of syllables in a word with /η/, /νι/, etc.) ended up pronounced as /i/ in SG, a phe- some distortions of vowels). Especially well known in nomenon sometimes referred to as iotacism. SG has also this regard is καλιαρντα′ /κalia
rθ da/, the dialect of the lost pitch accent, which has been replaced by lexical gay community that is characterized especially by a (dynamic) stress. Aspiration /h/ has also been lost, and large number of Turkish loanwords and divergent mean- voiced stops have been developed anew, usually arising ings for SG words (see Petropoulos 1971). One might from voicing of voiceless stops when found after nasals also mention here conventionalized child-language or from borrowings (e.g. AG /pente/ ‘five’ giving SG forms (e.g. with s/z for SG dental fricatives T/∆ , and /pe(n)de/ or /duvari/ ‘wall’ from Turkish). In terms of various lexical items, as for bodily functions) that all syntax, the modern language retains the free word-order (adult) speakers know and are able to use in appropriate capabilities of AG since most of the morphology and the situations (e.g. talking with young children). inflections have been retained. Adjectives normally pre- Also of importance for the issue of the mixing of cede nouns, and definite articles have to accompany varieties in Greek is the presence of other languages in proper nouns like in Ancient Greek (e.g. o giorgos ‘the Greece and in the territory surrounding Greece in the George’). The dative case has been completely replaced Balkans down through the ages and even into modern by genitive and accusative and many AG monolectic ver- times. These circumstances have led to the steady bal forms have been replaced by periphrastic construc- entry of numerous foreign words into Greek over the tions (e.g. AG eureka ‘I have found’ has become SG exo years, from Balkan, Middle Eastern, and more recent- vri). The infinitive form of the verb has also been lost, ly Western European languages, setting the stage for reflecting perhaps a Balkan pattern of language contact. variation in the use and integration of loanwords on the part of Greek speakers. In the modern era, there have been periods of reaction against the influx of References loanwords, with sometimes Italian but especially Browning, Robert. 1983. Medieval and modern Greek. Turkish words being the prime targets for purging and Cambridge: Cambridge University Press. 412  GREENBERG, JOSEPH HAROLD Holton, David, Peter Mackridge, and Irene Philippaki- Mirambel, André. 1939. Précis de grammaire élementaire du Warburton. 1997. Greek. A comprehensive grammar of the grec moderne. Paris: Société d’éditions ‘Les Belles modern language. London/New York: Routledge. Lettres’. Horrocks, Geoffrey. 1997. Greek. A history of the language and Newton, Brian. 1972. The generative interpretation of dialect. its speakers. London/New York: Longman. Cambridge Studies in Linguistics, Vol. 7. Cambridge: Joseph, Brian D., and Irene Philippaki-Warburton. 1987. Cambridge University Press. Modern Greek. London: Croom Helm Publishers. Petropoulos, Elias. 1971. Kaliarnta: Erasitechnike Glossologike Mackridge, Peter. 1985. The modern Greek language. Oxford: Ereuna. Athens: Digamma. Clarendon Press. GIORGOS TSERDANELIS Greenberg, Joseph Harold Joseph H. Greenberg was one of the most original and Eurasia. He published only preliminary results at that influential linguists of the twentieth century. Educated time, but he continued to collect evidence, and assem- at a time when there were virtually no linguistics bling that evidence took up the last 20 years of his life. departments, he pursued studies in classics, anthropol- In 1971, he published an article presenting evidence ogy, Indo-European comparative linguistics, and phi- that the languages of Papua New Guinea that do not losophy. Greenberg also shaped the empirical belong to the Austronesian family—the so-called scientific study of linguistics, emphasizing the exami- Papuan languages—and other non-Austronesian lan- nation of a broad range of languages across the world guages spoken as far afield as the Andaman Islands in the development of theories of language structure, and Tasmania form a single large family, Indo-Pacific, language change, and language classification. with 14 major branches and a number of languages of Greenberg’s earliest major work was on the genetic uncertain affiliation. The Indo-Pacific hypothesis was, classification of the languages of Africa. Prior to however, largely ignored by specialists in the Papuan Greenberg’s work, the languages of Africa had been languages. Although few linguists working in the area classified into five families, but the classification used currently believe that all Papuan languages belong to a inappropriate linguistic evidence and also nonlinguis- single family, a large grouping called the Trans-New tic evidence. Greenberg established three fundamental Guinea Macrophylum has been proposed by some, principles in classifying languages. The first is to and a number of its branches, as well as other accept- exclude typological linguistic features. Typological ed Papuan families, correspond to branches of features are patterns of sound only—such as the pres- Greenberg’s Indo-Pacific family. ence or absence of tones—or of meaning only—such In 1987, Greenberg published a book presenting as the presence or absence of sex gender distinctions. evidence that all of the indigenous languages of the The most reliable evidence for genetic classification is Americas other than the Na-Dene family in the Pacific the pairing of sound and meaning, especially in gram- Northwest and the Eskimo-Aleut family in the Arctic matical inflections and basic vocabulary. The second belong to a single family, Amerind, with 11 major principle is to exclude nonlinguistic evidence, such as branches. This proposal has also been strongly criti- skin color or cultural traits. The third principle is the cized and stimulated a lengthy debate on the methods simultaneous comparison of the vocabulary and and principles of linguistic genetic classification, to inflections of a large number of languages (mass com- which Greenberg contributed until his death. parison, later called multilateral comparison). The central criticisms of Greenberg’s methods are Greenberg used these principles to classify the lan- that a genetic family cannot be established without guages in Africa into four families: Afroasiatic, Nilo- reconstructing the ancestral language of the family; Saharan, Niger-Kordofanian, and Khoisan. His without doing so, it is impossible to determine whether classification was strongly criticized by established resemblances in inflections and vocabulary are due to African scholars in Europe at first, but it was ultimate- chance or borrowing, and whether the results are affect- ly accepted by virtually all African linguists. ed by errors in the transcription of the data. Also, since While working on the African classification through reconstruction is progressively more difficult as time the 1950s, Greenberg also turned his attention to lan- depth increases, one cannot prove genetic hypotheses for guage classification in Oceania, the Americas, and families older than about 8,000 years. The conclusion of 413 GREENBERG, JOSEPH HAROLD most of Greenberg’s critics, therefore, has been not so and the relative clause follows, and for both modifiers to much that Greenberg’s hypotheses are wrong, but that precede or follow; it only disallows the type in which the they are not provable in principle. relative clause precedes and the demonstrative follows. Greenberg argued in response that one cannot Greenberg’s first major publication on language reconstruct a family without a prior hypothesis as to universals proposed a series of universals of word which languages form the family, since the number of order and morphological categories. The paper used a possible groupings of even a small number of lan- sample of 30 languages and the proposed universals guages is astronomically high; and that genetic classi- were inferred from the distribution of languages in the fication is only the first step, to be followed by sample among possible structural types. Among other reconstruction. Greenberg also argued that the proba- things, the paper introduced the language types based bility of many languages simultaneously displaying on the relative order of subject (S), verb (V), and similarities in form and meaning is so high that chance object (O), such as SVO, SOV, etc. This paper is one resemblances (and the interference of errors) decrease of the most cited papers in linguistics, and Greenberg’s in likelihood. Greenberg also argued that borrowing word-order types have been used widely in all theoret- displays specific probabilistic patterns that can be ical approaches to linguistic analysis. detected, and that differences in the relative stability of Greenberg’s method of deriving universals became basic vocabulary and grammatical inflections mean the fundamental method of the typological approach to that one can identify families older than 8,000 years. grammar, and the typological approach was often The debate is not yet resolved, in part because it compared to the generative approach of Noam requires the use of quantitative techniques not yet Chomsky. The typological approach to grammar is applied to the problem. However, one can safely say characterized by: cross-linguistic comparison of a that Greenberg’s work has revived a subfield of linguis- sample of genetically and geographically diverse lan- tics that had been dormant for several decades, and has guages; classification of the languages into types stimulated new research in language classification, lan- based on their structural properties; the inductive der- guage contact, and historical linguistics. Greenberg’s ivation of language universals by examining attested work has also brought linguistics into contact with vs. unattested (or very rare) language types in the sam- anthropology and genetics. Greenberg’s three-family ple; and explanation of the universals, usually by linguistic hypothesis for the Americas found parallels in appeal to functional motivations (semantic, pragmatic, research on human dentition and genetics, and geneti- discourse-functional, or processing principles). cists such as Luigi Luca Cavalli-Sforza have compared Greenberg’s paper on word order was only one of Greenberg’s deeper linguistic classifications to genetic many typological studies that he produced. Many of evidence on prehistoric human migrations. these studies pertained to sound structure: on the conso- Greenberg published evidence for a family in nant–vowel dichotomy, consonant clusters, glottalized Eurasia consisting of seven branches: Indo-European, consonants, and word prosodic systems. Others were Uralic-Yukaghir, Altaic, Japanese-Korean-Ainu, Gilyak concerned with grammar: besides word order, (Nivkh), Chukotian, and Eskimo-Aleut. This work, is as Greenberg produced studies of numeral systems and of controversial as his Amerind hypothesis. a typological interpretation of the concept of markedness Greenberg’s other major line of research, also begun in and markedness hierarchies in inflectional categories. the 1950s, is in the area of typology and universals of lan- In the 1960s, Greenberg became more interested in guage. Typology began as the study of the range of differ- universals of language change, which he christened ent structural types of languages, according to their diachronic typology. Greenberg produced a series of grammar (morphology and syntax) or their sound system seminal methodological papers in diachronic typology, (phonology). Languages vary considerably in structure. and also a number of studies of universals of language Greenberg linked typology to the study of universals of change, including voiceless vowels, numeral classifier language through the discovery that structural variation constructions, and gender marking. The last study across languages was limited, and those limits can be helped begin the revival of the study of grammatical- described in terms of implicational universals. An impli- ization—the evolution of grammatical categories and cational universal is a universal of language formulated as constructions from independent words and syntactic an if–then statement, as in ‘If a relative clause precedes the patterns. Grammaticalization theory is now a central noun in a language, then a demonstrative adjective also area of research in historical linguistics and typology. precedes’. An implicational universal describes a relation- Greenberg’s seminal methodological and empirical ship between two structural properties of a language; but papers in typology and universals, including diachronic it also allows for cross-linguistic variation in language typology, created the foundations of several major con- type. For example, the aforementioned universal allows temporary strands of linguistic research. Greenberg also for language types in which the demonstrative precedes made significant contributions to African linguistics, 414  GRICE, H. PAUL sociolinguistics, and psycholinguistics. Although the Prize for Social Science in 1997. Greenberg died in fate of his genetic classifications is still being debated, Stanford, California on May 7, 2001. Greenberg’s legacy in the empirical, cross-linguistic study of language universals and language change is References undoubtedly secure. Comrie, Bernard. 1989. Language universals and linguistic typology, 2nd edition. Chicago, IL: University of Chicago Press. Biography Croft, William. 2001. Joseph Harold Greenberg (1915–2001). Joseph Harold Greenberg was born in Brooklyn, New Language 77. 815–30. _________. 2003. Typology and universals, 2nd edition. York on May 28, 1915; he married Selma Berkowitz Cambridge: Cambridge University Press. on November 23, 1940. He received his B.A. (1936) Croft, William, Keith Denning, and Suzanne Kemmer (eds.) 1990. from Columbia University and Ph.D. (1940) for his Studies in typology and diachrony for Joseph H. Greenberg. dissertation on the influence of Islam on the Hausa, Amsterdam and Philadelphia, PA: John Benjamins. tutored by Melville Herskovits, Northwestern Greenberg, Joseph Harold. 1957. Essays in linguistics. Chicago: University of Chicago Press. University. He was with the US Army Signal _________. 1963. The languages of Africa. Bloomington: Intelligence Corps, 1940–1945; Assistant Professor, Indiana University Press. University of Minnesota—Minneapolis, 1946–1948; _________ (ed.) 1963. Universals of grammar. Massachusetts: Assistant Professor, 1948–1953, Associate Professor MIT Press Cambridge. 1953–1957, Professor, 1957–1962, Columbia _________. 1966. Language universals, with special reference to feature hierarchies. The Hague: Mouton. University; Professor, Stanford University, 1962–1985; _________. 1968. Anthropological linguistics. New York: and Ray Lyman Wilbur Professor of social sciences in Random House. anthropology, 1971. He was also Director of African _________. 1971. Language, culture and communication: essays Languages and Area Center, 1967–1978 and co-editor by Joseph Greenberg. Stanford: Stanford University Press. of Word, 1950–1954. He was a Fellow of the National _________. 1974. Language typology: a historical and analytic overview. The Hague: Mouton. Academy of Sciences, the American Philosophical _________(ed.). 1978. Universals of human language, 4 vols. Society, and the American Academy of Arts and Stanford: Stanford University Press. Sciences, and a Guggenheim Fellow in 1954–1955, _________. 1987. Language in the Americas. Stanford: 1958–1959, and 1982–1983. He was also Fellow, Stanford University Press. Center for Advanced Study in the Behavioral Sciences _________. 1990. On language: selected writings of Joseph H. Greenberg. Stanford: Stanford University Press. in 1959–1960 and 1965–1966, and Stanford _________. 2000/2002. Indo-European and its closest relatives: Humanities Fellow in 1982–1983. He was the First the Eurasiatic language family, 2 vols. Stanford: Stanford Distinguished Lecturer, American Anthropological University Press. Association in 1970 and Hermann Collitz Professor, _________. 2005. Genetic linguistics: theory and method. Summer Institute of the Linguistic Society of America Oxford: Oxford University Press. Heine, Bernd, and Derek Nurse (eds.) 2000. African languages: 1987. He was President of the African Studies an introduction. Cambridge: Cambridge University Press. Association, 1964–1965; Chairman of the West African Newman, Paul. 1995. On being right: Greenberg’s African lin- Linguistics Association, 1965–1966; and President of guistic classification and the methodological principles the Linguistic Society of America, 1977. He received which underlie it, Bloomington, IN: Institute for the Study the Haile Selassie Award for African Research in 1967, of Nigerian Languages and Cultures, African Studies Program, Indiana University. the New York Academy of Sciences Behavioral WILLIAM CROFT Sciences Award in 1980, and the National Academy of Sciences/Academy of Arts & Sciences Talcott Parsons See also Afroasiatic; Grammaticalization; Typology Grice, H. Paul In the early period of his career, H. Paul Grice partici- ses of ordinary language philosophers, such as pated in Oxford’s ordinary language philosophy Strawson’s analysis of logical connectors (e.g. ‘and’, school. Some of his work is critical of particular analy- ‘or’, and ‘if’) and Austin’s ‘no modification without 415 GRICE, H. PAUL aberration’ thesis. Nonetheless, Grice retained a dis- ventionality as the basis for a general theory of lan- tinct interest in ‘conceptual analysis’ (i.e. the analysis guage in a behaviorist mold (Bennett 1976) or as the of the possible uses of a linguistic expression) as a foundation of a cognitive theory (Sperber and Wilson privileged, albeit not exclusive, method of philosophi- 1986). cal analysis. Grice’s intentionalism is not limited to meaningnn; The most significant contributions of Grice to lin- on the contrary, he distinguishes between what is said guistics revolve around a broad attempt to reduce in an utterance (roughly, literal meaning) and what is semantics to intentionality, via the notions of mean- implied by it. For the latter notion, Grice introduced ingnn and implicature. Meaningnn (nonnatural) is the neologism implicature. Recent work has ques- opposed to natural meaning. While natural meaning is tioned the division said/implied and introduced a third independent of any intentionality (e.g. the presence of term (variously labeled explicature/impliciture) to smoke means that there is a fire), meaningnn is defined denote the aspects of the meaning of an utterance that specifically in terms of the speaker’s intention that the while not explicitly stated, and hence derived prag- hearer recognize his or her intention to mean some- matically, are nonetheless not part of implicature. thing. The original formulation of the idea underwent In its simplest form, the notion of implicature can several reformulations under the pressure of several be defined as some proposition that while not explicit- counterexamples (Schiffer 1972), but the following is ly stated by the speaker is nonetheless intended by him particularly perspicuous: the utterer (U) of an utter- or her and is understood by the hearer as such. ance x, meant something, if U uttered x intending, Implicatures may be context-free (generalized) or con- given an audience (A), text-sensitive (particularized). In the former case, they apply in any utterance of a sentence, regardless of the (1) A to produce a particular response r circumstances/context. In the latter, they are sensitive (2) A to think (recognize) that U intends (1) to the context. Both types of implicatures are governed (3) A to fulfill (1) on the basis of his fulfillment of by the cooperative principle (CP), arguably Grice’s (2) (Grice 1989:92). best-known contribution to linguistics. The CP con- This M-intention (Grice 1989:105) or ‘reflexive sists of a general statement: intention’ (Searle 1969:47) is taken to be the defining make your contribution such as is required, at the stage feature of meaningnn. A significant issue, which seems at which it occurs, by the accepted purpose or direction to have received very little attention (Neale 1992), is of the talk exchange in which you are engaged that all reformulations of the M-intention fall victim to counterexamples in which the speaker’s motives are and of four maxims: deceptive. Grice suggests (1989:302–3) blocking out ● The maxim of Quality ‘sneaky’ intentions by stipulating that one cannot mean deceptively. This provides an interesting link between try to make your contribution one that is true; the concept of meaningnn and the cooperative principle, specifically specifically the maxim of Quality (see below). do not say what you believe to be false The conventional nature of the speaker’s reliance do not say that for which you lack adequate on meaningnn has been pointed out (Lewis 1969). A evidence potentially problematic infinite regression of ‘mutual ● The maxim of Quantity knowledge’ (Lewis 1969, Schiffer 1972) has also been noted. The issue revolves around the fact that the hear- make your contribution as informative as is er knows that the speaker intended a proposition P, but required for the current purposes of the the speaker must know that the hearer knows that the exchange speaker intended P, and so on. Many have challenged do not make your contribution more the original definition of the nature of the M-intention informative than is required on the basis of this alleged infinite recursion. Various ● The maxim of Relevance solutions have been proposed to this problem. Others have challenged the exclusive reliance on intentions, make your contribution relevant without use of the conventional (literal) meaning of ● The maxim of Manner the utterance (Searle 1969). Grice, in fact, argues that conventional linguistic meaning can be reduced to be perspicuous and specifically intended speaker meaning. The meaningnn thesis avoid obscurity remains controversial and has in fact been repudiated avoid ambiguity by some of its early proponents (Schiffer 1987). be brief Others have used Grice’s meaningnn and Lewis’s con- be orderly (Grice 1989:26–27). 416  GRICE, H. PAUL The exact relationship between the general princi- more maxims, or even to proliferate principles. ple and the maxims has been the object of controver- Recently, a proposal to augment the CP with a set of sy. The generally accepted opinion is that the maxims heuristics, based on the maxims, has been presented instantiate the principle. Implicatures can be drawn on (Levinson 2000). These heuristics generate general- the basis of the CP by either following its rules or by ized conversational implicatures, which are default deliberately and conspicuously not doing so (called implicatures that hold in most contexts. flouting). If the speaker follows the maxims, then the Overall, however, the critical voices have been hearer can assume that he or she is being truthful, rel- drowned out by the sheer mass of those who have silent- evant, etc. For example, the utterance of ‘Mary won ly accepted and incorporated a more or less faithful the Nobel prize’ implicates that the speaker has ade- Gricean distinction between what is said and what is quate evidence that she did so. On the contrary, if the implicated by an utterance. It is fair to say that the dis- speaker is obviously not following the maxims, and tinction has entered the realm of the basic linguistic makes no attempt at concealing this, then the hearer notion that is part of the curriculum taught to beginners. can seek an explanation. For example, as the phone In the Californian phase of his career, Grice focused rings, a speaker says ‘I am not here’. Since the viola- his attention toward ethics (1991). This aspect of his tion of the CP is obvious, the hearer considers the work is not unrelated to the research on the problems utterance relevant to the context (the incoming phone of meaning, although most linguistic applications have call) and therefore assumes that the speaker does not neglected to assess the connection between the two wish to talk to whoever is calling. aspects of Grice’s thought. For example, Grice Reactions to Grice’s CP have ranged widely. There attempted to provide a nonrelativistic, metaphysical have been total rejections based on accusations of foundation for value judgments, such as those of naiveté or on the grounds that, simply put, people do morality. This led Grice to the adoption of a substan- not generally behave as Grice’s CP would predict. tialist view of rationality. A substantialist (i.e. nonin- These objections can be countered by the observation strumental) rationality is concerned not only with the that Grice does not claim that people are always coop- rationality of the means to an end but also with the erative. Others have claimed that the CP is prescriptive rationality and morality of the ends themselves. The (i.e. tells speakers how they should behave). This view connection with the cooperative principle is thus clear. is in error, as stated in the literature, but deserves Another example of connection between Grice’s ethics reconsideration in light of Grice’s views on morality. It and his semantics is his insistence that the reflexive has been claimed that the CP has been empirically fal- intention of meaningnn is rational and hence evaluative sified by anthropological evidence and that therefore (1989:283–303). its universal nature has been falsified, but these claims have been repeatedly refuted. More moderate, but still Biography negative, reactions have embraced the general impetus of the theory but claimed that the CP and/or the max- Herbert Paul Grice (1913–1988) began his career in ims were too specific and/or too general and/or too his native England where he graduated from Oxford’s vague. A related criticism is that no explicit procedure Corpus Christi College in 1935 and received an M.A. is provided to calculate the implicatures. Several in philosophy from Merton College in 1938. He authors have remarked that the CP in its original word- became a fellow of St. John’s College, where he taught ing is limited to declarative statements and have from 1939 until 1967 when he moved to Berkeley, offered rewordings that broaden its scope to non- California. In 1967, Grice delivered the William James indicative and nonassertive utterances. Considerable lectures (published partially in 1975 and fully pub- effort has gone into attempts at formalizing the indi- lished in 1989), which were widely circulated in man- vidual maxims, especially the maxim of Relevance, uscript. The widespread interest in the idea of within the framework of Relevance Theory, and the implicature dates back to these lectures. He taught at maxim of Quantity. The judgment on the effectiveness Berkeley until his retirement in 1979. Between 1980 of these attempts is also mixed. Many have tried to and 1983, he taught at the University of Washington. determine the criteria for deriving conversational In 1983, Grice delivered the Carus lectures on the implicatures and distinguishing them from other conception of value, published posthumously as Grice semantically related propositions, such as presupposi- (1991). Significant parts of Grice’s writing remain tions. It has been suggested to reduce the maxims to unpublished. The move from Europe to the United two or even one principle, often based on processing States parallels a shift of interests in his work. In constraints, thus linking to the functionalist tradition the early part of his career, Grice adhered to and (e.g. the principle of least effort). Conversely, there contributed to the development of the ordinary lan- have been many proposals to augment the CP with guage philosophy of Austin, Ryle, and other Oxford 417 GRICE, H. PAUL philosophers, although he always maintained a dis- 16–19, 1990. General Session and Parasession on the tance from this school. During this period, he worked Legacy of Grice. Berkeley, CA: Berkeley Linguistics Society. primarily on problems of meaning. In the second Levinson, Stephen C. 1983. Conversational implicature, Ch. 3. phase of his work, he moved toward issues of ethics Pragmatics, 97–167. Cambridge: Cambridge University and metaphysics. Grice’s influence on linguistics is Press. largely restricted to the first phase of his work. Levinson, Stephen C. 2000. Presumptive meanings: the theory of generalized conversational implicature. Cambridge, MA: MIT Press. References Lewis, David K. 1969. Convention: a philosophical study. Bennett, Jonathan. 1976. Linguistic Behaviour. Cambridge: Cambridge, MA: Harvard University Press. Cambridge University Press. Neale, Stephen. 1992. Paul Grice and the philosophy of lan- Cosenza, Giovanna. 1997. Intenzioni, significato, comuni- guage. Linguistics and Philosophy 15. 509–59. cazione. La filosofia del linguaggio di Paul Grice. Bologna, Schiffer, Stephen R. 1972. Meaning. Oxford: Clarendon. Italy: Clueb. Schiffer, Stephen R. 1987. Remnants of meaning. Cambridge, Grandy, Richard E., and Richard Warner. 1986. Philosophical MA: MIT University Press. grounds of rationality: intentions, categories, ends. Oxford: Searle, John. 1969. Speech acts: an essay in the philosophy of Clarendon. language. Cambridge: Cambridge University Press. Grice, Paul. 1989. Studies in the way of words. Cambridge, Sperber, Dan, and Deirdre Wilson. 1986. Relevance: communi- MA: Harvard University Press. cation and cognition. Oxford: Blackwell, 2nd edition, 1995. _________. 1991. The conception of value. Oxford: Clarendon. SALVATORE ATTARDO Hall, Kira, et al. 1990. Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society. February See also Meaning Grimm, Jacob Together with his brother Wilhelm, Jacob Grimm is other ancient Indo-European languages. For this rea- perhaps the most influential scholar in the field of son, his work acquired primary importance for Indo- Germanic linguistics in the nineteenth century. The European linguistics, and he is nowadays regarded as sons of a lawyer, both brothers studied law in one of the founders of Indo-European comparative Marburg, and became interested in the history and tra- studies. ditions of Medieval Germany. Jacob Grimm’s first Although the Deutsche Grammatik included four publications, all of which issued from close collabora- volumes (after the first, the others followed in 1826, tion with Wilhelm (a constant companion to his broth- 1831, and 1837), what earned Jacob Grimm his fame er’s life and work), were editions of Medieval texts, in the field of Indo-European linguistics is constituted and the famous collection of fairy tales (Kinder-und by the first chapter of the first volume, a later addition Hausmärchen, published starting with 1812). to the 1822 second edition, devoted to sound struc- Research on ancient texts and on oral tradition tures. In this chapter, called in the German original finally led Jacob Grimm to develop an interest in lan- text Von den Buchstaben, Jacob Grimm set out to pro- guage change, which at his time mostly meant tracing vide a description of correspondences in the pronunci- the history of particular words. Very soon, Jacob ation of words, based on comparative research and Grimm turned to grammatical description. In 1819, he following his conviction that pronunciation is published the first volume of his most important work, autonomous from the rest of grammar. Grimm decid- the Deutsche Grammatik. The title should not be taken ed to add this important chapter after reading Rask’s as meaning ‘German grammar’ in a strict sense: the Icelandic grammar, in which Rask described what was word deutsch here is used by Grimm as including all later to be known as the ‘first sound shift’ (erste Germanic languages, at all their documented stages. Lautverschiebung) or ‘Grimm’s Law’. Grimm conceived of language as an ever-changing Although Rask must be credited with having been phenomenon, which can be studied only through the first to understand the sound change that now bears empirical observation. Consequently, he did not only Grimm’s name, Grimm had already demonstrated his include all early stages of the Germanic languages in own insight into sound change in the description of his grammar, but also referred to correspondences in particular vowel alternations (Umlaut and Ablaut) he 418  GUAYMÍ AND CHIBCHAN LANGUAGES gave to G.F. Benecke, the editor of many Medieval Latin elements, which make it perhaps the most pow- texts, in 1816. Grimm observed that, much in the same erful human language. way as certain Gothic consonants had undergone a Jacob Grimm did not lead the life of a retired schol- one-step change away from the related Greek and ar: he was active in the political events of his age, and Sanskrit, certain High German consonants had moved in 1848 he was, for a short time, member of the par- one step more in the same direction, as exemplified in liament of the German Republic. Especially because the following table: of his research on the history and traditions of the Germanic peoples, he was considered, already in his Greek p b ph t d th k g kh times, one of the fathers of German culture. became Gothic f p b th t d h k g which Biography became High Jacob Grimm was born in Hessen (Germany) on January German b/v f p d z t g ch k 4, 1785. He pursued juridical studies in Marburg under Hence, Grimm described consonant shift as a recur- Friedrich von Savigny (1779–1861), then assistant of rent tendency of Germanic, and called the sound shift Savigny in Paris in 1805. After various appointments in from Gothic to High German zweite Lautverschiebung, Kassel, he returned to Paris as a representative of the or ‘second consonant shift’. Prussian diplomacy and took part in the Vienna Research on the history of Germanic and other Congress in 1815. From 1816 to 1829, he worked in Indo-European languages did not exhaust Grimm’s Kassel as a librarian. From 1830 to 1837, his first interest in language. Apart from further writings in the appointment was as a professor in Göttingen. In 1837, field (among which it is worth remembering the he lost his position as a consequence of having protest- Deutsches Wörterbuch, started in 1854, again with the ed against the revocation of the constitution and returned collaboration of Wilhelm, and never finished), Jacob to Kassel. In 1840, he became a member of the Prussian Grimm also published an essay on the origin of lan- Academy of Sciences in Berlin. He moved to Berlin in guage (Über den Ursprung der Sprache) in 1851. As 1841. He died in Berlin on September 20, 1863. habitual in the first part of the nineteenth century, Grimm conceived of language evolution as a process References of decay from the ‘perfect’ ancient Indo-European Collinge, N. E. 1985. The laws of Indo-European. Amsterdam languages with complex systems of inflection to the and Philadelphia, PA Benjamins. ‘degraded’ modern languages that use simpler struc- Grimm, Jacob. 1822–1837. Deutsche Grammatik I-IV, 2nd edi- tures with independent words instead. However, his tion. Göttingen; reprinted Hildesheim: Olms, 1967. deep empirical knowledge of language led him to –––––––––. 1970. Geschichte der deutschen Sprache, 2 vols. understand that language change did not necessarily Leipzig: Weidmann, reprinted Hildesheim: Olms. –––––––––.1864–1890. Kleinere Schriften, 8 vols. Berlin: impoverish languages, but that, on the contrary, the Gütersloh, Dümmler, Bertelsmann. loss of old categories is compensated by the creation Morpurgo Davies, Anna. 1996. Nineteenth century linguistics. of new ones, and that the latter are not necessarily vol. 4, London: Longman. History of linguistics, ed. by ‘worse’ than the former, even if their implementation Giulio Lepschy. does not involve the same type of complexity. Thus, Sebeok, Thomas A (ed.) 1966. Portraits of linguists. Bloomington and London: Indiana University Press. Grimm, in a very original way, and partly contradict- SILVIA LURAGHI ing his own initial statements, ends up praising English as the perfect combination of Germanic and See also German; Gothic Guaymí and Chibchan Languages The Chibchan languages are spoken in a wide area Panama, Colombia, and the West of Venezuela. This lan- extending from Northeastern Honduras, through the guage family constitutes the largest family in Central Atlantic Coast of Nicaragua, most of Costa Rica, America and Colombia, and the best-studied family of 419 GUAYMÍ AND CHIBCHAN LANGUAGES the linguistic area known as Intermediate. The name Bocotá Panama 2,500 Chibcha stems from the civilization established around Cuna Panama, 70,000 Colombia Bogotá (Colombia) at the time of the Spanish conquest; Ika its language, also known as Muisca, was a member of (Bíntucua, Arhuaco) Colombia 7,000 this family. The term Chibcha has been used in many Damana (Arsario, Colombia 1,500 cases as a ‘ragbag’ for languages whose affiliation to Malayo, Guamaca) other (larger) families is uncertain. In the past, languages Kogi (Cágaba) Colombia 8,000 Tunebo (UwCuwa) Colombia 1,800 as disparate as Paez (Ecuador), Tarasco (Mexico), mem- Chimila Colombia 900 bers of other families (e.g. Carib, Aztec), and even lan- Barí (Motilón) Colombia- guages from as far as Chile (Atacama) and Argentina Venezuela 2,000 (Allentiac) have been labeled ‘Chibcha’. Recent, modern glottochronological studies, in which parallelisms There seem to be light differences between the lan- between languages are used to date a potential ancestor guages of Colombia and those of Central America in a language, have yielded a less extensive family. few respects. Languages in Colombia seem to use The geographic distribution of the family has given more auxiliaries and agreement markers. Clause con- rise to some controversy. The fact that only a few nectors as part of verbal morphology are absent in members are spoken in the peripheral north has given Central America (except in Rama), but present rise to two views: the Theory of North Migration and in Colombia. All these differences point to a more the Centrifugal Expansion Theory. The former posits elaborate verb structure in Colombia than in Central Colombia as the epicenter of the Chibchan languages, America. from where northbound migratory movements took Central America is a convergence zone. Two cultur- place in relatively late times (the name Chibcha al and linguistic areas meet there, that known as applied to the whole family mirrors in this view). Mesoamerica and the so-called Intermediate Area. In Glottochronological studies, however, do not seem to general, the Mesoamerican languages (Mayan, Aztec) back up this position (which is not based on linguistic exhibit a verb-initial basic word order pattern of data anyway), but tend to favor the latter, according to V(erb)–S(ubject)–O(bject) or VOS, they have preposi- which the original Chibchan territory was the present- tions, a possessed–possessor and an adjective–noun day border zone between Costa Rica and Panama, at order; the languages of the Intermediate Area (includ- the Talamanca mountain range, which cross-cuts that ing Chibchan), on the other hand, are SOV, they border; from there, migratory movements both north- have postpositions, and possessor–possessed and bound and southbound took place. Recent anthropo- noun–adjective orders. logical research tends to back the Centrifugal Expansion Theory. The Chibchan languages are thus Central American languages. Structural Features There are 24 members in the family, eight of which Basic Word Order are now extinct: Huetar (Costa Rica); Dorasque and The default word order in Chibchan is exceptionless: Chánguena (Panama); Muisca, Duit, Atanques, Catío, Subject–Object–Verb. The other possible orders occur and Tairona (Colombia). Others are facing imminent under the following conditions: (a) use of pronouns, (b) extinction, e.g. Paya (Honduras), Rama (Nicaragua), subject occurring in postverbal position; and (c) implied Guatuso, and Boruca (Costa Rica), having less than subject. Option (a) provides the widest range of possi- 300 speakers. The majority of the living languages bilities. If the verb carries agreement markers identify- have an average of 3,000 speakers and are, at least ing both subject and object, the word order is relatively temporarily, not seriously threatened by extinction. flexible: s–o–V (see (1) below), o–s–V (see (2) below), The living members, their geographic distribution, or o–V–s (see (3) below), but not s–V–o: and the number of speakers follow (with alternative names in parentheses): (1) mu-nu-pash-ka DAMANA you-me-hit-FACTUAL Paya (Pech) Honduras 600 ‘You hit me’ Rama Nicaragua 24 (2) ma ni-sung-u RAMA Guatuso (Maleku) Costa Rica 365 you I-see-PAST Bribri Costa Rica 6,000 ‘I saw you’ Cabécar Costa Rica 2,500 (3) mi-tšua-na-rua IKA Boruca (Brunca) Costa Rica 4 you-see-far-I Teribe (Naso) Costa Rica, ‘I saw you’ Panama 1,500 Guaymí Costa Rica, As for alternative subject position (b), in some lan- (Ngäbére,Movere) Panama 150,000 guages it can be placed postverbally; the object 420  GUAYMÍ AND CHIBCHAN LANGUAGES TIJUANA MEXICALI NORTH AMERICA CIUDAD JUARAZ HERMOSILLO CHIHUAHUA TARREON CULIACAN BAHAMAS HAVANA GUADALAJARA MEXICO SANTA CLARA CANCUM MEXICO MERUDA MORELIA VERACRUZ CUBA GUANTANAMO CUERNAVACA ORIZABA YUCATAN DOMINICAN PENNINSULA HAITI REPUUBLIC JAMAICA Port-AU-Prince santo Domingo ACAPULCO SAN SALADOR BELIZE KINGSTON BEOMOPAN SAN PEDRO SULA GUATEMALA HONDURAS TEGUCIGALPA GUATEMALA SAN SALVADOR EL SALVADOR NICARAGUA MANAGUA SAN JOSE PANAMA SOUTH COSTA RICA PAMAMA AMERICA Geographic Location of the Chibchan Languages. appears in sentence-initial position for highlighting mal status of the class markers either as prefixes, as in purposes, as in Cabécar: Teribe, or as suffixes, as in Bribri, Kogi, or Cabécar. (4) tkabë su-wá ét-aba yís të Most languages with noun classifiers have around five snake classes—some have more: animate, long, round, thin, see-PERFECTIVE human; in most languages, those classifiers are free one-long I forms, i.e. words that can stand independently. Teribe, ‘A snake I saw’ for example, has six classes: (a) prototypically animate The implied subject realization (c) is illustrated in objects; (b) round objects; (c) wide objects; (d) long (7) below. objects; (e) long and wide objects; and (f) objects that can be counted in plots. The expression of each of the classes is effected through prefixes attached to a Word Classes and Parts of Speech numerical base. For example, the base for the number There are two basic types of lexical words in one is ara, to which the class marker prefix is added: Chibchan, around which most of the grammatical fea- kl-ara (animate), kw-ara (round), k-ara (wide), pl-ara tures and categories revolve; these are nouns and (long), kwan-na (long-wide), and kri-na (plots). verbs. A second set of word classes is formed by The Chibchan languages use demonstrative deter- adjectives and adverbs; these are second to the former miners, which are free forms, to express spatial rela- in that although there are clearly identifiable adjectival tionships. The languages can be grouped into those (colors and basic qualities) and adverbial (time and that have three degrees of spatial distance (Teribe, manner) roots, their number is not that large, and most Rama, Cuna, Kogi, Damana) and those that have two other adjectival and adverbial concepts are expressed degrees (Boruca, Ika, Tunebo). These languages can by other means (deriving adjectives/adverbs from also be divided in terms of the determiner’s position noun or verb roots, using phrasal expressions, etc.). relative to the noun, as prenominal (Rama, Guatuso, Boruca, Cuna) and postnominal (Teribe, Bocotá, Nouns and Nominal Morphology Bribri, Cabécar); one language, Guaymí, allows both The most common type of determiners are numeral orders. classifiers, forms used to categorize nouns exclusively The Chibchan personal pronouns exhibit two basic in the context of quantification; these are found in the traits; (a) with the exception of Rama, the free pronoun languages of Costa Rica and Panama. The existence of sets are formally indistinguishable for the core cases numeral classifiers in the Chibchan languages is the (subject and object), and (b) free personal pronouns result of the evolution of certain classes of nouns used are generally used for emphasis. This rather limited to form lexical compound nouns in the assumed com- use of personal pronouns has to do with two other per- mon ancestor (proto) language; the position of the vasive features of these languages, namely the use of numeral form in relation to the noun (prenominal vs. implied subjects/objects and, in most members, cross- postnominal) in Proto-Chibchan determined the for- referencing mechanisms. 421 GUAYMÍ AND CHIBCHAN LANGUAGES Where the word forms do not identify subjects vs. In most Chibchan languages, the most basic TAM objects, verbal affixes help to disambiguate, as in (5), distinctions are expressed by bound morphology (affix- where the verbal suffix marks the subject: es), not by free-standing auxiliaries. Auxiliaries are (5) lanyo kwe TERIBE mainly found in the languages of Colombia, as in Kogi: midë-rwa llëme (9) Nac gu-ngu-cu-aini-ki, yatai story DEMONSTRATIVE know-we not na-gu-cu-a ‘We don’t know that story’ come do-PAST-I-and-then, sit CERTAINTY do-I-PAST Another strategy is to use particles identifying the ‘I came and then I sat down’ subject or topic, in combination with word order: In some Central American languages, there are (6a) At ki ba ishd-ra BORUCA I SUBJECT YOU see-PRESENT positional verbs, a closed set of linking verbs specify- ‘I see you’ ing location and manner of the location (standing, (6b) Ba ki at ishd-ra lying), as in Bribri: you SUBJECT I see-PRESENT ‘You see me’ (10) ie? dur âula a (6c) At ishd-ra ba-ng he POSITION.STAND classroom in I see-PRESENT you-TOPIC ‘He’s standing in the classroom’ ‘You see me’ Teribe has seven positionals (löng, shäng, pang, A very conspicuous feature of Chibchan is the exis- buk, sök, jong, lok). The remarkable fact about these tence of forms expressing the status of participants as positionals is that sentences expressing actions and topic (what the sentence is about) or focus (new infor- movement are often ungrammatical if they lack such a mation). Such markers combine with word order to positional: create a wealth of foregrounding structures similar in (11a) Tawa shro-no löng na function to such mechanisms as passives or left/right we arrive-PERFECTIVE dislocations in other languages. In Teribe, the topic POSITION.BE here marker is li (7); there are two focus markers, om, omgo ‘We came here’ (ungrammatical without löng) (8a, b), and one contrast marker ra (8a): (11b) ëng wle) -no löng e wlo (7) tlε) lok ga ‘walë kuzong ga walë li’ each other meet-PERFECTIVE say PLURAL that POSITION.BE DEMONSTRATIVE ‘woman then that woman TOPIC in order to yo-y-dë’ ‘for us to meet’ (ungrammatical appoint-we-will’ without löng) ‘Then they said, ‘well then, if it’s a woman, a [this] woman we will appoint’ Person Markers (Agreement) (8a) Kone kone om wuë, pero kone The Chibchan languages fall into three groups in terms kone om wuë un llëme... of agreement: (a) those that have no agreement (Bribri, Some FOCUS eat, but some FOCUS Cabécar, Boruca, Bocotá, Guaymí, Cuna, Tunebo); (b) eat all not ... those with subject agreement (Rama, Teribe); and (c) tawa ra om wuë llëbo un. we CONTRAST FOCUS eat thing all. languages with both subject and object agreement ‘Some eat THEM, but some (Guatuso, Ika, Kogi). A special case is Damana, which [others] do not eat THEM... has verbal agreement markers for the indirect (dative) WE DO eat THEM ALL’ object in addition to agreement for the subject and the (8b) E tle) ga e omgo twa-ydë direct object. DEMONSTRATIVE say CONNECTIVE DEMONSTRATIVE FOCUS come-will ‘He said HE is the one Grammatical Relations who is coming’ Seven languages have nominative–accusative case- marking, where nominative case marks the subject and Verbs and Verbal Morphology accusative the (direct) object: Paya, Rama, Boruca, T(ENSE)A(SPECT)M(OOD) systems are abundant and Teribe, Cuna, Chimila, and Barí. in some cases highly elaborated. Some languages do not overtly distinguish tenses; tense-marking lan- Ergative Systems guages are Rama, Guatuso, Cuna, Kogi, and Tunebo. Eight languages (Guatuso, Bribri, Cabécar, Guaymí, Ika, Some tense-marking languages also have aspectual Kogi, Damana, and Tunebo) use ergative–absolutive distinctions (Rama, Kogi); others (Tunebo) have only case-marking, where the subject of an intransitive verb tense. patterns with the direct object of a transitive one (abso- 422  GULLAH lutive), while the subject of a transitive verb is marked I burn-PERFECTIVE with ergative case. ‘I burned’ (13c) Gliá gugé-du (12a) ie?-r dìwö sa-we BRIBRI leaves burn-PERFECTIVE she-ERGATIVE sun(ABSOLUTIVE) see-PERFECTIVE ‘Leaves burned’ ‘She saw the sun’ (12b) dìwö mìchò sun(ABSOLUTIVE) go References ‘The sun goes’ Constenla, Adolfo. 1991. Las lenguas del Área Intermedia. San José: Editorial Universidad de Costa Rica. An important property of some of these languages Constenla, Adolfo. 1995. Sobre el estudio diacrónico de las is that the ergative marker may only be used when it is lenguas chibchenses y su contribución al conocimiento del needed for emphasis. pasado de sus hablantes. Boletín del Museo del Oro (Colombia) 38/39. 13–35. Quesada, J. Diego. 1999a. CHIBCHAN: with special reference Active/Nonactive Systems to participant-highlighting. Linguistic Typology 3(2). 59–108. One language, Bocotá, appears to follow an Quesada, J. Diego. 1999b. Ergativity in Chibchan. active/nonactive pattern. Direct marking is used to Sprachtypologie und Universalienforschung 52(1). 22–51. Quesada, J. Diego. 2000a. A grammar of Teribe (Chibchan, mark the active subject by no: Panama). Munich: LINCOM-EUROPA. (13a) Cha no gliá gúg-le Quesada, J. Diego. 2000b. Word order, participant-encoding I ACTIVE leaves burn-PERFECTIVE and the alleged ergativity in Teribe: International Journal of ‘I burned leaves’ American Linguistics (IJAL) 66(1). 98–124. (13b) Cha gugé-du J. DIEGO QUESADA Gullah Gullah, sometimes also identified as Geechee or Sea ery). No such vernacular evolved on the tobacco and Island Creole, is an American English creole spoken cotton plantations, where the Africans were generally in coastal South Carolina and Georgia. It is not the in the minority and were not segregated from the only variety of American English that developed from European majority until the late nineteenth century. contact with another language, but it is the only one Geographical and social isolation from mainland that has been singled out as a creole. It shares several America until the mid-twentieth century helped pre- structural features with Caribbean English vernaculars serve Gullah to this day as quite distinct from other that have also been called creoles. Like the latter, it American offspring of English. developed on plantations where African slaves out- It took coastal South Carolina colonists 30 years or numbered the European indentured servants and other so from the foundation of the colony in 1670 to shift colonists with whom they interacted. from the homestead type of residence, characterized Gullah differs from African American Vernacular by no race segregation, to a partial plantation society, English (AAVE, ‘Black English’) in the socioeconom- in which segregation was institutionalized as early as ics of its development as well as in some of its struc- the mid-eighteenth century. (Virginia, which was tural features (including a speech melody that is closer founded in 1607, had not shifted to a largely agricul- to that of Bahamian and Caribbean Englishes). For tural economy, with tobacco plantations, until the end instance, in Gullah one can form the progressive with of the seventeenth century, either.) It also appears that, the preverbal marker [də] (spelled duh or da in ‘eye except for coastal Georgia (which patterned its eco- dialect’), as in he duh/da talk(in) ‘he is talking’, but nomic and social development on South Carolina), in not in AAVE. Gullah developed on the coastal rice the rest of the southeastern English ex-colonies segre- fields, where the African slaves were in the majority gation was not institutionalized on a large scale until from the early eighteenth century to the mid-nine- the late nineteenth century, after the passage of the Jim teenth century (the Civil War and the abolition of slav- Crow laws. 423 GULLAH This difference in the timing of segregation have resulted from the kind of plantations on which explains partly why Gullah has developed more non- they developed and the kinds of social contacts that standard characteristics than AAVE. It had a separate were obtained across race boundaries on these planta- evolution, in which African languages exercised a tions. The coastal vernaculars are rice-field phenome- greater influence on the integration of English materi- na, distinct from those that developed on the tobacco als into the emerging dialect. With a few exceptions, and cotton plantations. the vast majority of structural peculiarities associated From one evolutionary point of view, Gullah is con- with AAVE are shared by other American dialects, and sidered the most extensively restructured variety of the most significant differences between black and North American English (i.e. it shows the greatest white forms of American English are statistical. In influence of its creators’ ancestral languages). contrast, few of the features associated with Gullah are However, this judgment reflects a traditional bias attested in other North American English vernaculars, against all colonial vernaculars that took their vocabu- although the features have definite English origins too. lary from European languages and were appropriated In areas inland from the coast, segregation was institu- by descendants of Africans. It also reveals how little tionalized after over two centuries of intimate coexis- we yet know about vernaculars such as Old Amish tence of (descendants of) Africans and low-class English and Cajun Vernacular English, which devel- Europeans, after many structural features had become oped from appropriations of English by isolated entrenched in both their communities’ dialects. Later groups of proletarian continental Europeans. The changes associated today with differences between opinion is also undermined by the absence of a reliable them remain minimal. yardstick for assessing the extent of restructuring. The development of vacation resorts in coastal Every North American English variety has developed Georgia and South Carolina and the urbanization of from the contact of English dialects with each other some of the Sea Islands has fostered residential pat- (hence they too are restructured varieties) and with terns similar to those that have preserved AAVE in other languages, especially languages from Europe. mainland cities. The affluent in-migrants, who are pre- This restructuring-through-contact puts other dominantly white, have settled in their own parts of the American English vernaculars in the same category as Islands. They have not interacted regularly with the Gullah and AAVE. It makes the often-invoked argu- Gullah populations who preceded them there by two ment of mutual intelligibility (according to which and a half centuries. Gullah is not intelligible to most European Americans) There is arguably a geographical and structural con- only one-sided and largely a reflection of the fact that tinuum in African American speech from the South Gullah developed from contact with languages that Carolina and Georgia coast through the mainland happened not to be European. There is little wonder- Southeast. Accordingly, Gullah represents the variety ing about why the varieties that developed from the that is the most different from mainland white vernac- contact of English with other Western European lan- ulars. After all, although their sentence melodies are guages (before the massive nineteenth-century immi- clearly distinct, Gullah and AAVE share quite a few grations from Central and Eastern Europe) are to some features, such as constructions without the copula be extent mutually intelligible, even though familiarity (e.g. she Ø pretty ‘she [is] pretty’); use of the negative with a particular variety is an important factor in all ain(t) not only to mean ‘be not’ (e.g. he ain crazy) but such claims. Contacts of typologically similar vari- also ‘have not’ or ‘did not’ (e.g. he ain go/gone there eties in similar ethnographic settings produced similar ‘he has not gone there’, ‘he did not go there’); habitu- dialects, although listeners’ attitudes toward particular al be (e.g. they be havin company ‘they are/were USU- speakers affect their ability to understand them. Future ALLY IN THE PROCESS of having company’); and research may very well reveal that Gullah holds the emphatic PERFECT with done (e.g. she done gone ‘she most retentions from colonial English, although not has/had ALREADY gone’). (This is not to say that these necessarily in their pristine forms. An important factor peculiarities definitely did not grow out of English in comparing vernaculars is that their features often usages; but they have clustered in a way that makes originated in different English varieties. Neither they African American speech distinct from other nor Gullah, as systems, could be claimed to be com- American nonstandard vernaculars.) prehensive retentions of any particular British English The geographical continuum hypothesis does not dialect. apply to Gullah alone, but also to White Coastal There are no records to show how distinct Gullah Plantation English, which is distinct from American was from AAVE before the nineteenth century. White Southern English. One could thus also argue However, there is literary documentation of undiffer- that Southeastern American English vernaculars are entiated, stereotypical black speech, with pidgin-like basically plantation varieties. Differences among them features, since the early eighteenth century. This sug- 424  GULLAH gests that a distinct variety of English was spoken by The features that distinguish Gullah from AAVE are some African slaves in English colonies, but it does not necessarily of non-English origin. For instance, the not confirm that Gullah as a distinct ethnic variety had HABITUAL marker [dəz] (spelled duhz) is well attested already developed. Advertisements concerning run- in some British nonstandard vernaculars and in away slaves in the mid-eighteenth century reveal that Newfoundland Vernacular English. Even the fact that English among the African slaves varied from varieties Gullah has the option of expressing the PROGRESSIVE close to those spoken by white colonists to others that with [də] before the verb, and with a when the verb were quite different, which may be assumed to be the also has the ANTERIOR marker bin (as in he bina talk beginnings of today’s Gullah. Reports of ‘fluent’ or ‘he was talking’), does not necessarily suggest that ‘acceptable’ command of colonial English generally these characteristics stem from African languages. correlated with whether the slave was American-born, While African languages must have exerted some was imported young to the colony, or, if imported as an influence in the selection of specific particles that adult, had lived here long enough. The absence of lit- occur before the verb to mark tense and aspect, the erary or other records of a divergent speech of forms themselves have typically been inherited from Africans in the seventeenth century can be interpreted English. to suggest that no singularly African American ver- It seems very helpful indeed to make a distinction nacular was evident then. The generally nonsegregat- in scholarly debate between the origins of features and ed living conditions of the time, with the colonial the influence that determined their integration into the populations growing more by birth than by importa- developing dialect. This is more obvious in Gullah’s tion, lead to this conclusion. sound system, where one comes across sporadic uses The lack of reports of divergent speech suggests in of [ß] as a variant of [v] and [w], as in [ßεri ßεl] ‘very fact that Gullah must have followed an evolutionary well’, which also alternates with [vεri wεl]. path similar to those of other colonial English vernac- Typologically, the bilabial fricative is highly unusual ulars, differing from them essentially in the details of in the sound systems of sub-Saharan African lan- the restructuring process. The differences can be cor- guages. [ß] is not attested in Caribbean English cre- related to the following factors: The homestead phase oles, where African influence would be more during which the nonnative, non-European speakers extensive, nor in Caribbean French creoles, which had were in the minority did not last long, only about 30 similar influences. Its counterpart in Caribbean years after the founding of the South Carolina colony English creoles is [b], as in [bεks] ‘angry’ < English in 1670. This is a period during which the Africans [vεkst]. This suggests not necessarily the absence of were integrated into small family farm communities [v] in some African languages, but primarily the and interacted regularly with Europeans. Both inability of those who made the creoles to reproduce European American and African American children [ß], which is attested in Irish English and must have spoke the same colonial koiné as vernaculars. By occurred in colonial English. The fact that Gullah has 1720, the population of African descent was double [ß] in alternation with the English variants [v] and [w], the population of European descent, and on the emer- just like the fact that it has the schwa (unlike gent plantations on the coast, the Africans became the Caribbean English creoles), reflects peculiarities of overwhelming majority, often at a ratio of 9 to 1 by the the sociohistorical conditions under which it evolved. late eighteenth century. The early institutionalization African influence can, on the other hand, be identi- of race segregation forced the Africans to socialize pri- fied in serial verb constructions, in the associative marily among themselves. The shift to rice cultivation noun plural an dem, as in Easter (an) dem ‘Easter and as the primary colonial industry brought about some her family/friends/associates’, and in some other con- demographic changes, chiefly rapid population growth structions that it shares with AAVE and other creoles. by importation of new labor while the mortality rate Although serial verb constructions are more diversi- was high. Increasingly, language was transmitted to fied in Gullah, these features are also attested in other newcomers by nonnative ‘seasoned’ slaves, a factor American and British nonstandard vernaculars. that enabled more restructuring under dominant influ- However, their statistical predominance and some of ence from African languages. This evolution may their usage patterns reflect the influence that must account for why Gullah has certain grammatical char- have been exerted by some African languages during acteristics not found in other American dialects, the development of Gullah. including AAVE. As noted above, the development of Gullah must have started in the eighteenth century. References However, it is not clear what its actual grammatical Cunningham, Irma Aloyce Ewing. 1992. A syntactic analysis system was like then, especially how much it diverged of Sea Island Creole. Tuscaloosa: University of Alabama from other colonial English vernaculars. Press. 425 GULLAH Goodwine, Marquetta, and the Clarity Press Gullah Project. Turner, Lorenzo. 1949. Africanisms in the Gullah dialect. 1998. The legacy of Ibo landing: Gullah roots of African- Chicago: University of Chicago Press. Reprinted in 2002 American culture. Atlanta: Clarity Press, Inc. with a new introduction by Katherine Wily Mille and Jones-Jackson, Patricia. When roots die: endangered traditions Michael B. Montgomery. Columbia: University of South on the Sea Islands. Athens: University of Georgia Press. Carolina Press. Montgomery, Michael (ed.) 1994. The crucible of Carolina: SALIKOKO MUFWENE essays in the development of Gullah language and cutlure. Athens: University of Georgia Press. Pollitzer, William S. 1999. The Gullah people and their African See also African American Vernacular English; heritage. Athens: University of Georgia Press. Pidgins and Creoles Gumperz, John Joseph A blend of scientist, anthropologist, and linguist, John misunderstanding could result and in some cases the Joseph Gumperz’s background is manifested both in communication line may break down even if the speak- his theoretical and applied research. ers shared other linguistic features. Learning another His research focused mainly on the complex nature language does not guarantee effective communication of intercultural and interethnic communication or what unless the speakers are able to interpret the contextual- came to be known as interactional sociolinguistics. ization cues appropriately to keep redefining the con- Contrary to structuralists, who presuppose that human text. Contextualization cues, then, are interpretive cues communication takes place in an ideal and nonprob- that guide the speakers in their inferential effort in any lematic context where there is almost no risk of misun- communicative event similar, in this sense, to the derstanding, Gumperz gives substantial importance to Gricean Maxims of conversation (Grice 1975). While the five elements involved in any verbal communicative the Gricean Maxims are theoretically applicable to uni- event (the speaker, the listener, the context, the message, versal communication, the contextualization cues may and the channel) using the macrosociological factors to exist across cultures but they are still partly culture- interpret the microconversational effects. After his bound and even conversation-bound. fieldwork in India, he shifted from purely dialectologist Gumperz’s interpretation of code-switching as a research to conversational analysis, yet without over- contextualization cue has been very significant in looking his early research interests. One of his most shaping later research on this complex sociolinguistic influential books, Discourse strategies, which appeared phenomenon. He claims the existence of two types of in 1982, is a product of both theoretical research in con- code-switching: situational code-switching and versational analysis and his long years of fieldwork in metaphorical code-switching. The first is produced India, Europe, and the United States, where he scruti- when the codes used by the speakers change according nized both interracial and interethnic conversations. His to the redefinition of the situation in which they are main concern was with how the participants in a certain involved. A common example is when a monolingual conversation behave according to their interpretation of speaker joins a conversation already started by bilin- the communicative intent. To answer this concern, he gual speakers; they would switch to his or her lan- developed the theory of ‘contextualization cues’, which guage so as not to exclude him or her. Metaphorical he defines in Discourse strategies (p. 131) as ‘any fea- code-switching, on the other hand, is caused by the tures of linguistic form that contributes to the signaling change in the topic of the conversation without chang- of contextual presuppositions’. These features may be ing the situation. In the following example (Gumperz prosodic, morphological, syntactic, lexical, or phono- 1982a:77), the switch to Hindi by speaker B shows the logical. They may include other conversation behaviors rupture with the formal conversation that he was car- such as turn-taking or code-switching. In the majority rying out with speaker A. The change in the topic of of the cases, these cues indicate how the speakers the conversation stands behind the switch and not the understand every part of the talk as being sequentially change in the speakers since speaker C was present connected to both what has been said before and what from the beginning. is going to be said after. When the speakers fail to grasp ‘A group of Hindi speaking graduate students are dis- and use the contextualization cues appropriately, cussing the subject of Hindi–English code-switching: 426  GUMPERZ, JOHN JOSEPH A: Sometimes you get excited and then you speak Biography in Hindi, then again you go on to English. Born in Germany in 1922, Gumperz moved to the B: No nonsense, it depends on your control of United States in 1939 where he received a Bachelor of English. Science from the University of Cincinnati. He later B: [shortly thereafter turning to a third participant, earned a Ph.D. in German linguistics from the who has just returned from answering the University of Michigan. In 1955, he conducted field- doorbell] Kɔn hai bai (who is it)?’ work in India. And in 1965, he was appointed as an One of the main explanations for the occurrence of anthropology professor at Berkeley where he is still an code-switching is the in-group vs. out-group theory as emeritus Professor today. elaborated by Gumperz (1972, 1982a). He refers to Since the appearance of his first article early in 1955 code-switchers as the individual speakers who, for dif- (‘The phonology of North Indian Village Dialect’), ferent reasons, find themselves living in an ethnically Gumperz has contributed consistently to different publi- and culturally diverse setting. This includes the case of cations (Berkeley Cognitive Science Report, The immigrants who usually manage to keep their first lan- American Journal of Sociology, American Speech, etc.). guage (L1) to communicate among themselves in their In addition he has written and co-edited several other restricted speech communities, but who find them- influential books including The Ethnography of selves using the majority’s language, generally the Communication of (1964), Directions in Socio- official language of the host country, basically at work linguistics (1972), and Language and Social Identity and education. Within the boundaries of the ‘in- (1982). group’, the individual speaker tends to use the ‘we code’, i.e. the L1 of his or her own speech communi- References ty. An example is Spanish for the Puerto Rican com- Grice, H.P. 1975. Logic and conversation. Syntax and seman- munity in New York extensively studied by Gumperz tics: Speech acts, ed. by In P. Cole and J.L. Morgan (eds). and others. During the interaction with members of the Volume 3, 41–55. New York: Academic Press. ‘out-group’, i.e. members of the majority or any other Gumperz, John Joseph. 1955. The phonology of a North Indian speech community, the tendency is to use the ‘they village dialect. Indian Linguistics 16. 283–95. _________. 1964. The ethnography of communication (co-edit- code’, which is the majority language. ed with Dell Hymes). American Anthropologist 66(6, part 2, In some cases, the continuous switch back and forth Special Issue). between two codes can be an indicating strategy of the Gumperz, John Joseph, and Dell Hymes (eds.) 1972. Directions speaker’s neutrality with regard to contending parties in sociolinguistics. New York: Holt, Rinehart, and Winston. or ideologies that are associated with different codes. _________. 1972. Social meaning in linguistic structures: code switching in Northern Norway(cowritten with Jan-Petter On the other hand, religious or political issues, among Blom). Directions in sociolinguistics, ed. by J.J. Gumperz others, are often discussed in the code that is most and D. Hymes. New York: Holt, Rinehart, and Winston. often associated with them. _________. 1982a. Discourse strategies. Cambridge: In his later book edited with S.C. Levinson, Cambridge University Press. Rethinking linguistic relativity, Gumperz revisits his the- _________. 1982b. Language and social identity. Cambridge: Cambridge University Press. ory of contextualization cues to comment on the relation- _________. 1984. Communication and social identity. Identity: ship that language has with thought and culture and the personal and socio-cultural, ed. by A. Jacobson-Widding. way it shapes our conception of the world. The editors Uppsala Studies in Cultural Anthropology, Vol. 5. signal the changes in the understanding of the Sapir- _________. 1990. Contextualization and understanding. Whorf hypothesis and its implication for the theories of Rethinking context, ed. by A. Duranti and C. Goodwin. New York: Cambridge University Press. linguistic relativity. Here again, Gumperz manages to _________. 1996. Rethinking linguistic relativity (co-edited give equal importance both to the macrosociological and with Stephen C. Levinson). Cambridge: Cambridge cultural factors and the very specific microconversation- University Press. al features in each communicative event. LOTFI SAYAHI 427  H Haas, Mary Rosamond The life and growth of language: an outline of linguis- dissertation; it was later published with revisions in tic science (1875) by William Dwight Whitney intro- Franz Boas’s Handbook of American Indian languages duced Mary Haas to linguistics, and in 1930, following (1941) and remains an outstanding exemplar of her undergraduate studies at Earlham College, she set Sapirian anthropological linguistics. Meanwhile, Haas off for the University of Chicago intending to study spent her postdoctoral years as a fieldworker, gathering comparative philology. A class in historical Germanic data on three Muskogean languages, Creek, Koasati, linguistics taught in German by Leonard Bloomfield and Hichiti, as well as on Natchez. Her classic study of did not appeal to her, but when she ventured into the gender and language, ‘Men’s and women’s speech in anthropology department where Edward Sapir was Koasati’, appeared in 1944 in Language, journal of the elaborating upon the topics he had covered in his book Linguistic Society of America. Other studies drawn Language (1921), she encountered the kind of linguis- from the data she had collected were published later in tics to which she would devote much of her career. her career; much remained unpublished. She trained as an Americanist, a term used by World War II brought a new opportunity for paid anthropologists and anthropological linguists for the work, and Haas became the first linguist supported by study of the native languages and cultures of North the Intensive Language Program of the American America. Her first fieldwork was undertaken with her Council of Learned Societies (ACLS). The goal here then-husband Morris Swadesh. With support and was to develop descriptions and teaching materials for direction from Sapir, the couple traveled to Vancouver lesser known languages of Europe and Asia. She was Island in Canada in the summer of 1931 to collect assigned to work on Thai and when she located native texts and data on Nitinat. Their joint report was Mary speakers of the language studying at the University of Haas’s first publication. Michigan, she moved there to work with them. Her In Fall 1931, Sapir joined the faculty at Yale article ‘The linguist as a teacher of languages’ University. Haas, Swadesh, and several other graduate (Language 1943) discussed how she applied her students followed him, becoming the nucleus of the descriptive linguistic techniques to the development of ‘Sapir School’ of linguistics, focused on the collection materials and classroom experiences in foreign lan- of texts and the preparation of descriptive grammars of guage learning. Haas went on to publish more than 20 native American languages, with a strong concurrent articles and books on the Thai language, including a interest in the extension of concepts and techniques textbook and a reader prepared with her second hus- from comparative Indo-European linguistics to the his- band, Heng R. Subhanka. Her analyses of various torical and comparative study of American languages. aspects of the Thai language, including its phonology, As a graduate student, Haas worked on Tunica, a morphology, and writing system, are notable for their language spoken in Louisiana. ‘A Grammar of the clarity and insight and remain valuable descriptive Tunica Language’ became her 1935 Yale doctoral resources even today. 429 HAAS, MARY ROSAMOND Following a brief ACLS-sponsored pilot class in American Indian languages. Not particularly effective Thai at the University of Michigan, Haas moved to the as a classroom teacher, Haas was an inspiring mentor. University of California, Berkeley as a lecturer in what Her students and their students can still be found in the was there termed Siamese. She also taught linguistics linguistics departments of American and Canadian courses and was instrumental in the development of colleges and universities, many of them carrying on the Berkeley graduate department of linguistics. In her Americanist tradition. 1953, her colleague Murray B. Emeneau obtained funding for the establishment of the Survey of Biography California Indian Languages, and Mary Haas was appointed as director. This program provided the Mary Haas was born in Richmond, Indiana, on January research center of the department, with graduate stu- 23, 1910. She earned the AB at Earlham College in dents expected to focus their long-term work on a sin- 1930 and began her graduate study at University of gle language, writing a descriptive grammar for a Chicago in 1930–31. She carried out fieldwork on dissertation and preparing a dictionary and set of texts Nitinat in the summer of 1931 with Morris Swadesh. for publication, often in the University of California She went to Yale University in the fall of 1931, and Publications in Linguistics series. completed her Ph.D. in 1935 with a dissertation on With the first half of her career devoted to the doc- Tunica. She pursued postdoctoral independent research umentation of languages, Haas returned in the second on Muskogean languages and Natchez. In 1941–1943, half to the historical comparison of languages that had she pursued research, materials preparation, and pilot initially drawn her to the field. A lengthy 1966 article teaching of Thai at the University of Michigan under the on ‘Historical linguistics and the genetic relationship Intensive Language Program of the American Council of languages’ became The prehistory of languages of Learned Societies. In 1943–1944 and 1946–1947, (1969), her only book on a general linguistic topic. she was a lecturer in Siamese (Thai) at the University of Here, she used her earlier fieldwork on the Muskogean California, Berkeley, and in 1947–1952, she was assis- language family, along with data from other native tant professor of Siamese and linguistics. In 1952–1957, American languages, to illustrate types of language she was associate professor; in 1957–1977, professor of change, problems in genetic classification, and the linguistics; department chair in 1958–1964; and direc- reconstruction of protolanguages. tor Survey of California Indian Languages (later called Throughout her career, Haas remained steadfast in Survey of California and Other Indian Languages) in her commitment to the model of Americanist linguistics 1953–1977. She retired in 1977. She was visiting pro- developed originally by Franz Boas and then by Edward fessor or fellow at University of Washington, Ohio State Sapir. She rejected what she felt was the rigidity of post- University, Barnard College and Columbia University, Bloomfieldian structuralism and was openly dismissive Australian National University, Northwestern of some of the linguistic work of her east coast contem- University, University of California, San Diego, poraries. In the later 1950s and thereafter, Haas virtual- University of Georgia, University of Kansas and ly ignored the developments in generative grammar that University of Calgary. She was also Linguistic Society dominated American linguistics. She had little interest of America vice president in 1956 and president in in studies drawing on English or the languages of 1963. She was Guggenheim Fellow in 1964–1965 and Europe, and she believed that linguistic fieldwork was concurrently fellow of the National Endowment for the threatened by the use of their own language by genera- Humanities and the Center for Advanced Studies in the tive linguists who relied on their native knowledge for Behavioral Sciences in 1967–1968; she was also elect- linguistic insights. Haas continued on her own path, ed to National Academy of Sciences in 1978. She documenting the languages of native America, compar- received honorary degrees from Northwestern ing them, and proposing genetic classifications. Toward University in 1975, University of Chicago in 1976, and the end of her career, she began to look back at the his- Earlham College in 1980. Haas died in Berkeley, tory of her specialization. A set of articles on the histo- California on May 17, 1996. ry of American Indian linguistics is reprinted in a collection of her essays (Language culture, and history 1978) and remains a valuable account, sometimes first- References hand, of how that field developed in the United States. Campbell, Lyle. 1997. American Indian languages: the histori- Mary Haas was not interested in linguistic theory, cal linguistics of native America. New York and Oxford: but she was fascinated by the data of languages and Oxford University Press. she conveyed this to her students. Her most lasting Goddard, Ives. (ed.) 1996. Languages, Handbook of North American Indians, Vol. 17, series editor William C. legacy is the cadre of some 40 graduate students who, Sturtevant. Washington, DC: Smithsonian Institution. under her direction or with her assistance in the 1950s, Golla, Victor, James A. Matisoff, and Pamela Munro. 1997. 1960s, and 1970s, earned doctorates with work on Mary R. Haas. Language 73. 826–37. 430  HAITIAN CREOLE Haas, Mary Rosamond. 1932. A visit to the other world, a Historical linguistics and the genetic relationship of languages. Nitinat text (with translation and grammatical analysis), as Current trends in linguistics, Vol. 3, ed. by Thomas A. Mary Haas Swadesh with Morris Swadesh. International Sebeok. The Hague and Paris: Mouton (1966). Journal of American Linguistics 7. Hinton, Leanne. 1993. Women in southwestern linguistic stud- ———. Tunica. 1941. Handbook of American Indian lan- ies. Hidden scholars: women anthropologists and the native guages, Vol. 4, ed. by Franz Boas. New York: J. J. Augustin. American Southwest, ed. by Nancy J. Parezo. Albuquerque: Haas, Mary Rosamond, and Heng R. Subhanka. 1945, 1948. University of New Mexico Press. Spoken Thai, 2 vols., 1945 (Book I), 1948 (Book II). New Hinton, Leanne.(ed.) 1997. The Hokan, Penutian, and J. P. York: Henry Holt. Harrington Conferences and the Mary R. Haas Memorial, ———. 1956. The Thai system of writing, Washington, DC: 28–29 June 1996, University of California, Berkeley American Council of Learned Societies. (Survey of California and Other Indian Languages Report Haas, Mary Rosamond, with George Grekoff, Ruchira C. 10), Berkeley: Department of Linguistics, University of Mendiones, Waiwit Buddhari, Joseph R. Cooke, and Søren California. C. Egerod. 1964. Thai–English student’s dictionary, Parks, Douglas R. (ed.) 1997. Mary R. Haas: a memorial issue: Stanford, CA: Stanford University Press. a collection of essays. Anthropological Linguistics 39. Haas, Mary Rosamond. 1969. The prehistory of languages. The Shipley, William. (ed.) 1988. In honor of Mary Haas: from Hague and Paris: Mouton. the Haas Festival Conference on Native American The following articles, as well as many others, are reprinted in Linguistics. Berlin, New York, and Amsterdam: Mouton de Language, culture, and history: essays by Mary R. Haas, Gruyter. selected and introduced by Anwar S. Dil. Stanford, CA: JULIA S. FALK Stanford University Press (1978): The linguist as a teacher of languages. Language 19 (1943) See also Sapir, Edward; Swadesh, Morris; Thai and Men’s and women’s speech in Koasati. Language 20 (1944) Tai Languages Haitian Creole Haitian Creole is one of the many French-based words were relexified or changed into the lexical items Creoles that were established throughout the world of the colonizing or superstrate language. The polyge- during France’s colonizing period. Creole languages nesis theory suggests that whenever and wherever a are commonly accepted as being a language that was contact situation arose between two or more languages, developed from an earlier, simpler version of that lan- a pidgin was created, which subsequently led to the guage, called a Pidgin. establishment of a Creole. This theory, however, fails to Pidgins arose in situations of linguistic need (such account for the amazing amount of similarity between as for trade or other economic purposes) between two Creole languages in widely disparate geographical or more groups of people who spoke mutually unintel- areas of the world. In addition, Bickerton (1981) pro- ligible languages. After a generation or two, the off- motes the ‘bioprogram’ as the basis for the genesis of spring of the original Pidgin speakers tended to develop Creoles. Building on the idea of language universals a more complex language, with a wider range of and language simplification processes common to first- expressive and social functions. These languages are language acquisition, Bickerton suggested that humans referred to as Creoles, and their vocabulary is general- are genetically programmed to acquire a specific set of ly based on the lexicon of the colonizing group. linguistic structures, and that these structures are mod- Although Haitian Creole is French-based, it also ified later in life as children are exposed to their native contains elements of African influences from Wolof, language. The commonality among these theories is Fon, and Ewe. The basic word order is Subject, Verb, the acceptance that the superstrate language (French, Object (SVO), but typical Creole markers appear in English, Spanish, etc.) provided the lexical base, while the use of postpositions, articles, and possessives com- the syntactic elements were organized according to the ing after the noun (e.g. nom-la, ‘man’ ‘the’). substrate or original language used by Creole speakers. Two theories have been proposed to explain the ulti- For the origin of Haitian Creole French, Valdman mate origin of Creole languages. The claim for the (1997) claims that the earliest form of a French-based monogenetic theory is that Creoles all over the world creole appeared first on the island of St. Kitts; this lan- have a common origin; a Portuguese-based pidgin was guage was subsequently exported to Guadeloupe and created in Africa in the fifteenth century. During the Martinique, and then to Haiti. He extrapolates from establishment of the Creole, the Portuguese and native this diffusion theory to claim that the newest, most 431 HAITIAN CREOLE innovative form of a Creole language will be on those upheavals are the norm, the language appears to have islands from which it first came—Martinique, been cemented in the local conscience, and therefore Guadeloupe, and Haiti—and that areas that were later plays an important role in identification as Haitian. recipients of the language (such as Louisiana) will Haitian Creole is spoken on the western portion of the exhibit more conservative forms of the language. island of Hispaniola; the other part of the island is the One reason for language change (from conservation Spanish-owned Dominican Republic. There are estimat- to innovation, for example) is due to the process of ed to be over five million Haitian Creole speakers at decreolization. This occurs when the use of the standard present, who represent about 90% of the island popula- form of the superstrate is viewed as prestigious, and the tion, and constitute the largest number of Creole speak- use of the Creole is given low status. Because of the his- ers anywhere. In addition, there are many substantial tory of colonization and slavery often associated with Haitian enclaves in the Bahamas, Canada, and the Creole languages and the people who speak them, this United States. Ninety-five percent of Haitians are of low prestige is exacerbated, and causes a shift toward African origin, but the extremely small mulatto (mixed- the use of the standard form of the language. This race) and French population constitutes the upper class. results in a continuum of varieties: the basilect, which is This is also the group that has continued to use and speak farthest from the superstrate language, the mesolect, an standard French. Recent language attitude studies and intermediate variety between the two, and the acrolect, the political climate of the island suggest that linguistic, the code closest to the substrate language. This postcre- social, and cultural mores are moving in a direction of ole continuum exists in Haiti, and, just as in many areas viewing Creole as a positive force, while simultaneous- of the world where Creoles are spoken, speakers often ly rejecting or lowering the status of the standard French have control over several of the varieties. When and language and culture, which was imposed on the island how they use them is often dependent upon social fac- so many centuries ago. tors such as age, gender, ethnicity, and geographical origins of the speaker and hearer. In addition, identifi- References cation as Haitian often drives the use of the Creole Bickerton, Derek. 1981. Roots of language. Ann Arbor: Karoma. rather than a more Standard French code. Haitians take Ferguson, Charles. 1957. Diglossia. Word 15. 325–40. much pride in the fact that their island was the site of the Holm, John. 1988. Pidgins and Creoles, vols. 1 and 2. first slave revolt in modern history (1791–1804), which Cambridge: Cambridge University Press. abolished slavery and established Haiti as the second Hymes, Dell. (ed.) 1971. Pidginization and creolization of lan- independent nation in the Americas (Hall 1992). As a guages. Cambridge: Cambridge University Press. Lefebvre, Claire, Hélène Magloire-Holly, and Nanie Piou. result, they have clung tenaciously to their (Creole) lan- (eds.) 1982. Syntaxe de l’haïtien. Ann Arbor: Karoma. guage, their history, and their culture. Le Page, R.B. (ed.) 1961. Creole language studies II: Proceedings This linguistic strength was shown in 1961 when of the Conference on Creole Language Studies, University of Haitian French was granted legal and educational sta- the West Indies, Mona, 1959. London: Macmillan. tus in Haiti. It was accepted as the official language of Mühlhäusler, Peter. 1986. Pidgin and Creole linguistics. Oxford: Basil Blackwell. the island in 1987 (along with Standard French), and Romaine, Suzanne. 1988. Pidgin and Creole languages. London: has since developed a standardized written code and a Longman. growing literature. There are several dictionaries of Valdman, Albert. 1977. Pidgin and Creole linguistics. Haitian Creole, and newspapers, articles, fiction, and Bloomington: Indiana University Press. poetry are being written in the language. Radio and Valdman, Albert. 1978. Le créole: structure, statut et origine. Paris: Klincksieck television broadcasts are in Haitian Creole, and there MEGAN E. MELANÇON was a bible printed in the language in 1985. Although economic conditions are abysmal, and political See also Pidgins and Creoles Hale, Kenneth Like his teacher Carl Voegelin, Kenneth Hale was a Navajo and O’odham (Papago). His training combined prolific documenter and avid student of languages. He linguistics and anthropology in the Americanist tradi- learnt Spanish in his youth in southern Arizona, and tion following Edward Sapir. 432  HALE, KENNETH His perennial interests covered many areas of lin- as diverse as Chukchi, Dagur, English, Hopi, Miskitu, guistics, including language typology, lexicography, Igbo, Irish, Japanese, Navajo, O’odham, Warlpiri, and syntactic theory, historical linguistics and language West Greenlandic, often in collaboration with native reconstruction, and applied linguistics in school edu- speakers of the languages. cation especially in support of minority languages. Hale’s reputation as a polyglot was well-deserved, Hale’s early major work comprised several careful as he commanded dozens of languages. He had an comparative studies of language groups based on data extraordinary linguistic memory: he was able to he collected himself, first Uto-Aztecan, then sub- absorb intensely not only the grammar and vocabulary groups of Australian languages (Arandic, Northern of a language but also the precision of pronunciation. Pamic). From his 1959–1960 fieldwork, Hale per- He downplayed this ability, since he believed that full ceived that the phonologically aberrant languages of knowledge of a language also included knowing its northern Cape York Peninsula were related by regular ‘phrase book’ with thousands of common expressions sound changes to other Australian languages; he fur- and their appropriate use. His annual class at MIT ther showed the implications of this for prehistory. explored the grammar of one language each year, usu- Hale’s 1961 coinage ‘Pama-Nyungan’ has since ally with a native-speaker linguist, and included ses- been generally applied to name the largest classifica- sions of monolingual investigation. Over 30 years, the tory grouping of Australian languages, the outline of class ranged from several North American languages, which appeared in 1966 (with Geoff O’Grady and the to the likes of Maori, Warlpiri, Finnish, Amoy, Voegelins). Hale’s extensive notes, analyses, transcrip- Malayalam, Basque, Tamazight Berber, Abkhaz, tions, and sound recordings of about 70 Australian Zarma, Igbo, Yoruba, Korean, and American Sign languages, archived at the Australian Institute of Language (ASL). Aboriginal and Torres Strait Islander Studies (AIAT- In language investigation and description, Hale SIS) in Canberra, are a great accomplishment, and drew on his anthropology training in sensitivity to var- have provided a basis for much subsequent work ious linguistic registers (appropriate to gender, age, by others. kin relations, ceremonial situations), and to the lan- Hale was inspired by Noam Chomsky’s Aspects guage of songs. He did not just draw on his own abil- of the theory of syntax and arranged to return to ities, but showed how the native speaker could be a Australia and reexamine some Australian languages collaborator, and, in particular, he demonstrated the (Warlpiri and Lardil) with this new framework. While value and multiple uses of good ‘folk definitions’ or remaining sensitive to intricate details of complex ‘ethnosemantic oral essays’. Many other linguists phenomena (such as pronominal marking in clauses) enjoyed collaborating with him; he coauthored several at the same time Hale believed that linguistic analysis works with each of Geoffrey O’Grady, LaVerne needed to ‘strip away complexity’ and reveal simpler Masayesva Jeanne, Paul Platero, Samuel Jay Keyser, principles and modules at work in combination, relat- and Maria Bittner. ing the universal and the particular. Thus, Hale was Hale campaigned for action to stem the endanger- able to bridge from his Sapirian training to ment of languages. Conscious that knowledge was Chomskyan universal grammar. One of Hale’s influ- empowering, he sought to put the results of his lin- ential proposals was a later reaction (1983) to the guistic and anthropological research at the disposal of Aspects model, in which he argued, on the basis of his the speakers of those languages. He encouraged bilin- studies of languages such as Warlpiri and O’odham gual education, coauthoring a 1974 recommendation (Papago), that in some languages, the mapping from with Geoff O’Grady on which bilingual education (logical) argument structure to syntactic configuration programs in the Northern Territory of Australia were works quite differently than it does in English. Such founded. He also helped speakers of local languages ‘nonconfigurational’ languages do not reserve a spe- obtain training in linguistics both at workshops in cial syntactic position for clause subject (or object), their communities, and at MIT, so that they could and are expected to exhibit a contrasting cluster of study the language themselves and get some control syntactic properties, notably clausal word order inde- of its destiny. From 1986, he made several visits to pendent of syntactic relations. This subsumed another work with local people to study Misumalpan lan- of Hale’s related influential proposals, that there are guages (Miskitu, Sumu) of the Atlantic Coast of languages (such as Warlpiri) where relative clauses are Nicaragua and Honduras. He was instrumental in the not embedded (as they are in languages like English) development of dictionaries, including for Warlpiri, but rather simply added (adjoined) to their host clause. Lardil, and Ulwa and Miskitu (two Misumalpan For the rest of his life, he continued to explore how languages native to Nicaragua), and during his last conceptual argument structure relates to grammar, years, he mentored the Wampanoag revival project in word order, and word structure, looking at languages Massachusetts. 433 HALE, KENNETH Under Hale’s direction in the 1980s, the Lexicon and Sumu (Programa Educativo Bilingüe-Bicultural, Project at the Center for Cognitive Science, MIT car- Puerto Cabezas, Zelaya Norte, Nicaragua, January ried out comparative lexicology (research on the nature l986 and 1994, and Bluefields, Region Autónoma of the lexicon) with a view to discovering general Atlántico Sur, Nicaragua, January, 1988 and 1990); lexico-semantic themes, or principles that recur sys- and Mayan languages (Proyecto Lingüístico Francisco tematically in the definitions of lexical items. The lan- Marroquín, Antigua Guatemala, July 1988, and guages Hale himself focused on in this project were the CIRMA, Antigua Guatemala, August 1994. Central Australian language Warlpiri, Miskitu and He was a member of the Linguistic Society of Southern Sumu (Ulwa), and English. Later, Hale’s America (LSA), and its President in 1994–1995. He ideas on argument structure and the nature of thematic was also a member of the American Academy of Arts roles were the focus of his major book (with S.J. and Sciences from 1989, and of the National Academy Keyser) Prolegomenon to a theory of argument struc- of Sciences from 1990. Hale died in Lexington, ture, which was brought to completion as he suc- Massachusetts on October 8, 2001. cumbed to cancer. References Biography Carnie, Andrew, Eloise Jelinek, and Mary Ann Willie. (eds.) 2000. Papers in honor of Ken Hale. Working papers in Kenneth Hale was born in Evanston, Illinois on endangered and less familiar languages, Vol. 1. Cambridge, August 15, 1934; his family moved to Canelo, MA: MIT working papers in linguistics. Arizona by 1941. In 1952–1955, he received his B.A. Hale, Keneth and Carl F. Voegelin, and Florence M. Voegelin. (1955) in Anthropology at University of Arizona, 1962. Typological and comparative grammar of Uto- M.A. in 1956, with a thesis on Class II Prefixes in Aztecan: I (Phonology). Indiana University Publications in Anthropology and Linguistics, Memoir 17 of the Navajo, and Ph.D. in l959, with a dissertation on A International Journal of American Linguistics. Papago Grammar, both at Indiana University, Hale, Kenneth. 1964. Classification of Northern Paman lan- Bloomington, Indiana under Carl Voegelin. He was guages, Cape York Peninsula, Australia: Research Report. part of the linguistic survey of Australian languages Oceanic Linguistics 3(2). conducted in 1959–1960, affiliated with University of Hale, Kenneth, and Joseph B. Casagrande. 1967. Semantic relationships in Papago folk-definitions, Studies in Sydney, part of CF and FM Voegelin’s survey of Southwestern ethnolinguistics: meaning and history in the Languages of the World. He was Assistant Professor languages of the American Southwest, ed. by D. Hymes and in Anthropology and Linguistics, University of W. Bittle. The Hague: Mouton. Illinois, Urbana, Illinois in 1961–1964 and University ———. 1972. Some questions about anthropological linguis- of Alberta, Edmonton, in the summer of 1964, and tics: the role of native knowledge Reinventing anthropology, ed. by Dell Hymes, 382–97. New York: Pantheon. was also Associate Professor in Anthropology, ———.1973. A note on subject–object inversion in Navajo. University of Arizona, Tucson, in 1964–1966. His sec- Issues in linguistics: papers in honor of Henry and Renée ond stay in Australia was in 1966–1967, again sup- Kahane, ed. by B.B. Kachru et al., 300–9. Urbana: ported by National Science Foundation, focusing on University of Illinois Press. Warlpiri and Lardil languages. He was recruited by ———. 1973. Deep surface canonical disparities in relation to analysis and change: an Australian example. Current trends Morris Halle in 1966 to Massachusetts Institute of in linguistics, ed. by T.A. Sebeok, 401–58. The Hague: Technology, where he was Associate Professor Mouton. from 1967, and was promoted to Professor in 1972, ———.1976. The adjoined relative clause in Australia. and finally became Ferrari P. Ward Professor of Grammatical categories in Australian languages, ed. by Modern Languages and Linguistics in 1981–1999. R.M.W. Dixon, 78–105. Canberra: Australian Institute of Aboriginal Studies, and New Jersey: Humanities Press. His visiting appointments included University of ———.1976. Linguistic Autonomy and the Linguistics of Carl Arizona 1976–1977; Katholieke Hogeschool Tilburg, Voegelin. Anthropological Linguistics l8. Netherlands, 1983–1984; Australian Linguistic ———.1982. Some essential features of Warlpiri verbal claus- Institute, La Trobe University, Melbourne, July 1994; es. Papers in Warlpiri grammar: in memory of Lothar Jagst, Linguistic Society of America Linguistic Institutes, ed. by Stephen Swartz, 217–315. Work Papers of SIL-AAB, Series A, Vol. 6. Berrimah, NT: SIL-AAB. University of Arizona, Tucson, in the summer of Hale, Kenneth.1983. Warlpiri and the grammar of non-configu- 1989; and University of New Mexico, Albuquerque, in rational languages. Natural Language and Linguistic Theory the summer of 1995. 1(1). 5–47. He was active in training native speakers of indige- ———.1986. Notes on world view and semantic categories: nous and minority languages: Warlpiri (Yuendumu some warlpiri examples. Features and projections, ed. by Muysken and H. van Riemsdijk, 233–54. Studies in School, Central Australia, mid-1974); Navajo (Kin Li Generative Grammar, Vol. 25. Dordrecht: Foris. Chee, Arizona, summer 1975, Navajo Community ———.1992. Subject obviation, switch reference, and control. College, Tsaile, Arizona, summer 1982); in Miskitu Control and grammar, ed. by R.K. Larson, S. Iatridou, 434  HALLE, MORRIS U. Lahiri and J. Higginbotham, 51–77, Studies in Kenstowicz, Michael (ed). 2001. Ken Hale: a life in language. Linguistics and Philosophy, Vol. 48. Dordrecht: Kluwer. Cambridge, MA: MIT Press. Hale, Kenneth, and Albert Alvarez. 1970. Toward a manual of Keyser, Samuel Jay. 2003. Obituary: Kenneth Locke Hale. Papago grammar: some phonological terms. International Language 79. 411–22. Journal of American Linguistics (Memorial Issue for Laughren, Mary. 2001. Obituary Kenneth Locke Hale HansWolff, Fascicle II) 36(2). (1934–2001). Aboriginal History 25. 265–75. Hale, Kenneth, Colette Craig, Nora England, Laverne Simpson, Jane, David Nash, Mary Laughren, Peter Austin, Masayesva Jeanne, Michael Krauss, Lucille Watahomigie, and Barry Alpher (eds.) 2001. Forty years on: Ken Hale and Akira Yamamoto. 1992. Endangered languages. and Australian languages. Canberra: Pacific Linguistics Language 68(1). 1–42. 512. Hale, Kenneth, and Leanne Hinton (eds.) 2001. The Green Yengoyan, Aram. 2003. Kenneth L. Hale (1934–2001). Book of language revitalization in practice. San Diego: American Anthropologist 105(1). 222–5. Academic Press. DAVID NASH Hale, Kenneth, and Samuel Jay Keyser. 2002. Prolegomenon to a theory of argument structure. Linguistic Inquiry See also Chomsky, Noam; Configurationality; Monograph 39. Cambridge, MA: MIT Press. Sapir, Edward; Thematic Structure Halle, Morris Morris Halle can be called the father of the modern While working on his dissertation, Halle began study of speech sounds (i.e. phonological and phonet- working with researchers at the Research Laboratory ic theory). He was the central figure behind the devel- of Electronics (RLE) at the Massachusetts Institute of opment of Generative Phonology, and has had a Technology. He later began teaching in the Modern significant impact on the study of word structure (mor- Languages department there. In the mid-1950s, Halle phology). In addition to his intellectual stature, Halle’s was a driving force behind the move to develop a lin- influence as an educator is profound. He has taught guistics program at MIT. Part of building up this pro- and advised many of the leading figures in phonology gram included hiring Halle’s collaborator and today, including those who have proposed competing colleague Noam Chomsky. Taking a page from theories. Chomsky’s syntactic work, Halle claimed that phono- Halle began his study of structural phonology under logical information is not merely perceptual discrimi- Roman Jakobson, first at Columbia University, and nation and articulatory action, but is a cognitive later at Harvard. In his early work with Jakobson and system with its own set of principles. These principles in collaboration with Gunnar Fant, Halle started his form the foundation of Generative Phonology, so- career by questioning the idea that phonemes, i.e. called because it was assumed that all speech sounds sound units roughly corresponding to alphabet letters, could be generated on the basis of a finite set of prin- function as the basic cognitive units in the processing ciples or rules. The classic book of Generative of speech sounds. Jakobson et al. (1951), in the spirit Phonology is Chomsky and Halle’s (1968) The sound of work done in the Prague School of linguistics, pattern of English (SPE). Classical Generative claimed that phonemes are themselves divisible into Phonology assumes that there are two levels of phono- smaller units, so-called ‘distinctive features’, which logical information: the underlying or base form of a capture articulatory and acoustic characteristics of word and the surface or phonetic form. The underlying speech sounds. For example, the distinction between form is mapped to the phonetic form by a series of the word pat, pronounced [pæt], and bat, pronounced ordered generative rules. [bæt], does not lie in some categorial distinction Consider, for example, the pronunciation of the plu- between the cognitive units /p/ and /b/, but in the fact ral suffix –s in the words dogs [dagz], cats [kæts], and that only the latter is pronounced with vibrating vocal glasses [glæs z]. The underlying form of this plural cords. In later work, Halle and his students developed suffix is posited to be /z/. The phonetic form of the hierarchically organized models of feature systems to noun–suffix combination is generated via two rules. capture the fact that some features seem to be more The first rule states that a consonant is devoiced, i.e. prominent than others in the sound systems of specif- pronounced without vibrating vocal cords, when it fol- ic languages. lows a voiceless consonant. The second rule, called 435 HALLE, MORRIS epenthesis, inserts a schwa /ə/ sound between adjacent morphological form can occur before the words are coronal sibalents. inserted. The theory is “distributed” because what was The effect of these rules is seen in the following previously thought to occur in the lexicon is distributed derivations of the words mentioned above: into three grammatical systems: (i) a set of syntactic primitives, (ii) a list of phonological forms and the syn- underlying /dag/ + /z/ /kæt/ + /z/ /glæs/ + /z/ tactic environments in which they occur and (iii) a list representation of sound-meaning pairs. The first system comes into epenthesis —— ——— glæs z play as the syntactic rules construct the sentence, the devoicing —— kæts ——— second inserts the words into the sentence and the third phonetic [dagz] [kæts] [glæs z] is the tie between the surface phonological form and representation the semantic structure of the word or phrase. The influence that Morris Halle has had on the field It is argued that epenthesis must apply before the of linguistics is extensive. He has personally trained an devoicing rule, since applying them in the other order enormous number of today’s leading scholars in would result in an incorrect form *[glas s], as the final phonology, phonetics, and morphology, and his vision /s/ of glass is a voiceless sound, and the /z/ would be and research have led the field to new levels of rigor devoiced when adjacent to it: and understanding. underlying representation /glæs/ + /z/ Biography devoicing glæss epenthesis glæs s Morris Halle was born in Liepaja, Latvia on July 23, phonetic representation *[glæs s] 1923; he emigrated to the United States in 1940. He attended the City College of New York in Engineering Besides applying the theoretical framework of gen- (1941–1943), served in the United States Army in erative grammar to sound systems, The Sound Pattern World War II (1943–1946). He received his M.A. of English also provides one of the most complete dis- (1948) in Linguistics from University of Chicago and cussions of English phonology so far. did Graduate work in Slavic Languages at Columbia Metrical theory (or the study of stress) forms anoth- University (1948–1949) under the tutelage of Roman er important strand in Halle’s research program. SPE Jakobson. He did his Ph.D. (1955) on Slavic contained the first serious description of English word Languages, Harvard, in a dissertation entitled the stress in terms of predictable rules. Building upon the Russian Consonants: a Phonemic and Acoustical success of autosegmental phonology, in more recent Investigation, supervised by Roman Jakobson. He was years, Halle applied a notion of autosegmental structure an instructor in Russian, North Park College, Chicago, to the study of stress. Throughout this research, Halle 1946–1947; Teaching Fellow in German and Russian, has not only examined stress by examining individual University of Chicago, 1947–1948; Teaching Fellow in words, but also words in phrasal context and in verse. Russian, Harvard University, 1947–1948; Assistant In generative phonology, the sound system closely Professor of Modern Languages, Massachusetts interacts with the structure of the containing words. Institute of Technology (MIT), 1951–1956; Associate Thus, Halle is an influential figure in the subdiscipline Professor of Modern Languages, MIT, 1956–1961; of morphology as well. By applying the familiar gen- Fellow, Center for Advanced Study in the Behavioral erative model to words, he developed the first explicit Sciences, Stanford University, 1960–1961; Founder, theory of generative morphology. This theory is a lex- doctoral program in linguistics, MIT, 1961; Professor icalist theory of word formation (i.e. all the operations of Modern Languages, MIT, 1961–1976; Head, of word formation occur in the speaker’s lexicon or Department of Foreign Languages and Linguistics, mental dictionary). Generative morphology consisted MIT, 1976; Head, Department of Linguistics and primarily of a list of morphemes and a set of word for- Philosophy, MIT, 1976–1977; Ferrari P. Ward mation rules (WFRs) which built words out of the Professor of Modern Languages and Linguistics, MIT, morphemes. Work in generative morphology eventual- 1976–1981; Institute Professor, MIT, 1981–1996; and ly evolved into the theory of Lexical Phonology. Institute Professor Emeritus, MIT, 1996–present. He In the 1990s, Halle moved away from the lexicalist was a J.S. Guggenheim Fellow in 1960–1961 and J.R. position his work had previously taken, and developed Killian Jr. Faculty Achievement Award Lecturer at MIT the theory known as Distributed Morphology (DM), in in 1978–1979. He received the Science Prize from collaboration with Alec Marantz. Distributed Union d’Assurances de Paris in 1991; he also received Morphology holds that words are not constructed in the D.Sc. (hon), Brandeis University in 1988, and DHL, lexicon. Instead, syntactic operations which influence University of Chicago, 1992. He is a member of the 436  HALLIDAY, MICHAEL ALEXANDER KIRKWOOD Linguistic Society of America 1951–present, and was Halle, Morris, and William Idsardi. 1994. General properties of Vice President, Linguistic Society of America in 1973; stress and metrical structure. The handbook of phonological theory, ed. by John Goldsmith, 403–43. Oxford: Blackwell President, Linguistic Society of America in 1974; and Publishers. Fellow, American Academy of Arts and Sciences Halle, Morris, and Samuel J. Keyser. 1971. English stress: its 1963–present. Morris Halle is also a member of the form, its grows and its role in verse. New York: Harper and National Academy of Sciences from 1988 to present. Row. Halle, Morris, and Alec Marantz. 1993. Distributed morpholo- gy and the pieces of inflection. ed. by Kenneth L. Hale and References Samuel J. Keyser, 11–176. The view from building 20: essays in linguistics in honor of Sylvain Bromberger. Anderson, Stephen R, and Paul Kiparsky (eds.) 1973. A Halle, Morris, Bert Vaux, and Andrew Wolfe. 2000. On feature Festschrift for Morris Halle. New York: Holt, Rinehart and spreading and the representation of place of articulation. Winston, Inc. Linguistic Inquiry 32. 387–444. Aronoff, Mark, and Richard T. Oehrle (eds.) 1984. Language Halle, Morris, and Jean-Roger Vergnaud. 1987. An essay on sound structure: studies in phonology. Presented to Morris stress. Cambridge, MA: MIT Press. Halle by his teacher and students. Cambridge, MA: MIT Harley, Heidi, and Rolf Noyer. 1999. State-of-the-article: dis- Press. tributed morphology. Glot International 4.4. 3–9. Chomsky, Noam, and Halle Morris. 1968. The sound pattern of Jakobson, Roman, C.G.M. Fant, and Morris Halle. 1952. English. Cambridge, MA: MIT Press. Preliminaries to speech analysis. Cambridge, MA: MIT Goldsmith, John A. 1990. Autosegmental and metrical phonol- Press. ogy. Oxford: Basil Blackwell. Jakobson, Roman, and Morris Halle. 1956. Fundamentals of Halle, Morris. 1959. The sound pattern of Russian. The Hague: language. The Hague: Mouton Publishers. Mouton Publishers. Kenstowicz, Michael. 1994. Phonology in generative grammar. ———. 1962. Phonology in generative grammar. Word 18. Oxford: Blackwell Publishers. 54–72. Kenstowicz, Michael, and Charles Kisseberth. 1979. ———. 1973. Prolegomena to a theory of word formation. Generative phonology: description and theory. San Diego: Linguistic Inquiry 4. 3–16. Academic Press. ———. 1975. Confessio Grammatici. Language 51. 525–35. Kiparsky, Paul. 1982. From cyclic phonology to lexical phonol- ———. 1977. Tenseness, vowel shift and the phonology of ogy. The structure of phonological representations, Vol. 1. back vowels in modern English. Linguistic Inquiry 8. ed. by Harry van der Hurst and Norval Smith, 131–75. 611–25. Dordrecht: Foris. ———. 1983. On distinctive features and their articulatory Mohanan, K.P. 1982. Lexical Phonology, Ph.D. Dissertation implementation. Natural Language and Linguistic Theory 1. MIT, Cambridge, MA. 95–105. Newmeyer, Frederick. 1986. Linguistic theory in America. San ———. 1989. Why phonology is different. Linguistic Inquiry Diego: Academic Press. 20. 51–70. ———. 1995. Feature geometry and feature spreading. ANDREW CARNIE Linguistic Inquiry 26. 1–46. ———. 1997. On stress and accent in Indo-European. See also Chomsky, Noam; Jakobson, Roman; Mor- Language 73. 275–313. phology; Phonology Halliday, Michael Alexander Kirkwood In his thesis (‘The secret history of the Mongols’) and semantics, discourse analysis and stylistics, phonolo- in other research, M.A.K. Halliday mainly drew on gy, sociolinguistics, computational linguistics, lan- what he had learned in China, but for his linguistic the- guage education, and child language development. ory he drew on the work of the British teacher J.R. The most distinguishing feature of Halliday’s work Firth, who had drawn from his in-depth knowledge of is its holistic character. Halliday interprets language Indian Linguistic tradition. Firth was also influenced as a network of relationships and locates it in a socio- by the work of Polish anthropologist Bronislaw cultural context. For him, ‘language’ is a resource for Malinowski. All these strands inform Halliday’s work, making meaning—a semeogenic system. He suggests which is unique in its multinational range of influence. that language is a higher-order semiotic system—one Halliday’s work provides a theoretical base, offers that has a lexicogrammatical level of organization descriptive illustration, and projects applied values in (i.e. a system of wording) and one that is metafunc- almost all branches of linguistics, grammar and tionally organized (i.e. organized functionally into 437 HALLIDAY, MICHAEL ALEXANDER KIRKWOOD simultaneous strands of meaning: ideational, interper- Southern California. In the early 1990s, Halliday sonal, and textual). Our experience gets transformed began to interact with researchers developing fuzzy into meaning and this is effected by grammar. Hence, theory and logic in Japan. Halliday says that the grammar of every natural lan- Halliday’s commitment to research applications is guage is a theory of human experience: ‘a theory that related to his notion that linguists are socially account- we hold unconsciously, but that is all the more potent able. This is also reflected in his concern, since the for that very reason.’ Grammar ‘construes our experi- 1950s, with giving value to all varieties of language— ence’ but also ‘enacts our interpersonal relationships’— showing, for example, that spoken language is highly sharing experiences with others, giving orders, making systematic and complex in spite of the prevailing myth offers, and so on. In other words, ‘grammar transforms that it is scrappy and unorganized. The Hallidayan experience into meaning’ and it is this that constitutes mode of intellection recommends developing multiple what is called ‘understanding’. ‘Understanding and views on a phenomenon and advises engendering syn- knowing are semiotic processes—processes of the thesis when faced with a choice between a thesis and its development of meaning in the brain of every individ- antithesis. He wants us to move in helixes, not in circles. ual; and the powerhouse for such processes is the gram- mar.’ In his special lecture on July 19, 1999 at the Biography Central Institute of English and Foreign Languages in Hyderabad (India) — ‘Language and the Reshaping of Michael Alexander Kirkwood Halliday was born in Human Experience’—Halliday said that Leeds, Yorkshire (England) in 1925. He received his Ph.D. in Chinese Linguistics from Cambridge Children of around four to six years old are just reach- ing the stage, in their language development, when they University in 1955. can handle meanings that are abstract: they can con- He was earlier a student in China at Peking strue entities that have no perceptual correlate, like University and Lingnan University, where he assisted worth and clue and habit and intend and price; and this Professor Wang Li in surveying the dialects of Chinese has two important consequences. First, it means that in the Pearl River Delta. He taught at Edinburgh they can cope with abstract symbols, like letters or char- University, University College London, Nairobi acters, and the abstract concepts that go with them (Kenya), Chicago, Stanford, and Essex University. In (including, the critical distinction between writing and 1976, he joined the University of Sydney as the found- drawing); so they can now master this new medium. ing Professor of Linguistics. He retired in 1987 and Secondly, it means that they can cope with abstract cat- went on to serve as Visiting Professor at NUS, egories, and so are ready to explore new forms of Singapore and at ICU Tokyo. knowledge. In other words they are ready for a reshap- ing of their previous experience. Halliday has received honorary doctorates from the University of Nancy, the University of Birmingham, Halliday’s work on the internal organization of York University, and Macquarie University and was Language as a Semiotic system has also been holistic. conferred the Honorary Fellowship of the Central He describes semantics, lexicogrammar, and phonolo- Institute of English and Foreign Languages gy as resources for the language user, representing (Hyderabad, Lucknow, and Shillong), India. He has them as networks of options in what the user can published over 150 books and articles. His seminal mean. In his work ‘system’, comprising choices on works are collected and edited by Jonathan J. Webster the paradigmatic dimension, is the central organizing and Published by Continuum (London and New York), concept. as a ten-volume series. Systemic Functional theory has been applied with a wide range of contexts of research application. Halliday has pioneered the development of ‘educa- References tional linguistics’, interacting with teachers since the Halliday, Michael A.K. 1957. Some aspects of systematic 1960s in the context of various projects. He has made description in grammatical analysis. Studies in linguistic significant contributions in other areas, including liter- analysis, 54–67. London: Basil Blackwell. _________. 1961. Categories of the theory of grammar. Word ary studies and computational linguistics. He was 17. 241–92. involved in work on machine translation already in the _________. 1967. Intonation and grammar in British English. 1950s. The Hague: Mouton. His work has been particularly influential in the _________. 1967/1968. Notes on transitivity and theme in development of text generation as one aspect of natu- English. Journal of Linguistics 3. 71–81, 199–244; 4. 171–216. ral language processing—particularly through the _________. 1975/1977. Learning how to mean: explorations in Penman system that was launched in 1980 at the the development of language. London: Edward Arnold; New Information Sciences Institute of the University of York: American Elsevier. 438  HANDWRITING _________. 1976. System and function in language, ed. by Halliday, Michael A.K., and Christian M.I.M. Matthiessen. G.R. Kress. London: Oxford University Press. 1999. Construing experience through meaning. London and _________. 1978. Language as a social semiotic: the social New York: Continuum. interpretation of language and meaning. London: Edward Halliday, Michael A.K., and Ruqaiya Hasan. 1976. Cohesion in Arnold. English. London and New York: Longman. _________. 1985. An introduction to functional grammar. VENNELAKANTI PRAKASAM London: Arnold; New York: Oxford University Press (sec- ond revised edition 1994). See also Firth, John Rupert Handwriting Despite the common use of electronic communication every cognitive process takes a measurable amount of with keyboards, handwriting is still a major mode of time. Under certain assumptions, it is also possible to written language production. The study of handwriting determine the nature of the intervening processes, e.g. is concerned with the various mechanisms and letter-shape selection or force implementation. processes that intervene between the intention to write Likewise, the duration of the movement itself (move- a message and the movements of the body that execute ment time) may be assessed; this is indicative of the the intention by instrument gestures on a writing sur- current mental load on the limited-capacity informa- face. Some of the processes, such as conceptualization tion-processing system, because of the preparation and and formulation, are the same for spoken and written motor programming of further parts of the message. utterances. This paper will therefore concentrate on Because of the large forces of the involved muscles the specific movement-control aspects of handwriting: relative to the little writing movements and the small its planning, preparation, and production. implied forces, such as friction and gravitation, it Furthermore, it will be restricted to cursive Latin script seems justified to consider the kinematics (i.e. the by skilled right-handed adults. When fully developed, movement features) in the two-dimensional writing the skill enables fluent, semiautomatic writing ges- plane as a function of time as an adequate description tures, permitting cognitive and linguistic processing to of the dynamics of the handwriting trajectory. The continue in parallel. standard two-dimensional method is to let the partici- The aim of handwriting research is to arrive at a pant in the experiment write on a graphics tablet (dig- comprehensive insight into how all aspects of skill and itizer) that samples the x and y coordinates of the pen task are acquired, represented, planned, and organized tip at a high rate (100 times per second or more). by the writer, seen as a human information-processing Dedicated software analyzes the writing trace, yield- system, and how the eventual kinematic, or movement, ing relevant parameters such as size, slant, curvature, sequences are physically performed. In particular, velocity, acceleration, jerk, pen pressure, and pen tilt there is interest in mutual interactions, i.e. in how the as a function of time. Apart from reaction time and act of handwriting influences linguistic production on movement time, fluency measures may also reveal the one hand, and in how kinematic performance local or global processing load. Many laboratories measures reveal cognitive processes on the other. The now have a three-dimensional motion-tracking appara- latter aspect has received the most attention in recent tus as well, to monitor the individual joints (from investigations in the multidisciplinary field of quanti- shoulder to fingertip) involved in writing movements. tative and formal handwriting research, which, in con- By geometric transformation, the recorded data may trast to the art of graphology, is known as be represented, e.g. as a time function of the joints’ ‘graphonomics’. Contemporary research is primarily rotation, or as the phase relations between several experimental and is executed in laboratories equipped moving effector segments (e.g. hand and fingers). with timing devices and two- and three-dimensional The writing limb, with its multitude of segments, movement-recording instruments. joints, and muscles in the arm, hand, and fingers, is Exact timing allows an estimate of the delay (reac- capable of moving in many different ways. tion time) between the instruction to write and the Technically speaking, it has a large number of degrees actual start of the movement; this duration reflects the of freedom. However, this number is often reduced— intervening amount of cognitive processing, because although flexibly and not always in the same way—by 439 HANDWRITING the coactivation of several groups of muscles acting as duced ballistically, i.e. in a single burst, with only one single functional synergies or coordinative structures. acceleration and one velocity peak. The normal dura- In very general terms, the shoulder and elbow take tion of such a fluent stroke is 100 to 125 milliseconds. care of the global rightward progression across the A continued series of loops (each considered as two page, while the hand rotating in the wrist is responsi- strokes, one down and one up) will thus result in a ble for the small horizontal movements within and repetitive digitized signal with a frequency of 4 to 5 between letters, and the thumb and fingers are mainly hertz. An allograph may contain between two and six in charge of the vertical letter strokes. strokes, so that on average, two or three letters are Since 1965, mechanical and computational models written per second. Allographs and strokes have both of handwriting have been developed. The earliest ones been proposed as the basic units of handwriting pro- were concerned with the biomechanical formation of cessing (long-term storage, planning, and execution); strokes and their concatenation into letters and words. more recently, it has become clear that the size of the Later models, since 1980, accounted for the cerebral basic unit depends on familiarity, i.e. on learning and and neural organization of handwriting and assumed practice. information-processing stages in the cognitive-psy- A representative information-processing model chology tradition. Examples of these stages are the combines serial with parallel processing. Handwriting selection and ordering of the orthographically required is seen as a multicomponent task involving various letters (graphemes), the retrieval of a specific letter independent modules. From high to low, these deal shape (allographs; see below) from long-term motor with semantics (meaning), syntax (sentence structure), memory, and the setting of the required muscles and words, speech sounds, graphemes, allographs, strokes, their force levels to execute that letter by a selected task space, joints, and muscles. The order given here effector (e.g. hand or fingers) in the current task-space reflects the serial processing sequence: each module context at a specified size and speed. Recently, the transmits its processing result to the next-lower level. models stemming from the biomechanical tradition The parallel features of the model originate from the have gained in complexity and scope. They incorpo- principle that concurrently with the lower-order writ- rate the control of handwriting gestures at higher lev- ten performance of an earlier part of the message, later els in the brain, so that, e.g. the selection of the most parts will be planned and prepared by modules higher suitable effector combination for smooth performance in the system. Each module requires a small amount of is accounted for. Thus, the two types of model are processing time (milliseconds) and capacity drawn increasingly covering common ground. from a common source. If a message contains a pro- From a psycholinguistic viewpoint, however, the cessing difficulty for one of these modules, this may neurocognitive and information-processing models be reflected by a delayed initiation (increased reaction have made the most interesting contributions by gener- time) or by a slower ongoing writing movement ating testable hypotheses on the handwriting process. (increased movement time). Because of the processing These models assume that handwriting involves a hier- priority of larger units, increased cognitive load at archical procedure in which to-be-written larger units higher levels is reflected earlier in the writing trajecto- (representing the entire message, whole phrases, and ry than an extra load at lower levels. An unusual allo- words) are activated first, followed step-by-step by the graph combination will thus be ‘announced’ by a activation of their even more detailed components (sin- slight delay several letters in advance, whereas a diffi- gle allographs and individual strokes) at increasingly cult stroke will result in a slower production of the lower levels. Allographs are the (often idiosyncratic) immediately preceding stroke or even of the very shape variations of a grapheme with a certain letter stroke itself. identity (e.g. a specific cursive lowercase <r>). Research over three decades has shown that lin- Different allographs are required or preferred in differ- guistic factors of various kinds and levels do indeed ent contexts. Allograph representations are assumed to influence the handwriting process in a modular way, as be stored as generalized motor programs in long-term indicated here. Initiation and trajectory formation are memory, and—when required in the message—ready affected by the features of units with high, medium, or to be retrieved and executed in any size and orientation low linguistic status, which are assumedly processed (e.g. on a large vertical blackboard). at corresponding levels in the proposed handwriting- Strokes—effectively near-vertical trace segments— movement hierarchy. As indicated, high levels involve are defined in the spatial domain by the upper and words, morphemes (subword units of meaning), and lower turning points of the pen-tip excursions. These syllables; medium levels encompass phonology turning points are characterized by sharp bends (high (speech-sound features), orthography (spelling), and curvature), which are associated with low speed letter clusters; low levels imply graphemes, allo- (velocity minima). In fluent writing, a stroke is pro- graphs, and strokes. In many cases, these findings are 440  HARRIS, ZELLIG SABBETAI corroborated by clinical observations involving double Ellis, Andrew W. 1993. Reading, writing, and dyslexia, 2nd edi- dissociation between writing disturbances: different tion. Hillsdale, NJ: Erlbaum. Faure, Claudie, Paul Keuss, Guy Lorette, and Annie Vinter patients seem to suffer from different exclusive writing (eds.) 1994. Advances in handwriting and drawing: a multi- failures, presumably mediated by specific neural struc- disciplinary approach. Paris: Europia. tures at different levels. Application of the current Graham, S., and N. Weintraub. 1996. A review of handwriting insights into the neurocognitive and motor-control research: progress and prospects from 1980 to 1994. aspects of handwriting has resulted in advances in Educational Psychology Review 8. 7–87. Kao, Henry R. S., Gerard P. van Galen, and R. Hoosain (eds.) educational and remedial practice, in medication for 1986. Graphonomics: contemporary research in handwrit- motor disorders, in written-document examination, in ing. Amsterdam: North-Holland. the analysis of ancient writing, in forensic expertise, Levelt, Willem J.M. 1989. Speaking: from intention to articula- and in automatic handwriting recognition. tion. Cambridge, MA: MIT Press. Martlew, Margaret (ed.) 1983. The psychology of written lan- guage. New York: Wiley. References Peck, M., E.N. Askov, and S.H. Fairchild. 1980. Another Askov, E., W. Otto, and W. Askov. 1970. A decade of research decade of research in handwriting: progress and prospects in in handwriting: progress and prospect. Journal of the 1970s. Journal of Educational Research 73. 283–98. Educational Research 69. 86–98. Plamondon, Rejean, Ching Y. Suen, and Marvin L. Simner Bouwhuisen, Chris F., Ruud G.J. Meulenbroek, and Arnold (eds.) 1989. Computer recognition and human production of J.W.M. Thomassen. 2002. A 3D motion tracking method handwriting. Singapore: World Scientific. in graphonomic research: possible applications in future Thomassen, Arnold J.W.M., Paul J.G. Keuss, and Gerard P.van handwriting recognition studies. Pattern Recognition 35. Galen (eds.) 1984. Motor aspects of handwriting: approaches 1039–47. to movement in graphic behavior. Amsterdam: North-Holland. Bullock, Daniel, Stephen Grossberg, and Christian Mannes. Van Galen, Gerard P. 1991. Handwriting: issues for a psy- 1993. A neural network model for cursive script production. chomotor theory. Human Movement Science 10. 165–91. Biological Cybernetics 70. 15–28. ARNOLD J.W.M. THOMASSEN Harris, Zellig Sabbetai The scientific legacy of Zellig S. Harris can be divid- He then moved into a more general study of lan- ed into five areas: (1) Northwest Semitic linguistics; guage theory, even though he continued to maintain an (2) Structural analysis; (3) Transformational analysis; interest in Semitic languages as material for theoreti- (4) Operator grammar; and (5) Politics. cal investigations. Other semitists regretted his defec- While Harris was an undergraduate student at the tion from the field, and even expressed resentment at University of Pennsylvania in 1929, epoch-making what he was doing. The Semitics scholar J. Cantineau, texts in Ugaritic were published and promptly deci- in a review of Harris’s book Methods in structural lin- phered. This turned his attention to new and exciting guistics, claimed that Harris placed facts on a pro- possibilities of research in Northwest Semitic, and he crustean bed to fit his theories. He added that Harris subsequently worked in this area with his teacher J.A. brought nothing new or useful to linguistics, despite Montgomery, publishing with him a work on Ugaritic the specious appearance of being scientific. mythological texts. He already had a background in Harris shook off all criticism and entered his new Hebrew from his family. During the 1930s, he estab- phase of scholarship, the cornerstone of which was the lished a solid reputation as a semitist, publishing a book that Cantineau reviewed unfavorably. Although Phoenician grammar in 1936, and a study of Canaanite he was not formally a student of Edward Sapir, he dialects in 1939. Both works have remained useful. It acknowledges his great debt to him. Harris stresses the was here that he developed the style of writing that importance of describing language in terms of how lasted throughout his career. It was direct and sparse, items fit together, and how they are distributed. While with a wealth of detail and documentation when these he admits that meaning is the goal of most language were called for. But he made no concessions to the (poetry may sometimes be an exception), it is not sus- reader. Harris did not pretend to be a popularizer, and ceptible to scientific quantification or description, one may have to struggle to follow his technical, tight- and it, along with psychological and mentalistic ly constructed argumentation. notions, must be strictly avoided. He tried to isolate 441 HARRIS, ZELLIG SABBETAI the elements of a language, and the relationships that first argument (‘I know’) and the second argument is were possible between them. Harris adhered to the the same, giving Oon. Harris recognizes that this may structuralists’ credo that linguistics best studies lan- not give an analysis of pedagogic value by way of giv- guage synchronically, and on its own terms. ing a useful order of the findings; but then he was not Up to his time, analysis of language, whatever its particularly interested in pedagogy. Like Everest theoretical basis might be, ended at the conclusion of climbers, he explored the claimed mathematical struc- the sentence. Harris had the ambition of setting up an ture of language simply because it was there. Harris analysis of texts, sentences that were strung together, also explores the unreduced sentences of a particular and this led to his concept of Discourse Analysis, a language. We have no evidence of the existence now subject on which he taught a course regularly at the of a ‘me Tarzan, you Jane’ type of prehistoric lan- university. This endeavor led him to set up normal guage; so-called primitive languages may have richly forms of sentences, so that they could be compared complex syntactic structures, and are not fundamen- within texts, and this brought him, more or less acci- tally different in nature from other better known lan- dentally, to the notion of transformations that would guages. Maybe he considered these unreduced connect, for example, a sentence in its active and pas- sentences to belong to an earlier stage of language, but sive form (the dog bit the man/the man was bitten by he does not state this expressly, even though toward the dog). Thus evolved the idea of ‘transformations’, the end of his life in his masterly A theory of language which led in turn to ‘transformational grammar’, and information, he gingerly explores the nature and which has acquired great prominence, mainly on development of language. A simple example may be account of the research and publications of Noam seen in sheep eat green grass. This is a reduced form Chomsky, who was Harris’s student. This gave rise to of two sentences sheep eat grass and grass is green, a school of thought of continuing prominence, which i.e. sheep eat grass which is green, where the first two was subjected to obloquy similar to that meted out letters of which (wh… ) form a link, and the remainder once to the structuralists, being described by the struc- represents the word grass. An element of economy turalist Charles F. Hockett (died: 2000) as ‘a theory plays a part here, which has long been recognized as a spawned by a generation of vipers’. possible factor in phonetic change. Harris believes But Harris did not go down the same path as that the operator-argument method is fully adequate Chomsky. Already in the 1960s, he was teaching a for unreduced sentences at least. A theory of language course in Linguistics and Mathematical Logic, in and information, represents the end of his life work, which he was mulling over his aim to produce a math- which had remarkable consistency and tenacity. ematical model for language. It should be stressed that Thereafter, he devoted himself to a substantial politi- Harris, in moving to his new approach to language, cal tract, published posthumously by a group of his which was distinct from the path he had previously admirers. followed, and distinct from that of Chomsky, did not thereby reject his previous work, even though he may Biography have felt that he had taken it as far as he reasonably could. Harris did not belong to a ‘God’s truth’ mode of Zellig Sabbetai Harris was born in Balta, Ukraine, on looking at scientific description. Various approaches October 12, 1909. He was brought to Philadelphia at might have their own particular validity. He was the age of four years. His father was a mohel (Jewish remarkably unpartisan, and spent no time in engaging ritual circumciser) and cantor. Harris received the in public relations on behalf of himself or his views, a degree of A.B. at the University of Pennsylvania in fact that may account for the relative obscurity of his 1930, followed by a Ph.D. four years later, and spent viewpoints, and his failure to establish a school devot- his entire academic career at that institution as a stu- ed to his principles. dent and faculty member, until his retirement at the His new approach is best understood by a simple age of 70. For much of his life he commuted between example. If we consider the sentence sheep eat grass, Philadelphia and Israel, where he was a member of the acceptable to any English speaker, Harris observes communal settlement Kibbutz Mishmar Haemek, near that the word eat imperatively requires two other Haifa, and reportedly was known there simply as words, which do not themselves require anything else. Zellig the carpenter. This kibbutz was part of the He expresses this as Onn, where O represents the extreme left-wing Hashomer Hatzair (‘the Young operator (in this case the word eat) and this operator Watchman’) movement. He participated in setting up requires two arguments (here sheep and grass) which an adult education program for members of such set- can be considered to be operators with null arguments, tlements. His wife Bruriah was a researcher in physics hence n. If we expand this sentence to I know that at the Weizmann Institute in Rehovot, Israel. Harris sheep eat grass, then eat now has an operator as its kept his academic and political interests separate, and 442  HAUGEN, EINAR in his scientific work never broached his political _________. 1951. Methods in structural linguistics. Chicago: views, which were closest to anarchism in the spirit of University of Chicago Press. xv, 384p. Published in 1960 under title: Structural linguistics. Kropotkin. In 1941, he took over the editorship of the _________. 1952. Discourse analysis. Language 28. 1–30. Journal of the American Oriental Society, one of the _________. 1957. Co-occurrence and transformation in lin- oldest learned societies in the United States, from guistic structure. Language 33. 283–340. Norman Brown, a colleague at the University of _________. 1982. A grammar of English on mathematical prin- Pennsylvania. He served in that capacity until 1948. ciples. New York: Wiley, xvi, 429p. _________. 1991. A theory of language and information: a He was president of the Linguistic Society of America mathematical approach, Oxford: Clarendon Press; New in 1955. He published numerous articles in the journal York: Oxford University Press. xi, 428p. Language until 1957; after that time, most of his books _________. 1997. The transformation of capitalist society. and papers were published in Europe. He died on May Lanham, Maryland: Rowman and Littlefield. [Completed 22, 1992, just after publishing an uncharacteristically manuscript prepared for publication posthumously by Murray Eden, William Evan, and Seymour Melman.] readable account of his views in his A theory of lan- Hiz´, Henry. 1994. Zellig S. Harris. Proceedings of the American guage and information. Philosophical Society 138. 509–22. Koerner, E. F. Konrad, 1993. Zellig Sabbetai Harris: a compre- hensive bibliography of his writings, 1932–1991. References Historiographia Linguistica 20: 2/3 pp. 509–22. An updated Chomsky, Noam. 1957. Syntactic structures. The Hague: version of this bibliography is on the World Wide Web at Mouton. Janua Linguarum, Series minor, nr. 4, 117p. http://www.dmi.columbia.edu/zellig/ZSHbibliography.html. Harris, Zelig Sabbetai. 1939. Development of the Canaanite This includes a bibliography of evaluations of Harris’s work. dialects; an investigation in linguistic history. New Haven: Matthews, Peter. 1999. Zellig Sabbetai Harris. Language 75. American Oriental Society. American Oriental series, v. 16. x, 112–9. 108p. ALAN D. CORRÉ Haugen, Einar Einar Haugen was one of the foremost Scandinavian Norwegian dialectology, publishing, among other linguists of his time. His parents both immigrated works, a two-volume survey of varieties of Norwegian from central Norway to the United States, settling in spoken in the United States (The Norwegian language Sioux City, Iowa, in 1899. Haugen grew up there, in a in America: a study in bilingual behavior; originally Norwegian neighborhood, and learned Norwegian as a published in two volumes in 1953 and reissued as one native language and English in school. He in fact volume in 1969) and numerous articles on the topic. learned two dialects of Norwegian, the more standard He was also interested in numerous aspects of soci- dialect spoken by his schoolteacher mother and the olinguistics. His interest in language planning and lan- more rural dialect spoken by his father. He spent two guage standardization goes back to his dissertation; he years in Norway as a child, from 1914 to 1916, continued to work in this field, and eventually pub- between the ages of eight and ten, which allowed him lished a book dealing with language planning and to master his father’s dialect in its natural setting. This Norwegian. Bilingualism, another interest stemming stay reinforced his already strong ties to Norway and from his childhood experiences, was also a prominent Norwegian; the influence of his Norwegian immigrant theme in his work. The Norwegian language in community and his childhood visit to Norway cannot America, as indicated by the subtitle, also dealt exten- be underestimated when analyzing his life’s work. sively with bilingualism, focusing largely on the adap- Haugen contributed to many areas of linguistics. He tation of English words into American Norwegian. A was deeply concerned with the Norwegian language, later volume, Bilingualism in the Americas: a bibliog- writing several textbooks of Norwegian, and directing raphy and research guide, was also a seminal contri- the compilation of a Norwegian–English dictionary. bution to the field. Finally, Haugen was an important During World War II, he directed a project at the contributor to the field of ‘language ecology’ (and in University of Wisconsin to teach servicemen fact coined the term), the study of interactions between Norwegian. He also contributed extensively to a language and its environment. His 1972 work, 443 HAUGEN, EINAR The ecology of language, reprints selected essays on Scandinavian Study, 1938; Linguistic Society of topics such as the relationship between language and America, 1950; Ninth International Congress of immigration and communication difficulties between Linguists, 1962; American Dialect Society, 1962. He speakers of various languages. was also a member of Permanent International Haugen’s contributions to historical Scandinavian Committee of Linguists in 1966–1972. He was visit- linguistics are critical to the field. He prepared an edi- ing lecturer (1938) and Fulbright Research Professor tion and translation of The first grammatical treatise, (1951–1952), University of Oslo; Visiting professor, an anonymous twelfth-century Icelandic work on Old University of Iceland, 1955–56; Adviser to the English Norse phonology and orthography. His monumental Language Educational Council of Japan, 1958–1960; work on the history of the Scandinavian languages, and Fulbright lecturer, University of Uppsala, The Scandinavian languages: an introduction to their 1976–1977. He received grants and fellowships from history, originally published in 1976 and translated the Guggenheim Foundation (1942), National into German in 1984, remains a standard handbook. Endowment for the Humanities (1967), Center for His later monograph, Scandinavian language struc- Advanced Study in the Behavioral Sciences at tures: a comparative historical survey, offers an acces- Stanford University (1963–1964), and National sible discussion of the development of the various Science Foundation (1967–1969). Haugen was a components (phonology, morphology, syntax, and the member of the American Academy of Arts and lexicon) of the Scandinavian languages. Sciences, the Norwegian, Icelandic, Swedish, and Haugen also helped shape the field of structuralist Danish academies of science, and the Linguistic linguistics. He authored two influential articles on Research Group of Pakistan. He also received the structuralist phonology, ‘Facts and phonemics’ (with Order of St. Olaf, Norway in 1940 and Order of the W.F. Twaddell 1942) and ‘Phoneme or prosodeme’ North Star, Sweden in 1961. He received several hon- (1949). Other phonological topics he dealt with orary degrees, including University of Michigan included the vowel systems of Norwegian and (1953), St. Olaf College (1958), the University of Oslo Icelandic, tone in Norwegian, and the role of the syl- (1961), and the University of Iceland (1971). He also lable in phonology. Yet, he was never parochial in his received the Nansen Award in Oslo in 1970 and the scholarship; in his presidential address to the Jancke Prize in Uppsala in 1976. Haugen died in Linguistic Society of America (published in 1951), he Cambridge, Massachusetts, on June 20, 1994. urged American linguists to avoid scholarly isolation- ism and take developments from European scholarship References into consideration. Haugen also contributed to literary studies. He Appel, René, and Pieter Muysken. 1987. Language contact and bilingualism. London: Arnold. translated a number of Icelandic sagas and plays, a Eliasson, Stig, and Ernst Håkon Jahr (eds.) 1997. Language and history of Norwegian literature, and a biography of its ecology. Essays in memory of Einar Haugen. Berlin: Henrik Ibsen. He prepared word counts of a number Mouton de Gruyter. of Scandinavian authors, and also published studies of Firchow, Evelyn Scherabon, Kaaren Grimstad, Nils Hasselmo, various Icelandic sagas. and Wayne O’Neil (eds.) 1972. Studies for Einar Haugen. The Hague: Mouton. Fishman, Joshua. 1995. Einar Haugen. Language 71. Biography Haugen, Einar, and W.F. Twaddell. 1942. Facts and phonemics. Language 18. Einar Haugen was born in Sioux City, Iowa on April Haugen, Einar. 1947. Spoken Norwegian. 19, 1906. He attended Morningside College in ———. 1949. Phoneme or prosodeme. Language 25. ———. 1950. First grammatical treatise: the earliest Germanic 1924–1927. He received his B.A. in 1928 from St. phonology. An edition, translation, and commentary; 2nd Olaf’s College; M.A. in 1929, and Ph.D. for a disser- revised edition 1972. tation on the development of the New-Norwegian ———. 1951. Directions in modern linguistics. Language 27. dialect of Norwegian, University of Illinois (Urbana) ———. 1953. The Norwegian language in America: a study in in 1931. He was Assistant Professor, University of bilingual behavior. Vol. I: The bilingual community. Vol. II: The American dialects of Norwegian; republished in one Wisconsin (Madison), 1931–1936; Associate volume, 1969. Professor, 1936–1938; Torger Thompson Professor of ———. 1956. Bilingualism in the Americas: a bibliography Scandinavian Languages, 1938–1962; Vilas Research and research guide. Professor of Scandinavian Languages, 1962–1964; ———. 1972. The ecology of language. Victor S. Thomas Professor of Scandinavian and ———. 1972. Studies by Einar Haugen. ———. 1976. The Scandinavian languages: an introduction to Linguistics, Harvard University, 1964–1975; Cultural their history. relations attaché, US Embassy in Oslo, 1945–1946; ———. 1982. Scandinavian language structures: a comparative and President, Society for the Advancement of historical survey. 444  HAUSA AND CHADIC LANGUAGES ———. 1987. Blessings of Babel : bilingualism and language Romaine, Suzanne. 1989. Bilingualism. Oxford: Blackwell. planning. Problems and pleasures. MARC PIERCE Naess, Harald S., and Sigmund Skard (eds.) 1971. Studies in Scandinavian–American interrelations: dedicated to Einar See also Bilingualism; Language Planning; Struc- Haugen. Oslo: Universitetsforlaget. turalism Hausa and Chadic Languages The Chadic language family includes an estimated The accepted internal classification groups Chadic 140–150 languages spoken in the sub-Saharan area to languages into three major coordinate branches, West, the west, south, and east of Lake Chad. The best known Central (= Biu-Mandara), and East, plus an isolated and most widely spoken Chadic language is Hausa, Masa cluster (with sub-branches and smaller group- with upward of 30 million first-language speakers, ings) as follows: more than any other language in Africa south of the West-A, e.g. Angas, Bole, Fyer, Galambu, Gera, Sahara. The remaining languages, some of which are Goemai, Gwandara, Hausa, Kanakuru, Karekare, Kirfi, on the brink of extinction, probably number little more Kulere, Kwami, Mupun, Pero, Ron, Sha, Sura, Tangale than several million speakers in total, varying in size West-B, e.g. Bade, Boghom, Duwai, Geji, from less than half a million to just a few dozen speak- Guruntum, Miya, Ngizim, Pa’a, Saya, Warji ers, and new languages are still occasionally discov- Central-A, e.g. Bachama-Bata, Bura-Pabir, Daba, ered. Most Chadic languages remain unwritten, and for Ga’anda, Gisiga, Glavda, Gude, Guduf, Higi-Kapsiki, many only brief word-lists are available, although for Kilba, Lamang, Mandara, Margi, Matakam, Mofu- some, e.g. Bidiya, Guruntum, Kanakuru, Kera, Kwami, Gudur, Ouldeme, Sukur, Tera Lamang, Margi, Miya, and Mupun, good descriptive Central-B, e.g. Buduma, Kotoko, Logone, Mbara, grammars now exist, and several high-quality diction- Musgu aries have appeared, e.g. Dangaléat, Lamé, Ngizim, Central-C, e.g. Gidar and Tangale. Hausa has four up-to-date and compre- East-A, e.g. Kera, Lele, Nancere, Sarwa, Somrai, hensive reference grammars (all produced within the Tumak last few years) and two first-class dictionaries, and East-B, e.g. Bidiya, Dangaléat, Jegu, Mahwa, these works together make it the best documented lan- Migama, Mogum, Mubi, Mukulu, Sokoro, Toram guage in sub-Saharan Africa. Masa (isolated cluster), e.g. Lamé, Marba- Chadic is a member of the Afroasiatic language Monogoy, Masa, Musey superfamily, which includes Semitic (e.g. Amharic, Chadic languages have complex consonant invento- Arabic, Hebrew), Cushitic (e.g. Oromo, Somali), ries. A typical pattern is for a language to distinguish Omotic (e.g. Maale, Wolaytta), Berber (e.g. Tamahaq, voiceless stops such as /t/ from voiced /d/, in addition Tamazight), and (extinct) Ancient Egyptian/Coptic. to implosive stops // and prenasalized stops /nd/. The inclusion of Chadic within Afroasiatic was first Vowel inventories vary from two vowels, high /ə/ proposed almost 150 years ago, based on specific and low /a/, as in Bachama and Mandara, to seven points of resemblance between individual Chadic lan- vowels /i e ε a ɔ o u/, as in Dangaléat, with /i (e) a ə guages and other Afroasiatic languages. These features (o) u/ a regular pattern, and the diphthongs /ai/ and /au/ include an mV- prefix (V stands for an unspecified are found. It is also common throughout the family to vowel), which occurs on nouns that refer to the agent of distinguish between long and short vowels, especially an action, a location, or an instrument, e.g. [Hausa] ma- in word-medial position. hàifii ‘parent’, ma-fakaa ‘shelter’, ma-kullii ‘key, lock’. Chadic languages have contrastive tones (distinctive (Transcription: aa = long vowel, a = short; à(a) = intonational pitch levels), ranging from two (e.g. Hausa) L(ow) tone, â(a) = F(alling) tone; H(igh) tone is or three (e.g. Angas) level tone systems. Although tone unmarked.) Despite this, the family relationship has can function in lexical discrimination, its key role is only been generally accepted as a fact over the last 40 normally grammatical, e.g. [Hausa] taa/tâa/tà zoo ‘she years or so, following Joseph Greenberg’s (1963) com- came/will come/should come’, zàunaa! ‘sit down!’ vs. prehensive reclassification of African languages. zaunàa ‘to sit down’, nân ‘here (near me the speaker)’ 445 HAUSA AND CHADIC LANGUAGES vs. nan ‘there (near you the hearer)’. In a number of completed action), e.g. (transitivization) yàw-tu ‘take (West) Chadic languages, phonological tone is sensitive down’ from yàwwu ‘go down’ [Bole], (perfectivity) pet- to syllable structure in certain word categories. In Bole, inà ‘he came out’ [Pero]. Another characteristic feature for example, tone on bisyllabic verbs is predictable: if is the formation of so-called ‘pluractional’ verbs, denot- the verb is C(onsonant)V(owel).CV (light first syllable) ing an action repeated many times or affecting a plural- it will be H-H tone, e.g. tonu ‘sharpen’; if it is CVC.CV ity of subjects (if intransitive) or objects (if transitive), or CVV.CV (heavy first syllable) it will be L-H tone, e.g. [Hausa] tunàa/tun-tùnaa ‘remind/remind many e.g. ràamu ‘repair’, wùndu ‘call’. people and/or often’, [Bole] dàppu/dàdàppu ‘repair’, Many Chadic languages have masculine/feminine [Angas] fwin/fwan ‘untie’, [Pero] lofò/loffò ‘beat’. grammatical gender (an inherited Afroasiatic feature), In a number of languages, intransitive verbs (verbs with no distinction in the plural, and typically distin- without objects) are followed by an ‘intransitive copy guish gender in second- and third-person singular pro- pronoun’, which copies the person, number, and gender nouns, e.g. (independent pronouns) kai/kee ‘you features of the subject, e.g. [Kanakuru] nà pòrò-no ‘I (masc./fem.)’, shii/ita ‘he/she’ [Hausa], fiy/macə ‘you went out’ (lit. I went out-I). Verb stems can (masc./fem.)’, tə/njə ‘he/she’ [Miya]. Overt gender also be overtly inflected for tense–aspect–mood by seg- marking on nouns has also evolved in a number of West mental and/or tone changes, e.g. (perfective/ Chadic languages, e.g. [Western Bade] mângaa-n subjunctive/imperfective) kə`ma/kə`mì/kə`mà ‘hear’ ‘my friend (masc.)’ vs. mânga-n ‘my friend (fem.)’ [Ngizim], and grammatically-required number agree- (-aa = masculine, -a = feminine), [Hausa] jàakii ‘don- ment with the subject is also found sporadically, e.g. key’ vs. jàakaa ‘she ass’ (-ii = masculine, -aa = femi- [Karekare] mù ngàt-ʔan kò ‘we fell’ (with -ʔan plural nine). Some also exhibit the characteristic Afroasiatic suffix), cf. nà ngàtaa kò ‘I fell’. In some Chadic lan- n/t/n (masc./fem./pl.) noun/agreement marking pattern guages so-called ‘logophoric’ pronouns are used to (where the masculine and plural markers are phono- denote the person whose utterance is represented in logically identical), cf. [Masa] vèt-na ‘rabbit’, vèt-ta reported speech, e.g. (tones not marked) wur sat ni i n ‘female rabbit’, vèdai-na ‘rabbits’. nas an ‘hei said hei will beat me’ [Mupun], where the 3rd Noun plural formations are varied and elaborate, masculine singular logophoric pronoun i signals that and some plural suffixes can be reconstructed for the the subject of the subordinate clause refers to the same presumed ancestor language Proto-Chadic, e.g. *-Vn, person as the main clause subject. Some languages also *-aki, *-i, and *-ai (with *-Vn and *-ai probably even use impersonal forms (equivalent to implied [null] sub- reconstructable for Proto-Afroasiatic, the superfamily jects) with arbitrary, often human, reference, to express ancestor language). Examples: (-Vn) kùmən/kùmənən impersonal subjects, e.g. [Miya] a sənà-tlə`n ‘one will ‘mouse/mice’ [Bade], miyò/mishan ‘co-wife/co-wives’ spend the night’, and impersonal constructions contain- [Kanakuru], (-aki) goonaa/gòonàkii ‘farm(s)’ [Hausa], ing transitive verbs typically correspond to agentless (-i) duwimà/dùwìmi ‘guineafowl(s)’ [Gera], (-ai) passives in English, e.g. [Hausa] an dafà àbinci ‘the mùtù/mutai ‘sore(s)’ [Dangaléat]. food has been cooked’ (lit. (some)one.perf. cook food). Verbs in many Chadic languages preserve the lexi- Word order is normally S(ubject)-V(erb)-O(bject), cally arbitrary Proto-Chadic distinction between final e.g. [Hausa] Audù yaa sàyi mootàa ‘Audu bought a -a and final -ə verbs (where the final schwa vowel is car’ (lit. Audu he.perf. buy car), although VSO order is often pronounced as [i], [ə], or [u]), cf. [Tera] na ‘see’ found in a few Central Chadic/Biu-Mandara languages and dlə ‘get’, [Guruntum] daa ‘sit’ and shi ‘eat’, [Bole] spoken in the Nigeria-Cameroon border area, e.g. pata ‘go out’ and poru ‘say’, and in many West Chadic (tones not marked) kə bilə Musa tə bwaja ‘Musa has languages final -a(a) stems are typically limited to killed the leopard’ (lit. perf. kill Musa leopard) [Gude], bisyllabic verbs with an initial light syllable, e.g. [Bole] daxtaatə` -l marakw ‘he/one will take the woman’ (lit. kù.maa ‘hear’, [Hausa] nù.ka ‘ripen after storing’. take.fut.-he woman) [Lamang]. Pronominal indirect Verbal meanings are modified by the addition of one or objects usually attach to the verb like an affix, where- more suffixes. These extensions encode notions such as as nominal indirect objects occur as prepositional action ‘in the direction of’ (centripetal) or ‘away from’ phrases to the right of the direct object, cf. [Kanakuru] (centrifugal), or action ‘partially completed’ or ‘totally à jò-rò landài ‘he washed the robe for her’ (lit. he completed’, e.g. (centripetal) fit-oo ‘come out’ from washed-for her robe), and à jòè landài ə`n tamno ‘he fìta ‘go out’ [Hausa], (totality) sà-nyà ‘drink up’ from washed the robe for the woman’ (lit. he washed robe sà ‘drink’ [Margi]. Singular imperatives are often for woman). (Hausa is atypical in placing both overtly signaled with an -i or -u suffix, e.g. [Bachama] pronominal and nominal indirect objects immediately gur-ì! ‘sing!’, [Ga’anda] còk-u! ‘break (it)!’. after the verb.) In some cases, the verbal suffixes also have a syntac- Wh-questions, focus, and relativization often pat- tic function, signaling, inter alia, transitivization (allow- tern together with special (focus) marking. In Hausa, ing a verb to take an object), or perfectivity (indicating the wh-expression, focused element, or relativized head 446  HAUSA AND CHADIC LANGUAGES occur at the left periphery of the clause and focus mark- few languages signal plurality only by modifying the ing appears on the preverbal inflectional element, invariant noun with a demonstrative, quantifier or e.g. (wh-fronting) wàa kukà anii? ‘whom did you (pl.) numeral, cf. [Pero] landàa mijè ‘that shirt’ (lit. shirt see?’ (lit. who 2pl.focus.perf. see), (focus-fronting) that) and landàa mè ‘those shirts’ (lit. shirt those). yaarònkà mukà anii (‘it’s) your boy (that) we saw’ (lit. Reflexive pronouns (‘...self’) and reciprocals (‘each boy.your lpl.focus.perf. see), (relativization with other’) are typically formed with the body-part nouns fronting) yaaròn dà mukà anii ‘the boy whom ‘head’ and ‘body’ respectively, e.g. [Hausa] kâ-n-tà we saw’ (lit. boy.the rel. 1pl.focus.perf. see). In Miya ‘herself’ (lit. head-of-her), [Kwami] kuu-nì ‘himself’ questioned and focused (preverbal) subjects remain in (lit. head-his), [Hausa] juuna-n-mù ‘each other (we)’ situ but require special inflectional marking, e.g. wàa (lit. body-of-our), [Miya] tuwatùwà-tlə`n ‘each other də bə` sàwà? ‘who washed (it)?’ (lit. who foc. washed), (they)’ (lit. body-their). tə` də bə` sàw ‘he washed (it)’ (lit.he foc. washed). Negation in Chadic is typically signaled with a sin- References gle marker in sentence-final position, e.g. [Guruntum] Alio, Khalil. 1986. Essai de description de la langue bidiya du tâa kyur shau dà ‘she will not cook the food’ (lit. she Guéra (Tchad). Phonologie—Grammaire. Marburger will cook food neg.), [Kera] wə ùsnə` hàra bà ‘he Studien zur Afrika- und Asienkunde, Serie A, Afrika, 45. didn’t buy her a goat’ (lit. he bought her goat neg.), Berlin: Dietrich Reimer. sometimes accompanied by an additional reinforcing Caron, Bernard. 1991. Le haoussa de l’Ader. Sprache und Oralität in Afrika, Vol. 10. Berlin: Dietrich Reimer. preverbal negative marker, e.g. [Hausa] bà tà zoo Frajzyngier, Zygmunt. 1993. A grammar of Mupun. Sprache ba ‘she did not come’ (lit. neg. she.perf come neg.), und Oralität in Afrika, Vol. 14. Berlin: Dietrich Reimer. [Margi] nì ndà wì mài ‘I did not run’ (lit. I neg. Greenberg, Joseph H. 1963. The languages of Africa. (Indiana run. perf. neg.), [Miya] fà tà bə`ta zhaak-uw ‘don’t untie University Research Center in Anthropology, Folklore, and the donkey’ (lit. you (masc.) neg. untie donkey-neg.). Linguistics, publication 25; International Journal of American Linguistics, 29:1, part 2.) Bloomington: Indiana Cognate complement constructions are also wide- University. spread in Chadic, where the verb is repeated in a mor- Hoffmann, Carl. 1963. A grammar of the Margi language. phologically related (verbo-nominal) form, e.g. London: Oxford University Press. [Hausa] taa màaree shì maarìi mài ciiwòo ‘she Jaggar, Philip J. 2001. Hausa. London Oriental and African slapped him painfully’ (lit. she.perf. slap him slapping Language Library, Vol. 7. Amsterdam and Philadelphia: Benjamins. with pain), [Miya] mən rà-ya ràw ‘I surpass him’ (lit. Jaggar, Philip J., and H. Ekkehard Wolff. 2002. Chadic and I surpass-him surpassing). Comparatives are normally Hausa linguistics: the selected papers of Paul Newman with verbal sentences using the lexical verb ‘exceed, sur- commentaries. Cologne: Rüdiger Köppe. pass, be more than’, e.g. [Hausa] Audù yaa fi Muusaa Jungraithmayr, Herrmann, in collaboration with Njeno Andirya wàayoo ‘Audu is more clever than Musa’ (lit. Audu Galadima and Ulrich Kleinewillinghöfer. 1991. A dictionary of the Tangale language (Kaltungo, Northern Nigeria). he.perf. surpass Musa cleverness), [Miya] mə`n jiy baa Sprache und Oralität in Afrika, Vol. 12. Berlin: Dietrich də` rà-tlən aa wàshàshàm ‘it was I who was older than Reimer. them’ (lit. I it is one who perf. exceed-them for year), Newman, Paul. 1980. The classification of Chadic within [Margi] nàjà à mdia- à də dzə`àm ‘he is taller than Afroasiatic. Leiden: Universitaire Pers. me’ (lit. he surpassed-me with tall). Newman, Paul. 2000. The Hausa language: an encyclopedic reference grammar. New Haven and London: Yale In noun phrase syntax, the normative order for con- University Press. stituents is head-initial, i.e. head noun followed by Newman, Paul. 2003. Chadic languages. International definite determiners, possessives, numerals, relative Encyclopedia of Linguistics, 2nd ed., Vol. 1. 304–7. Oxford clauses, adjectives, etc. The linear order in genitive and New York: Oxford University Press. constructions is possessee X (+ ‘of’) + possessor Y, Pawlak, Nina. 1994. Syntactic markers in Chadic: a study on development of grammatical morphems [sic]. Warsaw: e.g. [Hausa] yaarò-n Muusaa ‘Musa’s boy’ (lit. boy-of Instytut Orientalistyczny, Uniwersytetu Warszawskiego. Musa), [Margi] tagu gə Haman ‘Haman’s horse’ (lit. Schuh, Russell G. 1998. A grammar of Miya. (University of horse of Haman). Many Chadic languages also make California Publications in Linguistics, 130.) Berkeley and an overt distinction between alienable and inalienable Los Angeles: University of California Press. possession, whereby inalienable possession (items that Stolbova, Olga V. 1996. Studies in Chadic comparative phonol- ogy. Moscow: Diaphragma Publishers. belong to someone in a deeper sense—they cannot be Wolff, H. Ekkehard. 1983. A grammar of the Lamang language taken away) is expressed by direct juxtaposition (i.e. (Gwa’d Lamang). Afrikanistische Forschungen, 10. with no overt linker), cf. (inalienable) mənda Miyim Glückstadt: J. J. Augustin. ‘Miyim’s wife’ (lit. wife Miyim), and alienable (items Wolff, H. Ekkehard. 1993. Referenzgrammatik des Hausa. that can be taken away) gam ma tamnoi ‘the woman’s (Hamburger Beiträge zur Afrikanistik, 2.) Münster and Hamburg: LIT. ram’ (lit. ram of woman) [Kanakuru]. PHILIP J. JAGGAR Some languages restrict plural marking to a narrow range of nouns (typically humans and animals), and a See also Afroasiatic; Greenberg, Joseph Harold 447 HAWAIIAN CREOLE ENGLISH Hawaiian Creole English Hawaiian Creole English (referred to by linguists as use of Hawaii Creole English began to be perceived as Hawaii Creole English) is one of a number of English- a liability on the job market because it indicated low lexifier pidgins and creoles. Varieties of pidgin and social status. It was a reminder of the plantation ori- creole English in Hawaii are the outcome of contact gins that many wanted to leave behind. The campaign between Hawaiians, Europeans (primarily English to eradicate Hawaii Creole English goes back at least speakers, who contributed most of the vocabulary to to a federal report issued in 1920 on education in the emerging pidgin), and the various immigrant Hawaii (then a US territory). The report recommended groups (e.g. Chinese, Japanese, Portuguese, and a number of reforms, among them the ‘elimination of Filipinos) brought to Hawaii to work as indentured the jargon of the plantations and ‘pidgin English’ of laborers on plantations from the 1850s onward. the streets’. Creole speakers were discriminated Hawaii Pidgin English, formed between 1890 and against through education in a school system that was 1910, is spoken only by a few remaining members of originally set up to keep out those who could not pass the oldest generation of immigrants to the plantations an English test. In this way, it was hoped to restrict the and is now dying out. Its descendant, Hawaii Creole admission of nonwhite children into the so-called English, a more complex language, emerged between English Standard schools set up in 1924 and attended 1910 and 1930. Through Hawaii Creole English, the mainly by white children. By institutionalizing what culture, identity, and solidarity of the local working was essentially racial discrimination along linguistic class were created. It became a badge separating locals lines, the schools managed to keep creole speakers ‘in from the tourists as well as from the resident middle their place’ and maintain a distance between them and class English-speaking haoles (Hawaiian ‘foreigner’), English speakers until after World War II. a term referring to whites. As a result of the increasing contact between creole Varieties of Hawaii Pidgin and Creole English are speakers and speakers of mainland varieties of English locally called simply ‘Pidgin’, even though most of after World War II and the political incorporation of the them are technically forms of creole English because islands into the United States as the 50th state in 1959, they function as the native rather than the second lan- the boundaries between standard English and the creole guage of most of their speakers. Hawaii Creole have become fuzzier than elsewhere in the Pacific, e.g. English is the first language of the majority of locally Papua New Guinea. Contact with English has born children and the first language of somewhat less decreased the autonomy of Hawaii Creole English and than half the state of Hawaii’s population of just over has created a continuum of varieties resulting from a million. The present ethnic composition of the changes introduced through decreolization, a process Hawaiian islands reflects the diversity of the different that has made the language less creole-like and more waves of foreign labor imported to work on sugar like English in some, but not all, respects. Although plantations. Hawaii is the only US state in which an adjacent varieties of the continuum are mutually intel- Asian/Pacific population is the majority: 61.9% of a ligible, the two extreme endpoints may often not be. total population of 1.1 million. Currently, those of ai go liv om most creole-like white ancestry account for 33.4%, Japanese for 22.3%, autsaid fo yu Filipino for 15.2%, Chinese for 6.2%, and others for ai gon liv om 10.2%. Native Hawaiians make up another 12.5% of autsaid fo yu the population. ai going liv om Like most of the world’s pidgins and creoles, autsaid fo yu Hawaii Creole English has no official recognition, and aim gonna liv om its coexistence with standard English has not been autsaid fo yu peaceful. Many speakers were and still are corrected at ail liv om/it least creole-like school for speaking ‘bad’ or ‘broken’ English. autsaid fo yu Standard English was seen as a tool of assimilation to I will leave it/them most standard a common American culture, even before pidgin outside for you English-like formed and spread. As the emerging middle class increasingly adopted and identified with Standard The most decreolized varieties are found on English, at least within the context of the school, the the island of Oahu, where three fourths of the state’s 448  HAWAIIAN CREOLE ENGLISH population is located, along with the capital, Honolulu, huhu, ‘angry, annoyed’, and ono, ‘tasty’, are actually and the main US military base, Pearl Harbor. The outer Hawaiian, rather than Japanese or even English in ori- islands of Kauai and Hawaii are the least decreolized. gin. Many of these Hawaiian words are being replaced Although many opponents of Hawaii Creole by English ones, although in some cases, the English English believed and hoped that the language would variants are still different from mainland US English, gradually be absorbed into English through continuing e.g. grinds, ‘food’ (cf. kaukau, ‘to eat/food’). The effect decreolization, this is clearly not happening. On the of Hawaiian on standard English outside Hawaii is oth- contrary, Hawaii Creole English is being maintained, erwise negligible, consisting mainly of items such as and even strengthened, particularly among certain aloha, lei, hula, ukulele, and a few others. groups of young people, as a symbol of local identity Distinctive grammatical features include the lack of in the face of inundation from mainland norms. In the the copula be (now you da head man, ‘Now you are face of public opposition to a proposed policy of the head man’), use of get in both possessive and exis- allowing teachers and students to speak and write only tential constructions (get one wahine she get one Standard English, in 1987, the State of Hawaii daughter, ‘There is a woman who has a daughter’), use Department of Education acknowledged that Hawaii of stay for locatives and progressives (Kimo stay Creole English constituted a language in its own right inside da house, ‘Kimo is inside the house’, Kimo stay with a structure distinct from English. However, this eating, ‘Kimo is eating’), use of wen or had as a sim- recognition has not resulted in any concrete action in ple past tense marker (Charlene wen/had play with da the classroom. kids, ‘Charlene played with the kids’), use of pau Like most pidgin and creole languages, Hawaii (Hawaiian ‘done, finished’) as a completive marker Creole English has not been standardized. This gives (Come back when you pau, ‘Come back when you are rise to the popular belief that the language cannot be finished’), preverbal negation (Harold no mo rice, written or that it is not a language at all. Linguists have ‘Harold doesn’t have any more rice’), and use of for as for some time been using a phonemic orthography (i.e. a complementizer (She like know how fo write pidgin, a writing system whose letters closely correspond to ‘She wants to know how to write pidgin’). The word the actual speech sounds) developed by Carol Odo order is generally subject–verb–object, as in Standard (1975). However, it has no wider recognition. Despite English, apart from topic/comment structures such as the lack of written norms and standardization, there are big, da fish, ‘The fish is big’, in which the comment nevertheless some writers who have attempted to use appears first. Formerly, there was variation in basic the creole as a medium for poetry, short stories, and word order among pidgin speakers of different ethnic drama by adapting English spelling to represent some backgrounds, reflecting the influence of immigrant of the distinctive characteristics of speech varieties in languages such as Japanese, e.g. sometime good road Hawaii. Each writer has worked out his or her own ad get, ‘Sometimes there are good roads’. hoc spelling system. This burst of literary creativity can There is considerable variation in pronunciation likewise be seen partly as a manifestation of opposition among local residents of different ethnic and social to colonialism and as an affirmation of distinctive local class backgrounds. Generally speaking, however, the identity in which the use of creole plays a key role. phonology of Hawaii Creole English has a smaller Although most of the vocabulary of Hawaii Creole inventory of distinctive sounds than many mainland English is derived from English, a significant number varieties of English. The [r] after a vowel in words of words come from Hawaiian. As many as a thousand such as shark is usually absent in the most creole-like Hawaiian words may have been in use at one time dur- speech varieties (i.e. [shak]), and many of the diph- ing the plantation era, of which 250 to 350 were in fair- thongs (double vowels) found in mainland varieties of ly common use colloquially among the population. English in words such as boat [bout] and play [pleI] Although this number is now fewer, many still persist are single vowels in the creole (i.e. [bot], [ple]). in local English and many more in the Hawaii Creole Hawaii Creole English often has full vowels where English of older speakers. Some words, such as hemo, mainland varieties use reduced ones, e.g. creole [tude], ‘take off, remove’, and ho’omalimali, ‘flatter, curry ‘today’, vs. mainland [tədei], and creole [mauntən] favor’, are not as widely known among younger people ‘mountain’ vs. mainland [maunten]. English interden- now, whereas virtually everyone resident locally would tal fricatives (‘th’) in words such as they and think tend know haole, ‘foreigner’. Because of the general decline to become stops in the creole (i.e. [de], [tink]). The in the knowledge of Hawaiian, many people who use a stops in consonant clusters such as [tr] in words such great many of these Hawaiian words every day, either as try are affricated (i.e. [Jrai]), and initial [s] in clus- in Hawaii Creole English or in other local forms of ters such as [str] in words such as street sounds more English, do not realize that commonly used words such like ‘sh’ [š]. There are also some stress and intona- as akamai, ‘clever, smart’, pio, ‘to turn off, extinguish’, tional differences, such as the use of falling pitch for 449 HAWAIIAN CREOLE ENGLISH yes/no questions, which also do not show the sub- Romaine, Suzanne. 1994. Hawaii Creole English as a literary ject/auxiliary inversion typical of standard English, language. Language in Society 23. 527–54. Romaine, Suzanne. 1999. Changing attitudes towards Hawaii e.g. you like go Honolulu? ‘Do you want to go to Creole English: Fo’ get one good job, you gotta know ho fo’ Honolulu?’ The falling intonation pattern has been car- talk like one haole. Creole genesis, attitudes and discourse, ried over from Hawaiian into creole. Rising pitch ed. by John Rickford and Suzanne Romaine, 287–301. together with a final question particle are used as a Amsterdam: John Benjamins. confirmation check in utterances such as no mo job fo Sato, Charlene J. 1985. Linguistic inequality in Hawaii: the post creole dilemma. Language of inequality, ed. by Nessa you, aeh? ‘There isn’t a job for you, right?’ Wolfson and Joan Manes. Berlin: Mouton. Sato, Charlene J. 1991. Language attitudes and sociolinguistic variation in Hawaii. English around the World, ed. by Jenny References Cheshire. Cambridge: Cambridge University Press. Bickerton, Derek. 1977. Change and variation in Hawaiian Watson-Gegeo, Karen A. 1994. Language and education in English, Vol. I, Creole Syntax. Honolulu: Social Sciences Hawaii: sociopolitical and economic implications of Hawaii and Linguistics Institute, University of Hawaii. Creole English. Language and the social construction of Odo, Carol. 1975. Phonological processes in the English dialect identity in creole language situations, ed. by Marcyliena of Hawaii, Ph.D. Dissertation. University of Hawaii at Morgan. Los Angeles: UCLA Center for African American Manoa. Studies. Reinecke, John E. 1969. Language and dialect in Hawaii. A SUZANNE ROMAINE sociolinguistic history to 1935. Honolulu: University of Hawaii Press. See also Pidgins and Creoles Hebrew: Biblical Biblical Hebrew (BH) is a Semitic language related to Phoenician, Ammonite, Edomite, and Moabite. One Arabic, Ethiopic (such as Amharic, the national lan- other language, known as DeirAlla¯ , is controversially guage of Ethiopia), Akkadian (Assyro-Babylonian), considered by some to be Canaanite, but this author etc. It is called ‘the language of Canaan’ in the Old does not agree with that classification (see Kaye Testament (Isaiah 19:18). It is also called yəhu¯d¯ t, ‘the 1991). The Canaanite dialects are sometimes referred language of Judah’ or ‘Judean’, after the dominant to as languages; however, there existed, in all proba- tribe in the southern part of the Holy Land (II Kings bility, a high degree of mutual intelligibility among 18:26, 28). Hebrew is first attested more than 3,000 them, along the lines of, say, Egyptian Colloquial, years ago, and the oldest BH texts date back to the Syrian, Palestinian, and Jordanian Arabic. twelfth century BC. In all likelihood, Hebrew became extinct as a spoken tongue in the third century AD; The Stages of Biblical Hebrew however, in most places it was being supplanted by its kindred language, Aramaic, in the second and first Based on linguistic factors that occur in the Hebrew centuries BC. It has been used continuously by Jews Bible, BH is divided into three different stages: Old around the world as the language of their liturgy and BH (1100–1000 BC), Standard BH (1000–550 BC), also for literary purposes. and Late BH (550–200 BC). The standard language may be exemplified by the Pentateuch. Old BH is best represented by the archaic poetic dialect found in Classification Judges 5 and Exodus 15 (the Song of Moses). The for- More specifically, BH belongs to the Canaanite branch mer, known as the Song of Deborah, appears to be of the Northwest Semitic (NWS) group. Many schol- Israelian, since it uses a masculine plural in -¯ n for ars consider Ugaritic to be a member of the Canaanite Judahite -¯ m, and the relative particle is ša- for a˘šer. family as well; however, the present author considers Thus, BH definitely had regional dialects: the Judahite this cuneiform language, discovered in northern Syria dialect (i.e. Jerusalem) and the (northern) Israelian in 1929, to be a distinct branch of Central Semitic in dialect (e.g. Samaria and the Galilee). About four its own right (see Kaye 1991). Most authorities would fifths of the Hebrew Old Testament is written in the agree that the Canaanite dialects consist of BH, Judahite dialect. Late BH shows influence from 450  HEBREW: BIBLICAL Aramaic, the lingua franca of that era, and is repre- pronunciation of the Biblical text. Among the most sented by the Books of Esther, Ezra, etc. significant of all Masoretic inventions was the dot, called daghesh, which was used inside some conso- nants to indicate a geminated or a stop version of that Diglossia consonant (as opposed to a fricative, since like BH was a diglossic language. Diglossia is a sociolin- Aramaic, BH developed fricative allophones for /bgd- guistic situation where a high and a low variety of a kpt/, viz. [βγðxfθ]). Consonantal gemination was language coexist (see Rendsburg 1990, Kaye 1993). phonemic in BH as it is in Arabic; however, a charac- Each variety is used under a different set of circum- teristic of the Masoretic text was that five consonants stances, viz. formal or official discourse vs. informal could not be geminated (probably for phonetic rea- (the closest parallels are the diglossic cases of Arabic sons, all being back consonants). They are: /hh¯r/. and Modern Greek). Examples of colloquial usage Since they do not geminate, vocalic change was com- have, for a variety of reasons, entered the Biblical text, monplace with the preceding vowel via a process of since the Old Testament should not be regarded as a compensatory lengthening. compilation of speech from Biblical times, but rather as an edited collection of essays and literary versions Basic Characteristics of Biblical Hebrew of direct speech and oral transmissions. BH has two grammatical genders, masculine ( ¯i š ‘man’) and feminine (iššɔ ‘woman’). It also has a dual Orthography for nouns that occur in pairs, such as for many body BH uses the Phoenician alphabet of 22 letters, includ- parts (e.g. / e¯ynayim/ ‘eyes’). It has a triconsonantal ing some polyphonous graphemes, and is written from root system like the other Semitic languages (see fur- right to left. All traditions of Ancient Hebrew agree ther below); e.g. the root qdš has to do with ‘holiness’ that there are 23 consonant phonemes, including two or ‘sanctification’. Furthermore, it is well known for semiconsonants that developed from a system of 29 its seven major binyanim or verbal forms or stems Proto-Semitic consonants. Originally, only the conso- (Arabic has 10!). Illustrating with the aforementioned nants were represented, and the corpus had achieved root, they are as follows (in the perfect, third-person its current stable state by the second century AD. As masculine singular, the usual or unmarked citation the orthography developed, a system of matres lectio- form for the BH verb, following Steiner 1992): nis ‘mothers of reading’, or so-called vowel letters, I qal (‘simple’ or basic) qɔ daš ‘be holy’ emerged, viz., <h>, <w>, and <y> were used to indi- II nifal (passive of I) niqdaš ‘reveal oneself as cate their respective phonetically similar correspon- holy’ ding vowels in final position. This system, in existence III piel (intensive or causative) qiddeš ‘sanctify; since the tenth century BC, expanded over time to purify’ include medial vowels as well (attested since the ninth IV pual (passive of III) quddaš century BC), while initial vowels did not need to be V hif  ¯i l (causative) hiqdi¯š ‘consecrate; devote’ included, since no BH word could begin with a vowel ¯ /, an allomorph of /wə-/ VI huf al (passive of V) huqdaš ‘be consecrated; (with the exception of /u devoted’ ‘and’). The orthographic representations without VII hitpael (reflexive-reciprocal) hitqaddeš matres lectionis are older spellings than those with ‘sanctify or purify oneself’ them; <kl> ‘each; every’ is spelled only once as <kwl> (cf. Arabic <kl>, which is the cognate). There are also minor stems, such as another The Hebrew Bible’s oral tradition was preserved causative šaf el, which derives from a Proto-Semitic from the sixth to the tenth century AD by a group of and Proto-Afroasiatic s-causative. The root ktb ‘write’ Jewish scholars known as the Masoretes (< Hebrew may serve as illustrious of nominal and verbal mor- ma¯so¯ra¯h ‘tradition’). These scholars established a rec- phology (it occurs 203 times in the Old Testament). ognized text of the Hebrew Bible called the Masoretic Among the verbal forms to be noted are kɔtab ‘he Text. Three different systems developed: Babylonian, wrote’, kɔtu¯b ‘written’, niktab ‘it was written’, Palestinian, and Tiberian. The authoritative one since məkatbi¯m ‘write constantly (masc. pl.)’, and the nom- about AD 1000 has been the Tiberian, the best known inal patterns kətɔb ‘register’, kətobε t ‘tattooing’, mik- of which is the work of Aaron Ben-Asher (AD 915). A tab ‘writing’, and, possibly, miktɔm ‘psalm of system of diacritics, mostly for vowels, developed, in expiation’ via b > m. addition to accent markings, cantillation marks for Further, BH has a perfect and an imperfect, which liturgical recitation, and marginal notes. Its purpose eventually came to denote the past and the present- was to insure the accuracy of recording the traditional future (see below). 451 HEBREW: BIBLICAL Phonology Syntax Among the most interesting of the BH phonological The adjective follows the noun and agrees with it in developments from Proto-Semitic are the changes of number, gender, and definiteness. Thus, iššɔ ku¯ši¯ t ‘an the interdentals to corresponding fricatives, creating Ethiopian woman’ and hɔ iššɔ hakku¯ši¯ t is ‘the mergers; e.g. */ð/ and */z/ > /z/, and */š/ and */θ/ > /š/. Ethiopian woman’. Also noteworthy are the following: */s
/, */´s
/, and */ θ
/ The basic word order is VSO, but for emphasis > /
s/, */γ/ and */c/ > //, and */x/ and */h¯/ > /h¯/. The other word orders are possible. There is an accusative best-known BH phonological fact, in all likelihood, particle et that marks definite direct objects. Thus, has to do with a dialectological difference reported in Genesis 1:1 reads: bəre¯ši¯t bɔrɔ e˘/o¯hi¯m et the Book of Judges, where the Ephraimites were haššɔmayim wəet hɔɔrεs, lit., ‘in the beginning-cre- unable to pronounce shibboleth ‘flowing stream’ or ated-God-Acc.-the heavens-and-ACC.-the earth’. ‘ear of corn’ correctly as in the Gileadite dialect, say- ing sibboleth instead (Judges 12:6), and this erroneous pronunciation gave away their identities. It should be Loanwords pointed out that the letters ši¯ n and si¯ n are written with BH has borrowed vocabulary from a variety of lan- the same grapheme and thus were polyphonous, since guages, including Egyptian, Sumerian, Akkadian, /s´ / merged with /š/ over time. One also notes several Aramaic, Greek, and Persian. Two well-known exam- dialectal variants with sibilants, e.g. ´sh¯q and sh¯q ples include barzel ‘iron’ < Hittite barzillu, and par o¯ ‘laugh’, and zq and s
q ‘shout’. ‘Pharaoh’ < Ancient Egyptian pr-3 ‘the great house’. Examples of BH words that have come down into Emphatics English include many proper names (Saul, David, etc.) and terms relating to religion, such as amen and sab- The exact articulation of emphatic phonemes /s /, /t/, bath, not to mention the ubiquitous schwa, the mid- and /q/ are not really known. The evidence from the central vowel [ə] known to all linguists. cognate languages is mixed. If one compares with Arabic, the corresponding sounds are pharyngealized and velarized, except for the voiceless uvular plosive. References On the other hand, if one adverts to Ethiopic and South Blau, Joshua. 1976. A grammar of Biblical Hebrew. Wiesbaden: Arabian, the sounds are glottalized (ejective). Most Otto Harrassowitz Verlag. Semitists believe that the evidence favors the glottal- Gesenius, Wilhelm, and E. Kautsch. 1910. Gesenius’ Hebrew ization of the Proto-Semitic emphatics. grammar, 2nd English edition. Oxford: Clarendon Press. Kaye, Alan S. 1991. Does Ugaritic go with Arabic in Semitic genealogical sub-classification? Folia Orientalia. Vol. 28, Morphology pp. 115–28. Kaye, Alan S. 1993. Review of Rendsburg 1990. AJS Review. We have already mentioned the triconsonantal root Kaye, Alan S. (ed.) 1997. Phonologies of Asia and Africa. (see above). However, it is important to keep in mind Winona Lake, Indiana: Eisenbrauns. Khan, Geoffrey. 1997. Tiberian Hebrew phonology. In Kaye. that early Semitic also had many biconsonantal roots. Koehler, L., W. Baumgartner, and J. J. Stamm. 1967–1990. Unquestionably, many original biconsonantal roots Hebräisches und Aramäisches Lexicon zum Alten became triconsonantal via Systemzwang. Basically, the Testament, 3rd edition. Leiden: E. J. Brill. idea is that a root determinative was added to make an Kutscher, E. Y. 1982. A history of the Hebrew language, ed. by original biconsonantal triconsonantal. The details of Raphael Kutscher. Jerusalem: Magnes Press and Leiden: E. J. Brill. this process are not fully understood, and this area is Malone, Jopseph L. 1991. Tiberian Hebrew phonology. Winona rife with controversy. Lake, Indiana: Eisenbrauns. Perhaps the most controversial area in BH mor- Rendsburg, Gary A. 1990. Diglossia in ancient Hebrew. New phology has to do with the nature of the verbal system Haven, CT: American Oriental Series No. 72. (see above). Many conclude that the imperfect origi- Rendsburg, Gary A. 1997. Ancient Hebrew phonology. In Kaye. nally expressed actions and was used to denote the Sáenz-Badillos, Angel. 1993. A history of the Hebrew lan- durative aspect. The perfect originally expressed the guage. Cambridge: Cambridge University Press. stative. The preterite nature of the imperfect is pre- Steiner, Richard C. 1992. Ancient Hebrew. In International served in the BH ‘consecutive’ or ‘conversive’ tenses. encyclopedia of linguistics, ed. by William Bright. Many of the 5,280 occurrences of the root mr ‘say’ Ullendorff, Edward. 1977. Is Biblical Hebrew a language? Wiesbaden: Otto Harrassowitz Verlag. are with the wa¯w-consecutive: wayyo¯mε /ar ‘and he ALAN S. KAYE said’. As is the case in BH, the wa¯w-consecutive form also occurs in Ugaritic prose, not poetry. See also Hebrew: Modern ; Semitic Languages 452  HEBREW: MODERN Hebrew: Modern Modern Hebrew (MH) is the first language for the nonexistent verb *nigdal. It appears that there is no native Jewish population in the State of Israel and other language family that is claimed to have roots the second language for the new immigrants and the consisting of consonants only. For this and other rea- native Arab population. In its current form, MH is dated sons, the validity of the notion of the root has been back only as far as the late nineteenth century; prior to challenged (Bat-El 2002). The binyan may contribute that period, Hebrew was used for liturgical and ritual a syntactic property to the verb, but only in paradig- purposes for about 1,700 years. The genetic affiliation matic relations (Berman 1978). For example, B2 is the of MH to the Semitic family, although not uncontrover- passive counterpart of B1 (gamar ‘finish B1’—nigmar sial (Wexler 1990), is strongly manifested in its mor- ‘be finished B2’), B4 is the transitive counterpart of phology and lexical stock, which were drawn from B1 (lamad ‘learn B1’—limed ‘teach B4’), B3 is the Biblical Hebrew, the language of the Old Testament. causative counterpart of B1 (xatam ‘sign B1’—hextim The phonology of the language, on the other hand, as ‘have someone sign B3’), and B5 is the reciprocal, well as word order, do not reflect the characteristics of reflexive, or passive counterpart of B4 or B1 (niʃek Biblical Hebrew. ‘kissed B4’—hitnaʃek ‘kissed with someone B5’, raxac ‘wash B1’—hitraxec ‘wash oneself B5’, kibel ‘receive B4’—hitkabel ‘be received B5’). There are Verb Morphology two passive forms, often considered to be binyanim as MH morphology is characterized by the Semitic type well, derived from B3 and B4 by substituting the nonconcatenative structure, especially in the verbal vocalic pattern with {u, a} (B3 higdil ‘enlarge’—hug- system. A verb must belong to one of the five mor- dal ‘be enlarged’, B4 gidel ‘raise’—gudal ‘be raised’). phological classes called binyanim ‘constructions’ These forms, although not commonly used, have no (singular binyan). Each binyan has a specific prosodic exceptions and are always contingent, structurally and structure (in terms of syllables) and vocalic pattern, semantically, upon their active counterpart. and some also have a prefix; for some binyanim, these Verbs are also accompanied by affixes indicating properties must be specified for each tense. The verbs tense, person, number, and gender. (see Table 2) below are given in their third-person singular form; The values of the person–number–gender proper- stress is final unless otherwise specified (by an acute ties of the verb are drawn from subject noun or pro- sign). (see Table 1) noun with which the verb must agree (ha-yalda gadl-a While the verbs in table 1 differ in their binyan (i.e. ‘the girl grew’, ha-yeladim gadl-u ‘the children grew’, prosodic structure, vocalic pattern, and prefixes, if ʔanaxu gadal-nu ‘we grew’). The gender properties of any), they all share a discontinuous string of conso- the noun are natural in case of animate nouns (sus nants, traditionally called the ‘consonantal root’. Verbs ‘horse ms.’ - susa ‘horse fm.’), but serve in a purely sharing a consonantal root often share a semantic grammatical role in inanimate nouns (séfer ‘book ms.’, property, and it is thus traditionally claimed (for all délet ‘door fm.’). Subject–verb agreement allows Semitic languages displaying this type of morpholo- phrases without a subject, but only when the subject is gy) that the consonantal root carries the core meaning not the third person (karánu séfer ‘(we) read a book’). of the verb. Notice that not every root has a verb in There are also phrases without a verb, but only in the all the binyanim, as can be seen from the different present tense (ha-séfer al ha-madaf ‘the book (is) on B2 verb in the table in table 1, which replaces the the shelf’, ha-yalda xaxama ‘the girl (is) smart’). TABLE 1 Binyan/Tense Past Participle Future B1 paʔal gadal gadel yigdal ‘grow’ B2 nifʔal (nigmar) (nigmar) (yigamer) (‘be finished’) B3 hifʔil higdil magdil yagdil ‘enlarge’ B4 piʔel gidel megadel yegadel ‘raise’ B5 hitpʔael hitgadel mitgadel yitgadel ‘agrandize’ 453 HEBREW: MODERN TABLE 2 Gender/Person Past Future Participle Singular Plural Singular Plural Singular Plural M/F - 1 gidál-ti gidál-nu ʔa-gadel ne-gadel M-2 gidál-ta gidál-tem te-gadel te-gadl-u M me-gadel me-gadl-im M-3 gidel gidl-u ye-gadel ye-gadl-u F-2 gidál-t gidál-tem/n te-gadl-i te-gadl-u F-3 gidl-a gidl-u te-gadel ye-gadl-u F me-gadél-et me-gadl-ot Noun Morphology house’, ha-ʃomer-im ʃomer-im al ha-báyit ‘the guards guard (ms.pl.) the house’). Nouns, like verbs, exhibit prosodic and vocalic restric- tions called mishkal ‘weight’ (plural mishkalim). Beyond the Word Related nouns (and adjectives) differ in their prosodic structure and vocalic pattern, as well as affixation A noun can also be formed from two nouns (and an (gadal ‘to grow’—gódel ‘size’, mi-gdal ‘tower’, gadol adjective from an adjective plus a noun), where the first ‘big’; carax ‘to consume’—córex ‘need’, ti-cróx-et element determines the general meaning, thus consid- ‘consumption’, carx-an ‘consumer’). Not only are there ered the head, while the second adds specific informa- far more mishkalim than binyanim, a noun, unlike tion. For example, beit+séfer ‘school (house+book)’ is a verb, does not have to fit into a mishkal. Thus, a type of a house, naʔarat+rexov ‘street girl while loan-verbs (or verbs derived from loan nouns) (girl+street)’ is a type of a girl, and ʔiver+cvaim ‘color must belong to a binyan (tilfen ‘to phone’, kided ‘to blind (blind+colors)’ refers to a type of blindness. codify’), nouns can maintain the structure of the source There are two types of noun+noun structure, one called word, with a few phonological modifications (diéta a compound and the other a construct state, where only ‘diet’, télefon ‘phone’). The mishkal system is not the latter is accessible to syntactic modification (Borer rigid because there is nothing in the morphology of 1989). In the construct state, the non-head (second) ele- nouns that is contingent upon the mishkal. In verbs, in ment can be modified (gan+ʃoʃanim ʔadumot ‘a gar- order to be able to derive one tense from the other, it is den of red roses (garden+roses red)’), or conjoined crucial to know the binyan, which is identified on the with another noun (gan+ʃoʃanim ve-cipornim ‘garden basis of the prosodic structure and vocalic pattern (and of roses and carnations (garden+roses and carna- prefixes, if any). Such information is not relevant for the tions)’). However, in a compound such as gan+yeladim morphology of nouns, which involves mainly suffixa- ‘kindergarten (garden-children)’ the non-head yeladim tion: plural suffixes -im for masculine (xatul-im ‘cats is not accessible to the syntax. Despite the difference in ms.’) and -ot for feminine (xatul-ot ‘cats fm.’), various accessibility to syntax, both the compound and the con- feminine suffixes (xatul-a ‘cat fm.’, rakdan-it ‘dancer struct state display the same morphology and phonolo- fm.’, sofér-et ‘writer fm.’), a diminutive suffix (sus-on gy. In both, the plural suffix, which has the special form ‘little horse’), an agent suffix (taklit ‘record’—taklit-an -ei for masculine, is attached to the first (head) element ‘DJ’), various nominal suffixes (manhig ‘leader’—man- (gan-ei +yeladim, gan-ei+ʃoʃanim), the feminine suf- hig-ut ‘leadership’, mila ‘word’—mil-on ‘dictionary’, fix -a takes the form -at (naʔara ‘girl’ - naʔarat+rexov magav ‘wiper’—magév-et ‘towel’), and an adjectival ‘street girl’), and, in both, the second element bears the suffix (tarbut ‘culture’—tarbut-i ‘cultural’). The suffix- main stress. Occasionally, speakers may disagree as to es -a, -it, -et, and -ut assign a feminine gender and all whether the noun+noun form is accessible to the syn- the others assign masculine gender to the base. The plu- tax, i.e. whether it is a compound or a construct state. ral suffixes do not assign gender but rather agree with Word order in MH is usually subject–verb–object the gender of the base (with several exceptions, such as (SVO), but permutation of this order is relatively free. ʃoʃana ‘rose fm’, which takes the masculine plural suf- Thus, the phrase dáni raʔa ʃoʃana ʔaduma ‘Dani saw fix -im, ʃoʃan-im). These suffixes are not sensitive to a red rose’ can appear in some contexts as ʃoʃana the structure of the noun to which they are attached and ʔaduma dáni raʔa or ʃoʃana ʔaduma raʔa dáni. one can even find borrowed nouns with a native suffix However, adjectives always follow the noun they mod- (geográf-i ‘geographic’, dialóg-im ‘dialogs’, simétriy-ut ify (ʃoʃana ʔaduma ‘red rose (rose red)’), and numer- ‘symmetry’). The participle form of the verb also func- als, with the exception of the numeral ‘one’, always tions as a noun, and its morphology is that of the noun, precede it (ʔéser ʃoʃanim ‘ten roses’, ʃoʃana ʔaxat i.e. it has only number and gender specification (ha- ‘one rose (rose one)’). Definite nouns are preceded by ʃomer ʃomer al ha-báyit ‘the guard guards (ms.sg.) the the definite article ha- ‘the’, which also appears in the 454  HINDI-URDU modifying adjective (ha-ʃoʃana ha-ʔaduma); there is monosyllabic verbs (rác-u ‘they ran’). In nouns, it is the no indefinite article. All other function words also vowel in the penultimate stem syllable that undergoes appear as proclitics: prepositions appear before the deletion, but only in a lexically marked group of nouns phrase (ʔal-ha-séfer ‘on the book’, me-ha-báyit ‘from and adjectives (gamal - gmalim ‘camel sg.-pl.’ but the house’), the conjunctive marker ve- ‘and’ appears gamad - gamadim ‘dwarf sg.-pl.’). When the first con- before the last element in the list (séfer, niyar ve-ʔet ‘a sonant in the penultimate syllable is a sonorant (m, n, l, book, a paper, and a pen’), and the subordination r, y), the vowel subject to deletion is replaced by e marker ʃe- ‘that’ appears before the subordinate clause (lavan - levanim ‘white sg.-pl.’). (ha-baxur ʃe-xika la ‘the guy that waited for her’). There is a stop-fricative alternation in MH (p - f, b Question words (mi ‘who’, ma ‘what’, matay ‘when’, - v, k -x), reminiscent of the postvocalic spirantization ʔeyfo ‘where’) appear at the beginning of the phrase of Tiberain Hebrew (katav ‘he wrote’ - yixtov ‘he will (mi zarak kadur ‘who threw a ball?’) and can be pre- write’). However, this alternation is not entirely regu- ceded by a preposition (ʔal-mi zarákta kadur ‘at lar as there are postvocalic stops (bikeʃ ‘he requested’) whom did you throw a ball?’). as well as fricatives in non-post-vocalic positions (xipes ‘he searched’). At the current stage of the lan- guage, there is a great degree of free variation in the Phonology verb system (yixtov ~ yiktov), which may suggest a The sound system of MH consists of 20 consonants— change toward a uniform verbal paradigm with respect p, b, m, f, v, t, d, s, z, c, n, l, ʃ, y, k, g, x, r, h, and ʔ (and to spirantization. in some dialects also ʕ and ), and five vowels; i, u, e, o, a. The sound h, is often substituted by ʔ, and both References are deleted in casual speech. Bat-El, Outi. 2002. Semitic verb structure within a universal Stress in nouns can be ultimate or penultimate (and perspective. Language processing and acquisition in lan- rarely antepenultimate). When the plural suffix is guages of Semitic, root-based, morphology, ed. by Joseph added, stress shifts to the end in most nouns, whether Shimron. Amsterdam: John Benjamins. the singular form has ultimate stress (gamád - gamadím Berman, Ruth A. 1978. Modern Hebrew structure. Tel-Aviv: ‘dwarf sg.-pl.’), or penultimate (yéled - yeladím ‘boy University Publishing Projects, Ltd. Borer, Hagit. 1989. On the morphological parallelism between sg.-pl.’); in some nouns, the stress stays in the same compounds and constructs. Morphology yearbook 1, ed. by position on the stem (salát - salátim ‘salad’, tíras - Geert Booij and Jaap van Marle. Dordrecht: Foris. tírasim ‘corn sg.-pl.’). In verbs, stress is always final on Glinert, Lewis. 1989. The grammar of modern Hebrew. the stem (gadál ‘he grew’, gadál-nu ‘we grew’), but Cambridge: Cambridge University Press. shifts to the end when a vowel initial suffix is added Ravid, Dorit D. 1995. Language change in child and adult Hebrew: a psycholinguistic perspective. New York: Oxford (gadl-á ‘she grew’). When stress shifts, the vowel in University Press. the stem final syllable is deleted when the preceding Rosen, Haiim B. 1977. Contemporary Hebrew. The Hague: syllable is open (gadal-a —-> gadlá ‘she grew’), Mouton. and changes to e when the preceding syllable is closed Wexler, Paul. 1990. The schizoid nature of modern Hebrew: a (yigdal-u —-> yigdelú ‘they will grow’) or when slavic language in search of a semitic past. Wiesbaden: Otto Harrassowitz. flanked by identical consonants (garar-a —-> garerá OUTI BAT-EL ‘she dragged’). Stress shift and vowel deletion/change do not affect B3 verbs (higdíl-a ‘she enlarged’) and See also Hebrew: Biblical ; Semitic Languages Hindi-Urdu Hindi is a modern Indo-Aryan language spoken in people speak Hindi either as a first or second lan- South Asian countries (India, Pakistan, and Nepal) and guage. It is ranked among the two most widely spoken also in countries outside Asia (Mauritius, Trinidad, languages of the world. Along with English, it is the Fiji, Suriname, Guyana, South Africa, and other coun- official language of India. In addition, it is the state tries). Approximately six to seven hundred million language of Bihar, Haryana, Chattisgarh, Jharkhand, 455 HINDI-URDU Himachal Pradesh, Arunachal Pradesh, Madhya Persio-Arabic sources, whereas Hindi borrows from Pradesh, Uttar Pradesh, and Rajasthan. indigenous Sanskrit sources. This tendency has resulted Hindi, which is a descendant of the Sanskrit lan- in the admission of new sounds (e.g. z, x, G) into Urdu guage, is not strictly the name of any chief dialect of and in the development of its distinct literary style. the area but is an adjective, Persian in origin, meaning The literary history of Hindi goes back to the Indian. Historically, it was synonymous with Hindui, twelfth century. Some notable literary figures of Hindi Hindawi, Rexta, and Rexti. The term Urdu is also used are Surdas, Tulsidas, and Mira Bai. Urdu literature to refer to this language. All these labels denote a flourished both in India and Pakistan. The literary his- mixed speech spoken around areas of Delhi, North tory of Urdu can be traced back to approximately the India, which gained currency during the twelfth and thirteenth century. Its first poet was Amir Khusro thirteen centuries as a contact language between (1253–1325), who termed the language zabaan-e- Arabs, Afghans, Persian, Turks, and native residents. Dehlvi or Hindi. Amir Khusro included Hindi verses Hindi and Urdu have a common form known as among his Persian poems and romances. But the great- Hindustani. This was the variety that was adopted by est poet of Hindi-Urdu was Kabir, who was famous for Mahatma Gandhi and the Indian National Congress as his devotional poetry. a symbol of national identity during the struggle for Urdu is particularly well known for its romantic lit- freedom from Britain. erature. The two most famous genres of Urdu are Urdu is spoken by approximately 50 million speak- Masnawii and Gazal, which are the gift of the mixing ers in the subcontinent of India and Pakistan. It is the of the Hindu and Persian-Muslim cultures. Some national language of Pakistan and one of the 19 offi- notable literary figures of Hindi-Urdu literature are cial languages of India. The term ‘Urdu’ is Turkish in Inshah Alla Khan, Malik Muhammad Jayasi, Mir, origin ( i.e. ordu ‘camp’). As mentioned above, both Daya Shankar Nasim, Bahadur Shah Zafar, Iqbal, and Hindi and Urdu have their origins in the mixed speech Mirza Ghalib. spoken around Delhi. In time, this mixed speech called Hindi-Urdu has an approximately three-century- khari boli developed its own variety called ‘Hindi’ or old, well-attested, and rich grammatical tradition of its ‘Urdu’. own. It is a by-product of the colonial era and was born The first question that is immediately raised about shortly after the arrival of Europeans in India. Urdu is: How does it differ from Hindi? This is not an easy question to answer. In many respects, the two Phonology speech varieties are mutually intelligible and at anoth- er level they are quite unintelligible. Moslems often The two notable phonological features of the lan- report their speech variety as Urdu, and Hindus and guage are as follows: (1) Hindi still retains the origi- other non-Moslem speakers tend to report their speech nal Indo-European (1500 BC) distinction between variety as Hindi. Yet, one notable difference in Hindi aspirated and unaspirated consonants, which results and Urdu is script. in a four-way contrast as shown in the following Hindi is written in the Devanagari script, which is examples: kaal ‘time’, khaal ‘skin’, gaal ‘cheek’, and ranked as the most scientific writing system among the ghaal ‘to put into’. (2) It has the feature of retroflex- existing writing systems of the world. The Devanagari ion in its consonant inventory, cf. Taal ‘to put off’ and script is written from left to right and is a descendant taal ‘pond’. The retroflex consonant is transcribed as of the Brahmi script, which was well established in the capital T. India before 500 BC. The script is phonetic in nature The inventory of distinctive segments of standard and there is a fairly regular correspondence between Hindi-Urdu is shown in Table 1. Sounds that occur in the letters and their pronunciation. Persio-Arabic words are provided in parentheses ( ). Urdu is written in the Persio-Arabic script. Like Arabic, it is written from right to left and short vowels Stress are usually not written. It is due to Urdu that the Although stress (meaning loudness) is not a prominent Arabic script underwent great modification and some feature of Hindi-Urdu, nevertheless, it seems that its new symbols were added to it to represent sounds, existence cannot be denied. Stress can distinguish such as retroflex consonants (produced with the tip of between grammatical categories such as nouns and the tongue curled upwards), which are specific to the verbs, as in Indian subcontinent. In addition to Hindi and Urdu being written in two Nouns Verbs different scripts, the two languages also differ from each other in minor ways in their grammatical systems and galaa neck galaa cause to melt vocabulary. Urdu tends to borrow its vocabulary from talaa sole talaa cause to fry 456  HINDI-URDU TABLE 1 Consonants Labial Dental Retroflex Palatal Velar BackVelar Stop Unvoiced unaspirate p t T c k (q) Unvoiced aspirate ph th Th ch kh Voiced unaspirate b d D j g (G) Voiced aspirate bh dh Dh jh gh Nasal m n N ñ ŋ Fricative Unvoiced (f) s sh (x) Voiced (z) Flap Voiced unaspirate r R Voiced aspirate Rh Lateral l Semivowels w (v) y Vowels Front Central Back High ii uu i u Mid e o ai a [schwa] au Low aa The stressed syllable is shown in bold. However, verbs). Simple, participial, and vaalaa adjectives are of stress is not usually distinctive in Hindi-Urdu. two types: inflected and uninflected. Inflected adjec- Therefore, in general, whether one places stress on the tives agree with their following noun in number, gen- first syllable or on the second, the meaning will not be der, and case; they end in morpheme -aa (e.g. acchaa affected. For example, the meaning of word, sunaa ‘good’), which changes to e for masculine plural and heard will remain unchanged whether one places stress masculine oblique (acche), ii for feminine nouns on the first syllable or the second. Therefore, Hindi- (acchii). Uniflected adjectives remain unchanged. Urdu is often characterized as a ‘syllable-timed’ lan- Although the case system of pronouns is essential- guage like French, where the syllables are pronounced ly the same as that of nouns, pronouns have more case in a steady flow, resulting in a ‘machine-gun’ effect. forms than nouns. Case relations are essentially car- ried out by means of postpositions. Personal pronouns are similar to their English equivalents except that Morphology there are no gender distinctions (like he and she in Word formation in Hindi-Urdu primarily uses prefixes English) (Table 3). and suffixes to define inflectional and derivational Adverbs and postpositions are invariant, except for word classes. Nouns are generally inflected for num- the genitive postposition, which behaves like an ber, gender, and case. There are two numbers, singular inflected adjective. The postpositions mark case rela- and plural; two genders, masculine and feminine; and tions and adverbial functions. three cases, simple, oblique, and vocative. The oblique Hindi has no articles. forms occur when a noun or noun phrase is followed by a postposition, vocative case marks nouns referring Verbs to the addressee of the utterance. Nouns are inflected There are three tenses in Hindi: present, past, and according to their gender and word-final sound, as future. The tenses are formed by the suffixation process. exemplified by the four paradigms given in Table 2. Verbs are inflected for number, gender, and person. Adjectives are primarily of three types: (1) Simple (1) vo aa-t-aa hai. adjective, such as acchaa ‘good’; (2) Derived adjec- he come-present-mas.Sg is (3sg) tives using various parts of speech such as nouns: mar- ‘He comes’. daanaa ‘masculine’, adverbs: mandaa ‘slow’, and (2) vo aa-yaa. from agentive/adjectival particle vaalaa, e.g. dillii he come-past-mas, Sg. vaalaa ‘from Delhi’; and (3) Participial adjectives: ‘He came’. caltii ‘moving’, bhaagtaa ‘running’. Adjectives can be (3) vo aa-ye-g-aa used both attributively (immediately placed before he come-3sg-future-mas. Sg nouns) and predicatively (immediately placed before ‘He will come’. 457 HINDI-URDU TABLE 2 Hindi Noun Paradigms Case Paradigm-I Paradigm-II Masculine Nouns Ending in -aa Masculine Nouns Not Ending in -aa Singular Plural Singular Plural Direct laRkaa ‘boy’ laRke aadmii ‘man’ aadmii Oblique laRke laRkõ aadmii aadmiõ Vocative laRke laRko aadmii aadmio Case Paradigm-III Paradigm-IV Feminine Nouns Ending in -ii Feminine Nouns Not Ending in -ii Singular Plural Singular Plural Direct laRkii ‘girl’ laRkiyãã maataa ‘mother’ maataae~ Oblique laRkii laRkiyõ maataa maataaõ Vocative laRkii laRkio maataa maataao TABLE 3 Hindi Pronoun Paradigms 1st Person 2nd Person 3rd Person Singular Plural Singular Plural Honorific Singular Plural  Direct mãι ham tuu tum aap vo ve Oblique mujh ham tujh tum aap us un Genitive meraa hamaaraa teraa tumhaaraa aapkaa uskaa unkaa In addition to simple verbs, Hindi has two cate- Syntax gories termed ‘conjunct’ and ‘complex’ verbs. The Hindi is a Subject–Object–Verb (SOV) language with class of conjunct verbs is usually derived adding kar- a relatively fixed word order. Interrogative or other naa ‘to do’ or honaa ‘to be’ to noun, adjective, pro- sentence types do not introduce any changes in word noun, or adverb; e.g. order. In topicalization and focus structure, however, (4) kaam ‘work’, kaam karnaa ‘to work’, phrases may occur in a marked (i.e. exceptional) posi- (5) acchaa ‘good’, acchaa honaa ‘to recover’, tion, usually initial. The verb generally agrees with the (6) dhiire ‘slow’, dhiire karnaa ‘to slow down’, subject. In transitive perfective sentences, the subject (7) complex verb: likh ‘write’, lenaa ‘to take’ - is marked with the ne postposition, and the verb agrees likh lenaa ‘to write’ (for one’s own benefit)’. with the direct object. A rule of thumb is that the verb never agrees with any constituent that is marked with Hindi is also sensitive to stative/active and volition- a postposition. al/nonvolitional distinction; these four types of dis- Any sentence can be negativized by placing tinction are denoted by morphologically related verbs: the negative particle nahii ‘not’ in the preverbal position. (8) khulnaa ‘to be opened’, kholnaa ‘to open’,  (9) TuuTnaa ‘to be broken’, toRnaa ‘to break’. (10) vo nahι ι aa-ye-g-aa he not come-3sg-future-mas. Sg Compounding is an integral and very productive ‘He will come.’ process of word formation in Hindi. The noun–noun compounding involves 12 types of compounding. For example, khaanaa ‘eating’ and piinaa ‘drinking’ can be References compounded into khaan–paan, which means ‘life style’. From the viewpoint of morphological complexity, Beg, M.K.A. 1988. Urdu grammar: history and structure. New Delhi: Bahari. Hindi can be classified as an agglutinating language. Bhatia, Tej K. 1987. A history of the Hindi grammatical tradi- This means that derivation of words takes place by the tion. Leiden: E. J. Brill. addition of suffixes to simple or derived stems of Bhatia, Tej K. 1996. Colloquial Hindi. London: Routledge. major word classes. The process of prefixation is Bhatia, Tej K., and Ashok Koul. 2000. Colloquial Urdu. London: almost exclusively used with nouns and verbs; other Routledge. Dimock, Edward C., Edwin Gerow, C.M. Naim, A.K. word classes rarely participate in this process. The Ramanujan, Gordon Roadarmel, and J.A.B. van Buitenen. process of suffixation is productive equally with both 1978. The literatures of India: an introduction. Chicago: nouns and verbs. University of Chicago Press. 458  HIRI MOTU Kachru, Yamuna. 1980. Aspects of Hind grammar. New Delhi: Shackle, Christopher, and Rupert Snell. 1990. Hindi and Urdu Manohar Publications. since 1800: a common reader. New Delhi: Heritage. Ohala, Manjari. 1983. Aspects of Hindi phonology. Delhi: TEJ K. BHATIA Motilal Banarasidass. Rai, Amrit. 1984. A house divided. Delhi: Oxford University Press. See also India Hiri Motu Hiri Motu is a pidgin language that is of particular that is not inherent in the nature of the items con- interest because it is one of the few well-described pid- cerned, is expressed by means of pronominal suffixes gins with a vocabulary derived from a non-Indo- attached to an independent classifying element. The European language. Its lexical source is the Motu alienable possession of something to be eaten involves language spoken in a dozen villages along the south- the possessive classifier a-, while the alienable posses- ern coast of the ‘tail’ of Papua around Port Moresby, sion of anything else uses the possessive form e-, e.g. the capital of Papua New Guinea. a-gu aniani ‘my food’ (literally ‘foodstuff-me food’), Motu is a member of the huge Austronesian lan- e-gu ruma ‘my house’ (literally ‘thing-me house’). guage family, which extends throughout Polynesia, Many Oceanic languages exhibit a word order of Micronesia, much of Melanesia, most of Indonesia S(ubject)-O(bject)-V(erb), while others have SOV. and Malaysia, all of the Philippines, Madagascar, and A few languages also have VSO and VOS. Motu falls the interior of Taiwan. The Oceanic subgroup, to into the SOV class. SVO languages typically use which Motu belongs, comprises approximately half prepositions to mark certain functions of a noun in the the total number of Austronesian languages, all of sentence, e.g. the English preposition at marks a loca- which are located east of a north–south line through tion. Motu uses postpositions instead (i.e. particles that the Indonesian province of Papua. are equivalent to prepositions but occur after the noun Motu exhibits many of the structural features that they modify). Thus, Guhi amo e-moru ‘(S)he fell from are widely distributed in Oceanic languages. There is the roof’ (literally ‘roof from (s)he:NONFUTURE-fall’). a set of inflectional prefixes on the verb that provide With this brief background on Motu, we can turn to grammatical information about the subject, with the Hiri Motu. Because it is a pidgin language, it would be result that the subject can go unmentioned. These pre- expected to be simplified by comparison with its major fixes also mark tense and negation. Verbs that take source language. Pidgin languages typically eliminate objects are also marked by suffixes providing gram- (or sharply reduce) grammatical inflections. This ten- matical information about the object. Thus, e-ita-gu dency is clearly evident in Hiri Motu in that none of the ‘(s)he saw me’ means literally ‘(s)he:NONFUTURE-see- inflectional prefixes, e.g. Motu’s e- ‘(s)he:NONFUTURE’, me’ or baine-ita-gu ‘(s)he will see me’ (literally have been retained. Not surprisingly, the independent ‘(s)he:FUTURE-see-me’). subject pronouns are obligatory in Hiri Motu rather Although subject and object pronouns are not than optional as in Motu. Tense and negation are required in Motu, they can be present. When both sub- marked not by prefixes but by free forms. For example, ject and object are present, the subject must be fol- do(hore), occurring before the verb, marks future tense, lowed by ese, as in ia ese lau e-ita-gu ‘(s)he saw me’ while lasi ‘no’ occurs after the verb to mark negation. (literally ‘(s)he SUBJECT me (s)he:NONFUTURE-see-me’). The object suffixes have also been lost. Pronominal Motu nouns also reflect the widespread Oceanic objects are expressed with the same free pronouns that pattern of distinguishing between inalienable and are found in subject position. Subject and object pro- alienable possession. ‘Inalienable possession’ refers to nouns are thus distinguished only by word order. With an inherent relationship between the possessor and the pronominal subjects and objects, the word order is possessed (e.g. persons and their body parts). In Motu, OSV. Thus, contrast the examples above from Motu it is expressed with pronominal suffixes attached with the following Hiri Motu equivalents: lau ia itaia directly to the noun, e.g. mata-gu ‘my eye’ (literally ‘(s)he sees/saw me’ (literally ‘me (s)he see’) and lau ‘eye-me’), mata-mu ‘your eye’ (literally ‘eye-you’). dohore ia itaia ‘(s)he will see me’ (literally ‘me Alienable possession, i.e. a possessive relationship FUTURE (s)he see’). 459 HIRI MOTU With noun subjects and objects, the word order in In 1888, the British colony of Papua was formally Hiri Motu is SOV, as in Motu. For a pidgin language, proclaimed. Much of this colony was completely this is relatively rare, and even in Pidgin Fijian, for unknown to outsiders. The first colonial administrator which the lexical source is predominantly OSV, the immediately set out to establish a police force that word order is SVO. would serve to spread and strengthen colonial power. Other features have also been simplified in Hiri Police constables from a variety of locations were Motu. The distinction between inalienable and alien- appointed and trained and then posted to different able possession is lost on the grammatical level. parts of the colony. Some of these constables came However, the distinction is still there on the lexical from Fiji and the Solomon Islands, where some form level: inalienably possessed nouns such as mata- ‘eye’ of pidgin with an English-based vocabulary was include the original third-person singular suffix -na as already in use. These initial police trainees were then part of the noun root. In the grammar, simple posses- joined by Papuans from a variety of locations. While it sion is marked by a possessive pronoun that is a com- is not known precisely how these constables commu- bination of the independent pronouns and the original nicated, it seems possible that they made use of an nonedible alienable possessive markers. Thus, compare English-derived pidgin, along with the kind of simpli- the variety of Motu possessive constructions presented fied Motu that they would have acquired as a result of earlier with the following from Hiri Motu: lauegu contact with the local Motu-speaking population (with matana ‘my eye’, lauegu aniani ‘my food’, lauegu these two varieties possibly influencing each other). ruma ‘my house’. As these constables were posted to outlying areas in Hiri Motu has been discussed so far as a monolith- the late 1800s and into the 1900s, they took with them ic entity, although there is in fact considerable struc- a knowledge of whatever variety of Motu they had tural variation. In particular, speakers of Austronesian acquired in Port Moresby, and the language was also languages closely related to Motu—predominantly carried to outlying areas by released prisoners (who from the coastal area around Port Moresby—incorpo- themselves sometimes subsequently became police- rate more of the patterns from Motu than do Hiri Motu men). This language spread throughout Papua as a speakers from farther afield. These people, for exam- result of the establishment of colonial authority ple, typically maintain the distinction between inalien- through the police force. This is why this pidgin was able and alienable possession, as well as maintaining initially named ‘Police Motu’ (instead of Pidgin object suffixes on transitive verbs. Motu), a name that was shed, however, due to its colo- Papua has long been an area of language contact, nialist connotations (see below). and Tom Dutton (1985) describes in detail the devel- Hiri Motu is currently used as a lingua franca opment of a number of contact languages from lan- throughout much of the former British territory of guages of this area in association with the ‘hiri’ Papua, in the southern part of what is now Papua New trading expeditions conducted along the Papuan coast Guinea. This is an area of immense linguistic diversity, before colonial times. As the hiri expeditions were with about 200 languages—some Austronesian, but abandoned, so too were the languages that arose in this many also highly varied non-Austronesian lan- context, although Dutton was able to record some guages—spoken by several hundred thousand people. information from one-time participants. However, the English-derived pidgin language known Pidgin languages based on European languages as Tok Pisin is increasingly becoming the main lingua have typically evolved in language contact situations franca in areas surrounding the rapidly growing capital, in plantation economies (whether slave-based as in Port Moresby, which is located within the traditional the Atlantic or contract labor-based as in the Pacific). Hiri Motu area; it is also gaining speakers in some of The development of Hiri Motu is significantly differ- the rural areas where it was not used formerly. ent, in that plantation labor played little role in its ini- It is difficult to say at this stage what sort of future tial development. In the Port Moresby area in the there is for Hiri Motu. Although it still has well over 1870s and 1880s, a number of overseas visitors 100,000 speakers, competition with Tok Pisin is likely arrived, although, of these, only a relatively small to slow its growth and possibly even cause its numbers number of Christian missionaries stayed for long to shrink. However, there has also been a degree of periods. The missionaries set out to learn Motu, but active resistance to the spread of Tok Pisin by some, many other people did not share the same goal, and who regard the English-based pidgin as a variety of a somewhat simplified form of Motu came to be ‘broken English’ imposed by the former colonial over- widely used as a lingua franca (and even some of lords. Hiri Motu, in contrast, has been seen as a gen- the missionaries ended up learning this simplified uinely ‘Papuan’ language, given its clearly non-English variety rather than the variety normally used by origins. This kind of political sentiment toward Hiri Motuan villagers). Motu probably peaked in the late 1970s around the 460  HISTORICAL LINGUISTICS time that Papua New Guinea gained its independence Papuans could already see the spread of Tok Pisin from Australia (1975) with the development of the into areas that had traditionally made exclusive use of Papua Besena movement, and there has been increas- Hiri Motu as a lingua franca, and some Papuan nation- ing acceptance among Papuans since then of a gen- alists declared their separate independence (but with- uinely national role for Tok Pisin. out recognition from any other nation). Hiri Motu was In this political context, the original name ‘Police felt to represent Papuan nationalist aspirations in a Motu’ also became quite controversial, and this name way that Tok Pisin never could. Because of assumed is seldom used in Papua New Guinea today. Prior to associations with the colonial past, a government independence, the original British colony of Papua had committee in 1971 officially renamed the language largely been neglected, with little attention given to from ‘Police Motu’ to ‘Hiri Motu’, based on the socioeconomic development. This contrasted sharply assumption—albeit an inaccurate one, as emerged with the northern colony of New Guinea, which had from Dutton’s subsequent research—that this was the originally been under German control (although after language of precolonial hiri trading expeditions. World War I, Australia took over). A vigorous planta- tion economy was established in coastal areas, and it References was in this context that English-based Tok Pisin devel- oped and spread throughout that colony. Amini, Brian, Percy Chatterton, Raka Igo, Ron Lean, Api Leka, After World War II, Papua and New Guinea were and Frank Wood. 1976. The dictionary and grammar of Hiri Motu. Port Moresby (Papua New Guinea): The Papua New united into a single administrative entity, which was Guinea Government Office of Information. finally granted independence from Australia in 1975. Dutton, Tom. 1985. Police Motu iena sivarai. Port Moresby However, many Papuans were uneasy about the (Papua New Guinea): The University of Papua New Guinea prospect of becoming part of an independent country Press. because they felt that they would be economically and Dutton, T.E., and C.L. Voorhoeve. 1974. Beginning Hiri Motu. Canberra: Pacific Linguistics. numerically dominated by their northern neighbors, TERRY CROWLEY whose primary lingua franca, as it happened, was an English-based pidgin. See also Austronesian; Tok Pisin Historical Linguistics Languages are constantly in the process of change. languages known to have been derived from a single Spellings of English homophones such as meet and historical source, led to the observation that sound meat reflect vowel distinctions at an earlier stage in the change is, in fact, systematic. For example, there is a history of our language, before the two vowel qualities set of words starting with h in Spanish, such as hijo merged. Changes take place on the level of sound struc- ‘son’, hacer ‘do/make’, and harina ‘flour’, that have ture, grammar, vocabulary, and meaning. Historical cognates in Portuguese that start with f: filho ‘son’, linguistics is the study of language change by compar- fazer ‘do/make’, and farinha ‘flour’. Through compar- ison of a language at two (or more) points in time. ison with the words in Latin (such as filius ‘son’), for The field of historical linguistics developed signifi- which written records existed, it was possible to write cantly in the nineteenth century, after European schol- rules to account for the various changes that must have ars realized that there were systematic similarities taken place. Written records served two purposes: in between Sanskrit, the ancient language of India, and addition to confirming the validity of the methodology other well-known languages, such as classical Greek being developed by historical linguists, they gave and Latin. These similarities indicated that a genetic important information regarding the direction of sound relationship must exist; i.e. these three languages must changes. In the above examples, the Spanish h devel- stem from the same ancestor language. oped from Latin f and not the contrary. It was already well known that the Romance lan- The same methodology was then applied to the guages descended from a common language, Latin. A comparison of other languages that were obviously systematic comparison of cognates, words in related closely related, even though no written records existed 461 HISTORICAL LINGUISTICS . . for the original language, or proto-language. For used before a vowel, as in or! ‘he is happy’, na-or!-i example, the Germanic languages were compared and ‘he is not happy’. Through internal reconstruction, we classified as to how closely they were related to each can account for these exceptions by hypothesizing that other. The same was done for the Slavic languages. there must have originally been a consonant (indicated Comparison within and between such groupings led by a capital C) between the prefix and the following . to a classification of nearly all the languages in vowel (*na-Cor!-i). Europe, and some beyond, as members of the Indo- The study of the loss or retention of specific vocab- European family, all descended from a hypothetical ulary items is another method used in historical lin- language referred to as Proto-Indo-European, for guistics, often called lexicostatistics. This can be used which there are no written records. to study the degree of change at two stages of the same In English, as well as other Indo-European lan- language or to study the degree of relatedness of two guages, there are written records of the language at different languages. Usually, lists of basic core vocab- various points in time, making possible a comparison ulary are used for this type of study. of earlier stages of the language with the present-day Dialect geography is another method used by his- forms. The use of written records requires interpreta- torical linguists. Details of the pronounciation of tion of the symbols appearing in them, to discover the selected words are plotted on maps throughout the sound structure of a language at the time of writing. In area where a given language is spoken to determine the case of the Germanic languages, the earliest writ- the details and geographical boundaries of different ten records use a different type of script, called runes, dialects, such as Northern and Southern American which themselves changed over time. English. Comparison of the sounds in cognate words from When different languages come into contact with related languages led to the development of a tech- each other, some degree of lexical borrowing nique for phonological reconstruction. Sounds are inevitably takes place. The influence of French on studied systematically in their various positions with- English after the Norman invasion of the British Isles in a word to obtain a full picture of what the original in 1066 was a major factor affecting English vocabu- sound system must have been like that would account lary, with words of French origin, such as arise, join- for the individual differences in the various daughter ing the already existing combinations of a verb and an languages. Returning to the examples of Spanish and adverb, such as get up. Portuguese, it would have to be noted that not all One principle that motivates sound change is sim- words with initial h in Spanish begin with f in plification. This may involve the weakening of more Portuguese; some begin with h, as in the cognate pair complex sounds to less complex counterparts, such as hombre (Spanish) and homem (Portuguese) ‘man’. a change from the affricate t∫ to a simple fricative ∫ or Likewise, not all words beginning with f in Portuguese s. Loss of consonants at the end of syllables or words begin with h in Spanish; some begin with f, as in fuego is another example of simplification. On the grammat- (Spanish) and fogo (Portuguese) ‘fire’. All of these ical level, simplification may involve the elimination patterns have to be examined. Furthermore, the best of certain grammatical inflections of nouns. But sim- results require that attention be given to all descendant plification on one level may actually increase the languages; in the case of Romance languages, this complexity of a language on another level and trigger would include not only the languages with many additional change. speakers (Italian, French, Rumanian, Portuguese, and One way by which simplification is achieved, par- Spanish) but also those with fewer speakers (Catalan, ticularly on the level of word structure, is through Provençal, Rheto-Romance, and Sardinian). The analogy. The extension of use of the plural suffix -s in precedent for the rigorous detail required in the com- English to words that originally had other plural end- parative method was set by Jacob Grimm, in his work ings has taken place by analogy. For example, the relating the German consonant system to the systems word hippopotamus, which came from Latin, original- of other Indo-European languages. ly took the plural form hippopotami. However, hip- Another method used by historical linguists is popotamuses, based on the main pluralizing pattern in internal reconstruction. This involves the observation English, is now the more commonly used form. of patterns within an individual language to arrive at Analogy is also used in the creation of new words hypotheses regarding an earlier stage of that language, based on an already existing pattern in the language. without taking into account information from outside The development of terms such as chairperson took languages. For example, the prefix n- (‘not’) in the place by analogy with the words they were replacing, Wayampi language has a variant na-, which occurs in this case chairman. when the prefix attaches to a consonant. However, Languages may change with respect to their sound there are restricted cases in which the na- variant is structure or grammar. The meanings of words can 462  HISTORICAL LINGUISTICS change, and old vocabulary items may be lost or new A detailed comparison of individual languages ones may be added. showing a high number of cognate words provides the The change of word-initial f to h in Spanish is one information necessary for a classification and subclas- example of a sound change. Another change in this sification within a language family. A diagram, similar language is from o to ue in all words in which o had to a family tree, can then be used to show the internal intonational stress (Latin novem ‘nine’, Spanish classification of the family. This type of detailed clas- nueve). This change is conditioned; it occurs under sification was first done for Indo-European by the specific conditions, i.e. only in particular phonetic nineteenth-century historical linguists. environments. If all occurrences of o had changed to The methods of historical linguistics are now being ue, without reference to any phonetic environment, the applied to the study of many indigenous languages and change would be unconditioned. language families. It is possible to recognize lan- One type of grammatical change consists of the guages that are similar to each other, by identifying a change or loss of affixes. Latin had a set of case suf- significant number of cognates (words coming from a fixes that occurred on nouns to distinguish between single source). Through a systematic comparison of subject, direct object, and indirect object, as in the the similarities and differences in the sound structure word for ‘daughter’: fili-a, fili-am, and fili-ae, respec- of cognates, it is possible to reconstruct what the tively. In Portuguese, filha is used without such case sound system of the common ancestor, or ‘proto-lan- distinctions; i.e. Portuguese has lost the Latin case sys- guage’, must have been like to account for the present tem. Other grammatical changes may affect pronouns, form in the various languages. Protoforms are preced- as in the elimination of thee and thou from most pres- ed with an asterisk to indicate that they are hypotheti- ent-day dialects of English. The relative word order of cal. For example, in proto-Tupi-Guarani, of lowland the subject, object, and verb may also undergo modifi- South America, one of the reconstructed consonants is cations over time. *ts, which has reflexes of ts, s, h, and 0 (zero) in indi- Words may change their meaning or develop sec- vidual languages of the family. ondary meanings. For example, the word deer, referring Phonological reconstructions have been done for to a specific animal, developed from the more generic several indigenous language families of the Americas, Old English word de-or ‘beast’. In recent years, the word as well as in other parts of the world. The quality of a mouse has acquired an additional meaning, referring to reconstruction depends on the availability of adequate a computer accessory. and accurate data of individual languages within a The acquisition of new vocabulary may take place given family. As more data become available, the through innovation (such as byte), blending of already- reconstructions are revised as necessary, and subclas- existing vocabulary (brunch, from breakfast and sification is possible. lunch), or the use of derivational processes existing in The word and sentence structures of some indige- the language (computer from compute). It may also nous language families, such as the Tupi-Guarani fam- take place through lexical borrowing as the result of ily in South America, have also been reconstructed. language contact. Some examples of such borrowing Grammatical reconstructions require much more data into English include canoe (borrowed from Carib), than do reconstructions of sound and are sometimes jaguar (from Tupinambá), safari (from Swahili), and done in stages, as an increasing number of grammati- smorgasbord (from Swedish). cal descriptions of individual languages are made As changes take place in different geographical available. To give an example, the set of pronouns and regions where a single language is spoken, different personal prefixes was first reconstructed for proto- dialects develop. If speakers of different dialects are iso- Tupi-Guarani without reference to the particular gram- lated from each other over long periods of time, the matical contexts in which they were used. Additional dialects may undergo so many changes that speakers data were necessary before a more complete descrip- from the different groups no longer understand each tion could be made. other. At this point in time, what were originally differ- Summarizing, the methods of historical linguistics, ent dialects of the same language have become different which were developed and tested in the study of Indo- languages, which are members of a linguistic family or European languages, are now being applied to the possibly a subgroup within the family. For example, the study of languages in other parts of the world. Scandinavian languages (Danish, Icelandic, Norwegian, and Swedish) make up a subgroup within the Germanic References language grouping. They are much more closely related Bynon, Theodora. 1977. Historical linguistics. London: to each other than they are to English, German, or Cambridge University Press. Dutch, having descended from a common ancestor lan- Campbell, Lyle. 1999. Historical linguistics: an introduction. guage referred to as Old Norse. Cambridge, MA: MIT Press. 463 HISTORICAL LINGUISTICS Jeffers, Robert J., and Ilse Lehiste. 1979. Principles and meth- McCrum, Robert, William Cran, and Robert MacNeil. 1993. ods for historical linguistics. Cambridge, MA: MIT Press. The story of English. New York: Penguin USA. Jensen, Cheryl. 1999. Tupi-Guarani. The Amazonian lan- Trask, R.L. 2000. Dictionary of historical and comparative lin- guages, ed. by R.M.W. Dixon and Alexandra Y. Aikhenvald. guistics. Chicago: Fitzroy Dearborn Publishers. Cambridge: Cambridge University Press. CHERYL JENSEN Lehmann, Winfred P. 1993. Historical linguistics: an introduc- tion. London and New York: Routledge. See also Grimm, Jacob History of Linguistics: Overview The Greco-Roman World rudimentary understanding of speech as the effect of articulatory interference with the flow of air from the The European tradition of linguistics begins in Classical lungs. The descriptive framework for Greek phonetics Greece with the development of an active philosophical was the Greek alphabet, and statements took the form discourse on linguistic topics. Several of Plato’s dialogs of accounts of the pronunciation of the letters in it. touch upon linguistic issues, with one, the Cratylus, Plato grouped together vowels in contrast to conso- devoted exclusively to the subject of the conventionality nants and distinguished within consonants between of language. The issue is discussed along the lines of a continuants and stops, the latter being unpronounce- genealogy of language, with one side supporting a ‘nat- able without an adjacent vowel sound. The Stoics dis- ural’ origin for human communication, and another a tinguished three aspects of the gramma, thereby ‘conventional’ basis. In the same context, onomatopoeia confirming its phonetic-orthographic unity: its phonet- (sound symbolism) and the internal relation between ic value, its written shape, and the name by which it names and their referents are discussed, with mixed con- was designated. These three properties of letters con- clusions. From the outset of the Greek tradition, ques- tinued to be distinguished throughout Antiquity and the tions about language concentrated entirely on the Greek Middle Ages, their later Latin names being potestas, language and were considered within the terms of the figura, and nomen. The Greco-Roman world did its nature–convention debate and the somewhat related reg- best and most influential work in the field of grammar. ularity–irregularity (analogy–anomaly) controversy. Dionysius Thrax’s Techne Grammatike (most of it of Aristotle also discussed linguistic issues in his treatises debated authorship) and Apollonius Dyscolus’s Syntax on logic and rhetoric—without dedicating any of them are the two most significant works in the field. exclusively to the subject—with a discernible overall The Romans built largely on the Greek foundations. tendency to subscribe to conventionalist doctrine. Varro—with his De Lingua Latina, of which only a After Aristotle, the Stoics, with their founder Zeno part survives—is the first attested Latin writer on lin- (c. 315 BCE), formalized the dichotomy between form guistic questions who is of importance. One of the and meaning by distinguishing in language the ‘signi- main sources for the analogy–anomaly controversy, he fied’ and the ‘signifier’. They gave separate treatment may have misinterpreted it as a matter of permanent to phonetics, grammar, and etymology, devoting con- academic attack and counterattack, rather than as the siderable attention to the last. The Stoics accorded more probable coexistence of opposing tendencies. grammar-independent recognition within philosophi- Varro envisaged language developing from an original cal studies; it is, however, difficult to reconstruct with limited set of primal words, imposed on things so as to certainty the details of their linguistic theory. refer to them, and acting productively as the source of The main aspects of linguistic study that received large numbers of other words through subsequent specific attention among early Greek scholars were changes in letters, or in phonetic form, as civilization etymology, phonetics, and grammar. Etymology was advanced and human life became richer. Apart from understood as the unfolding of words, through which his innovative distinction between derivational and their true meanings are made plain. In phonetics, some inflectional word formation, Varro also showed origi- articulatory classifications of speech sounds were nality in his proposed classification of inflected words attempted, the syllable was introduced as a unit of in Latin, by setting up a quadripartite system of four description, and by the time of the Stoics there was a inflectionally contrasted classes. 464  HISTORY OF LINGUISTICS: OVERVIEW Priscian’s voluminous grammar (c. 500 CE) may be Hebrew, Arabic, and modern European languages reg- taken as representative of late Latin grammarians. The ister a shift in the linguistic interests of Europeans and Greek technical terms are given fixed translations with trigger creative thinking on language structure and the nearest available Latin word. Priscian draws heav- written representation. With the invention of printing ily on Apollonius and his son Herodian, acknowledg- and the emergence of a need for uniform orthography ing both in his introductory paragraphs. He organized within each language, pronunciation issues gain the description of the forms of nouns and verbs and of greater attention. the other inflected words, by setting up canonical or By the end of the sixteenth and the beginning of the basic forms; from these, he proceeded to the other seventeenth centuries, the indigenous languages of the forms by a series of letter changes, the letter being for Americas had captured the imagination of Europeans. him, as for the rest of Western Antiquity, the minimal Colonization of the Far East meant that the Chinese unit of both orthography and pronunciation. languages and their writing system became known to Europe and were largely admired. With very different word structures those of European languages, Chinese The Middle Ages and Native American languages served to stretch the Continuing from the Classical tradition, the Middle linguistic experience of Europeans far beyond what Ages had a distinct focus on grammar as a foundation the Classical and Medieval world had allowed them. of scholarship, both as itself a liberal art and as a pre- At the same time, the Royal Society in Britain (estab- requisite for reading and writing Latin correctly. lished 1660) and the Académie Française (established Linguistic studies, almost wholly pedagogical in their 1635) in France were both concerned with linguistic aims and largely derivative in their doctrine, were sub- research and with literary and linguistic standards. ordinate to theology, the Christian faith and Christian In the intellectual movements of the sixteenth and doctrine, and closely followed Donatus and Priscian. In seventeenth centuries, both rationalist and empiricist the history of Medieval linguistic science, the most influences may be seen at work. The notion of a uni- exciting period is from around 1100 to the end of the versal thought possessed by mankind, basically inde- period (and more specifically between c. 1200 and c. pendent of any particular language and therefore 1350), when scholastic philosophy gave rise to a num- expressible in a universal language, such as that of John ber of ‘speculative grammars’ or treatises De modis Wilkins, was a rationalist conception. Rationalism is significandi. Scholasticism itself was the result of the also responsible for the production of philosophical integration of Aristotelian philosophy into Catholic grammars—especially those associated with the French theology. The mere description of Latin was considered Port Royal schools—which succeeded the Medieval inadequate, and attention was paid to the mental repre- scholastic grammars.They attempted to reveal the unity sentations to which speech relates. The main distinc- of grammar underlying the separate grammars of dif- tion was between significatio (the meaning of a word) ferent languages in their role of communicating and suppositio (its point of reference in the real world). thought, itself comprising perception, judgment, and This basic distinction appeared repeatedly in different reasoning. forms and with different interpretations, in oppositions As part of English empiricism, work was carried out such as meaning–reference, denotation–connotation, on phonetic questions under the titles of ‘orthography’ and intension–extension. In the modistic system, enti- (right spelling) and ‘orthoepy’ (right pronouncing) dur- ties possess various properties (modi essendi). The ing the same period (the term ‘phonetics’ is first record- mind apprehends these in certain modes of understand- ed in the nineteenth century). Work by William Holder ing (modi intelligendi). In language, the mind confers (1669) and Francis Lodwick (1686) revolutionized meaning on successions of sounds (modi significandi), articulatory phonetics, and John Wilkins included in which thus become words. The demand that grammat- his Essay towards a real character and a philosophical ical description should be integrated into philosophical language (1669) a sound chart that can be compared theory and represent universal principles of cognition with early editions of the International Phonetic brought about a great change in linguistic studies. Alphabet. Unlike all previous grammatical doctrines, the modis- In the eighteenth century, linguistic issues again tae were theory- rather than data-oriented. became central for philosophy as part of its search for the traits of primitive society. Both Etienne Condillac and Jean-Jacques Rousseau considered that abstract The Renaissance and Modern Times vocabulary and grammatical complexity developed The adherence of linguistic analysis to the Greco- from an earlier individual concrete vocabulary with Roman archetype became less strict in the fifteenth very few grammatical distinctions or constraints. and sixteenth centuries. Grammatical studies of They also regarded reliance on tonal contrasts in the 465 HISTORY OF LINGUISTICS: OVERVIEW manner of Chinese as a survival of a primitive feature Indo-European family. Both the use of comparison as and considered poetry to have sprung from chanting as the clue to earlier history and the conception of change the earliest form of language. as degeneration from primitive integrity were common In 1772, Johann Herder supported a monogenetic property of the scientific thought of the time. Relations (common origin) theory of all languages, as of all cul- between the parent language and the known Indo- tures. His theory suffered from the restricted time per- European languages were set out according to the spective with which the eighteenth century viewed Stammbaumtheorie or genealogical tree model. man’s existence on earth, with a consequent attempt to August Schleicher, with his biological approach to see enduring characteristics of the early stages of lan- language and his quest for the Ursprache (the original guage in alleged ‘primitive languages’ still existing in language), initiated the practice of distinguishing the present. Work by James Harris (1751) and Horne reconstructed forms with an asterisk (whence the later Tooke (1786) demonstrate the vivid interest of the term ‘starred forms’). He assigned Sanskrit a place time in the origin of language. like any other language, in the ‘Arian’ (Indo-Iranian) Wilhelm von Humboldt—exceptionally for his group and expressed his belief that the three current time—did not concentrate predominantly on history. language types (isolating, agglutinating, and inflec- He did not distinguish sharply between systemic and tional) represent historical stages in the growth of lan- historical perspectives and followed Herder in assert- guages to their highest point of organization. ing the individuality of each different language as a The Neogrammarians also strove to ground their peculiar property of the nation or the group that speaks work on comparative–historical linguistics within the it, despite believing in the universality of speech. natural sciences, preferring physics over biology as Humboldt is best known in linguistics for popularizing their model. Hermann Osthoff and Karl Brugmann set a tripartite language typology (isolating, agglutinative, out the principles of the movement in a programmatic and flexional) according to the predominant structure article (1878): all sound changes, as mechanical of the word as a grammatical unit. processes, take place according to laws that admit no exceptions (ausnahmslose Lautgesetze) within the same dialect and within a given period of time; the Nineteenth Century: Comparative and same sound in the same environment will always Historical Linguistics develop in the same way; analogical creations and ref- In the last quarter of the eighteenth century, linguistic ormations of specific words as lexical or grammatical science received a very productive stimulus from entities are a universal component of linguistic change India. In 1786, Sir William Jones showed the histori- at all periods of history and prehistory. The influence cal kinship of Sanskrit, the classical language of India, of the Neogrammarians in linguistic practice has been with Latin, Greek, and the Germanic languages. India overwhelming. It is true, however, that they were was not just the source of linguistic evidence but also largely drawing out what had been implicit in the very offered an alternative linguistic tradition, dating back practice of the subject, and distinguishing it from to the second half of the first millennium BCE, with unnecessary and fallacious assumptions. outstanding achievements in general linguistic theory The critique of the Neogrammarian position came and semantics, phonetics and phonology, and descrip- from a branch of linguistics that they had been at pains tive grammar. to encourage, dialectology. The temporal and geograph- Early comparativists concentrated on Sanskrit and ical limits of a dialect are difficult to determine, espe- its relations with the other Indo-European languages cially when one scrutinizes the language closely. The (hence the term vergleichende Grammatik ‘compara- most important opponents were Hermann Schuchardt, tive grammar’). Rasmus Rask (1787–1832) was the who included in his works an article ‘On Sound Laws: first to bring order into etymological relationships by Against the Neogrammarians,’ and Jules Gilliéron, who setting out systematic comparisons of word forms, put together the linguistic atlas of France. matching a sound in one language with a sound in another, with such matches exemplified in a number of Twentieth Century different words. The correspondences now known as Grimm’s Law were in fact first discussed by Rask. The key figure in the change from nineteenth- to twen- The historicism and nationalism of the time charac- tieth-century attitudes was the Swiss linguist Ferdinand terized the work of Franz Bopp and Jacob Grimm. de Saussure, who after studying in Leipzig with mem- Bopp saw as the main purpose of his Conjugation bers of the Neogrammarian school made his first sig- system the reconstruction of the original grammat- nificant contribution in Indo-European comparative ical structure of the language whose gradual disinte- linguistics. Saussure’s doctrine on linguistics in the gration had produced the attested languages of the early twentieth century was recorded in the Cours de 466  HITTITE linguistique générale (1916; Course in General versions of the theory and the lexicon as a source Linguistics), a reconstruction of his lectures from notes attained greater importance. taken by his students, published posthumously. In the last quarter of the twentieth century, linguis- Saussure’s contribution lies in three major areas: (1) He tics has gained still greater status as a pioneering aca- formalized the two fundamental dimensions of linguis- demic discipline. An increased diversity—stemming tic study: ‘synchronic’, in which languages are treated from the emergence of branches such as computation- as self-contained systems of communication at any par- al and applied linguistics and psycholinguistics along- ticular time; and ‘diachronic’, in which the changes to side the established subdisciplines—is the most which languages are subject in the course of time are prominent and promising feature of the discipline at treated historically. (2) He distinguished the linguistic the turn of the millennium. competence of the speaker as a member of a speech community from the actual phenomena or data of lin- References guistics (utterances), as, respectively, langue and parole. (3) He showed that any langue must be envis- Baratin, Marc, and Françoise, Desbordes. 1981. L’analyse lin- guistique dans l’antiquité classique. Paris: Klincksieck. aged and described synchronically as a system of inter- Bursill-Hall, Geoffrey Leslie. 1971. Speculative grammar in the related elements, lexical, grammatical, and Middle Ages: the doctrine of Partes orationis of the phonological, and not as an aggregate of self-sufficient Modistae. The Hague: Mouton. entities (which he compared to a mere nomenclature). Chomsky, Noam. 1957. Syntactic structures. The Hague: The greatest impact of structuralism was in the domain Mouton. Eco, Umberto. 1995. The search for the perfect language. of phonology (the study of sound systems), where Oxford and Cambridge, MA: Blackwell. Nikolaj Trubetzkoy and Roman Jakobson developed Humboldt, Wilhelm von. 1988. Über die Verschidenheit des the theory of ‘distinctive features’. menschlichen Sprachbaues; as On language: the diversity of ‘Structuralism’ was widely preached in America human language-structure and its influence on the mental and was epitomized in the historical figure of Leonard development of mankind, translated by Peter Heath, with an introduction by Hans Aarsleff. Cambridge: Cambridge Bloomfield. American structuralism, however, dif- University Press. fered from its European ancestor in its insistence on a Law, Vivien. 1997. Grammar and grammarians in the early data-oriented approach which originated from the middle ages. London and New York: Longman. urgent need to collect data on the disappearing Native Lepschy, Giulio (ed.) 1994–1998. History of linguistics, Vol I: American languages. This persistence remarkably lim- The eastern traditions of linguistics, Volume II: Classical and medieval linguistics, Volume III: Renaissance and early ited the theoretical perspectives of American struc- modern linguistics, Volume IV: Nineteenth-century linguis- turalism and eventually brought about its downfall. tics. London and New York: Longman. The reaction came in the middle of the century and Matthews, Peter Hugoe. 1993. Grammatical theory in the resulted in a profound change in the outlook of theo- United States from Bloomfield to Chomsky. Cambridge: retical linguists worldwide. Noam Chomsky’s Cambridge University Press. Pedersen, Holger, and John Webster Spargo. 1962. The discov- Syntactic structures (1957) marks the first public ery of language, linguistic science in the nineteenth century, appearance of a change in outlook on the study of lan- 3rd edition. Bloomington: Indiana University Press. guage and the scientific status of linguistics. His theo- Robins, Robert Henry. 1967. A short history of linguistics. ry, ‘generative linguistics’, is still developing and has London and New York: Longman; 4th edition, 1997. become the dominant theory in linguistics at the end of Saussure, Ferdinand de. 1983. Cours de linguistique générale. Paris: Payot, 1916; as Course in general linguistics, translat- the twentieth century. It involves the notion of a lin- ed and annotated by Roy Harris. London: Duckworth. guistic ‘deep structure’, whence utterances are pro- Ziolkowski, Jan (ed.) 1990. On philology, University Park and duced through a series of ‘transformations’. London: Pennsylvania State University Press. Transformations became less crucial in subsequent CHRISTOS G. NIFADOPOULOS Hittite Hittite belongs to the Anatolian branch of the Indo- sources ranging from approximately 1650 to 1180 European family. It was spoken in inner Anatolia dur- BCE. It is the oldest attested Indo-European lan- ing the second millennium BCE, with its written guage. Other Anatolian languages include Palaic and 467 HITTITE Cuneiform Luwian (also attested in the second millen- dative/locative [D/L], directive, ablative, instrumental, nium BCE), Hieroglyphic Luwian, Lycian, Milian, vocative) and number (singular and plural); nouns Carian, and Lydian (attested in the first millennium belong to two genders: common and neuter. The direc- BCE); all these languages are relatively poorly known. tive and the instrumental case are mostly limited to the Knowledge of Hittite is a recent achievement: ancient language (Old Hittite); later (Middle and New although some Hittite texts were found much earlier, Hittite), they merge with the dative/locative and the they were deciphered starting from 1916 only. Hittite ablative, respectively. Directive, ablative, and instru- is documented by several thousands of clay tablets, mental only had endings for the singular; in Old mostly found in the archives of the Hittite capital, Hittite, they could never occur with animate nouns Hattusa, near the modern village of Bogazkahle, about (the animate vs. inanimate distinction cross-cuts gram- 210 km southeast of Ankara. The Hittites wrote their matical gender, inanimate nouns belonging partly to language in cuneiform, a syllabic system constituted the neuter, partly to the common gender). by signs for CV, VC, V, and, to a limited extent, CVC An original feature of Anatolian is the use of deriva- syllables, also used for numerous other languages in tional suffixes to serve syntax, rather than to enrich the the Ancient Near East. The cuneiform syllabary also lexicon, a function normally assigned to inflectional contains a number of ideograms, conventionally ren- morphology. Thus, in Cuneiform Luwian, there is no dered in transliteration with the Sumeric equivalent (in inflectional genitive; nouns used as modifiers of other capital letters, see e.g. (1) and (2) below); ideograms nouns take a derivational suffix, which allows them to were the key for the decipherment of Hittite. behave as adjectives, and agree with their heads. In Hittite, neuter nouns cannot function as subjects of Grammatical sketch transitive verbs, but must obligatorily be transposed to the common gender by means of the suffix -ant-: after Phonology suffixation, -ant- derivates take the nominative ending Partly for problems connected with syllabic writing, of the common gender and can be made the subject of partly on account of as yet poorly understood scribal transitive verbs (e.g. tuppi, ‘clay tablet’, neuter, tup- habits, our understanding of Hittite phonology is not piyanza, same meaning, common gender). completely clear. Among vowels, four separate Adjectives and pronouns inflect for case, number, phonemes certainly existed: /a/, /e/, /i/, /u/; the exis- and gender; attributive adjectives usually agree with tence of a phonemic /o/ is still discussed. Vowel length their head. There is no specific morphology for com- was probably distinctive, as shown by pairs as uddar, parison. Personal pronouns have stressed and ‘word’, udda¯r, ‘words’. unstressed forms; an unstressed nominative only exists Consonants include four series of stops, bilabial, den- for third-person pronouns. Besides, Old Hittite also has tal, velar, and labiovelar, all with a distinction between enclitic (i.e. unstressed, hosted by the preceding word) voiced and a voiceless (or lax and tense), which was possessives inflected to agree with their head, which apparently neutralized in the word-initial and word-final also functions as a phonological host: attas=mas, ‘of positions. Simple vs. double writing of stops and of the my father’ (lit.: father-GEN my-GEN; the sign = indicates fricative -h- renders the voiced ~ voiceless contrast. The enclisis); after Old Hittite, genitive forms of stressed fricative /s/, liquids and nasals could be distinctively pronouns are used instead of possessives: ammel attas, simple or geminated word-internally (Table 1). ‘(of) my father’ (lit.: I-GEN, father-NOM (or GEN)). Inflectional categories of verbs include tense (pres- Morphology ent [PRES]/future and preterit [PRET]; there is a peri- Hittite is an inflectional language of the fusional type. phrastic perfect with the auxiliaries har(k)-, ‘have’, and Inflectional categories of nouns include case (nomina- es-, ‘be’), mood (indicative and imperative; other types tive [NOM], accusative [ACC], genitive [GEN], of modality are expressed by the particle man and the indicative), diathesis (active and medio-passive), per- son, and number. Nonfinite verb forms are the infini- TABLE 1 Hittite Consonants tive, the participle, and the -wan- supine, only used in Voiceless stops p t k kw inchoative constructions; verbal nouns are built with Voiced stops b d g gw the suffixes -atar or -uwar/-mar: iya-, ‘to make’, Voiceless fricatives s h Voiced fricative 1 iyawar, ‘the act of making’. Verbs are divided into two Affricate ts inflectional classes, called -mi and -hi conjugation Nasals m n from the ending of the first person singular. Some other Lateral l verbal categories are expressed derivationally. Vibrant r Causatives are made with the suffix -nu-: huinu-, ‘to let Glides w j escape’, from huwai-, ‘to run’, huisnu-, ‘to rescue’, ‘to 468  HITTITE give life’, from huis-, ‘to live’, ‘to be alive’ (this suffix (d) nu= war=an ANA DUTU can also derive verbs from adjectives: daluganu-, ‘to CONN PTC 3SG.ACC to sungod lengthen’, from daluki-, ‘long’). The suffix -sk- is used URU TÚL-na AŠŠUMLÚ SANGA-UTTIM for iteratives: pai-, ‘to give’, pesk-, ‘to give habitually’. Arinna for priesthood Derivation is widely used for creating new words; tittanumi derivational affixes are by the most part suffixes. install:1SG.PRES (There are two verbal prefixes, pe- and u-, which ‘(a) because I promoted him, (b) I never mostly occur with motion verbs and express opposite handled him over to a bad tribunal or to a deixes; verbal reduplication is not productive: often, bad god; (c) and now I will take him (d) reduplicated verbs have the same meaning as their and make him priest for the sun goddess of nonreduplicated counterparts.) Composition is com- Arinna’, StBoT 24, IV 11–15. paratively infrequent. Note: NEG = negation; PREV = preverb; PTC = particle; SG = singular Syntax Hypotaxis is relatively poor, the most frequent type Hittite has nominative–accusative alignment for com- of subordinate clause being relative clauses. Relative mon gender nouns. As remarked in the preceding sec- clauses are corelative in the great majority of occur- tion, neuter nouns must be transposed to common rences: the relative clause, mostly preposed, contains its gender in order to function as the subject of transitive head, which may or may not be referred to by a demon- verbs. Since this kind of derivation has a systematic strative in the main clause. An example is given below: character, the derivational affix together with the com- GIŠ mon gender nominative ending functions as an erga- (2) TUKUL= ma kuin tive case. A further distinction between transitive and weapon PTC REL.ACC intransitive subjects, only involving third person, is apiya harkun n= an that transitive subjects can be omitted if they are there have-1SG.PRET CONN 3SG.ACC recoverable from the context, whereas intransitive sub- halissiyanun jects cannot. inlay-1SG.PRET Hittite has a rigid Object-Verb order, with a verb-ini- ‘the weapon that I had there I had inlaid’, tial variant sometimes found under specific pragmatic StBoT 17, obv. 46. conditions (see e.g. (1a)). Modifiers (attributive adjec- Note: REL = relative tives, genitives, relative clauses) regularly precede their head, but certain determiners follow. Adpositions fol- Circumstantial clauses include temporal, causal, low their complements (postpositions), and inflected and conditional clauses; they are mostly preposed. auxiliaries follow nonfinite forms of main verbs. All Complement clauses are virtually nonexistent; only in enclitics, except for possessives, the focalizer -pat, and New Hittite a limited number of postposed comple- conjunctions used in noun phrase coordination, follow ment clauses occurs, introduced by the conjunction Wackernagel’s Law very strictly; consequently, they kuit (nominative/accusative neuter singular of the rel- are placed after the first accented word in the sentence ative pronoun, already used in causal clauses with the (P2). Sentence connectives (CONN) occur in virtually meaning ‘because’, cf. e.g. (1a)). every sentence after the Old Hittite period; often, it Hittite is extremely rich in enclitic particles. Among seems that they are introduced for the only purpose of P2 enclitics, one finds adversative connectives (as -ma hosting enclitics. The following example serves as an in e.g. (2)), the direct speech particle -wa(r)- (cf. e.g. illustration of Hittite sentence structure: (1)), the reflexive particle -za, and the so-called local particles, most likely cognate to the Indo-European (1 a) sallanun= war= an kuit preverbs (the modal particle man can be enclitic, and promote:1SG.PRET PTC 3SG.ACC because occur P2, or sentence initial and host the other encli- ammuk tics; enclitic personal pronouns also occur in P2, cf. 1SG.NOM the form -an in the examples). (b) nu= war=an huwappi DI-esni CONN PTC 3SG.ACC bad:D/L tribunal:D/L huwappi DINGIRLIM-ni UL para Hittite and Indo-European bad:D/L god:D/L NEG PREV When Hittite was deciphered, it immediately became UL kuwapikki tarnahhun clear that it did not fit into the traditional reconstruction never hadle:1SG.PRET of Proto-Indo-European (PIE), mostly based on Greek (c) kinuna=ya= war=an karapmi and Sanskrit. The absence of a number of grammatical now and PTC 3SG.ACC take:1SG.PRES categories traditionally reconstructed for PIE (optative, 469 HITTITE aorist, dual, and the three-gender system, among others) uiuere). After a vowel, */h2/ is also preserved as /1/: has given rise to different theories. The most well- Hitt. newahh- ‘renew’, < *neweh2- (Latin noua¯re). known is the so-called ‘Indo-Hittite hypothesis’, put Another important confirmation came from the forward by Sturtevant in 1933. According to Sturtevant, placement of enclitics in P2: Wackernagel (1892) had Anatolian must be considered a separate branch of a formulated his law on word order mostly on the basis language family whose other branch is constituted by of Ancient Greek and Sanskrit, where enclitics often all other Indo-European languages: the nonoccurrence occur in P2, but much less frequently than in Hittite. of certain features in Anatolian can then be explained by The Hittite verb also preserves a number of archaic separate developments of the Indo-European branch. features, notably the athematic (-mi) conjugation; the - Since the publication of Sturtevant’s book, the under- hi conjugation is a Hittite innovation, related to both standing of Hittite has become deeper, partly thanks to the PIE perfect and the PIE middle. the availability of newly found material; a better insight Among indisputable innovations, one can list the into the position of Hittite among the other Indo- emergence of split ergative alignment, and the con- European languages has been gained after cuneiform straints on omission of third-person subjects. paleography has allowed a precise dating of the Hittite tablets. Many other theories have been worked out, References among which was the so-called Schwundhypothese (‘loss hypothesis’), according to which Hittite once had Chicago Hittite dictionary, Chicago, 1980. Friedrich, Johannes. 1952. Hethitisches Wörtebuch. Heidelberg: all the inflectional categories known from the other Winter. ancient Indo-European languages, but lost them, as a Friedrich Johannes. 1960. Hethitisches Elementarbuch. result of intensive contact with non-Indo-European lan- Heidelberg: Winter. guages in prehistoric times. Held, Warren. 1957. The Hittite relative sentence. Baltimore: However, the discovery of Hittite did not only come LSA (Language Dissertations 55). Hrozný, Friedrich. 1917. Die Sprache der Hethitier. Leipzig: to raise problems for the traditional reconstruction of Hinrichs. PIE: many important findings provided evidence for Keilschrifttexte aus Bogazköy. Leipzig, 1916–23; Berlin, 1954. theories that could scarcely be demonstrated before. Keilschrifturkunden aus Bogazköy. Berlin, 1921. The most striking confirmation was given by the Kronasser, Heinz. 1966. Etymologie der hethitischen Sprache. reflexes of the PIE laryngeals, reconstructed on the Wiesbaden: Harrassowitz. Luraghi, Silvia. 1990. Old Hittite sentence structure. London- basis of Saussure’s coefficients sonantiques (Saussure New York: Routledge. 1878). Indirect evidence for these phonemes had been Luraghi, Silvia. 1997. Hittite. Munich–Newcastle: Lincom found in the other languages, but only Hittite preserves Europa. two of them as phonemes, also providing evidence for Melchert, H. Craig. 1994. Anatolian historical phonology. their consonantic character. It is nowadays commonly Amsterdam: Rodopi. Saussure, Ferdinand de. 1878. Mémoire sur le système primitif thought that the PIE phoneme inventory included three des voyelles dans les langues indo-européennes. Leipzig: laryngeals, conventionally written as */h1/, */h2/, and Teubner. */h3/. While in the other Indo-European languages Studien zu den Bogazköy Texten, Wiesbaden: Harrassowitz, laryngeals only affected the following or preceding 1965. vowel, in Hittite reflexes of */h2/ and */h3/ are found as Sturtevant, Edgard A. 1933. A comparative grammar of the Hittite language. Philadelphia: Linguistic Society of America. -h-. Reflexes of laryngeals are found in the following Wackernagel, Jacob. 1892. Über ein Gesetz der indoger- examples: Hitt. harki ‘white’ (Latin argentum); Hitt. manischen Wortstellung. Indogermanische Forschungen 1. hastai ‘bones’ (Ancient Greek osteon); Hitt. hulana 333–436. ‘wool’ (Sanskrit urna); and Hitt. huis- ‘to live’ (Latin SILVIA LURAGHI Hjelmslev, Louis Among post-Saussurrean European structuralists, Louis matter, and system and use. The son of a mathematician, Hjelmslev is the one who took the most extreme posi- he always denied any influence from his father for his tions with respect to the separation between form and interest in abstract structures: in fact, the son’s attempt at 470  HJELMSLEV, LOUIS creating a system independent of experience was criti- founded five years before. The first result of collabo- cized even by the father, who studied the relation ration among the Danish scholars was presented at the between geometry and practical experience. international conference on phonetic sciences in In spite of his denial of the value of experience, London in 1935, where Hjelmslev, together with Hjelmslev had a deep knowledge of numerous lan- P. Lier and H. Uldall, illustrated a new phonological guages, in some of which he was also fluent. theory, called ‘phonematics’. Shortly thereafter, in According to an often quoted anecdote, he used to pay December 1935, Hjelmslev and Uldall presented a his parent’s maid in order to give her Italian lessons. unified theory of phonology and grammar, called Later on, his interest for language diversity led him to ‘glossematics’, which aimed at the study of mutual write some insightful pages on the purposes of lan- relations between phonemic and grammatical systems. guage typology in the last chapter of his book Sproget In spite of the ambitious program put forth by its (1963), and to explore the grammar of a number of founders, the theory of glossematics was never fully exotic languages (for example in his ‘La catégorie des developed. In its final shape, glossematics should have cas’, published in 1935 and reprinted in the ‘Essais developed a metalanguage, capable of describing all linguistiques’). existing languages in algebraic terms. Since all lin- Hjelmslev illustrated Saussure’s claim that lan- guistic elements must be described only in terms of the guage is formed by working out a model of language relations holding between one another, a typology of based on the co-occurrence of two elements: expres- possible relations needed to be set up prior to the study sion and content. Each element consists of form and of languages. In his Prolegomena to a theory of lan- substance, e.g. sound is the substance of utterances, guage, Hjelmslev describes the possible types of rela- but their form is articulated in phonemes. Although for tions. Crucial to the understanding of his classification Saussure, neither form nor substance is independent of is the notion of ‘function’. A function is a relation the other, and it is the conjunction of two forms, form between linguistic entities; since they are considered of expression and form of content, which generates a only for the sake of functions, concrete linguistic enti- linguistic sign, Hjelmslev’s treatment clearly implies a ties are called ‘functives’. In the case of (superordi- logical priority of form over substance, contrary to nate) functions holding between other functions, the Saussure’s view. Note that the interdependence of latter can become functives. In his Prolegomena to a form and substance reveals an Aristotelian view of theory of language, Hjelmslev defines three types of matter: as Hjelmslev explained in his Prolegomena to function: (a) interdependence, where A implies B and a theory of language, substance can be studied only B implies A; (b) determination, where A implies B, but when molded by form, and beyond form it consists of B does not imply A; and (c) constellation, where A an unanalyzable ‘purport’. does not imply B and B does not imply A. It is impor- On account of his interest being centered only on tant to stress that this use of the term function differs structured, Hjelmslev took a completely different path significantly from the way in which the same term was from the one taken by Saussure’s own pupils, as Bailly defined by the Prague Circle, which is certainly more and Séchehay, who pursued a ‘linguistique de la parole’, familiar. One of the basic tenets of the Prague Circle thus aiming at the description of language usage. He was was that language must be investigated in its nature of also critical of the views held on phonology by members ‘functional system’, whereby ‘functional’ refers to the of the Prague Circle, because they paid too much atten- communicating function of language, with an implica- tion to the phonetic substance of phonemes. His major tion of teleology. In the theory of glossematics, on the emphasis was on the nature of a possible theory of lan- other hand, there is no space for teleology, and no guage, which should be arbitrary and adequate, and also intrinsic interest toward communication. aim at simplicity (in the case that two available explana- According to the promoters of glossematics, a sec- tions yield the same results, the one that requires less ond stage in the development of the theory should be complex description must be preferred). based on the scrutiny of the largest possible number of As is often the case in the history of linguistics, languages, in order to assess what types of grammati- Hjelmslev’s thought found a fertile soil in his cultural cal elements can occur in language. After the metalan- environment, and he could develop his ideas in close guage had been worked out completely, it could be collaboration with a number of other scholars, who applied to the description of any given language. shared his beliefs about the strictly formal and syn- The theoretical foundations of glossematics, which chronic nature of languages as systems. The outcome should have been published by Hjelmslev and Uldall, of scientific intercourse was the creation of the never saw the light because of disagreement between Linguistic Circle of Copenhagen in 1931, on the the two, and the only published part, the introduction, model of the Prague Linguistic Circle, which had been was the work of Uldall alone (Uldall, 1957). The only 471 HJELMSLEV, LOUIS thorough description of a language following the the- the Linguistic circle of Copenhagen, he also founded ory of glossematics is Togeby’s Structure immanente the journal Acta Linguistica together with Viggo de la langue française. Although Hjelmslev’s own Brøndal. He died in Copenhagen on May 30, 1965. focus had been primarily on phonology and morphol- ogy, Togeby’s book shows that glossematics was by no References means limited to these parts of grammar, and takes the text as its starting point; it also offers a discussion of Hjelmslev, Louis. 1953. Prolegomena to a theory of language. syntax, consisting of a description of subordinate Baltimore: Waverly Press (first published 1943). ———. 1959. Essais Linguistiques. Travaux du Cercle clauses and VP and NP structure. Linguistique de Copenhague, Vol. 12. ———. 1963. Sproget. En introduktion. Charlottenlund: The Biography Nature Method Center. ———. 1973. Essais Linguistiques II. Travaux du Cercle Louis Hjelmslev was born in Copenhagen on October Linguistique de Copenhague, Vol. 14. 3, 1899, as Louis Trolle Petersen (the name Hjelmslev Lepschy, Giulio. 1996. Twentieth century linguistics, Vol. 5, History of linguistics, ed. by G. Lepschy. London: Longman. was added in 1903 by his father, after the village Malmberg, Bertil. 1983. Analyse du langage au XXe siècle. where he was born). From 1917 to1923, he studied in Théories et méthodes. Paris: P.U.F. Copenhagen (Master of Arts), where in 1920 he Siertsema, B. 1955. A study of glossematics. The Hague: Mouton. received an award for work on Oscan inscriptions. He Spang-Hanssen, H. 1961. Glossematics. Trends in European and pursued further studies in Prague (1923–1924) and in American linguistics 1930–1960, ed. by C. Mohrmann, A. Sommerfelt, and J. Whatmough. Utrecht-Antwerp: Spectrum. Paris (1926–1927), as a pupil of Meillet and Vendryes. Togeby, K. 1965. Structure immanente de la langue française, His first book, Principes de grammaire générale, was 2nd edition. Paris: Larousse. rejected as a dissertation because of its synchronic Togeby, K. (ed.) 1967. La glossématique. L’héritage de approach. He received his doctorate degree from the Hjelmslev au Danemark. Langages 6. university of Copenhagen in 1932 with a volume of Uldall, H. J. 1957. Outline of glossematics. General theory. Travaux du Cercle Linguistique de Copenhague 10. studies in Baltic linguistics. From 1934 to 1937, he SILVIA LURAGHI was professor in Aarhus, then in Copenhagen (chair for comparative linguistics). Among the founders of See also Saussure, Ferdinand de; Structuralism Hmong and Hmong-Mien (Miao-Yao) languages Hmong-Mien is a language family and broad cultural sometimes used to refer specifically to the Hmong out- grouping of people primarily of China, some groups side of China, while in western linguistics literature, having emigrated over the last two centuries to Laos, ‘Hmong-Mien’ is becoming the more generally Thailand, Vietnam, and Burma, and in the last 30–40 accepted term for the language family as a whole. years as refugees to France, French Guyana, Australia, Figure 1 shows the genealogical relationships gen- and the United States of America. There are widely erally accepted for the language family, together with differing uses of terms designating various languages, some of the more common alternative language or dialects and cultural groups within this broad group- dialect names. These languages or dialects are for the ing—from conflicting uses of everyday terms such as most part mutually unintelligible, and each subdivides ‘Blue/Green Mong’, ‘Flowery Miao’, etc., through to further into a continuum of distinct varieties. Hmong- the inclusion of different sets of people and/or lan- Mien is typically classified as a Sino-Tibetan language guages under the term ‘Hmong-Mien’ itself, and an group by Chinese linguists, while western scholarship, alternation between this term and ‘Miao-Yao’. The lat- debating possible affiliations rather with Austro- ter term is generally preferred in China, while many Asiatic, tend to leave the question as unresolved to Hmong outside of China reject the term ‘Miao’ as date. offensive (due to association variously with wild There are around 321,000 speakers of Bunu lan- plants as a metaphor for barbarians, or, by an appar- guages, predominantly Bu-Nao. Note that Bunu speak- ently orthographic route, cats). Two consequences of ers are culturally Mien (Yao), and at least another this are that with regard to the people, ‘Hmong’ is 180,000 Bunu people speak languages other than Bunu. 472  HMONG AND HMONG-MIEN (MIAO-YAO) LANGUAGES Hmong-Mien Hmongic Bunu Chuanqiandian Qiandong (Black Miao) Xiangxi (Red Miao) Ho Nte Younuo Mong Njua Northern Western She Wunai Guiyang Eastern Eastern Bu-Nao Huishui Southern Baheng (Red Yao) Mashan Jiongnai Northeastern Dian (A-Hmao) Hmong Daw (White Hmong) Chonganjiang Luopohe Mienic Biao-Jiao Mian-Jin Zaomin Biao-Jiao Mien Biao Mien Ba Pai IuMien KimMun Figure 1. Hmong-Mien family tree. Qiandong speakers number around 1.4 million, mostly as prenasalization, preglottalization, and aspiration. the Northern dialect, while the vast majority of For instance, Baheng includes a consonant series /m Xiangxi’s 770,000 speakers speak the Western dialect, mh mj mhj/, while Xiangxi Hmong consonants and only a very small number of cultural She are also include /t th nt nth/. speakers of She. Chuanqiandian Hmong, something of Although CV is the basic syllable structure, some a repository category for languages that do not fit easi- languages also allow final nasals. In Northern ly into other subfamilies, boasts 3.6 million speakers, Qiandong Hmong, for instance, the vowels /a/ and /o/ who constitute the bulk of Hmong living outside China. may be followed by a nasal /aŋ, oŋ/, while in Mong Of these, Hmong Daw and Mong Njua (with around 1.4 Njua, nasals always follow a long (and nasalized) million speakers between them) are the primary lan- vowel, as in the language name ‘Mong’ /mɔ˜ŋ/. guages of the ex-refugee diaspora in the west (well over 100,000, mostly in the United States of America). Mong Njua is the most common Chuanqiandian Hmong lan- Lexical Tone guage in China, with around 1 million speakers, fol- Lexical tone refers to the relative pitch or pitch con- lowed by Northeastern Dian with 200,000. Of the 1.7 tour of a given word, when it is part of how that word million speakers of Mienic languages, 1.3 million speak and its meaning are distinguished. A change in pitch Iu Mien. signals a different word, in the same way as a change in a consonant or vowel would change the word. In Baheng, for instance, the word /ko33/ (pronounced Typological Characteristics with a mid level tone) means ‘nine’, while /ko22/ (mid- Hmong-Mien languages share a range of characteris- low level tone) means ‘road’. tics, many of which are also evident in other languages Hmong-Mien languages have variously developed of the Southeast Asian region. The discussion below seven, eight, or possibly as many as 13 lexical tones, focuses primarily on Chuanqiandian Hmong lan- all of which can be traced back to four original tones guages, in particular Hmong Daw, to exemplify some in the reconstructed ancestor language (known as of the language structures typical of this family. Proto-Hmong-Mien). Most languages also display (Examples are from Hmong Daw except where other- tone sandhi—that is, a change in the tone of a given wise specified. For notes on the RPA writing system word under specific circumstances, typically in used in Hmong Daw and Mong Njua examples, see the response to the tones of surrounding words. Qiandong section on Writing below. Examples in slash brackets Hmong is a notable exception. are notated in the International Phonetic Alphabet.) Tendency Toward a Monosyllabic Word Structure Syllable Structure of C(onsonant) V(owel) This characteristic is necessarily associated with Hmong-Mien languages typically have a large assem- certain principles of grammatical structure and word bly of initial consonants, incorporating features such formation. 473 HMONG AND HMONG-MIEN (MIAO-YAO) LANGUAGES Firstly, the words of Hmong-Mien languages do not Two- and Four-syllable Expressives accrue prefixes or suffixes to indicate meanings such These poetic phrases, including but by no means as plurality, possession, case, or the comparative. restricted to onomatopoeia, display a distinctive, inter- Rather, where such meanings are overtly present, they nally repetitive form, and are scattered throughout are expressed by additional words in the sentence, or everyday conversation, as well as in more literary by word order. These two strategies, respectively, are styles: demonstrated for the possessive in Example 1: Example 4 Example 1 plib pleb (of wood crackling) Daim dawb yog kuv li dig dug (of the ponderous bubbling of thick liquid) CLassifier white be I POSSessive kev noj kev haus nyuaj heev ‘The white one is mine’ way eat way drink difficult EMPHatic Kuv tsis xav kom kuv tus ntxhais chim ‘Life is very difficult’ I NEGative want CAUSative I CL daughter upset ‘I don’t want to upset my daughter’ Serial Verb Construction Hmong-Mien languages feature strings of verbs or Secondly, the primary strategy for word formation verb phrases, which cumulatively express a complex is compounding. This is evident both in long-estab- verbal meaning—representing, for instance, an analyt- lished words in the languages, and in new words ic breakdown of minutely differentiated components developed in the context of immigration to the west: of an action, or different simultaneous roles of partic- ipants in the event. Example 2 In the following example, the three verbs npuav, qab qab tsev kawm ntawv nqa, and dua are combined to describe the complex sweet sweet house study letters action of carrying something somewhere. Also note ‘delicious’ ‘school’ that the action of picking up the cubs must first be hnub chiv nub hli (Mong Njua) specified by yet another verb: day begin day moon ‘Monday’ ‘Monday’ Example 5 (two of several alternatives current in the United States Ob tus dais muab cov menyuam of America) two CL bear take CL child npuav nqa dua nram kwj-hav carry carry pass down valley Noun Classifiers by mouth This is a set of words that organize nouns into classes, ‘The two bears picked up their cubs and carried them down such as ‘round things’, ‘animate things’, etc. Their into the valley’ presence or absence with a noun in a given sentence depends on the function they serve—or under certain Reliance on Context to Determine Functions Such conditions, they may stand in place of the noun. as Tense and Plurality Classifier functions include clarifying the meaning of A range of grammatical and pragmatic functions are the noun, in a way analogous to the use of radicals in not obligatorily specified in languages of this family. Chinese characters, indicating definite rather than In the first of the following examples, the tense is nonspecific reference, or they may be associated with implied by the aspect marker tau (attainment), while in enumeration of the noun: the second, it can be understood only in context: Example 3 Example 6 Specifying meaning: daim ntawv tsab ntawv Yuav tau ib lub hnab-looj-tes CL (flat) paper CL (written) paper buy ASPect one CL gloves ‘piece of paper’ ‘letter’ ‘I bought a pair of gloves’ With numeral: Thov ib rab diav Kuv yuav lub hnab-looj-tes please one CL (tools) spoon I buy CL gloves ‘Please could I have a spoon?’ ‘I am buying/bought/will buy some gloves’ In place of the noun: Puas muaj tus pab koj ua? Also note also the omission of the pronoun QUestion have CL (animate) help you do in the first example, and the assumed plurality of ‘Is there someone to help you?’ hnab-looj-tes. 474  HMONG AND HMONG-MIEN (MIAO-YAO) LANGUAGES Concluding Remarks Other tones are written as: The above sample characteristics of Hmong-Mien lan- guages are not a random combination. Rather, certain Example 8 characteristics, such as underspecification, classifiers, and serial verb constructions, co-occur in linguistical- m low glottalized j high falling v mid rising d low rising, lengthened ly logical and historically traceable combinations. s mid-low falling g low breathy This, in part, explains the commonality of these and (zero) mid level other characteristics to not only this language family but also others of the region. Other aspects of Hmong- The Pahawh Hmong, although restricted in use for Mien languages are specific to that family—for the most part to small communities located throughout instance, the ordering Possessor–Possessed illustrated Australia, the United States of America and in north- in Example 1 above. ern Thailand, is recognized by many as a significant symbol of Hmong autonomy and, particularly for its proponents, as a realization of the millenarianist Writing prophetic tradition threaded throughout the dab neeg. Writing and orthography are a key issue for many The Pahawh is based on the CV syllable structure of Hmong-Mien groups, and a good number of writing Hmong, so that each syllable requires only one conso- systems have been proposed and to varying extents nant and one vowel grapheme. A limited set of diacrit- have been implemented—either from within the com- ics combines with each grapheme type, to (i) form a munities, or by missionary linguists working with the vowel-tone composite graph, or (ii) specify one of three communities, or by government organizations of the consonants possible for a given grapheme—reducing various countries of residence. The question is not only the number of graphemes that would otherwise be one of practical literacy in a heritage language, but is needed for 59 consonants (in Hmong Daw). Example 9 also, at least for the Hmong, embedded in cultural nar- is in Pahawh Version Two, as current in major Pahawh ratives of national identity and mythic histories, or dab teaching organizations. (Version Three, also found in neeg. In these dab neeg are encoded a Hmong history older and linguistics texts, is not fully readable on the of oppression, migration, and ancient literacy, which is basis of this Version.) Note that, although the writing is in turn linked to strong intragroup cohesiveness and from left to right, the order within a word is from right long-cherished goals of self-determination. to left (Vowel+tone–Consonant): Writing systems have been developed on a base of Chinese (e.g. Xiangxi and Qiandong Hmong), Thai, Example 9 and Lao scripts, as well as the Roman—Vietnamese Pahawh base, Pinyin base (e.g. Mien languages), and Roman RPA Phaj hauj Hmoob Popular Alphabet (RPA—Hmong Daw and Mong ‘Pahawh Hmong’ Njua). In addition, there are at least four independent systems: Pollard script (Northeastern Dian), Pahawh This example demonstrates the use of zero diacritic Hmong (Hmong Daw and Mong Njua), Ntawv Paj , , representation of the same tone by different Ntaub (Hmong Daw and Mong Njua, used only in a diacritics depending on the base vowel grapheme small area of California), and the Ntawv Puaj Txwm , , and similarly use of the same diacritic for (an apparently ancient script). different consonants, depending on the base consonant The most widespread writing system now in use is grapheme , . Many features of Pahawh writing, the RPA. This is based on a featural representation of including word-internal ordering and the shapes of the the sounds of Hmong, and utilizes consonant symbols graphemes themselves, hold symbolic significance in syllable-finally to indicate tone. The word ‘Hmong the tradition of its emergence. Daw’/mɔ˜ŋ55 da˝55/ is written <Hmoob Dawb>: The Hmong who have resettled in the west after the refugee period have taken advantage of the situation by embarking on a new period of accelerated lan- Example 7 guage development. This includes not only ortho- hm oo b d graphic development but also ongoing creation of a preaspirated or long vowel high level alveolar stop wealth of new words, in order to ensure the useabili- voiceless nasal (nasalised) tone aw b ty of the heritage language(s) in all the domains now diphthong high level encountered by the rising generation in the cultures tone of immigration. A third area of development has (Note that the final nasal of ‘Hmong’ is not directly been a surge of publications in Hmong (Hmong Daw notated, being a predictable closure of the long vowel.) and Mong Njua)—language teaching and reference 475 HMONG AND HMONG-MIEN (MIAO-YAO) LANGUAGES materials including dictionaries and literacy primers, and language maintenance, ed. by David Bradley and Maya books on culture and history both traditional and Bradley. London and New York: RoutledgeCurzon. Enwall, Joakim. 1994. A myth become reality: history and devel- emerging, community magazines, novels, and a rapid opment of the Miao written language, Vol. 1. Stockholm: proliferation of websites and related communications Institute of Oriental languages, Stockholm University. media. Grimes, Barbara. (ed.) 2000. Ethnologue, Vol. 1: Languages of the world, 14th edition. Dallas: SIL International. References Mottin, Jean. 1978. Eléments de grammaire Hmong Blanc. Bangkok: Don Bosco Press. Bisang, Walter. 1993. Classifiers, quantifiers and class nouns in Ratliff, Martha. 1992. Meaningful tone: a study of tonal mor- Hmong. Studies in Language 17(1). 1–51. phology in compounds, form classes, and expressive phras- Bisang, Walter. 1996. Areal typology and grammaticalization: es in White Hmong. Northern Illinois: Center for Southeast processes of grammaticalization based on nouns and verbs Asian Studies, Northern Illinois University. in east and mainland South East Asian languages. Studies in Smalley, William, Chia Koua Vang, and Gnia Yee Yang. 1990. Language 20(3). 519–97. Mother of writing: the origin and development of a Hmong Court, Christopher. 1986. Fundamentals of Iu Mien (Yao) messianic script. Chicago: University of Chicago Press. grammar. Ph.D. Thesis. Berkeley: University of California. Tapp, Nicholas. 1989. Sovereignty and rebellion: the White Edmondson, Jerold. 1992. Pa-hng development and diversity. Hmong of northern Thailand. Singapore: Oxford University Papers from the first annual meeting of the Southeast Asian Press. Linguistics Society 1991, ed. by Martha Ratliff and Eric Wang, Fushi. 1985. Miaoyu jianzhi (Sketch of the Miao lan- Schiller. Tempe, AZ: Program for Southeast Asian Studies, guage). Beijing: Renmin Chubanshe. Arizona State University. CHRISTINA EIRA Eira, Christina. 2002. Language maintenance at the micro level: Hmong ex-refugee communities. Language endangerment See also China Hockett, Charles F. Charles F. Hockett, one of the leading neo- research material on Menomini, an American Indian Bloomfieldian linguists in the United States, authored language, and compiling an anthology of his most more than 130 publications covering a broad spec- important publications. trum: general linguistics, phonology, morphology, In the 15 years between Bloomfield’s death and semantics, sound change, historical linguistics, lin- Noam Chomsky’s transformational grammar, Hockett guistic universals, the origin of language, animal com- was one of the leaders of American linguistics known munication, Chinese, Fijian, and analyses of such especially for his role in the development of American American Indian languages as Potawatomi, Ojibwa, phonological theory. As a post-Bloomfieldian struc- and Arapaho. As a student of Edward Sapir and turalist, Hockett was convinced that, as he later said, Leonard Bloomfield, Hockett was trained both as a linguistic analysis must be done ‘from the bottom up’ linguist and an anthropologist and never wavered from beginning with the sound system, progressing to the his conviction that the disciplines are inseparably grammatical level of inflection and word formation, intertwined. At the height of his remarkable career in and only then to the syntactic level, the rules govern- the late 1950s, Hockett’s behaviorist view of language ing the arrangement of words in the sentence. Each of and structuralist approach to linguistic analysis was these levels of language was to be kept strictly sepa- eclipsed by transformational grammar, Noam rate. This meant that grammatical considerations Chomsky’s revolutionary linguistic theory that could not be used to determine the distinctive sounds attacked the very foundations of structuralism: behav- of a language, the phonemes. European phonologists iorism and the neglect of semantics. Hockett rejected of the Prague school viewed the phoneme as consist- the validity of transformational grammar and contin- ing of a set of features that could partially overlap ued publishing spirited rebuttals in defense of his own with those of another phoneme, so that the distinction views on the nature of language for two decades. As between two phonemes could be neutralized in cer- the literary executor of his mentor, Leonard tain positions. The phonemes /p/ and /b/ are distin- Bloomfield, he devoted much time and energy pub- guished by the feature of voicing, /p/ being voiceless, lishing posthumously the bulk of Bloomfield’s /b/ being voiced. If, as is the case in German, /b/ is 476  HOCKETT, CHARLES F. always pronounced as /p/ at the end of a word, the Indian and Austronesian languages, and his role as contrast between these two phonemes is neutralized Leonard Bloomfield’s literary executor. in the word-final position. American structuralists, Hockett foremost among them, considered each Biography phoneme to be unique and rejected partial phonemic overlap or neutralization outright. Together with the Charles Hockett was born on January 17, 1916 in strict separation of linguistic levels, the requirement Columbus, Ohio. He entered Ohio State University at ‘once a phoneme, always a phoneme’ became a defin- 16 years of age. He received a combined B.A. and ing characteristic of American structuralism and a M.A. degree in ancient history in 1936. He worked as mark of its theoretical rigidity. One of Hockett’s rea- graduate student in anthropology at Yale with the sons for rejecting transformational grammar was the anthropologist/linguist Edward Sapir and the struc- ‘top-down’ approach to linguistic analysis beginning turalist linguist Leonard Bloomfield. Hockett received with syntax and semantics rather than phonology. his Ph.D. in anthropology in 1939. He was drafted dur- According to Hockett, linguistic analysis from the ing Word War II and wrote an elementary Chinese text top-down is inappropriate because the investigator book for the War Department. Hockett joined the lin- has to isolate words or morphemes before work on guistics faculty at Cornell University in 1946 as an syntax and semantics can begin. assistant professor and was promoted to Professor of At the peak of his career, Hockett was perhaps Linguistics and Anthropology in 1956. In 1955–1956, best known for his textbook A course in modern lin- he received a fellowship at the Center for Advanced guistics (1958), which, as the author states in the intro- Studies in the Behavioral Sciences at Palo Alto, duction, was not written to explore frontiers and California. He was president of the Linguistics Society indulge in flights of fancy, but to present the gen- of America, 1964–1965 and became Goldwin Smith erally accepted facts and principles of the field. He Professor of Anthropology and Linguistics in 1970. In also included chapters on dialectology, historical 1974, he was elected to the National Academy of linguistics, and man’s place in nature. He compared Sciences. After retirement in 1982, he served as an animal communication systems, for example, the adjunct professor at Rice University. He died on dance of honey bees, with human language by means November 3, 2000 in Ithaca, NY. of seven key properties. The property that, according to Hockett, is unique to human language is duality References of patterning, a term Hockett invented to account for the fact that an infinite number of meaningful Agard, Frederik B., et al. (eds.) 1983. Essays in honor of forms are built from a small number of meaningless Charles F. Hockett. Leiden: Brill. Chevillet, Francois. 1996. The phoneme and American linguis- elements. tics: an interview with C.F. Hockett. Etitudes Anglaises 49. Hockett’s monograph The state of the art (1968) is 180–91. a spirited attack on the basic principles underlying Fillmore, Charles J. 1969. Review of the state of the art. transformational grammar. As a behaviorist who American Anthropologist 71. 711–3. insisted that only observable utterances could be the Harris, Randy Allen. 1993. The linguistics wars. Oxford: Oxford University Press. subject of linguistic analysis, he questioned Hocket, Charles F. 1942. A system of descriptive phonology. Chomsky’s distinction between a speaker’s linguistic Language 18. 3–21. competence, the innate knowledge of the rules of his ———. 1954. Two models of grammatical description. Word or her language, and performance, the speaker’s actu- 10. 210–34. al utterance, which may be flawed by mispronuncia- ———. 1955. A manual of phonology. Indiana University Publications in Linguistics and Anthropology, Memoir 11. tions, inadvertent errors, repetitions, and the like. For ———. 1958. A course on modern linguistics. New York: Hockett, performance was the subject of linguistic Macmillan. investigation; for Chomsky, it was competence. He ———. 1962. Posthumous publication, ed. by L. Bloomfield. also disagreed with the Chomskyan premise that a The Menomini language, New Haven: Yale University speaker has access to an infinite number of sentences Press. ———. 1968. The state of the art. The Hague: Mouton. of infinite length and argued that there are limitations ——— (ed.) 1970. A Leonard Bloomfield anthology. to sentence length that are part of the structure of lan- Bloomington: Indiana University Press. guage itself. When the State of the Art was published, ———. 1973. Man’s place in nature. New York: McGraw- it was already clear that transformational grammar had Hill. won the day, Hockett’s spirited rearguard action ———. 1974. Posthumous publication, ed. by L. Bloomfield. Menomini lexicon. Milwaukee Public Museum Publications notwithstanding. in Anthropology and History, no. 3. Hockett’s most significant and lasting contributions ———.1977. The view from language. Selected essays are, perhaps, his many publications on American 1948–1974. Athens: University of Georgia Press. 477 HOCKETT, CHARLES F. ———. 1980. Preserving the heritage. First person singular, ed. Joos, Martin. (ed.) 1957. Readings in linguistics I. The devel- by Boyd H. Davis and Ramond O’Cain. Studies in the History opment of descriptive linguistics in America 1925–56. of Linguistics, Vol. 21. Philadelphia: John Benjamins. Chicago: The University of Chicago Press. ———. 1987. Refurbishing our foundations. Elementary lin- RENATE BORN guistics from an advanced point of view. Philadelphia: John Benjamins. See also Bloomfield, Leonard; Sapir, Edward Humboldt, Wilhelm von The most distinguished linguistic scholar of the nine- framework of cognition through which one formulated teenth century, politician, diplomat, and humanist, one’s thoughts. According to Humboldt’s famous dic- Karl Wilhelm Freiherr von Humboldt is generally con- tum, language in its very essence is not a product sidered to be the founder of general linguistics as a sci- (ergon) but a process and activity (energeia) inherent to ence that aims at the quest for the essence of linguistic human nature, an organ that forms the thought. Its true phenomenon itself through the analysis and compari- definition can therefore only be a genetic one, with the son of different languages. Together with his younger ‘form of language’ as largely programmed by the inter- brother, Alexander von Humboldt, a natural scientist, nal disposition of a genetic program. It is the ‘inner he was given privileged education in classical studies form’, the pattern, or structure, of grammar and mean- by excellent private tutors that not only resulted in his ing that is imposed on the ‘outer form’, the sounds profound and life-long interest in languages, philoso- from which different languages are differentiated one phy, and humanities in general, but also motivated his from another. This repeated work of the mind enables later work on reforms of the Prussian educational sys- the articulated sounds to express thought, which makes tem and the foundation of the University of Berlin. In an inseparable unity between language and thought and addition to Greek and Latin, he studied Sanskrit and language and sounds. In accord with his theory, other Indo-European and non-Indo-European lan- Humboldt holds that language acquisition does not guages such as the ancient Indonesian language Kawi occur through mechanical learning but is a matter of and Basque, while his brother Alexander, a pioneering development of cognitive capacities under appropriate explorer of the Americas, often sent back to him data external conditions, just as all main human capacities collected from Amerindian languages with which he develop at certain definite growth stages of life. came into contact. Postulating that reason is inseparable from lan- As a contemporary of German philosophers guage, thought is inseparable from speech, Humboldt Immanuel Kant, Friedrich Hegel, and Johann infers, in accordance with his dynamic view of lan- Gottfried von Herder, and the Romantic writers guage as activity, that languages are not really means Wolfgang Goethe and Friedrich von Schiller, for representing already known truths but are rather Humboldt synthesized the main ideas of philosophical instruments for discovering previously unrecognized idealism and the Romantic emphasis on free, individ- ones. However, from his metaphysical concept of the ual creativity and subjective meaning into his philoso- ‘inner form’, he further concludes that thought does phy of language. He outlined his full theory in a not depend only on the universal human capacity of lengthy introduction to his life work on the Jawan lan- speech but also on the different realizations of this guage Kawi, Über die Verschiedenheit des men- ‘inner form’ in different languages. This blend of rela- schlichen Sprachbaues und ihren Einfluss auf die tivistic concept of the language role in intellectual geistige Entwicklung des Menschlichengeschlechts processes and the idealistic, universalistic interpreta- (1836; On language: the diversity of human tion of the essence of the linguistic phenomenon by language—structure and its influence on the mental means of conceptual categories, represents the theoret- development of mankind). Based on the Kantian epis- ical basis of Humboldt’s deterministic doctrine on lan- temological stance of cognition as activity, through guage as a worldview (Weltanschauung theory). which mental categories were imposed upon sensible Emphasizing the difficulties of full understanding in experience, Humboldt held that language was not the process of communication and translation from only a means of communication but also an a priori one language into another, Humboldt indicates that 478  HUMBOLDT, WILHELM VON ‘each language sets certain limits to the spirit of those grammar elaborated by one of the most influential who speak it; it assumes a direction and, by doing so, modern linguists, Noam Chomsky. excludes many others’. Moreover, language acts not only as an intermediary between man and the world but also between the individuals in a community and Biography the whole humankind, so that in social terms, differ- Wilhelm von Humboldt was born in Potsdam, Prussia ences in individual linguistic worldviews lead to dif- on June 22, 1767. He studied law at the University of ferences in group worldviews. Following the views of Brandenburg in Frankfurt (Oder) in 1787, moved to Herder, Humboldt maintains that each language and Göttingen University where he also began studies in each culture reflect the world in a particular way and classical philology in 1788. He joined Prussian civil that the mental quality of each nation and its culture service as a law clerk to the Supreme Court of Berlin determines the language of its people, which in its turn in 1789 and resigned from civil service in 1790. He determines the way they think and experience reality. resided on family estates in Thuringia and Jena and Analyzing a large number of world languages by a there he continued his studies with private tutors, comparative method, Humboldt also made original devoted his interests to literature criticism and esthet- contributions to language typology. His studies on the ics, and associated with Goethe, Schiller, and the evolution of language and classification of languages, Schlegel brothers between 1790 and 1797. He moved however, were closely related to his view on interde- with his family to Paris to observe the turbulent social pendency between language and the mental spirit of a development in France in 1797–1799, and then to nation or culture. He based his work on Friedrich Spain where he investigated the Basque language. He Schlegel’s genealogical scheme, which distinguished rejoined the Prussian civil service and became an between the most developed ‘inflectional’ languages, envoy to the Vatican in Rome in 1802–1808. He was i.e. the Indo-European languages, and those that have head of the Department of Education and Arts at the no inflection and are therefore called ‘isolating’ (e.g. Home Office in Berlin, and reformed Prussian educa- Chinese) or the intermediary type of the so-called tional system by elevating standards for teachers’ train- ‘agglutinating’ languages (as Amerindian), which put ing and founded the Friedrich Wilhelm University strings of forms together, but do not allow for a modi- (modern Humboldt University) in Berlin, 1809. In fication of the root. Somewhat controversially, 1809–1812, he retired to private life. He re-entered Humboldt believed that these differences in language diplomatic service as the Prussian ambassador in types reflected the developmental stages of human Vienna and later London in 1812–1819. He resigned intellect, claiming that the highest achievement of the from civil service in 1819 and continued his private human mind was that of the speakers of highly inflec- philological studies in the quiet atmosphere of the fam- tional Ancient Greek. Although he did not relate these ily manor in Tegel, investigating the Basque, Sanskrit, types to historical progress or developmental stages of American, and Oriental languages and writing his most culture, he has been sometimes criticized for implying important linguistic works on the philosophy of lan- the idea of superiority of one people over another, guage. He died in Tegel, Prussia on April 8, 1835. based on the differences of language structure. Humboldt’s complex and rich ideas on dynamic lan- guage concept, ‘inner form’, and linguistic worldview References have continuously influenced the subsequent course of Aarsleff, Hans W. 1988. Introduction. On language—the diver- linguistics in both Europe and the United States. In sity of human language—structure and its influence on the Europe, his legacy is evident not only in the work of his mental development of mankind. New York: Cambridge University Press. immediate successors, such as Franz Bopp, Heymann Basilius, Harold. 1968. Neo-Humboldtian ethnolinguistics. Steinthal, and the forerunner of psycholinguistics Readings in the sociology of language, ed. by Joshua Wilhelm Wundt, and of the later group of the so-called Fishman. The Hague: Mouton. Neo-Humboldtians, including Leo Weisgerber and Jost Brown, Roger L. 1967. Wilhelm von Humboldt’s conception of Trier, who advocated the majority of Humboldt’s theo- linguistic relativity. The Hague and Paris: Mouton. Cowan, Marianne (ed.) 1963. Humanist without portfolio: an retical ideas, but also in some basic concepts of anthology of the writings of Wilhelm von Humboldt. Ferdinand de Saussure’s structural linguistics. In Detroit:Wayne State University Press. American linguistics, the essentials of his linguistic rel- Humboldt, Wilhelm von. 1821. Über den Einfluss des ver- ativity and Weltanschauung theory found their clear schiedenen Charakters der Sprachen auf Literatur und expression in the works of the founders of linguistic Geistesbildung; Linguistic variability & intellectual develop- ment, transl. by George C. Buck and Frithjof A. Raven, 1971. anthropology Franz Boas, Edward Sapir, and Benjamin ———. 1836. Über die Verschiedenheit des menschlichen L. Whorf, while his distinction of inner and outer lin- Sprachbaues undihren Einfluss auf die geistige Entwicklung guistic form represents the basic notions of generative des Menschengeschlechts. Berlin: Dümmler; as Linguistic 479 HUMBOLDT, WILHELM VON variability and intellectual development, transl. by George Wissenschaften, 17 vols., Berlin: Behr; Berlin: De Gruyter, C. Buck and Frithjof A. Raven, Coral Gables, FL: University 1967–1968. of Miami Press, 1971; as On language: the diversity of ———. 1963. Humanist without portfolio: an anthology of the human language and its influence on the mental develop- writings of Wilhelm von Humboldt, ed. and transl. by ment of mankind, transl. by Peter Heath, Cambridge: Marianne Cowan. Detroit: Wayne State University Press. Cambridge University Press, 1988; as On language: on the ———. 1997. Wilhelm von Humboldt: essays on language, ed. diversity of human language construction and its influence by Theo Harden and Dan Farelly, transl. by John Wieczorek on the mental development of the human species and Ian Roe. Frankfurt and New York: Peter Lang. (Cambridge texts in the history of philosophy), ed. by Manchester, Martin L. 1985. The philosophical foundations Michael Losonsky, transl. by Peter Heath, Cambridge: of Humboldt’s linguistic doctrines. Amsterdam and Cambridge University Press, 1999. Philadelphia: Benjamins. ———. 1903–1936. Gesammelte Schriften (Collective works), Sweet, Paul R. 1978. Wilhelm von Humboldt: a biography, ed. by von Albert Leitzmann, Bruno Gebhart and Wilhelm 2 vols. Columbus, OH: Ohio State University Press. Richter, Königlichen Preussischen Akademie der ANITA SUJOLDZIC Hungarian and Ugric Languages Hungarian together with Mansi (Vogul) and Hanty Arpad completed their ‘settlement’ (honfoglalas) by (Ostyak) belongs to the Ugric group of the Finno- AD 896. In the fifteenth century, the flourishing peri- Ugric language family. In its turn, Ugric group is sub- od of national formation was interrupted by the divided into Hungarian subgroup and Ob’-Ugrian Turkish invasion and then the Austrian rule. Only subgroup, containing Mansi and Hanty. Hungarian is since 1918, after the collapse of the the official language of the Hungarian Republic Austrian–Hungarian Empire, did Hungarian begin to (Magyar Nepkoztarsasag). The Hungarian language develop without strong foreign influences. These (magyar nelv) is spoken by Hungarians or Magyar, as numerous language contacts influenced modern they call themselves. In other languages, they have dif- Hungarian enormously, leading to the gap between ferent names, e.g. German ‘Ungarisch’, Ukrainian Hungarian and Ob’-Ugrian languages (Mansi and ‘Ugor’ska’, Russian ‘Vengerskiy’, etc. A certain num- Hanty). Now, communication between a Hungarian, ber of Hungarians live in the southwest of the Ukraine Mansi, and Hanty is impossible. The oldest Hungarian (Uzhgorod area), in Romania, Slovakia, Canada, and literary relics are the ‘Funeral oration’ (Halotti Beszed, the United States of America. The total number of c. AD 1200) and ‘Lament of Mary’ (o - Maria speakers may be estimated as 12–14 million. Siralom), which is dated about a hundred years later. Hungarians are said to have arrived to their present The greatest Hungarian poets are Sandor Petofi, Janos positions around the lake of Balaton from their origi- Arany, and Endre Ady. (In Hungarian it should be Ady nal homeland in the South of Western Siberia and the Endre, because the family name comes first followed Urals. Actually, Mansi and Hanty still live in the by the first name of a person.) Hungarian uses Roman Northern Urals and Western Siberia. The Old alphabet without some letters (q, x, y) but with a sys- Hungarian tribes moved to the West from the Urals tem of diacritics for vowel length and quality. Long and lived for about a thousand years on the Volga and vowels are shown by [‘] over a letter, e.g. a. Two dots Oka river basins where they had close contacts with over [o] and [u] like o and u shows that they are pro- the Permian peoples (Komi) as well as the peoples of nounced mildly, that is, like ‘Umlaut’ in German. The the Volga region: Mari (Cheremis), Mordva, and the sign [‘] shows long vowels with umlaut, e.g. o. Many ancient Bulgars, the tribes of Turkic origin, who part- consonant sounds are given by combinations of letters, ly moved to the lands of modern Bulgaria, partly e.g. cs [t?], dz [dz], dzs [dz], gy [d’], ly [l’], ny [n’], sz merged with the Tatars, Chuvash, and partly joined the [s], ty [t’], zs [z]. Hungarian also has long consonants. ancient Hungarians. After living for some time on the The length of the consonant is denoted by doubling the Don river, the ancient Hungarian tribes moved again letter, e.g. pp, bb, dd, mm, ll, etc. It is phonemic, i.e. westward to conquer the territory of the Danube and they change the meaning of the word: hal ‘fish’ —hall Balaton basins from the Slavonic tribes of Slovaks in ‘hears’; var ‘wound’ —varr ‘(he) sews’. Long conso- the ninth century AD. The ancient Hungarian king nants built by letter combinations require doubling 480  HUNGARIAN AND UGRIC LANGUAGES only the first consonant letter, e.g. ccs, ddzs, ggy, ssz, each other, e.g. haz+am ‘my house’; ‘haz+ad ‘your nny, zzs. One can see that the letter ‘y’ is not pro- house’. The plural marker ‘-k’ is added to the conso- nounced. It shows that the preceding letter is milder, nantal stem according to the rules of vowel harmony a e.g. gy [d’], ny [n’], ly [l’], ty [t’]. haz+ak ‘the houses’; a kep+ek ‘the pictures’. Linear The phonology of Hungarian is rather simple. It has addition of suffixes is characteristic of agglutinative no diphthongs, but it has seven long vowels denoted languages (e.g. Finno-Ugric and Turkic languages). by [‘] over the vowel. In the phonetic transcription, we The Hungarian case system is rich. Nouns, pronouns, use [:], e.g. a [a:]. There are 14 vowels all in all: [i, i:, adjectives, and numerals are declined according to the u, u:, u, u:, e, e:, o, o:, o, o:, a, a:], among them eight same rules. There are 23 cases. Unlike many other are labialized [o, o:, u, u:, o, o:, u, u:]. According to the languages of the world, in Hungarian, numerals are vertical position of the tongue, three types of vowels followed by a noun in the singular, kilenc haz ‘nine are defined: high vowels: [i, i:, u, u:, u, u:], middle [e:, houses’. Verbs can be transitive and intransitive in the o, o:, o, o:], and low [e, a, a:]. According to the hori- indicative, conditional, or subjunctive [imperative zontal movement of the tongue, we define two types of moods. Definite and indefinite conjugation. vowels: front [i, i:, u, u:, e, e:, o, o:] and back [a, a:, o, Hungarian has only 600–700 words of Finno-Ugric o:, u, u:]. The length of the vowel is phonemic, i.e. it or Uralic origin. All the rest of the Hungarian words are changes the meaning of a word: lap ‘page’, lap ‘marsh, borrowings. Old Hungarian borrowed a small number bog’, etc. Like in Turkish, Tatar, and other Turkic lan- of words from Iranian (through Old Ossetian, e.g. tej guages, there is vowel harmony in Hungarian. It ‘milk’, tehen ‘cow’, hid ‘bridge’, var ‘fortress’). The means that in a Hungarian word, all vowels share a first larger layer of borrowings (about 300 words) came certain articulation pattern. They are produced with in the fifth to ninth centuries from the Turkic languages either the front or the back of the tongue, e.g. the loca- (especially Chuvash), which belong to cattle breeding, tive suffix ‘-ban’ is not changed when there is a back husbandry, crafts, house building, cloth making, public vowel in the word root:_varos ‘town’ - varosban - ‘in and spiritual life. The second Turkic word layer entered town’; a haz ‘the house’ - a haz+ban ‘in the house’; Hungarian from the Pecheneg and Polovets languages haz+hoz ‘to the house’. However, the suffixes change in the tenth to thirteenth centuries. Some part of the if the root vowel is in front, e.g. a viz ‘the water’ - a Pecheneg and Polovets joined the Old Hungarian at viz+ben ‘in the water’; kert+hez ‘to the garden’. The that time. A great number of words (more than 500) stress is always on the first syllable. If the word is were borrowed from the Slavonic languages, especial- long, then it is also on the third syllable. ly Slovak and Russian. They pertain to all spheres of The hungarian consonantal system is rich. There everyday life. German also influenced Hungarian are 50 consonants in Hungarian all in all. It is 3.6 greatly giving it about 400 words concerning crafts and times greater than the number of vowels. We can sciences. Religious, cultural, and school terms came define them by the work of the active organ of speech from Latin. Dozens of words were borrowed from (i.e. the place of articulation), the manner of articula- Italian, Romanian, or Spanish. So one can see that the tion (i.e. the type of obstruction), and the work of the major part of Hungarian is the words of non-Finno- vocal cords. According to the work of the active organ Ugric origin. of speech, we define the following consonants: labio- labial [p, p:, b, b:, m, m:], labio-dental [f, f:, v, v:] (for The Mansi Language the lack of place, we shall not show the long vowels further); front (alveolar and dental) [t, d, s, z, ?, ts, dz Mansi belongs to the Ob’-Uric subgroup of the Ugric t?, z, dz, n, l, r]; palatal (mediolingual) [t’, d’, j]; back group of the Finno-Ugric family. Anthropologically, (velar) [k, g, h], pharyngeal [h]. According to the man- they are said to belong to the Mongoloid type. Some ner of articulation, the following consonants are scholars believe them to have migrated to the North of defined: sonorant [m, n, n’, l, r, j]; occlusive [p, b, d, Siberia from the Sayan mountains of the South of d’, t, t’, k, g]; affricate [ts, dz, t?, dz]; and fricative [v, Siberia. They were called formerly Vogul, which is not s, z, ?, z, h]. According to the work of the vocal cords, correct because the people call themselves ‘Man’s’i’. we define voiced [b, v, d, d’, z, dz, z, dz, g] and voice- In 1989, there were 8,474 native Mansi, but only less [p, f, t, t’, s, ?, ts, t?, k]. As we can see further, the 37.1% of them considered Mansi as their native phonetical system of Hungarian is rather different language. May be, half of those 37.1% can actually from those of Mansi and Hanty. speak Mansi. They live on the territory of the Hanty- There is no grammatical gender in Hungarian. Its Mansi autonomous national district in Russia, which is morphology is of an agglutinative type, i.e. there are situated in the basins of the Ob’, Sos’va, and Konda no prefixes before the root; the root does not change; rivers and in the Northern Urals. There were four suffixes are added (‘glued’) to the root, and then to dialects in Mansi, whose speakers did not understand 481 HUNGARIAN AND UGRIC LANGUAGES each other. These dialects were so different that they between Northern and Konda Mansi is also impossi- could be called different languages. Actually, they ble. Actually, Konda Mansi is quite a different lan- were more different than German and English. Now, guage. This can be seen even from its phonemic only some of the Northern Mansi can speak their system. Konda Mansi has 17 (short and long) vowels native language, the rest of the Mansi dropped it and and 18 consonants. Nowadays, it is an extinct lan- began to talk Russian. Two dialects of Mansi, guage. Thus, one cannot state for sure if the long vow- Northern and Konda, are analyzed here to show how els are phonematic. There exists a strong doubt. It may different they are. Russian Orthodox missionaries tried be stress that makes Konda vowels long. to christen Mansi in the second half of the nineteenth According to the work of the active organ of century; therefore, they translated the New Testament speech, the following four groups of consonants are into Mansi. It was the first book in Mansi. Literary defined: labio-labial [p, w, m], (there are no labio-den- Mansi languages, based on the Northern dialect, was tal consonants); front [t, s, ?, l, ?, r]; palatal [t’, s’, n’, introduced into the mass of illiterate Mansi in the l’, j]; and velar [k, x, ?, ?]. According to the manner of 1930s only. Then, the script was Latin, which was articulation, the following four groups are defined: changed into Cyrillic in 1937. Modern Mansi bor- sonorant [m, n, n’, l, l’, ?, ?]; occlusive [p, t, t’, k]; rowed many words of everyday life from the language (there are no affricates); and fricative [w, s, s’, ?, x, ?]. of Siberian Tatars, who ruled Mansi before the There are only two voiced nonsonorant consonants [w, Russians came to Siberia in the sixteenth to seven- ?]. One can see that the phonemic nomenclature of teenth centuries. Mansi has many Russian borrowings Konda Mansi is more similar to Northern Mansi than in the field of culture, sciences, arts, and education. to Hungarian. Mansi is an agglutinative language. The Mansi is used in the first four forms of schooling and suffixes are added to the root: hap ‘boat’; hap+um ‘my also on the radio and in one newspaper. The phonolo- boat’; hap+ag+um+t ‘in my (two) boats; gy of Mansi is much simpler than that of Hungarian. It nepak+ag+men ‘our (two) two books’; kat+ag+um+til has six vowels: [i, e, I, a, o, u]. Some scholars consid- ‘by my (two) hands’; war+s+an+en -’ they (two) did er Mansi to have long vowel counterparts. Vowel these’; and jal+as+al+is ‘he traveled many times’, i.e. length is not depicted in the script. This problem has jal- ‘to go’, ‘-as’ the suffix showing contiguity of not been studied by the methods of experimental pho- action, ‘-al’ the suffix showing the interruption of netics therefore, one cannot state for sure that the action; ‘-is’ the suffix showing the past tense, and the vowel length is phonemic. It can distinguish only lack of personal ending shows the third-person singu- some words with [a] and [o]. Anyway, it needs more lar. Mansi nouns have no gender. The noun has num- investigation. According to the vertical movements of ber: singular, dual, and plural. The Mansi noun has the tongue three positions, thus, three vowel types are seven cases, Hanty, 10. It takes possessive suffixes and defined: high [i, u], middle [e, I, o], and low [a]. is definite or indefinite. There are a small number of According to the horizontal movement of the tongue adjectives, but mostly nouns can serve as adjectives if three positions, thus, three types of vowels are defined: they are in preposition to the other nouns, e.g. ker front [i, e], central [I], and back [a, o, u]. Unlike unsah ‘iron bridge’; being in preposition to the noun, Hungarian, there is no vowel harmony. The stress falls it does not agree with the noun in number or case, on the first, third, and fifth syllables. The stressed being a part of the predicate, it agrees with the subject vowels sound longer than the unstressed ones. The in number. number of consonantal phonemes in Mansi is not so There are several types of Mansi pronouns: person- great as in Hungarian. There are no long consonants. al, demonstrative, interrogative, negative, and indefi- By the work of the active organ of speech (place of nite. Most of them are declined. Mansi, like articulation), the following consonants can be defined: Hungarian, has definite and indefinite conjugations labio-labial [p, w, m], (no labio-dental whatsoever!); and five moods: indicative, imperative, subjunctive, front [t, s, n, l, r]; palatal [t’, s’, n’, l’, j]; and velar [k, conditional, and inferential. Word order is usually x, ?, ?]. By the manner of articulation, the following S(ubject)-O(bject)-V(erb) or OSV. Mansi consonants are classified into four types: sono- rant [m, w, n, n’, l, r, j, ?], occlusive [p, t, t’, k], (no Hanty affricates), and fricative [s, s’, x, ?]. By the work of vocal cords, nonsonorant Mansi consonants are divid- Hanty together with Mansi form Ob’-Ugric subgroup ed into two groups: voiced and voiceless. There is only of the Ugric group (plus Hungarian) of the Finno- one voiced [?], all the rest are voiceless [p, t, t’, s, s’, Ugric language family. They appear more Mongoloid k, x]. In this respect, as in many others, Mansi is quite than Mansi. They call themselves Hanty, although in different from Hungarian. Communication between a the West they are still wrongly called Ostjaks. They Mansi and a Hungarian is impossible. Communication live in the basins of the Ob’ River and Vas’ugan River. 482  HUNGARIAN AND UGRIC LANGUAGES The majority of Hanty live in the Hanty-Mansi nation- palatal [t’, s’, n’, ?’, j]; and velar [k, x, ?] and accord- al District (11,900 people), although some (7,300 peo- ing to the manner of articulation: sonorant [m, w, n, n’, ple) live in the Jamal-Nenets national autonomous l, ?, ?’, r, j, ?], occlusive [p, t, t’, k], (no affricates), and District or in Tomsk region of Russia. The Hanty lan- fricative [s, s’, ?, x]. There are no nonsonorant voice- guage is known for its dialectal diversity; nearly every less consonants in the Kazym subdialect of Hanty. The village speaks differently. Only 67.8% of Hanty con- vowel and consonant phonemic system of the Vach- sider it their native language. In the 1930s, an attempt Vas’ugan dialect are more similar to Hungarian than was made to set up a literary Hanty language, but the the rest. The following 14 vowels are defined [i, I, e, o, success was even smaller than with Mansi. The o, o, o, u, u, u, a, a, ?, a], and according to the work of Northern (Kazym) dialect was chosen to be the basis of the active organ of speech: labio-labial [p, w, m], (no the literary Hanty language. The first desk books were labio-dental), front [t, s, ?, t?, n, l, ?, l, r], palatal [t’, n’, published in that dialect. However, only the Kazym l’, l’, j], and velar [k, q, ?, ?]. According to the manner Hanty could use them, while the native speakers of the of articulation, the following can be defined: sonorant other dialects could not understand them. This is why, [m, n, n’, l, l’, ?, j, ?], occlusive [p, t, t’, k, q], affricate in the 1940s they also began publishing books in the [t?], fricative [w, s, ?, l, l’, ?], and voiced consonants Obdor (Salehard), Ber’ozovo, and Middle Ob’ [w, l, l’, ?]. Each of the described consonantal groups (Sherkal) dialects. In the 1950s, Vach and Surgut has its influence on the sound structure of the lan- dialects were added. Nothing has ever been published guages speech chain. The high frequency of sonorants in the Southern Hanty dialects (Dem’an or Konda). may give a melodic picture of a language, while the Much earlier, in the 1840s, some of the Hanty folklore high frequency of occlusive consonants gives a sharp texts were written down by the Hungarian A. Reguly and abrupt picture of it. Comparing the consonantal and Finn M.A. Castren. In the second half of the nine- sound pictures of these four Ob’-Ugrian languages teenth century, some parts of the Holy Bible were with the help of ‘chi-squared criterion’, one comes to translated by the Russian missionaries, who tried to the conclusion that the Northern (Sos’va) dialects of christen the Hanty people. Since 1937, the Hanty script Mansi and Northern (Kazym) dialect of Hanty resem- is Russian (Cyrillic). All the dialects mentioned here ble each other most in terms of the distribution of cer- have plenty of subdialects, which can be called sepa- tain types of consonants in their sound chains (75). The rate languages and whose names cannot be mentioned sound picture distance between Sos’va and Konda due to space constraints. Even the main branches are dialects of Mansi is greater (409). This distance so many that none of them are mentioned. The between Kazym and the Eastern dialect of Hanty is Northern Hanty dialects (Kazym, Nizjam, Sherkal, even greater (476). The distance between Konda Ber’ozovo) have a rather simple system of phonemes, dialect of Mansi and Eastern dialect of Hanty is 832. if compared to the southern or eastern dialects. They Hungarian is much closer to the Eastern dialect of have only three cases and a rather simple verb system. Hanty (151) than to the Kazym dialect of Hanty (1612) They have double negation. The eastern dialects (Vach, or both Mansi dialects: Sos’va (1018) or Konda (1115). Vas’ugan, Surgut, Salym, etc.) have vowel harmony, One can see that these distances are much greater than i.e. rigid opposition between front and back vowels. the similarity of Hungarian to Mordovian (174) or They also preserve the other archaic traits, e.g. a more Mari (226), the Finno-Ugric of the Volga region. complicated phonemic system, paradigmatic vowel change, plenty of cases, and semiergative construction of the sentence. The southern dialects are intermediate References between the northern and eastern dialects in their pho- Balandin A. N., Vahrusheva M.P. 1957. The Mansi language, netical and grammatical characteristics. The ergative 275. Leningrad: Prosveshch. [in Russian]. construction (as well as in the Northern dialects) is Tereshkin N. I. 1966. The Hanty language The languages of the much more seldom, than in the Eastern dialects. Let us peoples of the USSR, Vol. 3. Finno-Ugric and Samodij lan- consider the similarity of the phonetical systems of two guages, 319–42. Moskva: Nauka [in Russian]. Yuri Tambovtsev. 1998. Some phonological features as meas- Hanty subdialects to those of Mansi. We took up ure of closeness of dialects. Philologia Fenno-Ugrica 4. Kazym (Northern) and Vach-Vas’ugan (Eastern) 1–19. dialects. The vowel system of the Kazym subdialect of Peter Hajdu. 1985. The Uralic Languages and Peoples. Hanty is very similar to that of the Northern (Sos’va) Moskva: Nauka [in Russian]. subdialect of Mansi. It has nine (long and short) vow- The Hungarian language. Budapest, 1972. K. E. Majtinskaja. 1993. The Hungarian language Languages of els [i, I, e:, o, o:, u, u:, a, a:]. According to the work of the World. Uralic languages, 179–256. Moskva: Nauka [in the active organ of speech (place of articulation), the fol- Russian]. lowing four consonantal groups are defined: labio-labial J. Tompa. 1968. Ungarische Grammatik. Budapest. [p, w, m] (no labio-dentals); front [ t, s, ?, n, l, ?, r]; YURI TAMBOVTSEV 483 HYMES, DELL HATHAWAY Hymes, Dell Hathaway Dell H. Hymes’ life and work revolve around an inter- rooted in his mixed training and later interactions est in language, a commitment to social justice, and an with anthropologists, linguists and social scientists, at abiding love of folklore. The intertwining of these Harvard, and at Berkeley. He chose the term linguistic themes provided the fertile ground for his views: that anthropology, making it a branch of anthropology. language is embedded in culture; that the division This approach resonates with his view of language’s between language and culture is a disciplinary artifact, role in social life and by applying it to the examination its maintenance a result of protected disciplinary bound- of language, challenged both linguistics and anthro- aries; and that the only legitimate study of language is pology. Using ethnography, anthropologists uncover socially constituted linguistics. Much of his writing the rules and symbols that hold meaning for the com- therefore challenges current–traditional practices in munity being studied. Such an approach was new to anthropology, sociolinguistics, sociology, and formal linguistics, which instead considered descriptions of linguistics. His work reflects a critical and evolutionary phonology and grammar as the principal framework. Marxist view of knowledge, interpreting the current By choosing the term ‘speaking’, he also challenged dialectic as a teleological stage in the development of anthropologists who, when studying cultures of a knowledge, leading, ‘... to a humanism which can deal community, took speech for granted. with concrete situations, with the inequalities that actu- This paper launched a new perspective, providing ally obtain, and help to transform them through knowl- the parameters for analyzing and describing speech edge of the ways in which language is actually communities—grouping members who share one or organized as a human problem and resource’ (1996: 60). more of its ‘ways of speaking’. Members who know Hymes considers the essence of humankind not as not only the rules of speaking but also the sociocultu- an abstraction in the individual, but an actuality ral norms and values that guide interaction and cultur- entailed in specific social relations. He thus rejects al behavior within specific communities can be giving priority to an abstract category of language or considered to have communicative competence. His grammar, and the notion of nativism of language, both framework for description included speech communi- positions putting him in opposition to structuralism ty, speech event (activities governed by rules for and Chomskyan linguistics. He met the Chomskyan speaking), factors in a speech event (e.g. sender), and challenge head on with his formulation of ‘commu- the functions of a speech event (e.g. persuasive). nicative competence’, which deliberately creates an Over the next decade, the framework was revisited as irony. For Noam Chomsky, competence inheres in the colleagues and students conducted ethnographies in dif- individual, and linguistic rules are verified by the intu- ferent settings—nonwestern, Native American, and, itions of the ideal speaker/hearer. The idea that ‘com- increasingly, schools. In 1972, he elaborated on the fac- municative’ can be collocated with ‘competence’ turns tors in the speech event, devising the mnemonic Chomsky’s notion and resulting linguistic discourse SPEAKING: setting, participants, ends, act characteris- on its head. He also refutes the basic premise in lin- tics, key (tone), instrumentalities, norms, and genres. guistics that all languages are equal. Hymes claims This list has been compared to Michael Halliday’s field, that all languages may have equal potential, but tenor, mode, and was found to be unmotivated linguis- because human beings ascribe differing attitudes to tically. Hymes, however, only ever meant SPEAKING different languages and varieties, speakers of some to be a heuristic, a list of questions to ask when exam- languages and varieties have reduced opportunities. ining language in context. For both, the bundling of lin- His strongly held view that forms and meanings are guistic features in a community results in linguistic part of social life has influenced both his theoretical repertoires available to that community. Ethnography of writings and his life choices. He is probably most speaking was broadened to the ‘ethnography of com- widely known for his formulation of the ‘ethnography munication’, Hymes never intended to ignore written of speaking’, first presented at a meeting of the language by his use of ‘speaking’; rather, he used it as a Anthropological Society of Washington in 1962, the surrogate for all communication modes, including ges- seed for which was germinated in a 1952 debate about tures and, most recently, electronic modes. the role of linguistics in anthropology, one that drew Hymes was the driving force behind a new journal, scholars from several disciplines. That ethnography calling it Language in Society, not a journal of soci- would become his approach to language in society was olinguistics, so that its concern would focus on 484  HYMES, DELL HATHAWAY social/cultural life. Perhaps, his greatest source of of Virginia in 1990–1998, now Emeritus. He was also influence was his 20 years of editing the journal, many Fellow of the American Association for the of the articles owing much to his imaginative thinking Advancement of Science; Life Fellow, Clare Hall, as he dialogued with the author, sending detailed rec- Cambridge; member of the British Academy; Fellow at ommendations, comments, and further ideas. At the the Center for Advanced Study in the Behavioral University of Pennsylvania, he moved from Sciences, 1957–1958; Guggenheim Fellow, Anthropology to Folklore and Linguistics, because of 1969–1970; Senior Fellow, National Endowment for political division within the department, even though the Humanities, 1972–1973; Trustee, Center for he felt that anthropology was in a unique position for Applied Linguistics, 1973–1978; President, American studying the sociology of language from a compara- Folklore Society, 1973–1974; President, Council on tive perspective. Later still, he moved to the School of Anthropology and Education, 1978; President, Council Education as Dean, a move that was not just practical on Anthropology and Education, 1978; President, (to save the School from closure) but also philosophi- Linguistic Society of America, 1982; President, cal. He had already set a new direction for education- American Anthropological Association, 1983; and al research with his volume coedited with Cazden and President, American Association for Applied John (1972), by insisting on the need to examine actu- Linguistics, 1986. He was also founding coeditor of al classroom interaction in order to understand the role Language In Society with William Labov and Allen of language in teaching and learning. In education, his Grimshaw, 1972–1992. beliefs that understanding how language varies and what that variation means for speakers, especially in terms of inequality, influenced both theory and prac- References tice leading to the development of a program in edu- Bauman, Richard, and Joel Sherzer (eds.). 1982. Explorations cational linguistics. Education was also a platform for in the ethnography of speaking. Cambridge: Cambridge advocating that all human beings could benefit from University Press. an understanding of language. His most recent work Heath, Shirley Brice. 1983. Ways with words. Cambridge: has focused on analyzing oral narratives, bringing Cambridge University Press. together his love of folklore, his commitment to social Hornberger, Nancy. 1988. Bilingual education and language maintenance: a Southern Peruvian Quechua case. Berlin: justice, and his insistence that language should always Mouton. be examined in its social context. Hymes, Dell Hathaway. 1969. Reinventing Anthropology. ———. 1972. The scope of sociolinguistics. Sociolinguistics: current trends and prospects. 23rd annual Georgetown round- Biography table, ed. by Roget W. Shuy. Washington, DC: Georgetown University Press. Dell Hathaway Hymes was born in Portland, Oregon on ———. 1986[1972]. Directions in sociolinguistics: the ethnog- June 7, 1927. He received his B.A. in 1950 from Reed raphy of communication, ed. by John Gumperz and Dell College; M.A. in 1953 from Indiana University; and Hymes. New York: Basil Blackwell. Ph.D. in 1955 for a dissertation that was a linguistic ———. 1972. Functions of language in the classroom, ed. by analysis of Kathlamet Chinook Indian texts that had Courtney B. Cazden, Vera P. John, and Dell H. Hymes, New York: Teachers College Press. been collected by Franz Boas, Indiana University. He ———. 1974. Ways of speaking. In Bauman and Sherzer. became Instructor, then Assistant Professor, Social ———. 1974. Foundations in sociolinguistics: an ethnographic Relations Department, Harvard University in, approach. 1955–1960. Also, he became Associate Professor, then ———. 1983. Studies in the history of linguistic anthropology. Professor, Department of Anthropology, University of ———. 1996. Ethnography, linguistics, narrative inequality: toward an understanding of voice. California, Berkeley in 1960–1965; he became McDermott, Ron. 1977. The ethnography of speaking and read- Professor, Anthropology, University of Pennsylvania in ing: Linguistic theory: what can it say about reading, ed. by 1965–1972; Professor, Folklore and Linguistics, Roger Shuy. Newark, Delaware: International Reading University of Pennsylvania in 1972–1988; Professor, Association. Sociology, University of Pennsylvania in 1974–1988; Philips, Susan Urmston. 1983. The invisible culture: communi- cation in classroom and community on the Warm Springs Dean of the School of Education, University of Indian Reservation. New York: Longman. Pennsylvania in 1975–1987; Professor of Anthropology Saville-Troike, Muriel. 1989. The ethnography of communica- and English, University of Virginia in 1987–1990; and tion. New York: Basil Blackwell. Commonwealth Professor of Anthropology, University DENISE MURRAY 485  I Iconicity Many believe that language is arbitrary. Arbitrariness guage is a form of meaning depiction. The notions of refers to a random relation between a word and its up and down in various sign languages relate closely meaning. However, there are many examples of nonar- to the vertical directions of these concepts. The resem- bitrariness in language known as iconicity. Iconicity is blance between a raptor snatching a sitting-up rodent the opposite of arbitrariness because iconicity refers to and one hand snatching at an upright finger in a resembling relation between a word and its meaning. American Sign Language is a good example of dia- A word is directly related to its meaning in many ways. grammatic-iconic gesture. However, believers of lan- The form of words like cockatoo, cuckoo, buzz, guage signs must arbitrarily deny sign language as thump, crash, splash, thrash, clash, hiss, puff, and purr language (see Armstrong, Stokoe, and Wilcox). is connected to their meanings. These words imitate Motivation is the operating force behind diagram- the sounds created by the referents. While, ‘ha, ha, matic iconicity. Motivation in iconicity refers to per- ha’, captures the sarcastic laughter in English, ‘he, he, ceived similarity between the structure of the he’ is commonly used after a remark intended as a joke expression and the concept represented in the expres- in Malay e-mails as well as Singapore and Malaysian sion. The need to name things results in the expansion English e-mails. These words do not come into lan- of meaning motivated by resemblance in appearance. guage by chance. The derivation of these words is A schematic resemblance between a tall crane based on certain comprehension based on sound. The machine and the neck of crane motivates the deriva- words are sound images of the various meanings. This tion of a new meaning in the word crane. The creation type of iconicity is termed imagic iconicity. of new words may be motivated by functional similar- Imagic iconicity comes in the form of sound sym- ity, e.g. the shuttling of a shuttlecock and a space shut- bolism. Sound symbolism occurs if a particular sound tle in the air. informs a particular sense across various words. Quantity iconicity is another type of diagrammatic Examples in Malay, Cantonese, English, and Spanish iconicity. Quantity iconicity holds the basic notion that words suggest that the vowel [o] symbolizes a negative more form entails more meaning. Physical increase in sense across languages (Table 1). quantity is captured loosely by more words in redupli- Current bad English o-iconic words of the world are cation. Hence, plurals usually have more words. bomb, romp, and storm all of which had brought down Simple noun reduplication in Malay exemplifies quan- lives in all four corners of the world resulting in either tity iconicity (Table 2). death or disgrace. Diagrammatic iconicity records the flow of time Another type of nonarbitrariness is diagrammatic from its commencement until its completion. This iconicity. Diagrammatic iconicity suggests that a word sequence of time and the natural order of an event is a diagram of a meaning. Linguistic signs are used to unfolding in time are reflected in sentences in sequen- illustrate some structures of the meaning. Sign lan- tial iconicity, e.g. ‘Syukri went to school on Monday 487 ICONICITY TABLE 1 Malay Cantonese English Spanish hodoh (ugly) soh (slow in thinking) horror tonto (stupid) bodoh (stupid) lõw (man in a rude sense) sorrow zonzo (stupid) kotor (dirty) põr (woman in a rude sense) fool coco (funny looking) rogol (rape) mór (devil) loon memo (silly, foolish) bocor (leak) mõng (dumb) moron chato (flat nose) TABLE 2 reflect the iconicity of quantity decrease. This is due to the politeness convention that prefers a lesser or weak- Basic Singular Noun Basic Plural Noun in Malay in Malay er meaning before a stronger one. Freezes like sir and madam, father and mother, pelajar (a student) pelajar-pelajar (students) papa and mama, Adam and Eve, king and queen, boys sekolah (a school) sekolah-sekolah (schools) and girls, husband and wife, Mr. and Mrs., and broth- ers and sisters show that man syntactically precedes gunung (a mountain) gunung-ganang (ranges of woman. That man comes before woman is based on mountain) the perception of man as the center of the universe or sayur (a vegetable) sayur-mayur (vegetables of the me-first principle. This is quite true in the ancient all sort) and current real world. Contradictions like mom and dad, bride and groom, aunts and uncles, and ladies and gentlemen can be explained either by the polite- and left for Singapore on Tuesday in the same week,’ ness convention, the stronger phonological laws, rather than ‘*Syukri left for Singapore on Tuesday and whereby high vowels precede low vowels or the less- went to school on Monday in the same week.’ In fact, syllabic items precede more-syllabic items order (see sequential iconicity is a refection of common sense, Cooper and Ross). which is imperative to sensible sentences like: Jyh cooked the fish with ketchup before he ate it. References *Jyh ate the fish with ketchup before he cooked it. Armstrong, David F., William C. Stokoe, and Sherman E. Siti knocked the vase, and consequently it broke. Wilcox. 1996. Gesture and the nature of language. *Siti broke the vase and consequently it knocked. Cambridge: Cambridge University Press. Askedal, John O. 1995. Iconicity in the basic serialization rules The language of distance is iconic. The vivid sense of Modern German. In Landsberg. Cooper, William E., and John R. Ross. 1975. World Order. of near is easier to register in the mind than the distant Papers from the parasession on functionalism of the notion of far. These two senses are codified in phrases Chicago Linguistic Society. like now and then, this and that, and in and out. That Haiman, John. 1985. Natural syntax: iconicity in language. near comes before far in language, just like their real- Cambridge: Cambridge University Press. world relation illustrates syntactic iconicity. Syntactic Landsberg, Marge E. (ed.) 1995. Syntactic iconicity and linguis- tic freezes: the human dimension. Berlin: Mouton de Gruyter. iconicity refers to the fixed order in phrases. These Landsberg, Marge E. 1995. Semantic constraints on phonolog- phrases are known as freezes. Syntactic iconicity in ically independent freezes. In Landsberg. here and there has equivalents like ingum angum in Malkiel, Yakov. 1990. Diachronic problems in phonosymbol- Tamil, sini sana in Malay, li do guo do in Cantonese, ism: Edita and Inedita. Amsterdam: Benjamins. ci dau he dau in Hokkien, ci gou heo gou in Teochew, Sew, Jyh Wee. 1995. Sound meanings in Malay. M.A. Thesis. The National University of Singapore. and zhe li na li in Mandarin. Sew, Jyh Wee. 1998. Review of Marge E. Landsberg ed., Likewise, above precedes below, on precedes Syntactic iconicity and linguistic freezes: the human dimen- under, and up precedes down (see Landsberg). Freezes sion. WORD 49. like ups and downs, head to toe, head over heel, top- Sew, Jyh Wee. 1998. Review of Friedrich Ungerer and Hans- down, high and low are common examples. Quantity Jörg Schmid: introduction to cognitive linguistics. The Web Journal of Modern Language and Linguistics 3. decrease in freezes like more or less, and big to small http://www.ncl.ac.uk/~njw5/issue03/jyh-wee.htm are found in Mandarin and Malay. Unlike the Malay JYH WEE SEW freezes lebih kurang (more or less), the Indonesian counterpart kurang lebih (less or more) does not See also Onomatopoeia 488  IDENTITY AND LANGUAGE Identity and Language Throughout human history, language and identity have been concerned with cross-cultural communication. been closely linked. Political boundaries have been ‘Cross-cultural’ has been rather loosely defined and drawn along linguistic lines, establishing nationality as includes interactions between people from different well as ethnic division. Often, a single feature of pro- national backgrounds (e.g. American–Japanese busi- nunciation can provide a convenient means of identifi- ness communication), between migrants and natives cation. Pronunciation, grammatical and lexical markers, (e.g. South-East Asians in British job interviews), naming, and discourses may all be used to establish between people from different ethnic backgrounds (e.g. identity. The linguistic approach that is most centrally Black and White Americans), or between women and concerned with the interrelationship between identity men. This social psychological conception of identity, and language is sociolinguistics. Sociolinguistic however, has come under attack for at least two differ- approaches to identity and language have changed over ent reasons. In this approach, identity as group mem- time. At least three different strands of inquiry into bership is seen as essentializing, meaning identity is identity and language can be identified: variationist, regarded as internal and immutable. Further, this view interactional, and social constructionist approaches. of identity relies on the idea of a homogeneous group Variationist sociolinguistics, which is strongly asso- while, in fact, people are members of many different ciated with the pioneering work of William Labov, is groups simultaneously (e.g. American, Hispanic, pro- based on the premise that language reflects identity. fessor, female, heterosexual). The latter objection, that Assuming that a person’s speech reflects characteristics social identity is hybrid and heterogeneous, is raised by of that person, aspects of social identity—such as linguists who work in bi- and multilingualism studies. social class, sex, ethnicity, nationality, or age—were The language use of bilinguals clearly exemplifies that traced through forms of language. The bulk of varia- speakers do not only have one single identity but rather tionist work on identity and language concerned pro- a repertoire of identities. nunciation. Typically, certain pronunciation variables Contemporary approaches to language and identity were linked to aspects of social identity. However, in most often build on the framework of social construc- recent years, work in variationist sociolinguistics has tionism. Identity is now seen as relational, cultural, and been challenged on the grounds that social identity can- contingent. It is relational because it is located in con- not provide an explanation for language use because it nections made between people rather than in the minds is a concept that is itself in need of explanation. Critics of individuals. Identity is cultural because it is based on of variationist sociolinguistics argue that instead of shared understandings, and it is contingent because it is asking, ‘How do women and men speak differently?’ or a strategic performance that may fail or misfire. ‘How do Black and White Americans speak different- Speakers are seen as strategically deploying their lin- ly?’ it would be more important to find out how people guistic repertoires in order to project chosen identities. become gendered or racialized beings in the first place. However, their acts of identity are not exclusively in Interactional sociolinguistic approaches to identity their own hands as they depend (a) on the repertoires at and language, as pioneered by John Gumperz and used their disposal and (b) on the ways in which their speech by Deborah Tannen, draw largely on social psychology. partners choose to view them. Identities are construct- In this approach, identity is based on group member- ed or coconstructed depending on the power relation- ship, where the negotiation of identity in interaction is ship that pertains between the interactants. In many of central concern. Instead of focusing on pronuncia- societies, categories such as caste, gender, or race are tion, interactional sociolinguists study discourse, writ- imposed and coerced, leaving little or no room for indi- ten or spoken communication, preferring qualitative viduals to perform or explore alternative identities that methods over quantitative in data collection and analy- might deviate from prevailing ideologies. Three aspects sis. Interactional sociolinguistics assumes that identity of social identity that have played a central role in as group-membership is negotiated, challenged, or recent linguistic work will be discussed based on the upheld in conversation. When people come together following: gender, nationality, and bilingualism. from different backgrounds, misunderstandings are Within feminist and gender theory, identity has likely to occur, as they do not share the same conven- become the central question since, in the 1980s and tions for signaling identity and role relationships. 1990s, women of color and lesbian theorists started to Consequently, most of the work in this tradition has challenge the assumption that the experience of white 489 IDENTITY AND LANGUAGE heterosexual middle-class American women was univer- acknowledge that bilinguals and second-language learn- sal. Consequently, language and gender researchers have ers tend to engage in more than one community and that begun to explore the linguistic construction of fluid, they may have different identities in different languages. hybrid, and alternative identities. Recent research (see At present, the study of identity and language is a Bucholtz et al., 1999) explores how identity is invented lively and diverse field of research, encompassing a (e.g. through the creation of shared experience in the wide array of issues and implications. Given the over- coming-out stories of deaf and hearing lesbians) and all global climate of identity politics, the scope of soci- how the performance and perception of identity is con- olinguistic study, both theoretical and practical, is strained through belief systems (e.g. through the likely to increase in the future. ‘woman as dessert’ metaphor in English). Further foci include speakers’ individual creative responses to cultur- References al ideologies and identity constructions as ‘on-line’ events that may never be complete (e.g. callers on a Anderson, Benedict. 1983. Imagined communities: reflections shopping channel may position themselves simultane- on the origin and spread of nationalism. London: Verso; 2nd edition, 1991. ously as consumers with questions about the product and Bucholtz, Mary, A.C. Liang, and Laurel A. Sutton (eds.) 1999. experts who are familiar with consumption patterns). Reinventing identities: the gendered self in discourse. New Unlike gender, nationality has been seen as intri- York and Oxford: Oxford University Press. cately linked to language for centuries. To the present Eckert, Penelope. 2000. Language variation as social practice: day, many nations, particularly in Europe and North the linguistic construction of identity in Belten High. Oxford and Cambridge, MA: Blackwell. America, subscribe to a ‘one nation, one language’ Gumperz, John J. (ed.) 1982. Language and social identity. ideology. Many states have used the reification of a Cambridge: Cambridge University Press. standard language and the common identity that sup- Heller, Monica. 1999. Linguistic minorities and modernity: a posedly goes with a shared language as a means to nat- sociolinguistic ethnography. London: Longman. uralize political borders. In an intricate genre analysis, Le Page, Robert, and Andrée Tabouret-Keller. 1985. Acts of identity: Creole-based approaches to language and ethnicity. Wodak et al. (1999) describe how ‘the same’ national Cambridge: Cambridge University Press. identity (Austrian) is performed in significantly differ- Mühlhäusler, Peter, and Rom Harré. 1990. Pronouns and peo- ent ways and with significantly different meanings in ple: the linguistic construction of social and personal identi- public and private speech (e.g. newspaper articles vs. ty. Oxford: Basil Blackwell. one-to-one interviews). Norton, Bonny P. 2000. Identity and language learning: gender, ethnicity and educational change. London: Longman. For some time, researchers into multilingualism have Pavlenko, Aneta, and Adrian Blackledge (eds.) 2001. seen language choice as meaningful and as an ‘act of Negotiation of identities in multilingual contexts. (Special identity’ (see Le Page and Tabouret-Keller 1985), issue of The International Journal of Bilingualism 5). whereby speakers proclaim their allegiance to a particu- Pavlenko, Aneta, Adrian Blackledge, Ingrid Piller, and Marya lar speech community. Code-switching is often seen as Teutsch-Dwyer (ed.) 2001. Multilingualism, second lan- guage learning, and gender. Berlin and New York: Mouton a means to signal changes in the role relationship that de Gruyter. pertains between interactants. Furthermore, second-lan- Piller, Ingrid. 2002. Bilingual couples talk: the discursive con- guage acquisition theory has often viewed successful struction of hybridity. Amsterdam: Benjamins. language learners as good assimilators. The easier they Shotter, John, and Kenneth J. Gergen (eds.) 1989. Texts of iden- find it to shed their native identity and assimilate to the tity. London, Newbury Park and New Delhi: Sage. Wodak, Ruth, Rudolf de Cilia, Martin Reisigl, and Karin one related to the target language, the more successful Liebhart. 1999. The discursive construction of national iden- their learning outcomes were thought to be. Only recent- tity. Edinburgh: Edinburgh University Press. ly, linguists (e.g. Pavlenko et al., 2001) have come to INGRID PILLER Ideology and Language Ideology’s features are topics of complex debate. embedded in and reproductive of specific socioeco- Often defined as matrices of ideas and images inhabit- nomic structures. Ideology can empower or oppress. ing individuals’ minds, ideologies are also thought of Furthermore, ideology is related to language in numer- as neither ideas nor images, but as practices that are ous ways. 490  IDEOLOGY AND LANGUAGE An ideology may empower persons when it defines Californians write and speak numerous languages, their ethnic, gender, class, or other identity, organizes but enough Californians felt threatened by this lan- a ‘worldview’ that gives meaning and coherence to guage diversity to institutionalize an ideological their lives, or facilitates their entrance into social, monolingualism that values social unity as a function civic, or political life. Oppressive ideologies marginal- of linguistic homogeneity. The propositions’ success ize, stigmatize, or otherwise define persons to repro- arguably related to anxieties concerning California’s duce their exploitation or subordination in a growing population of Spanish speakers. These socioeconomic hierarchy. examples raise the issue of ‘linguistic racism’, in A focus on an ideology as empowering may down- which prejudicial attitudes about an ethnic group play the relations of ideology and knowledge. extend to that group’s language. Such racism often Sociologists comparing various cultures’ ideologies assumes that one language is inherently more beauti- may want to know the differences among ideologies ful or a better vehicle for thought than others. And, in rather than to judge their compliance to a normative the form of slurs, jokes, and ideologically coded dic- definition of knowledge. A barrier to this unbiased tion, language can influence racial ideology. evaluation is the sociologists’ own assumptions about Language also reflects ideologies of sexuality and knowledge. The ‘critique of ideology’ exposes such gender. For example, consider the following sentence: mystifying errors or beliefs by which an ideology rein- ‘A linguist, if he wants a successful career, masters forces structures of oppression. Often the same ideol- contemporary linguistic research on English’. This ogy empowers some, oppresses others, and gains sentence, by assuming that the typical linguist is male, hegemony. An ideology becomes hegemonic by defin- betrays sexist beliefs concerning women’s abilities ing the arguments contending groups use to debate and roles in society. Sexism in language use not only with each other; by prescribing the values to which signals an ideology; sexism as a relation of power clashing ideas must adhere to achieve legitimacy; or embedded in institutions reproduces itself through by projecting a totality in which diverse political such uses when they are uncontested. Similarly, actors imagine an open network of cultural, social, and racism in language use reveals an ideology and con- economic relations to be enclosed. tributes to the continuation of racist power relations. Debunking an oppressive ideology’s knowledge Even in the study of language, ideologies present claims contests that ideology. Theorists have long barriers to unbiased evaluation. Our sentence also debated whether ideology can be distinguished from assumes that English should be privileged in linguistic knowledge. The European Enlightenment’s attack on inquiry. The phrase ‘language ideology’ refers both to religious dogma claimed that ideologies are illusions ideologies found in language and to ideologies about that reason can dispel to foster an ideology-free socie- language. When power relations dictate which lan- ty. Karl Marx defined ideology as mystification in the guages may be studied or what constitutes their legiti- interest of a ruling class. Like Marx, the twentieth- mate study, we confront language ideology’s influence century communist philosopher Louis Althusser on the production of linguistic knowledge. claims that scientific knowledge is distinct from ideol- The study of language itself reflects a language ide- ogy. Because of ideology’s role in subject formation, ology that assumes languages are static entities unaf- Althusser concludes that ideology will always exist, fected by their ideological contexts. These contexts even in a future classless society. Contemporary ‘post- sometimes influence a language’s pattern of sound Marxist’ theories include ethnicity, gender, and culture change or even precipitate its demise. Examples of as important determinants of ideology. Questioning demise are the fates many indigenous peoples’ lan- the opposition between science and ideology, post- guages have suffered in the wake of Europe’s colo- Marxists doubt the theorists’ ability to contest oppres- nization of Asia, Africa, Australia, and the ‘New sive ideologies effectively. Post-Marxists define World’, direct victims of racist ideology. The recent political struggle as a battle for hegemony in which no trend to install Standard American English as a global participant holds the cognitive or ideological high lingua franca endangers ever more infrequently used ground. languages. Language reflects ideologies, oppressive Ideology intersects language when politics privi- and progressive, which in turn influence the evolution leges one language and denigrates, restricts, or of those same languages. Further, this reciprocal rela- excludes others. California voters have passed a tionship also holds in the political arena. Language not series of proposals, making English the state’s ‘offi- only influences ideologies but may also be used as a cial language’, requiring that government agencies tool to control groups or thought. The study of the use English to the exclusion of other languages, and intersection of language and ideology contributes to outlawing bilingual instruction in the state’s schools, our understanding of the political implications of glob- exemplifying ideology’s impact on language. al socioeconomic realities. 491 IDEOLOGY AND LANGUAGE References Eagleton, Terry. 1991. Ideology: an introduction. London and New York: Verso. Althusser, Louis. 1970. Idéologie et appareils idéologiques d’é- Edgley, Alison. 2000. The social and political thought of Noam tat (Notes pour une recherche). La Pensée 151; as Ideology Chomsky. London and New York: Routledge. and ideological state apparatuses: notes towards an investi- Joseph, John, E., and Talbot, J. Taylor (eds.) 1990. Ideologies of gation. Lenin and philosophy, translated by Ben Brewster. language. London and New York: Routledge. London: New Left Books, and New York: Monthly Review Kress, Gunther, and Robert Hodge. 1979. Language as ideolo- Press, 1971. gy, London and Boston: Routledge & Kegan Paul; 2nd edi- Barrett, Michèle. 1991. The politics of truth: from Marx to tion, London and New York: Routledge, 1992. Foucault. Cambridge: Polity Press, and Stanford: Stanford Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices: University Press. the extinction of the World’s languages. Oxford and New Calvet, Louis-Jean. 1987. La guerre des langues et les poli- York: Oxford University Press. tiques linguistiques. Paris: Payot; as Language wars and lin- Pêcheux, Michel. 1975. Les vérités de la Palice. Paris: F. guistic politics, translated by Michel Petheram, Oxford and Maspero; as Language, semantics, and ideology, translated New York: Oxford University Press, 1998. by Harbans Nagpal, London: Macmillan Press, and New Chomsky, Noam. 1989. Necessary illusions: thought control in York: St. Martin’s Press, 1982. democratic societies. London: Pluto, and Boston: South End Woolard, Kathryn, A. 1998. Language ideology as a field of Press. inquiry. Language ideologies: practice and theory, ed. by Derrida, Jacques. 1996. Le monolinguisme de l’autre: ou la pro- Bambi, B. Schieffelin, Kathryn, A. Woolard, and Paul, V. thèse d’origine, Paris: Editions Galilée; as Monolingualism Kroskrity. New York and Oxford: Oxford University Press. of the other; or, the prosthesis of origin, translated by Patrick Mensah. Stanford: Stanford University Press, 1998. ROBERT S. OVENTILE Idiomaticity The traditional and most widespread definition of struction that counts as an idiom. Hence, there is a dis- idiomaticity maintains that it is the result of a mismatch tinction between ‘transparent’ idiomatic expressions between meaning and form: the overall meaning of an like skate on thin ice, which have a clear literal origin, idiomatic expression is figurative and noncomposition- and ‘opaque’ ones like to pull someone’s leg, which do al, that is, it cannot be deduced from the literal meaning not display any logical relation with their source. of its individual components. Idiomatic expressions are Idiomaticity is thus the outcome of a historical process thus words, phrases, or even whole sentences that seem that must nonetheless be analyzed as a rather static phe- to function as an unanalyzable unit: egghead, to bury the nomenon. hatchet, the early bird catches the worm, and the like. But classic theories of idiomaticity reject this The study of idiomaticity has advanced greatly since distinction because, according to them, all idiomatic the 1980s because of a change in the discipline of lin- expressions are dead metaphors that lack any connec- guistics that was characterized by an increasing interest tion with their original sources. As a consequence, the in meaning and discourse. This universal grammatical relation between their form and meaning is tradition- phenomenon has begun to receive the attention it did ally said to be arbitrary. Recent research within the not receive during the long period when sentence struc- theoretical framework called ‘cognitive grammar’ has ture was the principal focus of linguistic investigation. shown, by contrast, that the form–meaning relation Although unpredictable from its form, the meaning underlying many idiomatic expressions is far from of an idiomatic expression cannot be considered purely arbitrary—that it is motivated not only by their primi- random; it is the direct consequence of a historical tive meaning but also by conventional images and process of extension, whereby the literal, composition- conceptual metaphors. Thus, the traditional analysis al meaning of that pattern, although not completely lost, asserts that nothing explicit in to spill the beans, for is gradually replaced by a new metaphorical, noncom- instance, justifies its meaning ‘to reveal a secret’; this positional meaning that becomes widely accepted in new approach, however, claims that such a figura- society. Contrary to what might be expected, however, tive meaning is clearly determined by the ‘conduit the connection between the two senses is not easily rec- metaphor’—that is, a metaphor that compares the mind ognizable at the present time in every linguistic con- with a container and ideas with physical entities. 492  IDIOMATICITY Apart from shedding some light on the semantic Moreover, new studies have shown that idiomatic motivation underlying idiomaticity, the afore-mentioned expressions are less exceptional than traditionally historical process is also noteworthy for two further thought: (a) Far from being structurally frozen reasons: (a) it is responsible for the institutionalized patterns, they present more structural variety than has meaning of idiomatic expressions; (b) it explains their been asserted. Specifically, the more semantically frequent coexistence with literal interpretations of the motivated and institutionalized the meaning of an same expressions. idiomatic expression, the more structural flexibility it (a) Along with the nonliteralness and non- shows. (b) The absolute syntactic rigidity that does in compositionality of an idiomatic expression, its sense fact characterize some idiomatic patterns also appears also has to be institutionalized; that is, it must be in a large number of nonidiomatic constructions: to regularly associated by the members of a speech kick the bucket, for instance, does not have a passive community with a specific form. The number of people form, i.e. the bucket was kicked (by him) is accepting such form–meaning asymmetry is considered unacceptable; however, the nonidiomatic construction essential in this slow process of adaptation. They can she has two sons does not have a passive either: two range from a small in-group to an entire linguistic sons are had (by her) is equally unacceptable. The community, hence, their number influences the degree traditional assumption that special rules apply to of institutionalization of a specific idiomatic pattern and idiomatic language must thus be rejected in view of consequently its degree of idiomaticity: that is, the more these facts. widely used in society an idiomatic expression is, the In sum, idiomaticity is a heterogeneous and more institutionalized and idiomatic it is, hence the measurable linguistic phenomenon, variable in contrast between those idiomatic constructions that are meaning, form, and discourse functions. It thus forms listed in dictionaries as approved by the entire society a continuum with different degrees of idiomaticity that and those others, like slang, jargon, and familiar cannot be straightforwardly defined according to one euphemisms that have a private meaning accessible single parameter. only to small social groups. Regardless, during the institutionalization process they all acquire a national and cultural flavor that makes understanding and References acquiring them highly difficult for the outsider. Cacciari, Cristina, and Sam Glucksberg. 1991. Understanding (b) The institutionalization process is also idiomatic expressions: the contribution of word meanings. significant because, although it does not imply the Understanding word and sentence, ed. by G.B. Simpson, 217–40. North-Holland: Elsevier Science. systematic loss of the original construction that serves Chafe, Wallace. 1968. Idiomaticity as an anomaly in the as input for the idiomatic pattern, it explains the Chomskyan paradigm. Foundations of Language 4. 109–27. frequent coexistence of many idiomatic expressions Fernando, Chitra, and Roger Flavell. 1981. On idiom. Critical with literal interpretations of the same expressions. The views and perspectives. Exeter: University of Exeter. commonly suggested hypothesis that idioms are Fraser, Bruce. 1970. Idioms within a transformational grammar. Foundations of Language 6. 22–42. usually ambiguous between a metaphorical and a literal Gibbs Jr., Raymond W., and Jennifer E. O’Brien. 1990. Idioms interpretation is thus reinforced. However, such and mental imagery: the metaphorical motivation for ambiguity can no longer be considered a distinguishing idiomatic meaning, Cognition 36. 35–68. property of idiomaticity, due to the following results of Jackendoff, Ray. 1997. The architecture of language faculty, recent investigations based on immense collections of Cambridge, MA: The MIT Press. Jaeger, Leon. 1999. The nature of idioms. a systematic actual language data: (a) a vast number of idiomatic approach. Berne, Berlin, Brussels, Frankfurt, New York and patterns, such as rain cats and dogs and by and large, Vienna: Peter Lang. lack literal interpretations, either because they would Lakoff, George. 1987. Women, fire, and dangerous things. violate truth conditions or because they would be What categories reveal about the mind. Chicago: The ungrammatical, (b) the literal interpretation of University of Chicago Press. Makkai, Adam. 1971. Idiom structure in English. The Hague: whatever idiomatic expressions do allow both readings Mouton. 171–87. is highly infrequent in actual language data, because Makkai, Adam. 1987. Idiomaticity and phraseology in Psot- the discourse where they occur generally favors the Chomskyan linguistics: the coming-of-age of semantics figurative meaning over the literal one. Even when no beyond the sentence. Semiotica 64(1/2). discourse element serves this clarifying purpose, the Moon, Rosamund. 1998. Fixed expressions and idioms in English. A corpus-based approach. Oxford: Clarendon alleged ambiguity can still be removed via Press. pronunciation: the idiomatic reading, in accordance Strässler, Jürg. 1982. Idioms in English. A pragmatic analysis. with its form–meaning unity, requires shorter pauses Tübingen: Gunter Narr Verlag. and word durations than its literal counterpart. BEATRIZ RODRÍGUEZ-ARRIZABALAGA 493 IDIOMS Idioms An idiom can be defined as a set phrase, semantically to isolate the different matters of a chemical opaque, whose meaning cannot be deduced from the combination. meanings of its constituents. Although it shows an This can be explained by the fact that the semantic expressive and particular value, and sometimes a (lex- features selected are not the same ones whether the ical or syntactic) archaic structure, it behaves as a sin- constituents are used when they are free (if they still gle semantic unit. Expressions such as a night owl, a exist outside of the phrase) or as parts of the idiom. tempest in a teapot, not enough room to swing a cat, to For this reason, most of these utterances can be poly- leave no stone unturned, to knock on wood, to put the semous—they may have multiple meanings. They can cart before the horse are called idioms. be interpreted in either a literal or a figurative manner. A set phrase is a sequence that has experienced fix- Expressions are exocentric when the total of meanings ation, the process by which a multiword expression does not enable the whole phrase to be understood. For whose constituents are free (or were free when the example, an expression such as to be in the doghouse phrase was created) becomes an expression whose con- has two possible interpretations. When the speaker stituents are integral parts of it. Therefore, the elements says: “he is in the doghouse”, the expression is endo- lose their own meaning in favor of the global meaning centric if the subject is a dog, and exocentric if the of the expression and the phrase becomes a new lexical subject is a man (it means the man is in trouble with unit, independent, with its own new meaning. his wife). Thus, the endocentric utterance, whose Thus, idioms only correspond to one signified— meaning is compositional, is to be interpreted in a lit- they have a noncomplex, simple meaning. This means eral way (it is a common utterance). The exocentric that they can be substituted by simple words as they statement is fixed and to be interpreted in the figura- are expressive alternatives to a neutral concept that tive way; it is therefore an idiom. already exists in the language. For example, when hell When the phrase is exocentric, the speaker does not freezes over can be replaced by never. Idioms consti- conceptualize systematically the meaning of each con- tute a luxury of the lexicon in as much as it is possible stituent. For example, in an expression such as to have to communicate basically without them. Therefore, a green thumb, he will not imagine he has a colored understanding idioms shows that the speaker has a finger. For a nonnative speaker (or somebody unfamil- very good knowledge of the language in which he iar with the set phrase and its global meaning), his first expresses himself (most of the time, idioms are the last interpretation will be literal even though the phrase aspect that speakers identify when they learn a foreign must only be interpreted as a unit. Because they are set language). phrases, idioms are often used to create puns. One Because they are fixed phrases, idioms cannot vary constituent replaces another, or the other meaning of in their form, with the exception of some elements that the polysemous word is selected. This unexpected enable them to adapt to the specific context. For exam- aspect of the new phrase or the new signification caus- ple, when the idiom is a verbal unit, the space of the es humor, by reverting back to the literal form. For subject is free and the tense of the verb is to be chosen example, using the idiom to work for peanuts can lead according to the context: Bart/My brother/The cap- to the following play on words: ‘Charlie Brown has tain… broke the ice. The possessive can also be adapt- spent his entire life working for Peanuts.’ ed: I will keep my fingers crossed for you. Thus, in idioms, words are used figuratively, Idioms are semantically opaque. That is to say that through two main figures of speech: metaphor and they cannot be understood literally, in a compositional metonymy. This matter of fact gives life and richness to way. They cannot be interpreted from the total of each the language by enabling it to absorb new concepts that word that constitutes it. Even if you know and under- need to be expressed linguistically in a new way. The stand each constituent, the meaning of the whole use of metaphors is motivated by the incapacity of the phrase remains confusing. For example, white and lie human mind to reach a complete state of abstraction. are words that are common and easily understood, but To make an abstract notion accessible and to give it put together, they do not have the same meaning. The more weight, the speaker compares it with a physical compound must be learned as if it were a new word. object he knows. He uses basic schemes in order to Thus, it is impossible to deduce the global meaning conceptualize his experience and the external world from the total of the signifieds just as it is not possible (even if the analogy appears a little illogical): to bear 494  IDIOMS one’s cross, to have a monkey on one’s back, to have a used in the source language. It is not always possible frog in one’s throat. Thus, metaphors are based on sim- to meet this requirement. If the utterance in the target ilarity: metaphoric idioms appear in their simpler form language is not a set phrase, the translation cannot be when they are built around the adverbs ‘as’ or ‘like’: as satisfying. When translating idioms, three different sit- innocent as a lamb, as solid as a rock, like (two) peas uations arise. First, the opaque expression can be in a pod, like a bolt out of the blue. For its part, translated literally because the same idiom exists in metonymy is based on contiguity: one word or phrase both the source language and the target language. This is substituted for another with which it is closely asso- occurs with languages that are similar, such as English ciated. But it is also an erroneous perception, which and French: to read between the lines/lire entre les consists of confusing two different things, for example: lignes, to praise to the skies/porter aux nues, a drop in to put bread on the table (bread for money), to give a the ocean/une goutte d’eau dans l’océan, to burn the hand (hand for help), heads will roll (heads for people). candle at both ends/brûler la chandelle par les deux Sometimes, only one interpretation is possible bouts. Second, the idiom in the source language can because the phrase appears archaic, through a form that have an equivalent idiom in the target language, like is obsolete or belongs to an older stage of the language. any other simple unit with its equivalent in a diction- Most of the time, the archaism is lexical: some words ary: to have other fish to fry/avoir d’autres chats à that have now disappeared as free units still exist inside fouetter, before you could say Jack Robinson/en deux some idioms. These words are called fossils. For temps trois mouvements, as alike as (two) peas in a instance, hue is no longer used outside of the phrase pod/comme deux gouttes d’eau, the last straw (that hue and cry; it is the same for fro in to and fro (ancient breaks the camel’s back)/la goutte d’eau qui fait from), or shrift (confession), which never appears with- déborder le vase. Third, the idiom can be translated by out short (to give somebody/something short shrift). a simple word or a paraphrase if neither the same But before becoming opaque, those expressions idiom nor the equivalent exists in the target language. were motivated. The creation of idioms is influenced In this case, a part of the signified will be lost. by two main heritages: social heritage and cultural her- The expression of the same ideas varies from one itage. Even if the speaker perceives the world through language to another: when there’s something in the his language, the latter is made up of things that exist wind in English, there’s a conger eel under the rock in in the speaker’s social environment. Animals (blind as French (il y a anguille sous roche), and something is a bat, as the crow flies, to take a hair of the dog that boiling in the pot in Italian (qualcosa bolle in pentola). bit one, to kill two birds with one stone, to play pos- If you are in someone’s good books in English, you are sum), insects (a fly in the ointment, to have a bee in in somebody’s papers in French (être dans les papiers one’s bonnet, to get ants in one’s pants), flowers and de quelqu’un) or in somebody’s sleeve in Italian plants (to gild the lily, to nip something in the bud), (essere nella manica di qualcuno). And if you call a vegetables and fruit (to spill the beans, the apple of spade a spade in English, you call a cat a cat in French someone’s eye) are very present in idiomatic imagery. (appeler un chat un chat), and you say bread to the In the same way, if language mirrors life, men, and bread and wine to the wine in Italian (dir pane al pane society, it also arises from culture. Several idioms orig- e vino al vino). Thus, idioms have a supplementary inate in mythology (in seventh heaven—the antique and particular aspect that simple words do not have: firmament is composed of seven vaults, seven skies because they are motivated signs, and thanks to the that one reaches successively, to carry the weight of the images that are used, they give information about world on one’s shoulders, to rise from the ashes), in conceptions of the world considered by linguistic the Bible (not to know someone from Adam, to turn the communities. other cheek, like a lamb to the slaughter, to cast pearls before swine, to wash one’s hands off (something), to separate the wheat from the chaff), in literature References (Achilles’ heel, to bite the dust, the emperor’s new Bally, Charles. 1951. Traité de stylistique française. Geneva and clothes, it’s all Greek to me), and in history (to have an Paris: Librairie Klincksieck. axe to grind, when the balloon goes up, to haul some- Lakoff, George, and Johnson Mark. 1980. Metaphors we live one over the coals, it’s like the Black Hole of Calcutta, by. Chicago: University of Chicago Press. to win one’s spurs, to cross the Rubicon). Nida Eugene Albert. 1964. Toward a science of translating. Leiden: E.J. Brill. Because idioms cannot be understood literally, they Vinay Jean-Paul, and Darbelnet Jean. 1963. Stylistique com- cannot be translated literally either. The translation of parée du français et de l’anglais, (1995; Comparative stylis- idioms is characterized by a double problematics, as it tics of French and English: a methodology for translation). is necessary to express not only the meaning and the Paris: Librairie Marcel Didier. form but also the aspect of fixation of the expression CARINE MARRET 495 IGBO AND IGBOID LANGUAGES Igbo and Igboid Languages Igbo (Ibo, in old orthographies) is one of the three lexical differences are also in evidence, of course; but as major Nigerian languages. It is spoken by some 30 may be seen from the grammars of these dialect/lan- million people in present-day Imo, Abïa, Ebonyi, guages (e.g. Meier and Bendor-Samuel 1975), the basic Enugu, Delta, Rivers, and Anambra states, and also in grammatical features remain the same for all Igbo parts of Cross River, Akwa Ibom, Bayelsa, and Benue dialects. states of Nigeria. Igbo is now classified as belonging The history of Igbo language scholarship has been to the ‘Benue-Congo’, previously and variously to the intimately tied up with the matter of its dialects. In its ‘Kwa’, the ‘Eastern Kwa’. The current Benue-Congo first efforts, the Church Missionary Society promoted group includes not only Yoruba, Edo, and other small- ‘Isuama Igbo’ as the future national Igbo dialect, com- er Nigerian languages but most of the ‘Bantoid’ lan- parable to Oyo-Yoruba. The plan failed. The Roman guages of eastern, central, and southern Africa. The Catholic Mission chose the Önïca dialect for its trans- basis for this classification has been seriously ques- lations and other publications. This also proved unac- tioned, and has been shown in many cases to be unten- ceptable, as did ‘Union Igbo’, being promoted by the able. The proposition advanced by Kay Williamson Colonial Education Department. A Standard Igbo (1989) that Igbo was a language of the lower Niger (Igbo Izugbe) has now emerged, which combines Delta from where it spread back north to present-day generalized features of vocabulary and grammar from Igbo heartland has also been largely rejected by stu- the Önïca and Owere dialects. This is the standard dents of Igbo history and linguistics such as M. currently in use in the school system, and for most Angulu Önwüejiögwü (1980) or E. Nölue Emenanjö publications, although not without dissent from some (2001). eminent quarters. Igbo is a single language, with several dialects. The standardization of the Igbo orthography was Following the Biafra/Nigerian war (1967–1970), there key to this development. The first orthographies of the has been continuing pressure on some Igbo speakers to language were devised by repatriate Igbo missionaries formally dissociate their dialects from Igbo proper. of the Church Missionary Society, notably Rev. John This has been particularly so with respect to Ikwere, C. Taylor, working under Bishop Samuel Crowther, Ahoada (Ekpeye), and Ïka. In fact, the term, ‘Igboid’, who adopted Richard Lepsius’s ‘Standard alphabet’ in was coined in 1982 on an analogy with the other a modified form for Isuama primer (1859, 1860), and Benue-Congo languages (Edoid, Defoid, Platoid, etc.), the Vocabulary of the Ibo language (1882). A ‘war of in the face of an attempt to eliminate the ‘Igbo’ lan- the orthographies’, however, ensued when the Roman guage name altogether and replace it with the nonde- Catholic Mission began promoting its own system script term ‘Lower Niger’. Still, the effect of the use of based largely on a pan-Nigeria alphabet developed by the term Igboid has been to make Igbo just one other the International Institute of African Languages and member of a cluster of ‘languages’ previously under- Cultures (London) and recommended by the stood to be dialects of Igbo—Agbo, Ekpeye, Ikwere, [Colonial] Education Board, Lagos in 1931. The Izi, Ïka, Ogbah, and Ükwüanï/Enuanï. For these rea- resulting conflict between the Missions was not sons, the term ‘Igboid’ is now disregarded as a strictly resolved until 1961, with the adoption of the Official linguistic term since it misrepresents and understates (Önwü Committee) Orthography. This orthography the remarkable internal linguistic unity of Igbo among has gained general acceptability, owing in large meas- all the other languages of the Benue-Congo group. ure to the leadership of F.C. Ögbalü and the Society Broadly speaking, allowing for the special dialects men- for the Promotion of Igbo Language and Culture, tioned above, there are two major dialect zones in Igbo: which also codified the basic elements of standard Owere and Önïca, although the early missionaries pro- written Igbo. A New Standard Orthography introduced posed three: Isuama, Önïca, and ‘Bonny’ (Samuel in 1998 adopts the Önwü Orthography, but replaces Crowther 1882). In each of these zones, however, native the subdotted vowels with their more readily available speakers can identify sub-dialects, each marked by umlauted versions, as follows: some distinctive detail of phonology, e.g. the glottal fricative in Ngwa; the voiced implosive /≺/ replacing A, B, C/CH, D, E, F, G, GB, GH, GW, H, I, Ï, J, K, KP, the alveolar plosive, or the voiced glottal fricative /←/ KW, L, M, N, NW, NY, Ñ, O, Ö, P, R, S, SH, T, U, Ü, in Übakala, the ninth vowel in Izi. Other syntactical and V, W, Y, Z. 496  IGBO AND IGBOID LANGUAGES Interestingly enough, although Igbo is a tone lan- movement over an entire word or utterance, irrespec- guage using intonational pitch to differentiate mean- tive of word boundary. ing, most books (other than textbooks for teaching the language) are rarely tone-marked. The situation does Syllablic Structure not show any signs of changing any time soon. What follows is a brief review of the phonological The Igbo SYLLABLE consists of a vowel (V), a nasal and syntactical features of Igbo based on Standard consonant (/m/, /n/), or a vowel preceded by a conso- Igbo, with references to dialect variants as necessary. nant (CV). Unbound forms (other than pronouns) are either CV (as in di ‘husband’; ji ‘yam’) or VCV (as in ada ‘daughter’; uru ‘grief’); or else, they are CVV Phonology (e.g. gaa ‘go’) where the final V is a marker for the Igbo has eight phonemic VOWELS /a/, /ï/, /ö/, /ü/; /e/, imperative and is sometimes parsed as a verb stem + /i/, /o/, /u/. A ninth vowel (sometimes /a/, other times vowel suffix combination. Polysyllabic CVCV forms, /ë/) is encountered in some dialects, notably Ïka, except for initial nasal consonants (mmanü ‘oil’) and Ükwüanï, Önïca, and Izi. These vowels occur in two borrowings (fada ‘priest’) are invariably inflected sets: Set (1): /e/, /i/, /o/, /u/, and Set (2) /a/, /ï/, /ö/, /ü/ verbs; polysyllabic VCV(V) forms are either simple whose vowels, like the ninth vowel, are produced with nouns, or nominalizations from verbs form. an advanced tongue root. Igbo vowel harmony rules do not permit the co-occurrence of vowels from both sets Syntax within the same word. The CONSONANTS include the stops (/b/, /p/, /d/, The noun and the verb are the only significant gram- /t/, /j/, /ch/, /g/, /k/, /gw/ and /kw/); the implosives matical class in Igbo, pronouns, adjectives, adverbs, (/kp/, /gb/); the fricatives (/v/, /f/, /z/, /s/, and /sh/); the and similar classes being either technically nonexist- nasals (/m/, /n/, /ñ/, /ny/, and /nw/); the trill (/r/); the ent, or having (in some instances) no more than two or lateral (/l/); and the approximants (/y/, /gh/, /w/, and three members. NOUNS and nominals (with a few /h/). In some dialects, the bilabial plosives, the exceptions) have an initial VCV syllable structure; fur- palatals, and labialized velars are often aspirated, the ther syllables may be added, in every case ending in a fricatives sometimes nasalized. Aspiration of the plo- CV. General rules govern how two vowels may be sives has phonemic significance in certain dialects. adjacent to each other, and the conditions for assimila- Some fricatives and approximants are palatalized tion and tone change between them. Nouns are not across dialects: /v/ > /f/ > /h/ > /hw/; or /h/ > /s/ > /sh/ inflected. Polysyllabic nominals with initial nonnasal > /r/, even /l/. Other variations include the /l/, /r/, and consonants (e.g. gburugburu ‘all around; on all /n/ interchange in ala/ana (= earth); ala/ara (= breast), sides’) are usually adjectivals, adverbials, or ideo- and the use of /nw/ for /w/, especially in Ngwa and phones formed from base verbs. Nouns are never Ikwere (e.g. nwanyï/wanyï = woman). inflected for number or gender. VERBS are the only inflected class of words in the language. The verb has a primary root, (usually Tone System derived from the infinitive form), and either one or Igbo is a two-tone (High, Low) language. The vowel more secondary roots from other bound verb forms, or bears the syllable tone. Tone is also a factor in the lex- from extensional suffixes. There are two verb types: (i) icon, morphology, and syntax. It is the only difference the regular verbs and (ii) the so-called auxiliaries ( dï, between akwa [LH] ‘egg’ and akwa [LL] ‘bridge’; or ga, ka, ma, na) and copulas (bü, dï, nö, and wü). In between saa [HH] ‘be wide or outspread’ and saa compound (sometimes called complex) verbs (e.g. [LH] ‘answer; give reply to’. For nouns, the pattern of dapuo [<daa + puo] ‘fall’ + ‘have a gap or opening’, the final syllable determines their inherent tones. Most the end-verb carries the appropriate inflectional end- verb inflexions involve variations of tone. The primary ing in affirmative imperatives. distinction is still between verb forms with High or TENSES include Present, Past, Future, or Past Low tones in their (first) root syllable. Since the infini- Perfect. The Present is actually an absolute present and tive prefix (i/ï) is always high, H-tone verbs may or only a few verbs can express it. The present tense form may not accept low-tone inflected forms, whereas L- is the same as the verb root, without inflexional affix. tone verbs remain low tone in all their inflected forms. The Simple Past is formed with an rV inflexional affix Downstep, a phonetic lowering of a high tone after after the root, V being a reduplicate or copy of the verb another high tone, is a recurring feature of the lan- root vowel. The Past Perfect is formed with an rV suf- guage, and so also Downdrift in which there is a low- fix attached not to the verb root, but to the affirmative ering of high tones after a high tone in a terrace-like imperative forms, almost always a CVV. The Future 497 IGBO AND IGBOID LANGUAGES tense is formed with an auxiliary -ga and a harmoniz- verbal prefix. This is followed by the verb stem and ing a/e suffix after the verb root. The initial prefix is the required inflections (mood, tense, negation/affir- high, the second always low, except for verbs with HH mation) together with enclitics that extend the mean- tones in the imperative. ings of these verbs. Aspect Situation Today Igbo is the ‘mother tongue’ of some of Africa’s emi- ASPECT is distinguished from tense. In the present nent writers and artists: Chinua Achebe, Cyprian tense, aspectual inflexion may be used, for example, to Ekwensi, Christopher Okigbo, for example. Still, orig- reflect the duration of an action, the fact that it is on- inal works in Igbo by these established writers are hard going, or that it is customary. Similarly, in the past to come by. A kind of pan-Igbo cultural revival appears tense, the harmonizing aspectual prefix a/e + verb root to be growing, however, and with it, a renewed effort to + inflexional suffix [la/le; ne/na/; go/we, depending speak, write, and teach the language at all levels. on dialect] may be used to indicate fully completed Already several fictional works have been published (perfected) action. for the school market and the general public. Two major dictionaries have been recently published Affixes, Suffixes, and Enclitics (Echeruo 1998, Igwe 1999); others are in preparation. INFLECTIONAL AFFIXES are always attached to In addition, a major research effort is under way for the the verb, either pre- or post-positionally, to specify production of an archival Igbo dialect dictionary, in tense or aspect with a harmonizing prefix a/e; a suffix addition to the effort to create an Igbo metalanguage. -we (or dialect variants). A harmonizing rV suffix with the so-called stative verbs indicates continuing present References time or the past tense. To confirm a negative status for Achebe, Chinua. 2001. Echi dï Ime; Taa bü Gboo. Odenigbo a statement, the verb uses a harmonizing ghï/beghï Lecture Series, no. 3. Owerri: Assumpta Press. and their dialect variants. EXTENSIONAL SUFFIX- Bennett, Patrick R., and Jan P. Sterk. 1977. South-Central ES (sometimes called lexical suffixes) are a major ele- Niger-Congo: a reclassification. Studies in African ment in the syntax of the Igbo verb. (En)clitics are also Linguistics 8. 241–73. Crowther, Samuel. 1859;1860. Isuama-Ibo Primer. Revised and usually affixed to verbs, but they are excluded from enlarged by Rev. John Christopher Taylor. London: Church suffixes because they can also occur independently, as Missionary Society. free morphemes in association sometimes with a noun Crowther, Samuel 1882. Vocabulary of the Ibo language. phrase (NP), and other times with a verb phrase (VP). London: Society for Promoting Christian Knowledge. Echeruo, Michael J.C. 1998. Igbo–English dictionary: a com- prehensive dictionary of the Igbo language, with an Word Order English–Igbo index. New Haven: Yale University Press. Emenanjö, E. Nolue. 1987. Elements of modern Igbo grammar: The basic WORD ORDER (in declarative sentences) is a descriptive approach. Ibadan: University Press Ltd. S(ubject)–V(erb)–O(bject)-1-O(bject)2, or more exactly, Hair, P.E. 1967. The early study of Nigerian languages: essays N(oun)P(hrase)1 + (VP [+ NP2]). The O1 and O2 nouns and bibliographies. Cambridge: Cambridge University Press. may be the direct object of the verb, or an inherent or Igwe, G. Egemba. 1999. Igbo–English dictionary. Ibadan: University Press, Plc. cognate complement. Transitivity in the verb is difficult Meier, Paul, and Inge, and John Bendor-Samuel. 1975. A gram- to establish because of the language’s extensive use of mar of Izi: an Igbo language. Norman, OK: Summer verb complements—inherent, cognate, or otherwise. Institute of Linguistics, University of Oklahoma. Nwachukwu, P.A. 1987. The argument structure of Igbo verbs. Lexicon Project Working Paper, 18. Cambridge, MA: MIT Noun Phrases Center for Cognitive Science. Nwachukwu, P.A. 1981. Towards an Igbo literary standard. The NOUN PHRASE, whether NP1 or NP2, is made London: Kegan Paul. up of the nominal head and the qualifiers that come Oluikpe, Benson. 1978. Igbo transformational syntax: the after it, with the exception of the numeral, otu ‘one’, Ngwa example. Onitsha: Africana-FEP Publishers Ltd. which always precedes the head. The modifier may be Önwüejiögwü, M. Angulu. 1977. Some fundamental problems a word, phrase, or clause, in that order. Qualifiers in the application of lexostatistics in the study of African languages. Oduma 3(2). 29–36. come after the Igbo noun head. Önwüejiögwü, M. Angulu. 1980. An Igbo civilization: Nri kingdom and Hegemony. London: Ethnografica. Verbal Phrases Williamson, Kay. 1989. Benue-Congo Overview. ed. by John Bendor-Samuel, The Niger-Congo languages, 247–74. The VERB PHRASE (VP) is a complex inflected unit, Lanham, MD: University Press of America. marked by an initial (nonmandatory) auxiliary or a MICHAEL J.C. ECHERUO 498  INDIA India India represents an astonishing array of linguistic Two branches of the Austroasiatic language family diversity with four language families, and more than are also found in India: the major members of the 600 languages as well as numerous geographical, Munda subfamily are Santhali, Mundari, and Ho, social, ethnic, religious, and rural varieties or which are spoken in southern Bihar, northeastern dialects. The subcontinent competes with Papua New Andhra Pradesh, and broader areas of West Bengal in Guinea in terms of the sheer numeric weight of its India. The Mon-Khmer branch is represented by languages and varieties, a rural–urban dichotomy, and Khasi, which is spoken in Meghalaya state of India. varied social structures. For these reasons, India is Tibeto-Burman is represented by many languages, often described as ‘the linguistic laboratory of the spoken in the subcontinent’s northeastern borderlands. World’. Consequently, one finds a complex array of While Tibetan or Bhotiya languages are spoken in linguistic situations in terms of the extent of bilin- India (in Laddakh and Sikkam), Bhutan, and Nepal, gualism, ranging from diglossic bilingualism to the major concentration of the Gurung-Tamang group bilingualism based on three of the five highest rank- is present in central to eastern Nepal. ing languages in the World (Hindi, English, and Nineteen languages are recognized by the Indian Bengali) according to the number of speakers. constitution. In addition, there are numerous dialects The magnitude and scale of linguistic diversity and and languages. The number of vernaculars/dialects parameters of language usage are often beyond the reported ranges from 216 to over 1,500 (1,652 mother comprehension of speakers who are accustomed to tongues self-reported in the 1961 census; 216 mother western-style monolingualism. tongues with more than 10,000 speakers each reported in the 1991 census). Linguistic Profile The linguistic situation and communication net- working in India can be represented diagrammatically Approximately one in every six inhabitants of the by a pyramid-type structure, shown in Figure 1. world (that is 1.0 billion people) speaks at least one of the Indian languages. Indian languages belong to four language families: Indo-European, Dravidian, Languages and Modes of Communication Austroasiatic, and Tibeto-Burman. Indo-European, with Hindi and English are the two national and link lan- 12 major and three minor languages, represents the guages of India. While Hindi is the language of the most influential language family both in terms of the masses in the northwestern and north-central part of number of speakers and social significance. India, English is the pan-Indian language of the edu- The 12 Indo-Aryan languages of the Indo-European cated elite. Thus, English and Hindi represent the peak language family are: Hindi, Assamese, Oriya, Marathi, of the pyramid. Konkani, Gujarati, Urdu, Punjabi, Sindhi, Bengali, As noted earlier, Hindi, even without combining Kashmiri, and Nepali. Hindi and Bengali are ranked with Urdu, has the second largest number of speakers second and fifth in the world, respectively, in terms of among the languages in the world and is the lingua the number of speakers, according to the World franca of India and of the subcontinent. The great Almanac (2002: 447). Approximately 80% of the popu- Moghul empire in the past, and Hindi films and the lation of India speak languages of the Indo-European mass media in the twentieth century, have played a sig- stock. Kashmiri (India and Pakistan) belongs to the nificant role in the spread of Hindi-Urdu, not only in Dardic sub-branch of Indo-Aryan. The three minor lan- India but also across the subcontinent. Hindi and Urdu guages are as follows: Shina, Dogri, and Bhili. The are mutually intelligible but are written in two different Indo-European languages not only dominate India but scripts—Hindi is written in the Devanagari script and also all the countries of the subcontinent (i.e. Pakistan, Urdu is written in Perso-Arabic script. There is also a Nepal, Sri Lanka, and Bangladesh). difference in literary affiliation—Urdu is associated The second major family are the Dravidian lan- with Persian and Arabic literary traditions, while Hindi guages, which include four major languages: Tamil, is tied to the Sanskrit tradition indigenous to India. In Telugu, Malayalam, and Kannada. These four lan- addition, Urdu borrows its technical and learned vocab- guages are spoken in South India. Tamil is spoken both ulary from Persian and Arabic, whereas Hindi borrows in India and Sri Lanka. the same kinds of terms from Sanskrit. 499 INDIA National Languages Hindi/English ‘Scheduled’ Languages Assamese, Bengali, Gujarati, Kannada, Kashmiri, Konkani, Malayalam, Manipuri, Marathi, Nepali, Oriya, Punjabi, Sanskrit, Sindhi, Tamil, Telugu, Urdu Languages with Widespread Currency 47 languages used for primary education 87 used in print media 71 used in radio 13 used in films 13 in state-level administration Local Vernaculars over 114 recognized varieties 216 “mother tongues” with more than 10,000 speakers were recorded in the 1991 census Figure 1. Linguistic situation in India. ‘Scheduled’ languages (henceforth, state lan- should be noted that there are a number of problems guages) are spoken predominantly in their respective with these figures (and other census figures as well). A states: see Table 1 below. Hindi, along with English, is claimed language may be a language or a nonlan- the only state language that is spoken in more than one guage—a reflection of regional, religious, caste, state. Urdu is the official language of the state of social, ethnic, literary and script affiliation, and even Jammu & Kashmir. However, not all ‘scheduled’ lan- occupation. For example, it is not unusual for an iron- guages are spoken in a particular state. Sindhi is not smith to report Lohari ‘ironsmith’ as his mother the official language of any Indian state. Kashmiri is tongue or for a villager to report a dialect as simply the spoken in the state of Jammu & Kashmir but the offi- name of his village or caste. Alternatively, a language cial language of the state is Urdu. Finally, rural might be a speaker’s perception of language/dialect on dialects (mother tongues or vernaculars) are shown at a language continuum, thus blurring language vs. the base of the pyramid. dialect; or language vs. language distinction (in lan- As pointed out earlier, Modern Tamil, Telugu, guage contact situation). Kannada, and Malayalam are Dravidian languages and The label ‘mother tongue’, which is the term used are primarily spoken in the South, while Indo-Aryan by the census, may also be ambiguous. An individual languages, the descendants of the oldest documented questioned may interpret this term as referring to his Indo-Aryan language, Sanskrit, are primarily spoken or her cradle language, parent tongue, language in the North. Dravidian languages are the descendants ordinarily used, or prestige language only. A prestige of the proto-Dravidian languages. language such as English is often over reported, Sanskrit is also recognized as an official language while stigmatized varieties (e.g. immigrant lan- of India and is the language of India’s cultural and guages such as Saraiki from Pakistan) are underre- intellectual tradition. English is the language of mod- ported. An illustration of these problems is the fact ern intellectual communication and is the official lan- that almost 50,000 speakers reported Sanskrit as their guage of three states and all ‘union territories’. mother tongue (see Table 1). It seems extremely Before continuing with a discussion of the language unlikely that Sanskrit is the cradle language of any- situation in India, it is worthwhile discussing the one. However, its prestige and daily use by some sources of figures in Table 1—the census of 1991. Indians have apparently led them to claim it to be Because the census asks for self-report of languages, it their mother tongue. 500  INDIA TABLE 1 Scheduled Languages of India Name Language Family State/spoken in Number of Mother Tongue Speakers (1991 Census: Govt. of India) Assamese Indo-Aryan Assam 13,079,696 Bengali Indo-Aryan Bengal 69,595,738 English1 Indo-European; Meghalaya, 178,598 Germanic Nagaland, Tripura Gujarati Indo-Aryan Gujarat 40,673,814 Hindi Indo-Aryan Hindi belt: 337,272,114 Bihar, Chattisgarh, Rajasthan, Haryana, Delhi, Himachal Pradesh, Jarkhand, Madhya Pradesh, Uttar Pradesh, Kannada Dravidian Karnataka 32,753,676 Kashmiri Indo-Aryan Jammu & Kashmir2 56,690 Konkani Indo-Aryan Goa 1,760.607 Malayalam Dravidian Kerala 30,377,176 Manipuri Tibeto-Burman Manipur 1,270,216 Marathi Indo-Aryan Maharashtra 62,481,681 Oriya Indo-Aryan Orissa 28,061,313 Punjabi Indo-Aryan Punjab 23,378,744 Sanskrit Indo-Aryan No state 49,736 Sindhi Indo-Aryan Metro areas of western India 2,122,848 Tamil Dravidian Tamil 53,006,368 Telugu Dravidian Andhra Pradesh 66,017,615 Urdu Indo-Aryan Jammu and Kashmir 43,406,932 1 The number of mother tongue speakers of English decreased in 1991! According to 1981 census, the mother tongue speakers of English were 202,000. 2 These figures are not complete as no census was taken in J&K in 1991. Urdu and other scheduled languages are also spoken out- side their respective states. Sources and Processes of Multilingualism: Sanskrit Role of Sanskrit, Persian, and English It is well known that Sanskrit language is the single major force responsible for the formation of Indian Three millennia of language contact have led to languages and literatures. So great is the influence of the convergence of four language families in the form Sanskrit on Indic languages that almost every literary of Aryanization of Dravidian languages and and language-related activity is affected by it. As the Dravidianization of Aryan languages. In addition, language of the Hindu scriptures, and of the epic and Persian, which was the official language of India dur- classical literature, Sanskrit has traditionally constitut- ing the Moghul empire, and the newest arrival, ed the medium of philosophical and technical litera- English, have added greatly to the mixed character of ture throughout South Asia. A number of other factors Indian languages, particularly Hindi. Hence, it is not such as past royal patronage and the promotion of uncommon for a Hindi speaker to switch between Sanskrit education by the Brahmins from the sixth Persianized Hindi, Sanskritized Hindi, and Anglicized century BC onward have reinforced its prestige. Hindi. In short, three contact languages have played a Consequently, Sanskrit has played a leading role not particularly important role in the history of India: only in the process of Aryanization of Dravidian and Sanskrit, Persian, and English. Each has served as a other languages but also in forming the ethos of India prime vehicle for language contact and bilingualism, and other South and South East Asian countries to a both historically and in the present day. varying degree. 501 INDIA For example, the effects of borrowings from Sanskrit source of designative cultural and technical innovations on the lexical and grammatical systems of Dravidian in South Asia and beyond. Consider, for example, that languages have been profound. Murray Emeneau and the technical vocabulary is drawn either from Sanskrit Thomas Borrow (1962) note the tendency ‘for all four or English: the terms for prime minister (pradhaan of the Dravidian literary languages in the South to make mantrii  pradhaan ‘head, principal’+mantrii ‘minis- literary use of the total Sanskrit lexicon indiscrim- ter’), president (raashTrapati  raashTra ‘country’  inately’. So massive has the influence been that it is pati ‘husband’), radio (aakaashvaaNii  aakaash ‘sky’ hard to utter more than a few sentences in these lan-  vaaNii ‘voice, speech’), and TV (duurdarshan  guages without using a word borrowed from Sanskrit. duur(Hindi) ‘far’  darshan ‘sight’). The words given Over a century ago, Western scholars subscribed to the in the parenthesis are drawn from Sanskrit. view that the Dravidian languages descended from Furthermore, Sanskrit plays an important role in the Sanskrit, although Robert Caldwell, in his classical process of language modernization of South East Asian work, dispelled this misconception. About 20% of the languages. noncultural part of basic vocabulary in literary Dravidian languages is loaned from Indo-Aryan mostly Persian from Sanskrit. This proportion peaks to 50–60% in Language contact among Persian, Indo-Aryan, and some languages due to shared cultural beliefs (e.g. Dravidian languages (Kannada and Telugu) formally Brahmanic values). This resulted not only in the began with the annexation of Punjab by the Turkish replacement of content and function words but also had ruler Mahmud Gaznavi (twelfth century). The lan- a substantive impact on the phonological, morphologi- guage contact situation with Persian was different from cal, and syntactic structures of the Dravidian languages. the Sanskritization of Indian languages. Persian was Furthermore, the study of VyaakaraNa (grammar) viewed as an imposed language in much of India; thus, in Sanskrit was viewed as an area of the highest lin- with Moghul patronage, Persian became a marker of guistic sophistication and a testing ground for various Islam. As a consequence, after independence in 1947 linguistic theories. The Sanskrit grammatical tradition when India and Pakistan became independent nations, resulted in grammatical works such as Panini’s gram- the influence of Persian continued to regenerate in mar, the Ashtaadhyaayii, which is still regarded as Pakistan, a Muslim nation, while it reached a state of ‘one of the greatest monuments of human intelligence’ fossilization in India. In contrast, after independence, not only in the history of Indic grammatical tradition Sanskrit began to regenerate its original impact in India but also in linguisitics. due to its association with Indian culture (particularly The Dravidianization of Indo-Aryan languages with Hindu religion). Just as extensive Sanskritization also took place, although the impact of Dravidian lan- played an important role in the separation of guages on Indo-Aryan languages was not nearly as Malayalam from Old Tamil, extensive Persianization extensive; nevertheless, this mutual influence (on lex- together with Arabicization led to the separation of icon as well as grammar) set the stage for leaky Urdu in Pakistan from Hindi-Urdu in India. Similarly, boundaries among the languages of India whether they heavy Persianization and Arabicization of Hindi led were genetically related or not. Western grammarians like George Hadley (1772) and Not only this, Sanskrit also become a marker of others (Schultz 1745) to claim that Persian must be the caste identity, which in turn gave rise to diglossic mother tongue of Hindi. Others thought that the ances- bilingualism through Brahmin speech in Dravidian tor of Hindi must be a Semitic language, perhaps languages, particularly in Tamil. The use of Sanskrit in Arabic or Hebrew. The concept of Indo-Aryan lan- topical domains such as religion, philosophy, poetics, guages and the genetic hypothesis of languages was science, technology, and mathematics in Dravidian, as still a distant reality in the nineteenth century. As point- well as in Indo-Aryan-speaking areas, is particularly ed out earlier, Hindi and Urdu in their spoken style are noteworthy. Hence, Sanskrit became the single most essentially the same language, approximately parallel important marker of Indian culture—both in the north to British and American English. and in the south. Persianization of Indian languages was not restrict- The role of Sanskrit in promoting bilingualism and ed to the lexicon (content and function words). The language modernization inside and outside of India (e.g. impact of Persian borrowing on the grammar of Indian in South East Asia) to this day is particularly striking. languages was as significant as the influence of Today, the All India Radio relays news in Sanskrit and Sanskrit. The morphological processes such as reverse even newspapers are published in the language. Just as compounding (sher-e-panjaab ‘the tiger of Punjab’ terms for technological innovations are created from rather than the unmarked pattern—panjaab kaa sher Latin and Greek for western languages (television= ‘Punjab’s tiger), inflectional morphology (plural Greek tele-‘far’, Latin –vision ‘seeing’), Sanskrit is the markers), and word compounding with Persian 502  INDIA became part of modern Indo-Aryan languages and so once belonged to either Sanskrit or Persian. The uses of did conjunct verb construction, complementation English, parallel to its predecessor contact languages— (with ki ‘that’), and conjunction with noun phrases. Sanskrit and Persian—has led to the Englishization of Domains such as law and regulation, sports, and Indian languages. On the other hand, English has business belong to Persian-mixed Hindi to this day in undergone significant changes locally to carry the com- India. A highly Persian-Arabic style is widely used in municative burden of Indian society. The process of the modern-day legal register. Also, newspaper report- nativization of Indian English continues to this day. ing, a business section that includes a Price-index English has colored the phonology, morphology, report, share market, and reports dealing with eco- and syntax of Indian languages. Hindi, for instance, nomics such as budgetary reporting, and so on, incline has borrowed several consonant clusters such as st-, toward the Persianized style. sk-, sl-, and phonemic segments such as /f/ and /z/ as the result of large-scale infiltration of English lexical items, such as station, school, slow , fail, etc. Even at English the level of language, which is less prone to change, Although English is a newcomer in the linguistic i.e. syntax, Indian languages have not managed to scenes of India, it has already registered some signifi- escape the influence of English. The preference for cant results from the viewpoint of the sociology of passive over active construction is attributed to the communication, in general, and cross-linguistic com- influence of English. munication, in particular. At present, it serves as one of Turning to the first layer of the pyramid, along with the most important sources of pan-Indian bilingualism English, Hindi provides a major contact language link or pan-Indian communication. Consequently, it has within the communicative network. The heart of this added greatly to the already mixed character of Indian link consists in the Hindi-Hindustani-Urdu-Punjabi languages. What is noteworthy is that mixing with core/axis. This axis forms a giant speech community English is an important linguistic feature of India. with direct links to Bengali in the East, Gujarati and When the introduction of English to the Indian lin- Marathi in the West, and Telugu and Kannada in the guistic landscape began with the dawn of the British South. In contact with these languages and other colonial era, English began to develop roots in Indian regional varieties, Hindi has developed its own region- education. A blueprint for India’s educational policy al varieties, for example: Mumbai Hindi (Bombay was laid down in Lord Macaulay’s Minute (Feb. 2, Hindi, Bindi), Kalkatiya Hindi (Calcutta Hindi), 1835). Macaulay’s stated mission for the British Raj of Madrasi Hindi (Madras, renamed Chennai, Dakkani creating ‘a class of persons, Indian in blood and color, Hindi ‘southern Hindi’). but English in taste, in opinions, in morals and intel- Other factors such as education, military, migration, lect’ introduced English education to India; his attitude literature (particularly devotional literature), religion not only toward Indian languages but also toward the (pilgrimages), popular and electronic media, and trade languages of Southwest Asia—particularly, Arabic. play an important role in societal bilingualism or I have no knowledge of either Sanskrit or Arabic. But I multingualism in India. have done what I could to form a correct estimate of their value...I am quite ready to take the oriental learn- Education and Government Policies ing at the valuation of the orientalists themselves. I have The language policies of the government of India are never found one amongst them who could deny that a very conducive to the promotion of the language rights single shelf of a good European library was worth the of minority languages and the advancement of linguis- whole native literature of India and Arabia. tic diversity and pluralism. This is evident from the More than one and a half centuries later, English has facts regarding languages represented in the third tier overcome its status as merely the language of the colo- of the pyramid. The number of languages used in pri- nial power, and has become an integral part of the mary education is 47; 87 in the print media; 71 in radio Indian linguistic mosaic. Contrary to the most popular and broadcasting; 13 in films; and 13 in state-level preindependence consensus that Hindi would dethrone administation. In addition, the introduction of the the English language after independence, English not Three Language Formula in education is yet another only continued to flourish in the educational and offi- notable feature of national policy. This formula calls cial network of India but also became one of the offi- for trilingualism or quadrilingualism in education. In cial languages of the nation and thus continues to enjoy addition to learning of the two national languages— the patronage of the Indian elite. Although the inci- Hindi and English—students are expected to learn a dence of bilingualism with English is still very low, third language apart from their native tongue. For English has acquired domains such as education, law, example, in the Hindi-Urdu-Punjabi belt, students are government, media, and science and technology, which expected to learn one of the four Dravidian languages 503 INDIA (Tamil, Telugu, Kannada, and Malayalam). Although that India ‘babbles’ chaotically in ‘hundreds of the debate concerning the effectiveness of the Three dialects’. Such labels or expressions reflect a lack of Language Formula is still brewing, the underlying understanding of the complex but structured network merit of the educational policy in the promotion of of linguistic communication and bilingualism in India. mulitilingualism is hardly questionable and best repre- The main focus of this section is to highlight the results sents the multilingual character of the nation. of prolonged and recent bilingualism in the region. Similarly, the National Academy of Letters (The Sahitya Akademi) of the Government of India and its Indian Multilingualism: A Natural Phenomenon regional counterparts promote literary activities at There is no doubt that there are linguistic rivalries and least in the 19 ‘scheduled’ languages. conflicts in India, even language vs. dialect conflict. Centuries-old coexistence and an ongoing process of Bridging the Rural vs. Urban Divide: Literature, convergence have led to an unmarked pattern of wide- Popular Media, and Trade spread naturalistic linguistic convergence, rather than Unchanged for centuries, rural populations constitute separation, dominance, and disintegration. Large-scale the heart of India. According to the 1991 census, for diffusion of linguistic features across genetic and areal instance, most of the population of India (about boundaries has resulted in mutually feeding relation- 77–78% of the total population) lives in more than half ships and reciprocity. Although the incidence of a million (627,000) villages and speaks in numerous ‘antagonistic bilingualism’ of the sort witnessed in vernaculars (see the base of the pyramid in Figure 1). Belgium and other parts of the world cannot be ruled The most obvious linguistic vehicles for reaching rural out, its incidence is very limited in degree and scope. India are either Hindi or the regional languages and In short, one of the defining features of India is one of their local vernaculars. The incidence of literacy in long-term stable bilingualism and linguistic accom- English is not significant in rural India. Thanks to liter- modation. This situation led Caldwell (1903) to pro- ature in the past and popular media at present, the pose implicitly that India as a ‘linguistic area’—that boundaries between rural vernaculars and Hindi have is, an area in which genetically distinct languages become very fluid. Historically, literature (e.g. the show a remarkable level of similarity and diffusion at devotional poetry tradition—poets such as Kabir, the level of grammar. For instance, retroflex sounds Tulasi, Surdas, and Meera Bai) has played a major role (produced with the tip of the tongue curled upward) in neutralizing such boundaries and bringing the influ- are the integral part of the languages of India and are ence of regional languages (from East to West and thus dative subject constructions: South to North; bidirectional) and rural varieties into Hindi. The consequence was a mixed speech, which is mujhe kitaab pasand hai termed ‘sadhukari bhaashaa’ (the language of saadhus me book liking is. and saints free from any prescriptive norms). At pres- ‘I like the book’. ent, the most powerful and vital force for bridging the gap between the urban vs. rural (and regional) divide is Language Attitudes and Linguistic Accommodation the Hindi film industry and the mass media. These Why has diffusion between genetically distinct lan- sources of influence offer a unique appeal in terms of guages taken place in India to a degree rarely witnessed cinematic techniques, dance, drama, and music, and in other parts of the world? The answer lies not only in even constitute a viable marketing alternative to language-internal developments but also in linguistic Hollywood in the world of entertainment in India and attitudes that are in turn associated with the ‘worldview’ outside India. The reach of the Hindi media indeed of Indians since ancient times. William Bright (1984) extends well beyond one sixth of the world’s popula- characterizes linguistic attitudes in India as an ‘accept- tion that inhabits India, reaching members of the Indian ing attitude, which has brought about the assimilation of diaspora worldwide. Consequently, mutual intelligibil- features from Dravidian, Indo-Aryan, Islamic, and even ity between regional languages and rural varieties of Christian and European cultures into a single system, Hindi is growing steadily. At least this is true of spoken complex, but integrated’. Thus, linguistic accommoda- Hindi—caltii Hindi ‘colloquial Hindi’ or bazaar or tion is another important feature of Indian bilingualism. Dakkini Hindi. Nevertheless, some barriers do remain. Stable multilingualism and multiple identities Salient Features of Indian Multilingualism Multilingualism is not a borderline phenomenon As is self-evident from the above discussion, it is clear in India nor is it restricted to either educated or busi- that linguistic diversity is a hallmark of India. India is ness communities. Multiple languages and multiple often labeled the ‘Tower of Babel’, since it is remarked language identities are defining features of Indian 504  INDIA multilingualism that reveal the dynamics of language notes leisurely travel with religious overtones, whereas usage and a constant negotiation of identities. For the latter marks any ordinary travel. The English-based instance, the simple act of a social encounter and asso- tour karnaa has added yet a new semantic dimension, ciated greetings is likely to begin with the choice of which expresses the concept of business travel which is one of three modes of greetings, which reveal the reli- contrary to what the verb tour conveys in English. gious affiliation of speaker, listener, or both in addition to conveying the social ritual of greetings: namaste or Freedom of choice namaskaar (Hindu), aadaab (Muslim), sat sri akaal Linguistic diversity highlights freedom of choice of (Sikh), drawn from Sanskrit, Perso-Arabic and speech, which is not just a recent, twentieth-century Punjabi, respectively. Not only this, a Hindu speaker phenomenon in India, but a centuries-old tradition. may decide to use raam-raam and jai maataa dii to This attitude has not been seen in many parts of the express regional (rural) and ethnic affiliation, respec- world. For example, French, Chinese, and some other tively, within Hinduism. Similar conditions hold true major languages of the world were imposed on others of Muslims and members of other religious and social as part of a movement to eradicate rural and regional groups. language varieties. Verbal Repertoire Code-mixing and code-switching are natural phenom- Scripts ena in the life of a bilingual and are thus all-pervasive Indic scripts constitute yet another example of ‘India in India. Bilingual societies organize their linguistic as a linguistic area’. Both Indo-Aryan and Dravidian repertoire differently from monolingual societies. No languages are written in various scripts, which one language is viewed as suitable for all communica- descend from the Brahmi script (third century AD). tive occasions. This has led to the emergence of mixed Indian scripts are phonetic in nature and there is a fair- languages such as Hinglish (Hindi+English). The ly regular correspondence between the letters and their extent of the impact of English is so great that it has pronunciation. A number of Indian languages such as caused a great deal of alarm among purist speakers of Hindi, Marathi, Nepali, and Sanskrit are written in the Indian languages and is seen as threatening their inde- Devanagari script, whereas it is not uncommon to pendent existence. write a language in different scripts. For example, The phenomena of code alternation have brought Punjabi is written in at least three scripts: Gurmukhi, profound changes to the formal grammar of Indian Devanagari, and Perso-Arabic. languages. Consider, for example, the case of Causatives in Hindi. Hindi Causatives are derived by the morphological process of suffixation. By adding References the suffixes -aa- and -waa-, the first and the second Annamalai, Elayaperumal. 2001. Managing multilingualism in causative verbs are formed as in: paR ‘to study/read’, India: political and linguistic manifestations. New Delhi: paRaa ‘to teach’ (lit. ‘cause to study’), and paRwaa Sage Publications. ‘cause x to cause y to study/read’. Mixing with Bhasham, Arthur L. 1954. The wonder that was India. New York: Grove Press Inc. English has resulted in the introduction of a new class Bhatia, Tej K. 1982. English and vernaculars of India: contact of transitive/causative verbs of a mixed type, i.e. study and change. Applied Linguistics 3(3). 235–45. karnaa ‘to research (on a topic); lit. study+ karnaa ‘to ———. 1985. Religious-colonial models of language and early do study’), teach karnaa ‘to teach’, study karwaanaa Hindi grammars. Lingua 65. 123–34. ‘to guide research’, teach karwaanaa ‘to guide teach- ———. 1999. Advertising in rural India: language, marketing communication and consumerism. Tokyo, Tokyo Press. ing’. The verb study karnaa ‘to research’ is not just a Bright, William. 1960. A case study of caste dialect in Mysore. paraphrase or translational equivalent of Hindi paR ‘to Indian Linguistics 21. 45–50. study/read’; the English-based verb expresses the Bright, William, A.K. Ramanujan. 1964. Sociolinguistic and meaning ‘to research or study a topic from the view- language change. Proceedings of the Ninth International point of research’ and the causative counterparts, i.e. Congress of Linguistics. Caldwell, Robert P.A. 1903. A comparative grammar of the study karwaanaa ‘to guide research’, teach kar- Dravidian or South-Indian family of languages. Madras: waanaa ‘to guide teaching’ highlight the ‘facilitative’ Oriental Books. (i.e. help the causee to do a particular act) while the Emeneau, Murray B., and Thomas Borrow. 1962. Dravidian Hindi causative marks ‘compulsive’ meanings. borrowings from Indo-Aryan. University of California Before the introduction of English, Hindi had two Papers in Linguistics, 26. Berkeley: University of California. conjunct verb expressions with the meaning ‘to travel’: Gumperz, John. J, and R. Wilson. 1971. Convergence and cre- yatraa karnaa and safar karnaa. The former is Sanskrit- olization: a case from the Indo-Aryan/Dravidian border in based and the latter is Persian-based. The former con- India. ed. by D.H. Hymes, Pidginiztion and creolization of 505 INDIA languages, 151–67. Cambridge: Cambridge University Pandit, Prabodh B. 1972. India as a sociolinguistic area. Poona: Press. University of Poona. Hadley, George. 1772. Grammatical remarks on the practical Schiffman, Harold, 1999. South and Southeast Asia. ed. by J.A. and vulgar dialect of Indostan language. Menston: The Fishman, Language and ethnic identity 431–43. Oxford, Scholar Press. Oxford University Press. Kachru, Braj B. 1983. The Indianization of English: the English Schultz, B. 1745. A grammar of the Hindostan language. language in India. Delhi: Oxford University Press. Saxony: Hall. Kachru, Yamuna, and Tej. K. Bhatia. 1978. The emerging Sridhar, S.N. 1981. Linguistic convergence: Indo-Aryanization ‘dialect’ conflict in Hindi: a case of glottopolitics. of Dravidian languages. Lingua 5. 199–220. International Journal of the Sociology of Language 16. 47–56. TEJ K. BHATIA Khubchandani, Lachman M. 1997. Revisualizing boundaries: a pluralism ethos. New Delhi: Sage. Krishnamurti, Bh., and Aditi Mukherjee (eds.) 1984. Moderni- See also Bilingualism; Dravidian; Farsi; Hindi- zation of Indian languages in news media. Hyderabad, India: Urdu; Indo-European 3: Indo-Iranian Languages; Department of Linguistics, Osmania University. Language: Contact—Overview; Panini; Sanskrit Indian Ocean Creoles In the Indian Ocean, several French-lexicon creoles are five different forms, some even having suppletives. spoken by between one and two million people. They The language, furthermore, has an attributive copula are usually divided into two mutually unintellibible and inherited from French, lacks bimorphemic interroga- probably (largely) historically unrelated groups: (1) tives, and has very little of the incorporation into Réunionnais (Réu) of Réunion island, and (2) the so- nouns of etymological articles, so characteristic of called Isle de France (IdF) creoles, which emerged in French creoles in general. The postverbal placement of Mauritius (Mau), and later (eighteenth to nineteenth the sentence negation also aligns it closely with its lex- centuries) spread to the Seychelles (Sey), Rodrigues, ifier. Moreover, Réu retains grammatical gender, as the Chagos Islands, and elsewhere. opposed to virtually all (other) creoles. Some varieties Réu differs both historically and structurally from of Réu also retain front rounded vowels. other creoles in several respects. Early settlement of the It should be noted, however, that claims regarding island was characterized by a relatively strong compo- the structure of Réu are difficult to make, and when nent of French speakers, and a rather slow influx of made, they should be treated with some caution, for alloglotts. Whites made up a majority of the Réunionnais the island is dialectally very fragmented, with remark- population until about 1715, i.e. half a century after the able variation in speech from one area to another. The beginning of permanent settlement of the island. As late fact that there is a continuum between regional French as in 1690, only one family in six did not include a native and Réu among the island’s almost 750,000 inhabi- Francophone. Moreover, in search of subsistence, many tants also makes it impossible to give the number of members of the white proletariat were forced to take up speakers of either. employment in the fields alongside (former) slaves. As In contrast to Réu, the IdF varieties are typological- opposed to other plantation societies, where the eco- ly rather similar to Caribbean creoles. There is some nomic power was always in the hands of a small white evidence that a stable IdF had emerged within 50 years élite, a large proportion of the white population on after the French settlement of Mauritius in 1721. Réunion has always been as poor as their dark-skinned Although later slave imports were predominantly compatriots. Together with a high rate of inter-racial from Madagascar, and later East Africa, West Africans marriages, this led to a language less distant from French (speaking e.g. Wolof, Bambara, and Fon) were among than most creoles. In fact, most non-Francophone lin- the first arrivals in Mauritius, who may have con- guists see Réu as a semicreole (comparable to, e.g. tributed to making the language so similar to Afrikaans) rather than as a creole proper, while in the Caribbean creoles. Other substrate languages include French-speaking scholarly tradition, it is treated as a Bengali and Tamil. fully fledged creole on a par with e.g. Haitian. Occupied by the British for 150 years, English is As opposed to IdF, Réu maintains significant parts Mauritius’s official language, but the position of of French verbal morphology. Réu verbs have up to French (L1 of a small group) remains remarkably 506  INDIAN TRADITIONAL GRAMMAR strong at all levels, and it may in practice be said to Pluralizing (see below) and overt copular (ete < été) be co-official. Mau, despite its lack of any official morphemes differ from those used in the Caribbean, recognition whatsoever, is nowadays the home lan- but both appear to be postformative developments. guage of three fourths of the total population of Another such example is the verb apocope (mãze vs. 1,200,000. It is gaining ground in the Indian and mãz ‘to eat’), which occurs wherever the verb governs Chinese languages introduced by nineteenth century an overt element. immigrants whose descendants now make up the Also noteworthy is the exceptionally high frequen- majority of the population. cy of agglutination of French articles to nouns (e.g. The Seychelles, now with a population of about lavi [< la vie] ‘life’, rather than specifically ‘the life’). 80,000 was colonized in the 1770s, with settlers coming Although probably genetically unrelated, Réu and both from Mauritius and Réunion. Ethnolinguistically, IdF do share a small number of features not found in the Seychelles differs from Mauritius mainly in having the New World French creoles, including the depalatal- no sizable minority of Indian descent, and so far as ization of postalveolar fricatives, the use of zot (< native language is concerned, it is virtually monolin- vous/les autres) as both 2pl and 3pl, the pluralizer bann gually creolophone. The Republic of the Seychelles is (< bande), and possibly also the predicate marker i. unusual in giving official recognition to its Creole lan- guage—the country is officially trilingual in Creole, English, and French. English is spoken as an L2 by References most, and French by a sizeable minority. Media are Armand, Alain. 1987. Dictionnaire kréol rénioné-français. available in all three official languages. Saint-André: Éditions Océan. The IdF varieties are all mutually comprehensible Baggioni, Daniel. 1990. Dictionnaire créole réunionnais/ français. Paris: l’Harmattan. without much effort. The most noticable differences Baker, Philip. 1972. Kreol. A description of Mauritian Creole. between Mau and Sey are the use of the predicate London: C. Hurst. marker i in the latter, and nasal vowels in Sey often Baker, Philip, and Chris Corne. 1982. Isle de France Creoles: corresponding to oral ones in Mau. In both cases, Sey affinities and origins. Ann Arbor: Karoma. is conservative vis-à-vis Mau. Lexically, Sey displays Baker, Philip, and Vinesh Hookoomsing. 1987. Diksyoner kreol morisyen. Paris: l’Harmattan. some influence from Réu in, e.g. the use of komela Bollée, Annegret. 1977. Le créole français des Seychelles: ‘now’ instead of aster as in Mau. Esquisse d’une grammaire, textes, vocabulaire. Tübingen: As mentioned above, IdF is rather similar to the bet- Max Niemeyer. ter-known French creoles of the Caribbean, which Carayol, Michel, and Robert Chaudenson. 1989. Atlas linguis- may or may not have something to do with the early tique et ethnographique de la Réunion, 2 vols. Paris: Édi- tions du CNRS, 1984–1989. presence of West Africans in Mauritius. The main Cellier, Pierre. 1985. Comparaison syntaxique du créole réunion- phonological difference is that postalveolar fricatives nais et du français. Saint-Denis: Université de la Réunion. are depalatalized in the Indian Ocean (i.e. /ʃ, /➝/s, Chaudenson, Robert. 1974. Le lexique du parler créole de la z/). IdF also has only five oral vowels, as opposed to Réunion. Paris: Éditions Honoré Champion. seven in the Caribbean creoles. Corne, Chris. 1977. Seychelles Creole grammar. Tübingen: Günther Narr. The system of preverbal tense–mood–aspect mark- Offay, Danielle d’, and Guy Lionnet. 1982. Dictionnaire créole ers is particularly close to that of, e.g. Haitian, with seychellois-français. Hamburg: Helmut Buske. progressive (a)pe (< après), past ti (< te < était/été), Stein, Peter. 1982. Connaissance et emploi des langues à l’Île completive ((f)i)n (< finir), and two different futures Maurice. Hamburg: Helmut Buske. (v)a (< va) and pu (<pour). MIKAEL PARKVALL Indian Traditional Grammar It is to the Indian past that linguists of all countries and we think, some two and half millennia ago. The great all theoretical persuasions still look for one of the grammarians of Sanskrit—Panini (phonetically, crowning glories of intellect in linguistic science—to Pa¯ nini), and his followers Katyayana and Patañjali— Panini’s grammar of the Sanskrit language, written, created in their grammars of the Sanskrit language 507 INDIAN TRADITIONAL GRAMMAR linguistic masterpieces, as awe-inspiring in the realm ers, as we might learn Latin or Greek today. Panini of language as, in other esthetic spheres, the Mona was taken to be infallible: any deviations in Sanskrit Lisa, the works of Euclid, the Taj Mahal, or the works compositions from the forms prescribed by Panini of Shakespeare. were errors. Sanskrit was the language of ancient northern India One of the crowning achievements of Indian tradi- and of the sacred writings of Hinduism. We believe tional grammar was phonetics. The Indian grammari- that Indo-European (also called Indo-Aryan) invaders ans possessed a control of the technical details of came into India from the northwest around 2500 BCE. phonetics that remained unmatched until well into the The Dravidian family of languages was indigenous to nineteenth century. Unlike most contemporary classi- India, and Sanskrit developed out of the interaction fications of sounds, which use different features for between the Dravidian substratum and the Indo- vowels (high, low, back, rounded, unrounded, etc.) and European adstratum. By about 350 BCE, Sanskrit had consonants (voiced, voiceless, continuant, etc.), begun to diverge regionally, as all languages spoken Panini’s phonetic classification of the sounds of over larger geographical areas do, given enough time. Sanskrit used a single grid for both consonants and The Hindu grammarians broadened their interest from vowels, as well as semivowels. Revised slightly and the Scriptures to the notion of ‘correct’ Sanskrit, mak- using modern phonetic terms, the phonemes of ing rules and lists of forms descriptive of the correct Sanskrit are as shown below: type of speech in every sphere: phonology, morpholo- Morphophonemics and morphology are described gy, and syntax. Sanskrit was a part of every Hindu rit- by means of rules, the exact nature and the interaction ual, a sacred language, and it was essential that it be of which has been the subject of much debate. pronounced and inflected correctly to the last detail. Typical—typical both of type and their intrinsic com- The culmination of this intense activity was the old- plexity—rules are: est treatise that has come down to us, the grammar of a. h —> d¢h, if h is sam¢ yoga¯ di, that is, if h is the Panini. This grammar (the Astadhyayi), which dates first component of a consonant group when the from between 350 and 250 BCE, is one of the great suffix begins with a jhal-consonant (any monuments of human intelligence. It describes in consonant with the exception of a nasal or a detail every usage of its author’s speech. Virtually no semivowel). Example: lih-dhi —> lid¢ h-dhi. other language, even today, has been so completely b. A pair of rules (simultaneously ordered): described. Sanskrit became the official and literary d¢ h-Deletion: d¢h —> zero / ___ d¢ h language of all of Hindu India, even to the farthest Lengthening: i —> ¯ before a d¢h, reaches of the south. It ceased to be spoken as any- which is dropped one’s native language probably around the beginning of the Common Era, but it remained (as classical Latin Certain aspects of Panini’s treatment of morphology remained in Europe and Hebrew in Judaism) the arti- seem queer to us today; there is, for example, no spe- ficial medium for all writing on learned or religious cific recognition of the notion ‘subject of a verb’. topics, the language of the pundits. There is no case marking as such, even though Upper-class education in ancient India began with Sanskrit has a richly developed case system. Perhaps the acquisition, as a child, of the Sanskrit language. the greatest accomplishment of the Indian grammari- This was done by discussion and above all by memo- ans in morphology is the notion that words are analyz- rizing Panini’s grammar or one of the grammars based able—can be broken down into roots and morphemes. on it. Well after Sanskrit had ceased to be a widespread Today, we think of the idea of ‘root’ as self-evident. It spoken language in India, classical Sanskrit literature is not: Greek and Latin grammarians did not analyze was being written by men who had learned their words into roots and morphemes. Certain aspects of Sanskrit from grammars and dictionaries, and teach- the Indian analysis of noun compounds are useful 508  INDO-EUROPEAN 1: OVERVIEW today: dvandva = copulative (bittersweet, flimflam, academic discipline. Nevertheless, what the Indian zigzag), bahuvrihi (‘much rice’) = exocentric (door- grammarians produced over 2,000 years ago can hold knob is a kind of knob, blackbird is a kind of bird; its own with the best of linguistic theory today. these contrast with endocentric compounds such as foul-up, which is not a kind of up, or turncoat, which References is not a kind of coat). Panini’s method has been summarized as follows. Allen, W. Sydney. 1953. Phonetics in ancient India. London: Oxford University Press. First, rules (sutras) are formalized. Second, metalin- Burrow, Thomas. 1965. The Sanskrit language. London: Faber guistic elements (e.g. natural classes) are introduced and Faber. for reasons of economy of description. Third, Cardona, George. 1987. Sanskrit. The World’s major languages, metarules (paribhasas), which explain how the rules ed. by Bernard Comrie. New York: Oxford University Press. work, are made explicit. Deshpande, M. 1993. Sanskrit and Prakrit, sociolinguistic issues. Delhi: Motilal Banarsidass. Linguistics in ancient India was the core of the Katre, S.M. 1954. Introduction to Indian textual criticism. intellectual and scientific tradition—it possessed an Pune, Deccan College. intellectual centrality and scholarly hegemony that Rocher, Rosane. 1975. India. Current trends in linguistics, Vol. beggar belief today. Linguistics was suffused with the 13 (Historiography of linguistics), ed. by Thomas A. light of sanctity, endowed with religious purpose. Sebeok. The Hague: Mouton. Staal, J. F. 1972. A Reader on the Sanskrit grammarians. Panini and Patañjali and later Indian grammarians Cambridge, MA: MIT Press. conceived their ministrations on behalf of the Sanskrit Whitney, William Dwight. 1889. Sanskrit grammar. Cambridge, language as devotional. Their activity was more akin MA: Harvard University Press. (often reprinted). to a priesthood, a calling, than it was a profession or an ROBERT D. KING Indo-European 1: Overview The Indo-European (IE) family is the most successful genders (masculine, feminine, and neuter), and case and widespread of the language families of the world. systems ranging from five cases (e.g. Greek, Seven out of the ten most widely spoken world lan- Germanic) to eight (Sanskrit), a fairly complex verbal guages are IE, and all of these seven belong either to system, characterized by distinctions of tense, voice, its Romance, Slavic, Germanic, or Indo-Iranian and mood, with some development of an aspectual branches. (These figures are taken from system, synthesis rather than analysis, and a tendency http://www.sil.org/ethnologue/top100.html (updated to SOV (subject–object–verb) word order, etc. Most of February 1999).) IE has spread over nearly all of the modern IE languages show greater or lesser Europe, plus large areas of Western Asia and the degrees of reductions of these systems, with some lan- Indian subcontinent. Over the past few centuries, it has guages showing remodelings. The most common further spread to nearly the whole of the Americas, reductions of these systems may be characterized by Australia, and New Zealand, plus most of Northern the following: Asia, and small pockets in the Pacific and Southern Africa. This expansion has not been entirely one- Grammatical Features sided: over the last 1,200 years, Hungarian and Turkic have gained at the expense of IE, specifically Slavic, Gender: Reduction of original three-gender system to Greek, Armenian, Iranian, and Tocharian. a two-gender system (mostly masculine/feminine, IE is often considered from the point of view of its although sometimes common/neuter). classical ancient languages (Latin, Greek, and Number: Widespread loss of the dual. Sanskrit), with supporting citations from Germanic and Slavic. These languages are very similar in overall Case system: General reduction of case, sometimes structure, with nominal systems mostly distinguishing leading to the complete loss of the category, or three numbers (singular, dual, plural) and three increased syncretism. 509 INDO-EUROPEAN 1: OVERVIEW Word order: In Europe, VSO and SVO orders have and Romans on language were mostly confined to developed, with vestiges of SOV, while in Asian IE Greek and Latin, and later to the interrelationships languages, SOV is still more widespread. between the two. For a long time, it was believed that Latin had descended from Greek, and this is reflected Most IE languages in Europe have also developed a in the writing of authors such as Varro. This was later system of definite and indefinite articles, with the shown to be erroneous, although there is still a tradi- definite article being more widespread than the indef- tion of writing comparative grammars of Latin and inite one. They have also carried through the develop- Greek. ment of the verb ‘have’ more fully than the classical IE As the Middle Ages progressed, followed by the languages. Renaissance, people became more aware of the lin- In broad outline, the following well-attested major guistic interrelations within Europe, and the major branches are considered IE: Anatolian (the ancient lan- families, and relationships between the coeval fami- guages of Asia Minor, e.g. Hittite, Luwian, Lydian, lies, and the Classical languages. Medieval writers etc., attested between about 1700 BC and the fourth such as e.g. Giraldus Cambrensis and Otto of Bamberg century BC), Indo-Iranian (e.g. Sanskrit (plus its noted several lexical items shared by Brythonic and descendants, e.g. Hindi, Bengali, Urdu, Marathi), Greek or Latin and Slavic, respectively. The general attested from around 1000 BC; Avestan and Old outlines of their linguistic interrelations slowly Persian plus the modern Iranian languages, e.g. became clear during and immediately after the Persian, Kurdish, Pashto, also attested from around Renaissance. In 1599, Scaliger grouped the languages 1000 BC), Hellenic (Greek and its dialects, earliest of Europe into four major and minor groupings. attestations from 1400 BC), Italic (Latin (and its Scaliger’s major groupings correspond to Romance, Romance descendants, e.g. French, Spanish, Italian), Greek, Germanic, and Slavic. possibly plus other ancient languages of Italy, e.g. But the start of serious scholarship, and the realiza- Oscan, Umbrian, attested from 500 BC), Celtic (e.g. tion that IE extended into Asia, is normally dated from extinct Continental Celtic languages, e.g. Gaulish, etc. Sir William Jones’ famous address to the Royal Asiatic attested in Roman times; modern Insular Celtic lan- Society of Bengal in 1786, where he first conceptual- guages, e.g. Irish, Scottish Gaelic, Welsh, Breton ized the idea of ‘some common source, which perhaps attested from AD 300), Germanic (e.g. Gothic and no longer exists’ to explain similarities between Latin, other extinct East Germanic languages, English, Greek, and Sanskrit, the language of the Indian German, Scandinavian languages attested from the Brahmins, of which Europeans had been slowly second century AD), Baltic (e.g. Lithuanian, Latvian, becoming aware for about three centuries. Jones also attested from the sixteenth century), Slavic (e.g. suggested that ‘both the Gothick (Germanic) and the Russian, Polish, Czech, Serbo-Croatian, Bulgarian, Celtick’ and the old ‘Persian’ might also be derived attested from the ninth century AD), Armenian (attest- from this ‘common source’. ed from the fifth century AD), Albanian (attested from At first glance, Sanskrit appears even more archaic the fourteenth century AD), and Tocharian (Central than Latin or Greek. For a while, it was even believed Asia, attested between the sixth and eighth centuries that Sanskrit was the ancestor of all IE languages. As AD). Other extinct IE languages have been glimpsed the reconstruction of IE progressed, it was revealed in outline through early attestations, such as, e.g. that in some respects Sanskrit was less archaic than Illyrian, Ligurian, Venetic, Pelasgian, Phrygian, and other IE languages (see below). Moreover, a later part Thracian. In addition to the extinct IE languages noted of Jones’ address deals with the modern languages of above, the ranges of Hellenic, Baltic, Albanian, and India, which he did not believe as descended from Celtic have been severely restricted. Sanskrit. In a lesser-known passage, he stated: ‘and In this context, it should be noted that frequently this analogy might induce us to believe, that the pure family trees drawn for IE give distorted pictures, allot- Hindi, whether of Tartarian or Chaldean origin, was ting nearly all the space to Germanic, Celtic, Italic, primeval in Upper India, into which the Sanskrit was Indo-Iranian, and Balto-Slavic (nowadays sometimes introduced by conquerors from other kingdoms in also including Anatolian), while bunching Hellenic, some very remote age...’; it was only later that it was Albanian, Armenian, and Tocharian together in a tiny discovered that Hindi stood in roughly the same rela- slot. Despite their much smaller number of speakers, tionship to Sanskrit as French to Latin. or lack thereof, it might be argued that together, the Much research followed, and modifications to the latter four groups encompass as much diversity as the theory came with time, but the basic edifice of IE former five. remained. In the 1860s, Schleicher even felt confident The discovery of IE was a gradual process, with enough to compose a fable in the reconstructed IE lan- several stages and false trails. Thoughts by the Greeks guage. Schleicher’s fable was later reworked by Hirt in 510  INDO-EUROPEAN 1: OVERVIEW the 1920s, and by Lehmann in the late 1970s, but as although much has been superseded. IE reconstructed the century progressed it became clear that IE was at this stage in the development of the field is some- unamenable to this sort of analysis. The second half of times called ‘Brugmannian IE’, after one of the most the nineteenth century witnessed numerous apparent prominent Neogrammarians. irregularities worked out. The formulations of the Attempts were also made to delineate subgroupings explanations for these irregularities were usually within IE. The major one normally proposed, and ele- called ‘laws’, and this usage has stuck, although cer- vated to considerable importance in the late nineteenth tain scholars were at pains to point out that they could century, was the centum/satem split (from the words not be equated to more universal laws, such as those of for ‘hundred’ in Latin and Avestan, respectively). In physics. These ‘laws’ were usually named after the certain forms, the centum languages retained an origi- scholars who formulated them. nal velar, while the satem languages developed the Some of the major such ‘laws’, each representing a velars into a palatal or sibilant. Throughout most of substantial step in the reconstruction of IE, and an this century, however, this postulated centum/satem advance in general linguistic theory, run as follows: distinction has steadily been losing importance in IE studies. (1) Grimm’s Law (1822) systematized the One ongoing area of IE studies is the attempt to Germanic sound shift already noted by several demarcate early dialectal divisions within IE. The only scholars, where voiceless stops developed to universally accepted one, however, is Indo-Iranian. It voiceless fricatives, voiced stops to voiceless was also once popularly believed that Italic and Celtic stops, and voiced aspirates stops to plain voiced were closely related, based on a few common items of stops (e.g. t > th, tráyas/three, d > t, e.g. da vocabulary and a few shared grammatical innovations a/ten, dh > d, e.g. vidhava/widow; English and archaisms. Certain scholars have seen a European forms contrasted with Sanskrit). Although IE grouping of languages, mainly based on shared lex- many previous scholars had noted individual ical features. There is an ongoing controversy about manifestations of this phenomenon, it is nor- whether Baltic and Slavic are two separate branches of mally named Grimm’s Law, as Jakob Grimm IE that have come to resemble each other through con- was the first to systematize it, grouping the vergence, or whether an intermediate Balto-Slavic voiceless stops (tenues), the voiced stops grouping may be reconstructed. (mediae), and the voiced aspirates (mediae In the early years of this century, new discoveries aspiratae) together in classes. and decipherments shed fresh light on IE, while at the (2) Grassmann’s Law (1863), which accounted for same time making the picture more complicated. The apparent irregularities in the distribution of first was the discovery in Western China in 1900 of voiced aspirates and voice stops, found in manuscripts in a previously unknown IE language, Greek and Sanskrit (sound changes could be Tocharian, split up into two heavily differentiated spread over more than one syllable). dialects, conventionally called Tocharian A and (3) Verner’s Law (1876), which explained the role Tocharian B. The structure of Tocharian, while recog- played by accent in apparent anomalies in the nizably IE, differed in several important aspects from distribution of the reflexes of Grimm’s Law the Brugmannian system. (the role of accent in sound change). A few years later, a massive corpus of material (4) The Law of the Palatals (1870s), which is diffi- written in cuneiform was unearthed in Anatolia. The cult to ascribe to any single scholar. It showed decipherment and linguistic interpretation of the that the system of five short vowels attested in material took some time, but finally it was recog- Latin and Greek (a, e, o, i, u) was actually more nized that here was another new IE language, identi- archaic than the system of three short vowels fied with the language of the Hittites of the Old attested in Sanskrit (a, i, u), previously seen as Testament. While definitely IE, the structure of the the original system. newly interpreted language was quite unlike what Shortly after all these laws had been formulated, the had been expected: in many ways, it was closer to so-called Neogrammarian school emerged. They pro- modern IE languages, including a fully developed posed that there were no exceptions to sound laws as a verb ‘have’. The verb morphology was far less com- working hypothesis and compiled numerous compara- plex than Greek or Sanskrit, and this caused some tive grammars, both of IE (Brugmann, Delbrück, scholars to give serious reconsideration to the recon- Osthoff) and of individual IE languages (e.g. Leskien struction of IE morphology. As research progressed, on Old Church Slavonic). Mainly centered in Germany, it was discovered that Hittite was only one member they were very influential on IE studies, and the vast of a whole family of IE languages, normally referred compilations of their material are still consulted today, to as Anatolian (see above). The most important 511 INDO-EUROPEAN 1: OVERVIEW contribution from Anatolian, however, was its appar- relations), and the degree of morphological complexity ent confirmation of the ‘laryngeal theory’. In the late of the protolanguage. 1870s, speculations about the IE vowel system had led de Saussure to postulate the existence of other- References wise unattested segments, which were later recon- structed as laryngeals. These segments were actually Beekes, Robert. 1995. Comparative Indo-European linguistics. attested in Hittite. Not all linguists accept the laryn- Amsterdam: John Benjamins. Birnbaum, Henrik, and Jaan Puhvel (eds.) 1966. Ancient Indo- geal theory, although the discovery and interpretation European dialects. Berkeley: University of California Press. of the Hittite material have given it a tremendous Collinge, N.E. 1985. The laws of Indo-European. Amsterdam: accession of support. John Benjamins. The discovery and interpretation of Tocharian and Ivanov, V.V., and Gamkrelidze, T.V. 1995. Indo-European and Hittite material also caused the centum/satem theory the Indo-Europeans, translated by Johanna Nichols. New York: Mouton de Gruyter. some problems, as both of them showed centum Lockwood, William.B. 1969. Indo-European philology. reflexes, despite being located to the east. It has London: Hutchinson. taken time for the results of research into Hittite and Ramat, Paolo, and Anna Giacalone Ramat (eds.) 1998. The Tocharian to be properly integrated into mainstream Indo-European languages. London and New York: Routledge. IE studies, and perhaps, the process is not yet The website http://www.dla.utexas.edu/depts/lrc/iedocctr/ ie-links.html at the IE Documentation Center at the complete. University of Texas at Austin is also an excellent resource. IE continues to be subject to modification and rein- ROBERT ORR terpretation. It has even been quipped that ‘no lan- guage has changed as much in the past fifty years as See also Indo-European 2: Germanic Languages; IE’. Areas where there has been much discussion Indo-European 3: Indo-Iranian Languages; Indo- recently include the glottalic theory (a reinterpretation European 4: Romance; Indo-European 5: Slavic; of the nature of the IE stops), ergativity (subject–object Jones, Sir William Indo-European 2: Germanic Languages The Germanic languages are spoken by more than 450 language we now call Old Norse, the ancestor of the million native speakers and display a very wide geo- modern Scandinavian languages. In the fifth century graphical distribution, partly due to the current status CE, three West Germanic tribes, the Angles, Saxons, of English as the world’s most important international and Jutes, crossed the North Sea into Britain, bringing language. with them a language that would later be known as It is generally assumed that by the first century English. And in the ninth century, Old Norse was car- BCE, Germanic people speaking a fairly uniform lan- ried far westward to Iceland. These wanderings left guage were living somewhere in southern Europe dotted with Germanic dialects. Scandinavia, the Danish Isles, and northern Germany, The Germanic languages offer extensive and homo- on both sides of the North and Baltic Seas, and they geneous documentation from a relatively early period. started to expand rapidly during the Roman period, The oldest monuments of the Germanic languages are primarily to the detriment of Celtic languages. In time, more than a hundred Scandinavian inscriptions in a spe- there developed the so-called West, East, and North cial alphabet called Runic. They date from the third or Germanic dialects. The West Germanic tribes settled fourth century CE and are representative of the North, in the lands between the Elbe and Oder rivers and to or Scandinavian, group of Germanic languages. the immediate south and west of that area (about the The Germanic languages have very clear-cut and area of modern Germany and the Benelux states), and readily identifiable features distinguishing them from it is here that the German language gradually evolved. the other Indo-European stock. The best known is The East Germanic tribes settled in the Baltic area described by the so-called Grimm’s Law (or the First east of the Oder River (the area of modern Poland), Sound Shift), referring to a systematic shift in the but their languages have long since become extinct. Indo-European consonant system during the develop- In Scandinavia, the North Germanic tribes spoke a ment of the Germanic ancestor language. Besides, the 512  INDO-EUROPEAN 2: GERMANIC LANGUAGES Indo-European verbal system was simplified. The and Icelandic. The oldest North Germanic language Germanic languages developed a preterite tense is referred to as Old Icelandic, or Old Norse, which (called weak or regular) with a dental suffix, -d or -t suffered a rapid change after 700, due to the restless (e.g. fish, fished, etc.). Germanic languages thus have movements of the Viking Age. However, Old Norse two types of verbs: weak (regular) and strong (irregu- is the language of an extraordinarily rich and varied lar). Strong verbs indicate tense by an internal vowel literature inspired by the traditions and manners of change (e.g. ring, rang, rung). the incomparable Viking Age. In terms of vocabulary, Germanic has a number of unique words that are not found in other Indo- Danish European languages. These words may have been lost in the other Indo-European languages, borrowed from Danish, the mother tongue of over 5 million people, is non-Indo-European languages, or perhaps coined in the official language of the kingdom of Denmark Germanic. (including Denmark, Greenland, and the Faroe Numerous groupings of the Germanic languages Islands). The modern language developed on the basis have been proposed during the last centuries, begin- of the written language of Reformation influenced by ning effectively with the grammar of Jacob Grimm in the seventeenth to eighteenth centuries language of 1819. Traditionally, the Germanic languages are divid- Copenhagen, the economic and cultural center of the ed into East Germanic (with Gothic as its most impor- emerging nation state. Of all Scandinavian languages, tant member), North, and West (sometimes ‘South Danish has differentiated the furthest from the Germanic’) subgroups. Common Scandinavian ancestor. Besides the living Germanic languages described below, there are several extinct languages such as Swedish Bastarnae, Burgundian, Frankish, Gothic, Herulian, Lombardic, Norn, Rugian, Scirian, and Vandalic. Swedish is the official language of Sweden and it is spoken by 8.5 million Swedish and 300,000 Finlanders. It was first recorded in about 2,000 runic East Germanic inscriptions from the eleventh and twelfth centuries. The Goths were the first Germanic tribe to leave their The modern language developed in the Mälar-Uppland Germanic homeland (presumably around 100 BCE) region, where Stockholm and Uppsala have been and due to this early migration their language devel- raised as the main centers of government and learning oped differently from those of the other Germanic since the Middle Ages. peoples. Gothic offers the oldest literary record of Germanic Norwegian languages: a translation of a Greek bible into Gothic by Bishop Wulfila in the fourth century CE. It is an Norwegian is the official language of Norway and the extremely regular and consistent translation and mother tongue of over 4 million people. It has two vari- includes a large part of the New Testament and some eties, Neo-Norwegian (nynorsk) and Dano-Norwegian parts from The Old Testament. (bokmål), both official languages used by national and Gothic has played an important part in the recon- local officials. Nynorsk is more prevalent in rural areas, struction of the Indo-European and Germanic ancestor Bokmål in the cities. Bokmål is spoken by approxi- language. Its structure is reminiscent of the other old mately 90% of the population. These two modern stan- Indo-European languages, with complex inflections of dards have a historical cause: during 1380–1814, noun, verb, and adjective. It was heavily influenced by Danish was the written language of Norway, while Greek in Wulfila’s translation of the Bible and this most Norwegians used their local dialects and pro- influence is evident with respect to the lexicon, word nounced Danish with their own Norwegian sounds. order and sentence structure. Neo-Norwegian is based on local dialects and was offi- cially recognized in 1885. Dano-Norwegian, or ‘book language’, is the first language of the majority of the North Germanic population. North Germanic is commonly divided into two sub- groups gradually splitting up from the Common Faroese Scandinavian toward the end of the Viking Age (800–1050): the eastern one, including Swedish, Faroese is spoken by 47,000 inhabitants in the Faroe Danish, and Gutnish (on the Baltic island of Gotland), Islands, where it is also the official language (together and the western one, including Norwegian, Faroese, with Danish). It is a very archaic variety of North 513 INDO-EUROPEAN 2: GERMANIC LANGUAGES Germanic isolated during the Viking expansion and it Frisian developed from the language of the Norwegians who Frisian is spoken in northern Holland (the province of colonized the islands in the early 800s. Faroese, a Friesland and some adjacent islands), and also in descendant of Old Norse, has few speakers, but a live- Germany (the low coastal areas). It is one of the two ly national literature for nearly 200 years. It has only official languages of the Netherlands. Of all Germanic come into relative prominence in modern times. languages, Frisian is most closely related to English. Structurally, Faroese corresponds more to Middle The oldest records are a series of legal documents dat- Norwegian than to Icelandic. ing from the tenth to the sixteenth centuries. Structurally, the language is very similar to Dutch. Icelandic Icelandic is the official language of Iceland where it Dutch has been spoken ever since the country was settled over a thousand years ago. It is the mother tongue of Dutch is the official language of the Netherlands, 250,000 people. Icelandic is the most conservative Suriname, and the Netherlands Antilles, and Flemish North Germanic language, due to its geographical iso- is the official language of Belgium. The language is lation and the fierce purism of its speakers, i.e. no also spoken in small parts of Germany and France, other Germanic language has remained so close to the Indonesia and elsewhere in Southeast Asia. It is native ancestral form. It has never shown any real tendency to to some 14 million people. The Dutch spoken in split into dialects. It has a rich literature dating from Belgium is called Flemish, although there is practical- the thirteenth century. ly no difference from Dutch in the Netherlands. Standard Dutch is based on the dialect of the Amsterdam region after it became the capital of an West Germanic independent nation. West Germanic is considered to be the least unified of the three major groups. The records are relatively late Afrikaans and very uneven in chronology and content. Each of the individual languages is highly inconsistent and dis- Afrikaans, sometimes considered a creole of Dutch, is persed into various dialects. However, two main an official language of South Africa and Namibia. It is branches can be identified: the English group, which is spoken by 10 million people. Afrikaans developed in usually considered with Frisian as Anglo-Frisian, and the seventeenth century from the Dutch brought to the German group, including High and Low German. South Africa by the first settlers from Holland. Originally, it was a popular dialect composed of Dutch with considerable borrowings from aboriginal lan- English guages of Africa, especially Malay (spoken by the English is an official language of Britain, Canada, the slaves in the seventeenth century). United States, Australia, New Zealand, Ireland, the Bahamas, Jamaica, Barbados, Grenada, Trinidad and German Tobago, St. Lucia, Belize, Guyana, Botswana, Cameroon, Dominica, Gambia, Ghana, Gibraltar, German is an official language in Germany, Austria, Lesotho, Liberia, Malawi, Nigeria, the Seychelles, and Switzerland; it is also spoken in eastern France Sierra Leone, Singapore, South Africa, Swaziland, (Alsace and Lorraine), Belgium, Luxembourg, Uganda, Vanuatu, Western Samoa, Zambia, and Liechtenstein, pockets of the United States, parts of Zimbabwe; it is the associate official language of Eastern Europe, and in former German colonies India. It is the native language of more than 300 mil- (Namibia, Congo, Cameroon). German is also used as lion people and the second language of another 300 a second business language for Continental Europe million. English is also the universal language of trade, and Scandinavia. communications, science, and transportation and the Following the so-called ‘High German Consonant most frequently acquired second language in the Shift’, the dialects of German are divided into Low world. German (Niederdeutsch, Plattdeutsch) and High English is characterized by a very large vocabulary, German (Hochdeutsch). Low German is spoken in the nonphonetic spelling, an almost total lack of inflection lowlands of northern Germany. The term ‘Low (most plurals of nouns are indicated), a syntax almost German’ is essentially a geographic term referring to totally dependent on word order, and a very compli- the coastal, or lowland, area of the German region, as cated periphrastic verb system. opposed to the High German area. Modern Standard 514  INDO-EUROPEAN 3: INDO-IRANIAN LANGUAGES German developed mainly from the language of the literature and its speakers are all bilingual, as the official late medieval chancery language of the court and language is English. Saxony and the East Central dialect area of Dresden. References Yiddish Abraham, Werner, Wim Kosmeijer, and Eric Reuland (eds.) 1991. Issues in Germanic syntax. Berlin, New York: Mouton Yiddish (also called Judeo-German) is a High German de Gruyter. language spoken throughout northern continental Braune, W., and E. Ebbinghaus. 1981. Gotische grammatik (Gothic grammar), 19th edition. Halle: Niemeyer. Europe (especially Germany, Poland, Lithuania, and Comrie, B (ed.) 1987. The World’s major languages. London: Russia) and North America by the Jewry. Originally, it Routledge. was the language spoken by East European Jews, who Haugen, E. 1982. The Scandinavian languages. A comparative migrated from Germany in the fourteenth to sixteenth historical survey. Tübingen: Niemeyer. centuries. There are two varieties of Yiddish: an east- Helbig, G., and J. Buscha. 1991. Deutsche Grammatik (German grammar), 14th edition. Berlin: Langenscheidt. ern one developed under Slavic influence and devel- Lippi-Green, Rosina L (ed.) 1992. Recent developments in oping a rich literature in a highly standardized Germanic linguistics. Amsterdam, Philadelphia: J. Benjamins. language, and a western one developed under German Lippi-Green, Rosina L., and Joseph C. Salmons (eds.) 1996. influence. Yiddish literary records go back to about the Germanic linguistics. Amsterdam, Philadelphia: John thirteenth century. Like all Jewish languages, it is writ- Benjamins Publishing Company. Rauch, Irmengard, Gerald F. Carr, and Robert L. Kyes (eds.) ten in Hebrew characters read from right to left. 1992. On Germanic linguistics: issues and methods. Berlin, New York: Mouton de Gruyter. Russ, C.V.J. (ed.) 1990. The dialects of Modern German. A lin- Pennsylvania German guistic survey. London: Routledge. Stephens, M. 1976. Linguistic minorities in Western Europe. Pennsylvania German, also known as ‘Pennsylvania Llandysul: Gomer Press. Dutch’ or simply ‘Dutch’, is spoken by 300,000 native Vikør, L. 1993. The Nordic languages: their status and interre- speakers, descendants of German colonists, mainly in lations. Oslo: Novus. the United States. It is the vehicle of a typical folk LAURA DANILIUC Indo-European 3: Indo-Iranian Languages Indo-Iranian languages are the easternmost subfamily branch of Indo-Iranian. Indo-Iranian languages also of the I[ndo]-E[uropean]family, spoken by almost a have a very old history. The Indic and the Iranian lan- billion people, chiefly in Iran, Afghanistan, Tajikistan, guages have passed through three historical periods, Bangladesh, India, Nepal, Pakistan, and Sri Lanka. namely old, middle, and new. The languages of this subfamily are among the oldest of the Indo-European group, and are well represented Iranian among the oldest records of Indo-European languages. However, there is debate about the relation of these The Iranian languages are mainly spoken in today’s languages to the Hittite languages of Anatolia. Some Iran, Afghanistan, Tajikistan, and some parts of references believe them to have originated around Pakistan, Turkey, Iraq, and former Soviet countries. It modern Afghanistan, and some others name Turkistan. is estimated that about 95 million people speak in This branch consists of three groups of languages, the these languages. Iranian, the Indo-Aryan (or Indic), and the newly added Dardic (also known as Kafir or Nuristani lan- Old Iranian guages), preserved in the Himalayan Mountains. Historically, the oldest Iranian languages of which The Indo-Iranian group is remarkable, both in terms there are records available are Old Persian and of its history and in terms of the number of its speak- Avestan, both highly inflected languages. Old Persian ers. Two of the ten most widely spoken languages on has survived in cuneiform inscriptions from the time of earth today, Hindi and Bengali, belong to the Indic the Achaemenid kings, who ruled ancient Persia during 515 INDO-EUROPEAN 3: INDO-IRANIAN LANGUAGES the sixth to fourth centuries BC. Avestan is the lan- and some of them are greatly in danger of extinction. guage in which Avesta, or the sacred text of the Some of the best-known languages of this group are Zoroastrian religion was composed. The Avesta proba- Pashto, Wakhi, Ishkashmi, Yoghnobi, Yazgulami, bly dates from about the seventh to the fifth centuries Shughni, Parachi, Bartangi, Rushani, and Sarikoli (the BC, but apparently was handed down orally and was easternmost Iranian language). not recorded in writing until much later. Both of these languages are extinct, although Avestan is still formal- Indic ly in use as the religious language of the Zoroastrians. The Indic languages form the largest group of the Indo-Iranian subfamily. This group contains about thir- Middle Iranian ty languages, some of which with numerous dialects. The Middle Iranian period, dating from the third cen- About 700 million people speak in these languages, of tury BC to the ninth or tenth century AD, is character- which the most widely used languages are Hindi-Urdu, ized by considerable grammatical simplification, as in Bengali, Punjabi, and Marathi (to these languages, we the reduced inflection of the noun and verb. Among should add the Gypsy or Romano languages, which are the languages surviving in written records that fall scattered in various parts of Europe, as well as in the within this period are Parthian, Middle Persian, Indian subcontinent). Having voiced aspirated plosives Khwarazmian, Sogdian, and Saka. like [bh], as well as retroflex sounds in the phonologi- cal system, are regarded as the two main characteris- New Iranian tics of these languages. Like the Iranian languages, The new Iranian languages are those languages used in Indic languages have a long history, and may be stud- the region after the emergence of Islam in the region ied within Old, Middle, and Modern Periods. (ninth or tenth century), and represent more phonetic and grammatical simplifications. For example, many Old and Middle Indic case endings have dropped, and have been substituted The oldest written form of the Indic group is Vedic, by adpositions. The new Iranian languages are which is the language of the religious hymns of Veda. considerably more in number in comparison with the The Veda represents an oral tradition preserved by the middle Iranian languages. What is interesting about ancestors of Hindu priests. Vedic has been described as these languages is that most of them do not have a the parent language of Sanskrit which by the fourth direct and specified ancestor, and only Persian and century BC had become the sacred and literary lan- Yaghnobi can be regarded as those having direct guage of the Hindus of India, and its classical form was ancestors, Middle Persian and Sogdian, respectively. in use (at least for literature) until c. AD 1100. Sanskrit, The new Iranian languages are now divided into as the most important source for all modern Indic lan- ‘eastern’ and ‘western’ subgroups. New west Iranian guages, has survived to this day as a liturgical language languages are basically spoken in today’s Iran, and in India. It is written in Devanagari, a development of include Persian (the widest used Iranian language, the Brahmi script. Most Indic languages are written in with Iranian, Afghan, and Tajik variations), Kurdish, some modified form of the Devanagari alphabet. Baluchi, Mazandarani, Lori, Gilaki, Ta¯ti, Ta¯leši, Prakrit (or ‘natural’ as opposed to Sanskrit ‘pol- La¯ resta¯ ni, Bashkardi, and some others. There are lan- ished’), which represents the Middle Indo-Iranian, guages, of course, spoken beyond this border, and yet refers to the state of the language between the third considered to be west Iranian, like Baluchi (which is century BC and the fourth century AD. The best- also spoken in Pakistan), Zazaki (spoken in Turkey), known language of this era is Pali, in which the and Kumzari (spoken in Musandam peninsula in Buddhist Ašoka scriptures were originally written. Oman). Most of these languages do not have a writing system, and make use of Persian alphabet. Iranian Persian is also written this way, but the Tajiks make Modern Indic use of Cyrillic alphabet. Modern Indic languages have undergone extensive Middle Iranian languages have been ergative lan- changes in the course of their evolution. Although the guages, and this feature has still been retained in many vocabulary of many Indic languages derives primarily modern Iranian tongues. Persian, Lori, Mazandarani, from Sanskrit, Muslim influence over the centuries has and Gilaki are, of course, among the nominative-accu- added loan words from Arabic and Persian to some sative Iranian languages. Indic languages such as Urdu and Sindhi. Besides The east Iranian languages are spoken in Pamir phonemic and vocabulary changes, the simplification region; Afghanistan, Tajikistan, and to a much lesser of the inflection of nouns and verbs is quite consider- extent, central Asia. They are numerous and scattered, able. Cases of the nouns have been reduced from eight 516  INDO-EUROPEAN 4: ROMANCE to two, and are mainly represented by prepositions, which are so many in number) are spoken by the peo- rather than the earlier case markers. ple in Nepal and North India. Pahari has two main Modern Indic languages are so many; thus their dialects, Garhwali and Kumaoni, and Western Pahari. classification is also diverse. They can be classified as ‘Northwest Indic languages’ (like Punjabi and Sindhi); Dardic ‘Central Indic languages’ (like Hindi and Urdu); East Indic languages (like Assamese, Bengali, and Oriya); There is no general agreement among the linguists on the West Indic language (Gujarati); South Indic lan- the status of Dardic group. Many references believe guages (like Marathi and Singhalese); and the northern them to be a subgroup of the Indic, but there are firm or Pahari languages. evidences that classify them as a separate category. As to the population of the speakers, Hindi, now the Dardic, or Pisacha, languages are spoken in national language of India, stands out, and is under- Afghanistan. These languages share certain distinctive stood by about half a billion people, mostly living in phonetic characteristics, feature the use of pronominal central India. Urdu is the other language, so similar to suffixes with various verb forms, and include in their Hindi (but recorded by Persian alphabet), spoken vocabularies a number of words that among the lan- mainly in Pakistan and India. Punjabi, spoken in guages of India are usually encountered only in Vedic Northwest India and Pakistan, is close to the Western Sanskrit. Kashmiri is the sole Dardic language that Hindi dialect and is written in an alphabet based on the both has a literature and is recognized in the Indian Devanagari script. Sindhi, the native language of peo- constitution of 1950. It is written in Persian letters by ple in Southeast Pakistan and West India, is the other Muslims, whereas Hindus use a script similar to the language of this group, recorded in a modification of a Devanagari alphabet. Persian script by Muslims, although Hindus use a vari- ety of the Devanagari alphabet as well. References Of the East Indic languages, Assamese is notewor- thy. Although reminiscent of Sanskrit in vocabulary, it Boyle, Grammar of modern Persian (1966); C.P. Masica, The is allied grammatically to Bengali, the national lan- Indo-Aryan languages (1989). Campbell, George L. 1991. Compendium of the World’s lan- guage of Bangladesh. Oriya, the language of about 32 guages. London: Routledge. million persons, chiefly in the Indian state of Orissa, is Chatterji, S.K., Indo-Aryan and Hindi (2nd edition 1960); A.M. closer to Sanskrit phonetically and lexically than any Ghatage, Historical Linguistics and Indo-Aryan Languages other modern Indic tongue. The leading West Indic (1962); J. Bloch, Indo-Aryan, from the Vedas to modern language is Gujarati, spoken chiefly in the states of times (revised edition, tr. 1965); T. Burrow, The Sanskrit language (2nd edition 1965). Gujarat and Maharashtra in India. Crystal, J.A., David. 1987. The Cambridge encyclopedia of lan- Of the South Indic languages, Marathi is spoken in guage. 2nd edition. New York: Cambridge University Press. the Indian state of Maharashtra. The other important Grimes, Barbara F. (ed.) Ethnologue. Summer Institute of South Indic language is Singhalese, the language spo- Linguistics, ongoing. ken by people of Sri Lanka. This language is geo- Lyovin, Anatole V. 1997. An introduction to the languages of the World. Oxford: Oxford University Press. graphically separated from the other Indic languages Ruhlen, Merritt. 1987. A guide to the World’s languages: clas- of North and Central India by an intervening region in sification, Vol. 1. Stanford, CA: Stanford University Press. South India, in which the Dravidian languages are spo- Schmitt, R. 1992. Compendium linguarum Iranicarum. ken. Thus, it is greatly influenced by the Dravidian Wiesbaden: Dr. Ludwig Reichert. languages nearby. The Pahari languages (or dialects, BEHROOZ MAHMOODI-BAKHTIARI Indo-European 4: Romance Spoken throughout the Roman Empire, around the More precisely, it was its late form by the end of the Mediterranean Sea, and even beyond, Latin, an Indo- Empire (fourth century CE), which was also called European language, was the starting point of a dialec- ‘Vulgar Latin’. It is difficult to pinpoint the moment tal differentiation that gave rise to Romance languages. when Romance languages first appeared, however, 517 INDO-EUROPEAN 4: ROMANCE since these idioms are extensions of the Latin language. which, starting with Alfonso X El Sabio (thirteenth Today, Romance languages in Europe extend from the century), became dominant and the basis for modern Atlantic to the Black Sea. From west to east, the fol- Spanish language, even though today there are lowing national languages can be found: Portuguese, attempts to rehabilitate Aragonese. Spanish is the Galician, Spanish, Catalan, Occitan, French, Rhaeto- Romance language that has grown the most: it can be Roman, Italian, Sardinian, and Romanian, making up heard in Israel where there is a large Judeo-Spanish what is known as Romania. The international expan- colony (about 350,000 speakers) and also almost all sion of these languages must also be taken into over Latin America since the conquest of the Aztec account. Linguistically, they are divided into two enti- and Inca Empires in the sixteenth century, when it ties: western Romania (Italo-Roman, Ibero-Roman, became the official language of all these countries. Its Gallo-Roman, Rhaeto-Roman) and eastern Romania expansion continues in US States contiguous to (Daco-Roman). Mexico (Arizona, New Mexico, California, Texas, etc., with at least 20 million speakers). The only note- worthy Creole that developed from contact with Italian Spanish is that of the Dutch Antilles, Papiamento, Italy is the epicenter of the Italian language: 95% of its which is spoken by 200,000 people and is an official population speaks it. Dialectical segmentation, howev- language. Spanish is not widespread in Africa and er, is very important there. Three key subgroups can be only 2 million Filipinos use it in Asia. Overall, 300 identified: in the north, Septentrional Italian with Gallo- million people are Hispanophones. Italic (Piemontic, Lombardic, Genoese) and Venetian, Tuscan in the center (but also in Sardinia and in Portuguese Corsica), and the Central-Meridional (ranging from Roman [Rome] to Calabrian and Sicilian [Calabria, The formation of modern Portuguese comes mostly Sicily]). This fragmentation slowed the standardization out of Gallego-Portuguese, which used to link of the Italian language: Venetian first played this part, Portuguese and Galician dialects. Political independ- but it was Tuscan (Florence) that triumphed from the ence allowed Portuguese to become a true Romance nineteenth century onward, even though all the region- language. It is the result of the harmonious fusion of al dialects still remain quite lively today. Italian the different dialects from the provinces that make up (Venetian) also remained a vehicular language up until the country: dialects from Coimbra, Lisbon, Beira, and the end of the nineteenth century. Although it still Estramadure. The existence of Azorean and Madeiran, retains traces of its past expansion in Africa (Somalia, however, can be noted as their differences remain Ethiopia, Erythrea), Italian now radiates mostly through somewhat stronger due to their insular character and the emigration of its speakers. There are thus large com- late colonization (fifteenth century). As with Spanish, munities of Italians in America (6 million), Australia (1 but to a lesser degree, Portuguese lexicon was influ- million), or Europe (2 million, mainly in Belgium). enced by Arabic. On the other hand, it has imported Overall, this Romance language is spoken by about 66 more words from its (former) African colonies and its million people, 57 million of whom are in Italy. (ancient) Asian trading posts. Furthermore, it can be said that the notable differences found in the Portuguese spoken in Brazil are contributing to the Sardinian creation of a new Romance language: Brazilian. The Still being used in Sardinia by 1.2 million people, differences are most obvious in the lexicon, with bor- Sardinian language has a conservative character com- rowings from indigenous languages (Tupi) and from pared to other Romance languages. Its lexicon is archa- the languages of the former African slaves (Bantu), ic, except for late borrowings from Catalan, Castilian, and in its syntax, which is at times archaizing. and Tuscan languages. There are two subgroups: Brazilians now make up the largest Lusophone group Logudorian (in the north) and Campidanian (in the in the world, with 150 million speakers. Further, south). The vast majority of Sardinian speakers are Portugal’s former African colonies have all retained bilingual Sardinian–Italian, although its recent status Portuguese as the official language, but only 5% of as an autonomous region is giving it new strength. their population (i.e. 2.5 million people) use it. Native people in these countries have often modified it, giv- ing way to Creoles, such as Capeverdean Creole (Cape Spanish Verde), which is spoken by 230,000 people. Lastly, the The Spanish Language has three main dialects: Astur- Portuguese Diaspora can be found in France as well as Leonese from Asturias to Salamanca, Aragonese in the in the United States of America (Massachusetts). All southern Pyrenees, and Castilian (ancient Castile), in all, it can be estimated that Portuguese is now 518  INDO-EUROPEAN 4: ROMANCE spoken by at least 170 million people all over the as the eleventh century; it then spread across its bor- world, with the great majority of speakers being from ders (conquest of England in particular), and beyond outside its country of origin. (Lebanon). In the eighteenth century, at a time when France excelled, Francization reached Flanders and Russia. Later, the colonization of America and Galician Africa (Maghreb, Black Africa) gave rise to Creole Geographically, Galician is located within a Spanish languages in America (Louisiana), Caribbean (Haiti, Province. Yet, it is a dialect from the Portuguese fam- Antilles, Guyana), and some parts of the Indian Ocean ily and can be found in northern Portugal also. Its loca- (Madagascar, Reunion, Seychelles) as well as to the tion made it lose its contact with Portuguese very early Québecker dialect (the official language of Québec) on. In fact, it mixes both Spanish and Portuguese char- and Acadian (New-Brunswick). There are 70 million acteristics. It began to reappear at the beginning of the francophone speakers in Europe, 72 million in African twentieth century and is today spoken by about 2.5 countries (Senegal-Djibouti axle), and 45 million in million people (or 90% of Galicians), but its use is the Maghreb. To this, the Caribbean (750,000), the strongly declining in cities. Galician has, within its Indian Ocean (600,000), and Lebanon must be added. Spanish Province, the status of an official language. There are about seven million Francophones in Québec and 280,000 more across the other Canadian Provinces. The Acadians in Louisiana number about Catalan 400,000 people to which 50,000 more speakers living Its resemblance with Occitan leads one to consider in the eastern part of Louisiana and in Texas must Catalan as part of the Gallo-Roman group rather than be added. Thus, French is actually spoken in at least Ibero-Roman. Its mixed lexicon, however, shows that 38 states and the total number of French speakers is this language is more like a ‘bridge’ between the two estimated to be a little under 200 million, with only groups. Its strength is that it is an international idiom, 30% residing in Europe. since this language can be found in Spain (Catalonia, Baleares), as well as in France (Roussillon), Andorra, Occitan or Italy (Sardinia). It does not show any strong dialec- tical segmentation. In Spain, it counts six million Occitan dialects can be found in the south along a line speakers and served as the official language of that splits France from Bordeaux to Briançon. There Catalonia between 1931 and 1939. It has regained this are three subgroups: Septentrional Occitan (Limousin, status since 1979. Although it has the status of official Auvergnat, etc.), Southern Occitan (Languedocian, language in Andorra, it does not have any real status in west of the Rhone, and Provencal in the east), and France, nor in Sardinia, where less than 20,000 people Gascon (in Aquitaine). Occitan is characterized by dif- use it. ferent phonetic and morpho-syntatic traits. Gascon, however, through its relation to Ibero-Roman, shows originalities to the point that it could be considered a French Romance language in itself. Literary and judicial lan- French is the one Romance language that has evolved guage until the Albigensian Crusade (1209–1229) and the most compared to the original Latin. French lan- the annexation of its lands by the Crown, the Occitan guage is located in the northern half of today’s France then experienced a decline. It would not reemerge (except for Brittany and Alsace), as well as in southern until the nineteenth century with the Felibrige Belgium (Wallonia). It is part of the Langue d’Oil Movement and F. Mistral. Since the second half of the dialects, which were thus named in opposition to the twentieth century, autonomist movements have Langue d’Oc languages (Occitan). These dialects are claimed it as a regional language. The number of many. Some remain in use, such as Norman speakers who master Occitan nowadays can be rea- (Normandy) and Gallo (west Brittany). Walloon has sonably estimated at about 10 million. even been recognized as a regional language in Belgium since 1990. Others have fully disappeared, Rhaeto-Roman such as Champenois or Berrichon, or are hardly sur- viving, such as Franco-Provençal (100,000 speakers in There is no real dialectical unity in this subgroup. Its France and Italy). Each has its own set of particulari- specificity is mostly historical. Linguistically, it is ties that would be difficult to detail herewith. hinging on both Romania, based on its lexicon in par- The expansion of French is multiple. It first grew ticular, but globally, it is close to Gallo-Roman. It is within the country, with the progressive eviction of all made up of three dialects, geographically fragmented: other dialects (be they Gallo-Roman or not) as early Romansch, spoken in Switzerland (Canton of Grison), 519 INDO-EUROPEAN 4: ROMANCE Ladin in Italy (Tyrolean Alps, north Trentino), and the Western Italo-Roman Italian 66,000,000 Friulan also in Italy (Province of Udine). Wedged Romania Sardinian 1,200,000 between Bavarian and Alemannic in the north Ibero-Roman Spanish 300,000,000 (Germanic languages), and Lombard in the south Portuguese 170,000,000 (Italian subgroup), this Romance language has frittered Galician 2,500,000 away as the centuries went by. Today, there are only Gallo-Roman Catalan 6,020,000 36,000 Romansch speakers, 30,000 Ladinophones, and French 200,000,000 about 600,000 Friulan speakers. Depending on their Occitan 10,000,000 geographical location, most are bilingual (German in Rhaeto-Roman Romansch 36,000 the north, Italian in the south). And even though Ladin 30,000 Romansch has been one of the national languages of Friulan 600,000 Switzerland since 1938, assimilation seems irre- Eastern Daco-Roman Romanian 24,000000 versible. Romania 780,386,000 Romanian One characteristic of the Romanian language is that it Speakers of Romance languages are superior in is very conservative on some points while at the same- number to that of Germanic languages (about 560 mil- time being very innovative on others. The influence of lion speakers) thanks, in particular, to Latin America. Slavic and Byzantine cultures explains its originalities. Spanish is nowadays the most spoken language, fol- The Romanian lexicon is under Slavic influence, as lowed by French and Portuguese. Still, they cannot well as under Greek, Hungarian, and even modern quantitatively vie with Chinese (at least 900 million) French influence. In the eighteenth century, Slavonic or English, which, although officially spoken by only stopped being the official language. Influenced by the 450 million people, has become the international lan- ideas of the French Revolution and by an awareness of guage for communication and information. the Latin origins of the language, modern Romanian appeared at the beginning of the nineteenth century. References The Latin alphabet became prominent in Romania Allières, J. 2000. Les langues de l’Europe. Paris: PUF, Que beginning in 1860, but had to be adapted. The Cyrillic sais-je ?. alphabet, however, was still used in Moldova up to the Bach, K.F., and G. Price. 1977. Romance linguistics and the dissolution of the USSR. Romanian language is subdi- Romance languages: a bibliography of bibliographies. vided into four groups: Daco-Romanian (found in London: Grant and Cutler. Bal, W. (ed.) 1992. Les langues régionales romanes de today’s Romania and official language and in Wallonie. Bruxelles, Traditions. Moldova), Istro-Romanian (north of Thessalia), Baldelli, I. (ed.) 1987. La lingua italiana nel mondo. Rome: Macedo-Romanian (or Vlach, dispersed in Greece and Istituto della Enciclopedia italiana. Macedonia), and Megleno-Romanian (south of Bec, P. 1980. La langue occitane. Paris: PUF, Que sais-je ?. Bulgaria and north of Thessalonica). Istro-Romanian Boléo, M. de P. 1974–1975. Estudos de linguística portuguesa e românica, 2 vols. Coimbra: Acta Univ. Conimbrigensis. is spoken by only about a 1,000 people and is an Chaurand, J. 1972. Introduction à la dialectologie française. endangered language. Macedo-Romanian is spoken by Paris: Bordas. about 100,000 people and Megleno-Romanian, the Entwisle, W.J. 1962. The Spanish language (together with dialect of a Muslim community, by 15,000 people. The Portuguese, Catalan and Basque). London: Faber & Faber. unified Romanian language (Daco-Romanian) counts Ferreira, J. de A. 1989. Bibliografia selectiva da língua por- tuguesa. Lisboa: Instituto de Cultura e Língua Portuguesa. overall 24 million speakers, 20 million of whom are in Foster, D.W., and V. R. Foster. 1970. Manual of Hispanic bibli- Romania and 3 million in Moldova. The language can ography. Seattle: University of Washington Press. also be heard in Ukraine (400,000 speakers), Griera, A. 1947. Bibliografía lingüística catalana. Barcelona: Macedonia, Voivodina, Albania, and Hungary. The Escuela de filología. largest Diaspora outside Europe is in the United States Hall R.A. Jr. 1958–1988. Bibliografia della linguistica italiana, 6 vols. Firenze: Sansoni. of America with over 200,000 speakers. Harris, M., and N. Vincent. 1977. The Romance languages. London: Routledge. Lleal, C. 1990. La formación de las lenguas románicas penin- Conclusion sulares. Barcelona: Barcanova. On the whole, 780 million people speak a Romance Lombard, A. 1974. La langue roumaine. Paris: Klincksieck. Martin, R., and M. Wilmet. 1973. Guide bibliographique de lin- language, which represent about 20% of the Indo- guistique française. Paris: Klincksieck. European language speakers and 8% of humanity, as Maxfield, M.F. 1941. Raeto-Romance bibliography. Chapel shown below. Hill: University of North Carolina Press. 520  INDO-EUROPEAN 5: SLAVIC Melo, G.C. de –. 1981. A lingua do Brasil. Rio de Jeneiro: Padrão. Wagner, M.L. 1951. La lingua sarda : storia, spirito e forma. Pei, M. 1941. The Italian language. New York: Columbia Bern: Francke. University Press. Walter, H. 1994. L’aventure des langues en occident. Paris: Picoche, J., and Chr. Marchello-Nizia. 1989. Histoire de la Laffont. langue française. Paris: Nathan. Wartburg, W. von –. 1967. La fragmentation linguistique de la Posner, R. 1996. The Romance languages. Cambridge: Romania. Paris: Klincksieck. Cambridge Univeristy Press. Ethnology Database: www.ala.doc.ic.ac.uk/~rap/Ethnologue/ Pottier, B. 1969. Grammaire de l’espagnol. Paris: PUF, Que wgt.html sais-je ?. Titus: www.fusl.ac.be/Files/General/BCS/Ling1.html Roca-Pons, J. 1971. Introducció a l’estudi de la llengua cata- THIERRY PONCHON lana. Barcelona: Vergara. Rohlfs, G. 1977. Le Gascon. Pau-Tübingen: Marrimpour See also Canada; Dialectology; France; French Niemeyer. Language; Indo-European 2: Germanic Lan- Tétu, M. 1987. La Francophonie. Histoire. Problématique. guages; Indo-European 3: Indo-Iranian Lan- Perspectives. Montréal: Guérin. guages; Indo-European 5: Slavic; Italian; Italy; Vasconcellos, J.L. de –.1970. Esquisse d’une dialectologie por- tugaise. Lisboa: Centro de Estudos filológicos. Latin; Meillet, Antoine; Mexico; Phonology; Veny, J. 1982. Els parlars catalans : síntesi de dialectologia. Romanian; Saussure, Ferdinand de; Spain; Span- Palma de M: Moll. ish and Iberoromance Languages Indo-European 5: Slavic The Slavic languages, spoken by some 288 million (Croatia), Bosnian (Bosnia and Herzegovina), Serbian people, constitute one of the branches of the Indo- (Serbia and Montenegro), Macedonian (Macedonia), European family. Until approximately 500 CE, before and Bulgarian (Bulgaria). Upper and Lower Sorbian it disintegrated into the various Slavic languages spo- are spoken in Germany by the ethnic Sorbian minori- ken today, the language of the Slavs was relatively uni- ty in and around the towns of Bautzen and Cottbus, form. This language, referred to as Proto-Slavic or respectively. Kashubian, a divergent dialect of Polish Common Slavic, was not written down; thus, it must that has gained literary status, is spoken in northern be reconstructed by using the evidence of later written Poland in and around the city of Gdansk. Significant and oral sources. The phase of the unattested Slavic minority populations of Slavs exist outside the borders language up to 500 CE is usually referred to as Proto- of the matrix nations, e.g. Russian is spoken by signif- Slavic, and from that point to approximately the tenth icant percentages of the population of Belarus, century, it is referred to as Common Slavic. Its closest Estonia, Latvia, and Ukraine; Slovak is spoken in relative is the Baltic family, made up of modern Hungary and Ukraine; and Slovene is spoken in Italy, Latvian and Lithuanian, as well as extinct Old Austria, and Hungary. Prussian. The modern Slavic languages are divided Slavic writing began in the late ninth century CE in into three branches: West (Czech, Slovak, Upper and connection with the conversion of the pagan Slavs to Lower Sorbian, Polish, and Kashubian), East Christianity and, specifically, the mission of the (Russian, Ukrainian, and Belarusian), and South Byzantine monks, Constantine and Methodius, who (Slovene, Croatian, Serbian, Bosnian, Macedonian, developed the first alphabet for the Slavs, called and Bulgarian). Today, the Slavic languages are spo- Glagolitic, and translated the Scriptures into Slavic. ken over a large part of Europe and parts of northern The earliest surviving texts are from the late tenth cen- and central Asia, as well as in diaspora communities in tury; this means that there are no extant texts produced North and South America and Australia. Most are during the lives of Constantine and Methodius. By the national standard languages: Polish (Poland), Czech eleventh century, the Glagolitic alphabet was largely (Czech Republic), Slovak (Slovakia), Russian (Russia, replaced by a modified version of the Greek alphabet, also as a second language in many former republics of which has come to be known as Cyrillic. This alphabet the erstwhile Soviet Union), Belarusian (Belarus), is still used in modernized forms by the East Slavs, as Ukrainian (Ukraine), Slovene (Slovenia), Croatian well as the Serbs, Macedonians, and Bulgarians. The 521 INDO-EUROPEAN 5: SLAVIC bulk of the earliest texts that have survived are from Slavs apparently shared religious beliefs: bogu˘ ‘god’, the eleventh century, the language of which is now raj ι ‘heaven’, sve tu ‘holy’, xvala ‘glory’. Germanic referred to as Old Church Slavic (or Old Church loanwords entered Slavic over a longer period, before, Slavonic, which is also the British term), because of presumably during, and after the migrations. Early their religious content and function. However, their loans from Proto-Germanic and Gothic include kupiti writing was also used for secular purposes, as evi- ‘to buy’, t’ud’ ι ‘foreign’, and st ιklo ‘glass’. Later denced by the Novgorod Birchbark Letters, which loans from Old High German, but still in the Common contain business records and personal correspondence. Slavic phase, include kral’ ι ‘king’ (from Karl, refer- The Letters are still being excavated in and around the ring to Charlemagne) and pene dz ι ‘coin’ (cf. German town of Novgorod, Russia. Latin script writing Pfennig). Borrowings from various Turkic languages developed among the Slavs who used the Roman rite, appear throughout Slavic, although they are arguably and it continues to be used in modified, modern forms from different sources, e.g. Russian kolpak ‘hat’ vs. by the Czechs, Slovaks, Poles, Kashubians, Slovenes, Czech klobouk, Slovene klobuk (cf. Crimean Tatar Croats, and Bosnians. The oldest use of the Latin kalpak ‘cap’). Greek and Romance loanwords entered alphabet for a Slavic language is attested by the Slavic substantially through the mediation of the Freising Folia—three brief texts in the Roman rite that Church, more or less at the beginning of the historical reflect an early eleventh-century Slavic dialect that period (i.e. from the tenth century), e.g., d ι javolu was a precursor to modern Slovene. ‘devil’, psalmu ‘psalm’, kr ιstu ‘cross’ (Greek); m ιša The Slavic languages are most closely related to the ‘Mass’, kriDι ‘cross’, Didu ‘Jew’ (Latin). Contact with Baltic languages, as evidenced by certain innovations varieties of Romance and other languages is responsi- that are common to the two families and absent in ble for many of the peculiarities of the Balkan lan- other Indo-European branches. For example, both guages. Contact with Baltic Finnic languages played a Slavic and Baltic have abstract nouns formed with the role in the formation of Russian, especially its north- component -iba: Old Russian druD bba ‘friendship’, ern dialects. Latvian draudz¯ι ba ‘friendship’ (the Indo-European The Slavic languages present a variegated picture root *dhreugh- ‘to keep together’ was derived as drug with respect to their sound structures. The West and ‘friend’ in Russian and other Slavic languages; it is East Slavic languages are characterized by complex also related to Old English gedre¯ag ‘pack [of ani- distinctions in consonants. Russian, for example, dis- mals]’). There is some debate as to whether these com- tinguishes between plain and palatalized consonants, mon innovations arose as a consequence of a the latter of which are characterized by a raised tongue continuous common dialect emerging from Indo- position, similar to that produced by English speakers European or a rapprochement resulting in intensive when pronouncing the letter y, viz. privedëš’ contact between Proto-Slavic and Proto-Baltic. [pryivyi2dyoʃ] ‘you will bring’. However, it should be Whether through common origin or contact, the close emphasized that Russians perceive this secondary relationship of Slavic and Baltic continued up to about articulation as a characteristic of the consonant itself, the fifth century BCE, a time depth at which it is very because they contrast such sounds with sequences of difficult to reconstruct the early forms of the language plain consonant plus y, e.g. sel [2syel] ‘sat’ and s”el in adequate detail. Before the Balto-Slavic period, the [2syyel] ‘ate’. Czech has intensified the development of precursor of Slavic seems to have emerged from Indo- palatalization so that what were formerly palatalized European in conjunction with the precursors of Indo- consonants have evolved into consonants articulated at Iranian, Armenian, Germanic, and Celtic. After the the middle of the palate; thus, the first consonant in the Balto-Slavic period, having migrated westward from word tichý ‘quiet’ is neither t nor k, but a sound in the Indo-European homeland, Proto-Slavic crystal- between the two; this sound contrasts with a plain t, as lized as a separate ethno-linguistic entity in what is in ty ‘you’. This intensification has resulted in the now Ukraine, in an area bounded by the rivers Bug in development of a trilled fricative, r (r eka ‘river’, cf. the west, the Dnepr in the east, and the Pripet’ in the Russian [rye2ka]), which is pronounced as the s in north (separating them from the Proto-Balts), although English pleasure and simultaneously rolled/trilled as the exact location of the Proto-Slavic homeland cannot in Spanish burro. In Polish, this sound existed histori- be determined. cally, but it has merged with the nontrilled fricative, Borrowings of words from non-Slavic languages rzeka [2Deka] ‘river’. Vocalic and accentual systems that are attested throughout the Slavic world give evi- vary considerably from language to language. Czech dence of the contacts that the Proto-Slavs had before and Slovak distinguish long and short vowels and have their migrations. Among the earliest of these contacts stress fixed on the first syllable of the word, e.g. Czech were with Indo-Iranian-speaking peoples, the Alans, dal [2dal] ‘he gave’, dál [2da:l] ‘further’. Slovene and Scythians, and Sarmations, with whom the Proto- Serbo-Croatian also preserve long and short vowels, 522  INDO-EUROPEAN 5: SLAVIC but in addition they distinguish between rising and formation is the rich array of alternations in the shape falling intonation (pitch) in stressed syllables, e.g. of words; these alternations have led to a well-devel- Slovene brati [b2rá:t] ‘to read’ (long low or rising oped investigation into the nexus of sound systems and pitch), brat [b2rà:t] ‘to go read’ (long high or falling word structure known as morphophonemics. An exam- pitch), bràt [b2ràt] ‘brother’ (short high or falling ple of a morphophonemic alternation is found in pitch). In general, there seems to be a complementary Russian viDu ‘I see’ vs. vidiš ‘you see’, where the basic relationship between consonantal and vocalic/accentu- form of the root vid- ‘see’ changes to viD- in the al systems: the more complex the consonantal system, context of the first-person singular nonpast. the simpler the vocalic and accentual systems, and Verbs in Slavic languages are distinctive among vice versa. European languages in that they overtly mark As regards word structure, the Slavic languages are aspectual contrasts. That is, they distinguish complet- characterized by inflection; that is, words change to ed and uncompleted actions/events, as well as express reflect grammatical relationships. For example, the other varieties of temporal and spatial manners of Russian word sobaka ‘dog (as a subject)’ has the fol- organization. For example, Slovene skocim ‘I jump lowing forms: sobaki ‘of the dog’, sobake ‘to the dog’, (once)’ (completed—perfective aspect), skacem ‘I am sobaku ‘dog (as an object)’, sobakoj ‘as a dog’, na jumping’ or ‘I jump habitually’ (e.g. as a professional sobake ‘on the dog’, sobaki ‘dogs (subject)’, sobak ‘of ski-jumper) (uncompleted—imperfective aspect). the dogs’ or ‘dogs (as an object)’, sobakam ‘to the Further aspectual distinctions may be derived through dogs’, sobakami ‘as dogs’, and na sobakax ‘on the suffixation and prefixation, e.g. Russian ona brosila dogs’. This characteristic allows word order to remain mjac ‘she threw the ball’ (perfective), ona brosala flexible, which in turn permits word order to take on mjac ‘she was throwing the ball’ (imperfective); functions other than indicating grammatical relations, ona podbrosila mjac ‘she tossed the ball up’ (perfec- such as, for example, emphasis: Ja ljublju sobaku ‘I tive), and ona podbrasyvala mjac ‘she was tossing the love the dog’ (normal word order) vs. Sobaku ljublju ja ball up’ The arrangement of tense (time) in Slavic ‘It is I who loves the dog’ (emphasizes the subject). In verbs has been considerably reorganized from the the Balkan Slavic languages, Bulgarian, and Indo-European starting point. Russian is an example Macedonian, most of these inflectional changes in of one of the more innovative tense systems. For nouns have been eliminated. Slavic verb forms mark example, the future is formed by using a perfective person (the subject of the verb) and number (singular, verb with the same formal properties as the imperfec- plural, and in some Slavic languages, dual); for this tive present: ja pojdu ‘I shall go’ (the corresponding reason, many Slavic languages omit the subject pro- imperfective ja idu means ‘I am going’ or ‘I go’). The noun in neutral speech, e.g. Czech vidím ‘I see’, vidíš future with the imperfective aspect is formed by ‘you see’, vidí ‘s/he sees’, vidíme ‘we see’, vidíte ‘you adding an auxiliary verb budu (which goes back to an see’ (polite or plural), vidí ‘they see’. Three genders Indo-European root *bheuH- meaning ‘to grow’) to are distinguished—masculine, feminine, and neuter— the infinitive: ja budu iti ‘I will be going.’ Past tense is although unlike many Western European languages, formed by a participle derived with the formant -l- these are not expressed by definite articles (‘the’, such that agrees in gender and number with its subject, e.g. as German die, der, das or French la, le), there being ja pošël ‘I left.’ This rather simple schema supplanted no articles in Slavic languages. Rather, noun endings a much more complex system of past tenses that was indicate gender, e.g. Russian Dena ‘wife’, kniga ‘book’ still in place in medieval varieties of Slavic, e.g. Old (both feminine); celovek ‘person’, stol ‘table’ (no end- Russian (twelfth to thirteenth centuries), i vu lny byša ing—both masculine); and okno ‘window’ (neuter), vyše korablä ‘and the waves were higher than the more ‘sea’. Gender serves to coordinate sentence ele- ship’ (aorist, a simple past narrated event); i rece s ι ments through agreement; for example, an adjective unoša roda velika jest ι bylu ‘and this youth said that modifying a noun must agree in gender (as well as he was of a great family’ (perfect tense, meaning that other grammatical categories) with the noun: Russian the narrated past event is of relevance to the moment krasivaja Dena ‘beautiful wife’ (feminine), krasivoe of narration). This complex system was preserved and more ‘beautiful sea’ (neuter). It may also agree with developed further in Macedonian and Bulgarian, the past tense of a verb, e.g. Dena stojala tam ‘the wife which have otherwise lost much of the complexity of was standing there’, celovek stojal tam ‘a person was noun and adjective inflection found elsewhere in standing there’. Moreover, the formation of gendered Slavic. Parallel to the complementariness between noun pairs is effected by adding a suffix to produce the consonants and vowels/accent, pointed out previously, feminine member of the relationship, e.g. Russian such a relationship seems to hold in Slavic languages kurd ‘Kurdish man’, kurdjanka ‘Kurdish woman’. A between the complexity of nominal and verbal typical characteristic of Slavic inflection and word inflection. 523 INDO-EUROPEAN 5: SLAVIC References Schenker, Alexander M., and Edward Stankiewicz (eds.) 1980. The Slavic literary languages: formation and development. Birnbaum, Henrik. 1979. Common Slavic. Progress and prob- New Haven: Yale Concilium on International and Area lems in its reconstruction. Columbus, OH: Slavica. Studies. x+287pp. xii + 436pp. Shevelov, George Y. 1965. A prehistory of Slavic. The histori- Birnbaum, Henrik, and Peter T. Merrill. 1985. Recent advances cal phonology of common Slavic. New York: Columbia in the reconstruction of common Slavic (1971–1982). University Press. xx+62pp. Columbus, OH: Slavica. vi+141pp. Comrie, Bernard, and Greville G. Corbett (eds.) 1993. The MARC L. GREENBERG Slavonic languages. London: Routledge. xiii+1078pp. See also Balkans; Baltic Languages; Polish and Goa b, Zbigniew. 1992. The origins of the Slavs. A linguist’s view. Columbus, OH: Slavica. 454pp. West Slavic Languages; Russian and East Slavic Schenker, Alexander M. 1995. The dawn of Slavic. An intro- Languages; Serbo-Croatian and South Slavic Lan- duction to Slavic philology. New Haven: Yale University guages; Soviet Union; Soviet Union: Successor Press. xviii+346pp. States Indo-Pakistani Sign Language Indo-Pakistani Sign Language (IPSL) is a visual-gestur- are multidialectal or become multidialectal very al language that uses movements of the hands, facial quickly as soon as they travel to other areas. Moreover, expressions, and head/body positions to convey linguis- since IPSL is a minority language in constant contact tic messages. Dialects of IPSL are used in deaf commu- with spoken languages, all IPSL users are to some nities in urban centers of the Indian Subcontinent. extent bilingual in the signed and the spoken medium, Figure 1 shows the extent of the geographic area as doc- the latter mostly in its written form. However, compe- umented to date. It is likely that dialects of IPSL are also tence in spoken languages, be it English, Hindi, or one used in other parts of India and/or Pakistan, maybe even of the regional languages, is often very low among in neighboring countries (Sri Lanka, Nepal, deaf people. The use of so-called ‘mouthing’ is also a Bangladesh), but this has not been fully documented. contact phenomenon resulting from the bilingual situ- In the deaf community, the sign language is simply ation. Mouthing means that sign language users called ‘sign/signing/sign language’, sometimes in com- accompany signs with mouth movements that corre- bination with the sign for the country (as in INDIA spond to words of a spoken language. For example, SIGN). Various names are used by hearing people. In one may imitate the mouth movements of the Hindi the Hindi/Urdu-speaking area, the sign language is word kaam ‘work’ while signing WORK. known as ishaaron kii zubaan (“language of signs”). The size of the language community has not been Official usage also refers to “Pakistan Sign Language documented reliably, but IPSL users definitely number (PSL)” and “Indian Sign Language (ISL)”, respectively, in the hundreds of thousands, possibly even over a mil- although this usage runs contrary to the linguistic facts. lion, thus representing one of the largest sign language The large area covered by one and the same sign communities in the world. IPSL is not known to be language is particularly noticeable in view of the great genetically related to any other sign language. A minor linguistic diversity of spoken languages in the region. influence from British Sign Language can be seen, for IPSL is used in both the Indo-European language area example, in the use of a two-handed manual alphabet (e.g. Hindi-, Nepali-, and Marathi-speaking areas) and (fingerspelling) for representing English words. No the Dravidian language area (e.g. Telugu-speaking manual alphabet for indigenous Indian languages is areas in Andhra Pradesh). All IPSL dialects have the widely in use. same grammar, but lexical variation may be consider- able. On average, IPSL dialects have about 75% of Word Classes and Sentence Structure shared vocabulary, with about 25% of the vocabulary different across dialects. IPSL has three main word classes: verbs, multifunc- Despite dialectal differences, IPSL users can com- tional words, and particles. There are no word classes municate freely across a large geographic area. Many of nouns, adjectives, and adverbs. The most important 524  INDO-PAKISTANI SIGN LANGUAGE Kashmir Islamabad Jammu Lahore Punjab PAKISTAN U New Delhi NEPAL tta r Pr Kathmandu ad es Sindh Assam h Karachi Madhya Pradesh BANGLADESH Bhopal West Gujarat Bengal Dacca Calcutta INDIA a iss Mumbai Or Bhubaneshwar Maharashtra h es ad Pr ra Hyderabad dh An Bangalore Chennai du ilna m Ta Colombo SRILANKA Figure 1. Geographic extent of IPSL. subclass among the particles is a set of nine functional IPSL has a subclass of signs related to pointing. particles. These appear at the end of a sentence or Pointing with the index finger is often function- clause and indicate sentence types such as negative ally equivalent to pronouns in spoken languages. sentences, questions, completed actions, commands, Pronouns have the full set of IPSL number distinc- and existentials (equivalent to English ‘there is’). tions, which include a form unspecified for number The class of verbs is rather small, consisting of that may have either singular or plural reference, a about 50 signs, such as the signs for ‘help’, ‘ask’, dual (for the number ‘two’), and several types of plu- ‘give’, ‘tell’, ‘teach’, ‘go’, ‘come’, and the like. These rals. Another type of pointing sign, with the index fin- are characterized by so-called ‘directional movement’, ger drawing a line between two locations in space, is which means that the movement of the verb varies used as an auxiliary. The auxiliary can express sub- depending on the relationship between the persons, ject–object-like relationships in a general way (trans- objects, or places involved. Figure 2 shows an exam- latable as ‘he to me’, ‘I to you’, and so on) in ple of a sign expressing a relationship between two combination with signs that do not have directional persons. movement, for example: SIGN AUXILIARY-you-me- Most signs in IPSL belong to the class of multi- you ‘You talk to me in sign language and I talk to you’ functional signs, including signs as diverse as (see Figure 3). ‘change’, ‘begin’, ‘difficult’, ‘black’, ‘three’, ‘self’, IPSL is a predicate-final language, the basic sen- ‘many’, ‘war’, ‘child’, and so on. One and the same tence structure being ARGUMENT – PREDICATE – multifunctional sign can function as the equivalent of FUNCTIONAL PARTICLE, as in SCHOOL FAR both a noun and a verb/predicate in other languages. NEGATIVE-PARTICLE ‘The school is not far.’ A sec- For example, one would sign BEGIN DIFFICULT to ond argument will usually be a pronoun and can appear express ‘The beginning is difficult.’ (nominal func- just about anywhere, e.g. INDIA YOU LIKE EXISTEN- tion), and WAR BEGIN COMPLETIVE-PARTICLE to TIAL-PARTICLE? ‘Do you like India?’ Sentences are express ‘The war has begun.’ (verbal function). often highly elliptic, leaving out information that can 525 INDO-PAKISTANI SIGN LANGUAGE Figure 2. ‘I give someone a gift’ and ‘Someone gives me a gift’. Figure 3. ‘You talk to me in sign language and I talk to you’. be recovered from the context, so that the predicate is Structural Features Shared with Other Sign the only obligatory constituent. Signs indicating time Languages (such as ‘tomorrow’, ‘before’, ‘in the future’, etc.) always appear at the beginning of a sentence. Like IPSL shares a number of aspects of its structure with most other sign languages, IPSL has no tense system. other sign languages. By modifying the movement Rather, time is expressed at the level of the text, with pattern of a sign, it is possible to express several aspec- time signs placed at the beginning of a new paragraph tual meanings, such as ‘do something repeatedly’, ‘be each time the time reference changes. about to do something’, or ‘something happening 526  INDO-PAKISTANI SIGN LANGUAGE Figure 4. ‘Develop’: single opening movement; and ‘develop gradually’: stepwise gradual opening. gradually’. Figure 4 shows an example, and it is obvi- ous how the form of the sign corresponds to its mean- ing. Such iconicity, both in the vocabulary and in grammatical processes, is also a characteristic feature of sign languages in general and is used extensively in IPSL. Iconicity also plays a role in some of the most pro- ductive IPSL constructions. These include a large set of signs denoting geometrical shapes, such as ‘square two-dimensional’, ‘round three-dimensional’, ‘flat surface’, and so on. They can be used in countless modifications and combinations to describe a wide variety of objects in terms of their visual-geometric properties. IPSL also has a subsystem known as ‘clas- sifiers’ in other sign languages. However, the IPSL classifiers are fewer in number, used less frequently, and refer to people and animals only, not to other cat- egories such as vehicles. In such a construction, one sign can express the equivalent of a whole sentence in Figure 5. ‘Two persons approaching each other’. English (see Figure 5). IPSL and other sign languages make use of a wide range of facial expressions for both adverbial concepts are extremely important in all sign languages docu- (such as ‘do with difficulty’, ’more and more’, ‘unfor- mented so far. tunately’, etc.) and grammatical categories. In IPSL, Finally, the important grammatical role of space, the latter include facial expressions that go with for example, with directional movement, is a unique yes/no-questions, questions with question words (e.g. feature of sign languages that has no correlate in spo- ‘what’, ‘how’), negative sentences, and subordinate ken languages. In a text, sign language users often clauses. For example, in questions with question assign particular points in space to the objects or peo- words, the eyebrows are raised and the head is tilted ple they are talking about and make continuous refer- backward. Such facial expressions, which are often ence to these locations, for example, by pointing at obligatory and have to follow specific syntactic rules, them. This process is known as ‘localization’ and 527 INDO-PAKISTANI SIGN LANGUAGE plays an important role in IPSL grammar as well as in area, one and the same sign may be accompanied by other sign languages. mouthing from different spoken languages in different areas. While in some sign languages, mouthing is functionally very important, it is much less significant Structural Features Different from Other in IPSL. The extent of mouthing is extremely variable; Sign Languages hence, signing with extensive mouthing and signing On the other hand, many structural features are pecu- with no mouthing at all is equally possible. liar to IPSL and differ from other sign languages. For Even the signing space in IPSL is structured differ- example, many known sign languages do not have an ently from other sign languages. The upper sign space auxiliary construction of the type described above for around the head and above the shoulders is particular- IPSL. In the negative, IPSL has two different func- ly significant and is associated with the concepts of tional particles for neutral negation (saying that some- distance and authority. Therefore, when place names thing is not the case) and contrastive negation or names of institutions with authority are mentioned, (rejecting something that has been said or implied an index finger pointing is added that has to be direct- before). Unlike in other sign languages, there is no ed upward, for example, GOVERNMENT INDEX-up, particular construction for conditionals (‘if-clauses’), CALCUTTA INDEX-up. and there are no conjunctions to link clauses. Instead, there is a general subordinating construction that is Social and Political Factors marked nonmanually by a combination of facial expression, head posture, and sign rhythm. The deaf community in India and Pakistan is a lin- IPSL has few genuine compounds, but there are a guistic and cultural community rather than an ethnic or number of striking types of sign combinations. Almost political entity. Focal points are the deaf schools and all signs for kinship terms are compositional, with the the deaf associations, of which there are many in all sign for ‘man/male’ or ‘woman/female’ preceding a urban centers. On the other hand, the rural areas are sign for the family relationship, for example, usually cut off from the deaf community due to a lack MAN+SIBLING ‘brother’, WOMAN+MARRY ‘wife’ of infrastructure, and isolated deaf people in these (see Figure 6). IPSL has a single question word with a areas use limited ad hoc gestural communication general interrogative meaning (see Figure 7). To ask (‘home sign’). The origins of the deaf educational sys- specific questions, the general interrogative must be tem date back to the late nineteenth century, when the combined with other signs, for example, first schools for the deaf were founded in India (c. FACE+QUESTION-WORD ‘who’, PLACE+QUES- 1882 in Mumbai, c. 1893 in Calcutta). However, this TION-WORD ‘where’, TIME+QUESTION-WORD does not necessarily coincide with the age of the sign ‘when’. language itself. So far, there is no reliable information The use of mouthing in IPSL is also markedly dif- about when and how IPSL originated. ferent from other sign languages. Since there are so IPSL is not an officially recognized language in any many different spoken languages in the IPSL-using part of the Indian Subcontinent, and there is a great deal of prejudice and misinformation about it. This Figure 6. ‘Daughter’: WOMAN+BORN. Figure 7. General question word. 528  INFLECTION AND DERIVATION may account for various recent attempts at ‘develop- recognition of the linguistic human rights of the deaf ing’ the sign language, for example, by inventing con- community in the Indian Subcontinent still continues. trived sign systems (‘Indian Sign System’, ‘Sign Urdu’, etc.). In these systems, which are not used by References deaf people among themselves, signed sentences are Ayjnihh and Ulrike Zeshan. 2003. Advanced course in Indian constructed to mirror the structure of a spoken lan- sign language. Mumbai: Ali Yavar Jung National Institute guage, disregarding IPSL structures and adding a for the Hearing Handicapped. number of newly invented signs. There have also been Jepson, Jill. 1991. Two sign languages in a single village in equally unsuccessful attempts at ‘standardizing’ IPSL India. Sign Language Studies 20(70). 47–59. dialects. Zeshan, Ulrike. 2000. Sign language in Indo-Pakistan: a description of a signed language. Amsterdam: Benjamins. Deaf people themselves have often adopted the ———. 2000. Gebärdensprachen des indischen Subkontinents negative views on sign language held by the hearing (Sign languages of the Indian subcontinent). Munich: LIN- majority. Use of the primary sign language, in particu- COM. lar, in the deaf educational system, is still stigmatized, ———. 2001. Mouthing in Indopakistani sign language: regu- although the situation is improving. Officially, most larities and variation. The hand is the head of the mouth: the mouth as articulator in sign languages, ed. by Penny Boyes schools for the deaf follow the policy of ‘oralism’, Braem and Rachel Sutton-Spence. Hamburg: Signum. with emphasis on speech training/articulation and lip ———. 2003. Indo-Pakistani sign language grammar: a typo- reading, but day-to-day practice is often more open logical outline. Sign Language Studies 3(2). 157–212. toward forms of manual communication. Over the last An annotated historical bibliography of deafness and sign lan- few years, the Indian government has supported the guage in the region, compiled by Mike Miles, is available at <http://www.sign-lang.uni-hamburg.de/bibweb>. development of sign language courses and the training ULRIKE ZESHAN of qualified sign language teachers, so that attitudes toward sign language and related policies are begin- See also American Sign language; British Sign lan- ning to change. However, the struggle for a genuine guage; Japanese Sign language; Signed Languages Inflection and Derivation Inflection and derivation are terms used in morpholo- katabta you have written gy, the study of word structure. They distinguish the (masculine singular) katabat she has written two main ways of forming words in fusional lan- kataba he has written guages such as Finnish, Italian, Arabic, and—to a less- katabna we have written er extent—English. Isolated languages such as katabtunna you (feminine plural) have written Mandarin, Thai, and Vietnamese tend not to use affix- katabtum you (masculine plural) have written es; i.e. these terms do not apply very well to these lan- katabnaa they (feminine plural) have written katabuu they (masculine plural) have written guages. An affix is a morpheme attached to a word to katabtumaa you two have written change its grammatical function or meaning: at the katabataa they two (feminine) have written start (prefix), around it (circumfix), in the middle katabaa they two (masculine) have written (infix), or at the end (suffix). Inflectional affixes signal grammatical relation- The appearance of inflectional suffixes in a lan- ships such as number, person, tense, and case, and they guage interacts with sentence formation. The more do not change the grammatical class of the word to highly inflected the language, the less important the which they are attached. In highly inflected languages order of words to convey meaning. In Latin, the word such as Arabic, much information is conveyed very order in the sentence parvum puerum magna puella economically, as shown by the paradigm of the perfect vidi could be changed without loss of grammatical form of yaktub: meaning because the inflectional suffixes indicate ‘the tall girl saw the small boy’and not vice versa: -a is katabtu I have written subject case, feminine, third-person singular, whereas katabti you (feminine singular) have written -um is object case, masculine, third-person singular. 529 INFLECTION AND DERIVATION Although Old English inflected all nouns, pro- fewer words. In most languages, derivational word nouns, and adjectives, there are only eight inflectional formation is less productive than inflectional. Since suffixes in Modern English, only one of which can be the Middle Ages, most changes in English word struc- attached to a word at a time: ture have occurred by attaching derivational suffixes to Latinate roots. However, a recent tendency toward -s Third-person singular present (wants) -s Plural noun (dogs) Anglicization may be noted: thus, contemporary addi- -ed Past tense (or past participle) (wanted) tions to the lexicon include compound verbs such as -’s or -s’ (dog’s, dogs’) Genitive case of noun downgrade, downsize, outpace, outsource, etc., rather -en Past participle (given) than their Latin-based equivalents. -ing Progressive aspect (giving) There is an interaction between affixation and -er Comparative adjective (bigger) -est Superlative adjective (biggest) sound structure. For example, in English, the pronun- ciation of -s depends on the previous consonant sound: These inflections are highly ‘productive’ —that is, -s is pronounced [s] in cats, because the preceding t is new nouns, verbs, or adjectives that enter the lan- voiceless, but it is pronounced [z] in kids, because the guage will be automatically assigned these markings. preceding d is a voiced sound. Sometimes, the word ‘Nonproductive’ patterns lack this potential; for stress shifts, for example, photograph/photography. In example, alternate plural markings such as -en (oxen) other cases, the vowel sound may alter, as in or -ren (children) would not be assigned to new divine/divinity. In some cases, a feature of the final nouns. consonant changes, for example, advice/advise; occa- However, the limited number of inflectional affixes sionally, there are changes in both vowel and conso- in English means that they are overworked and poten- nant quality, as in breath/breathe. tially ambiguous. For example, the suffix -s on trains The previous discussion has made two assump- can mark the third person of the verb train or the plu- tions. The first is that the two categories of affixation ral of the noun train. Similarly, train’s can indicate are easily distinguishable—but this is not always so. possession, as in the train’s noise, or it can be a con- As noted previously, some inflectional affixes may tracted form of a noun and auxiliary, as in the train’s have the same form as some derivational ones (e.g. -s: here. Furthermore, -ed, -ing, and -en may be deriva- noun plural or verb third-person singular). tional adjectival markers (interested, interesting, bro- Second, it is also assumed that the notion of a word ken), and -er may mark an agent of a verb (worker, is itself unambiguous—but this is not the case either. teacher). The English indefinite article (a) and the Kivunjo Derivational affixes are used to create new words expression näikìmlyìïa (‘he is eating it for her’) would and, unlike inflectional suffixes, they change the each be identified by respective native speakers as a grammatical class of the root word to which they are single word. Fully understanding what a word is, and attached. In most languages, derivational affixes are therefore how it is formed through affixation, requires usually found nearest to the root, whereas inflectional knowledge of sound structure, sentence grammar, and affixes tend to be on the periphery of a word, usually the rules of word formation. at the beginning or at the end. In English, most derivational affixes are suffixes, and—unlike inflectional suffixes—more than one can References be added to a root. Thus, the noun nation combines Bauer, Laurie. 1983. English word formation. Cambridge: with -al and -ly to produce an adjective and then an Cambridge University Press. adverb, respectively. Derivational suffixes can also Bauer, Laurie. 1992. Introducing linguistic morphology. change subclasses of words: thus, the abstract noun Edinburgh: Edinburgh University Press. Jensen, John, T.1990. Morphology: word structure in generative fatherhood derives from the concrete noun father. grammar. Amsterdam and Philadelphia: John Benjamins The rules of derivational word formation are com- Publishing. plex, but every adult native speaker of a language Katamba, Francis. 1993. Morphology. London: Macmillan. knows and uses them to understand words not previ- Matthews, Peter , H. 1991. Morphology, 2nd edition. Cambridge: ously encountered. Moreover, many new words can be Cambridge University Press. Spencer, Andrew. 1991. Morphological theory: an introduction coined by derivational affixes. For example, a British to word structure in generative grammar. Oxford: Blackwell. Prime Minister gave rise to thatcherism and Spencer, Andrew, and Arnold M. Zwicky (eds.) 1998. The thatcherite, and thatcherize and thatcherizable would handbook of morphology. Oxford and Malden, MA: be fairly readily understood: these two suffixes (-ize Blackwell. and -able) are particularly productive. Others (for ROGER BARNARD example, the prefix im-, as in impossible) are referred to as ‘semiproductive’ because they can be attached to See also: Affixation; Word 530  INFORMATION RETRIEVAL Information Retrieval Information retrieval (IR) is a branch of computer sci- documents in which the term GUITAR occurs and, at the ence (Baeza-Yates and Ribeiro-Neto 1999; Salton and same time, either the term JAZZ or ROCK MUSIC, or McGill 1983; van Rijsbergen 1979; Jackson and both, occur, too’. The IR program will then return all Moulinier 2002) that deals with content-based search- hits, i.e. documents that exhibit the required co-occur- es for relevant information on very large document rences of search terms. collections, usually composed of hundreds of thou- This is the classical Boolean approach to IR, in sands of documents. These documents—the IR term which search terms are combined by Boolean opera- for all kinds of written texts—are contributed by a tors, viz. ‘and’ or ‘or’, sometimes also ‘not’. Such a large variety of authors and cover a wide range of top- Boolean query must exactly match with the terms in a ics, genres, and writing styles, although scientific document to be retrieved. Despite its merits as the papers and newspaper articles prevail. The general baseline for almost all commercial IR systems, its goal of an IR system, in a narrow sense, is to locate all drawbacks are evident: Boolean logic is a formal lan- the documents in the underlying document collection guage most people are not familiar with and, even that are relevant with respect to a user query express- worse, Boolean queries for nontrivial search problems ing a specific search topic. Given these documents (or can become quite complex; the result set is often very bibliographic references to them), the user of an IR large and usually unordered, all search terms are system then has to access these documents, read them, equally relevant for the query evaluation, and, finally, and understand their contents in order to solve the the search partitions the entire collection into exactly problem that led him or her to formulate the specific two subsets: documents that satisfy the Boolean query search topic. More and more, this document-centered and documents that do not. view of IR systems is going to be broadened to include There are some disturbing linguistic phenomena the direct location of factual knowledge in documents that arise through such a surface-oriented approach, and procedures for content condensation (summariza- particularly concerning lexical variants due to inflec- tion) of documents as well. Still, searches are run on tion. A very simple approach for dealing with them is unformatted textual material. This marks the border- the use of the truncation operator ‘$’. Used as a suffix line to fact retrieval with database systems that per- such as in ‘SING$’, this expression retrieves all docu- form searches on prestructured, formatted data ments in which the string ‘SING’ appears followed by collections (e.g. relational tables). In particular, the an arbitrary string. Hence, it matches SING, SINGS, rise of the World Wide Web and the proliferation of SINGER, SINGING but also SINGULAR, etc. machine-readable texts have spurred a strong interest Linguistically more sophisticated alternatives are in IR techniques. Because documents in the WWW are stemmers, i.e. programs that strip off suffixes based on often multimedia aggregates, the goal of IR also grad- a small set of rules without the use of dictionaries ually expands to include the content-based search of (thus preventing e.g. the reduction of SINGULAR to audio, video, and graphical data as well, although the SING). In this case, the IR system takes over the ini- primary focus is still on textual data (and the textual tiative and the user no longer has to make ad hoc description of other media). guesses about morphological regularities (as required A typical IR scenario can be described as follows: by the proper placement of truncation operators). A person wants to find relevant documents on guitar However, linguistically more advanced morphological players, particularly those playing jazz or rock music. analysis used to cope with the term variation problem Rather than submitting this search topic to an IR sys- is usually not supported by IR systems. tem in colloquial English, it must be translated into a In the basic Boolean approach to IR, all search terms ‘formalized’ query format. First, the facets of the topic are equally important. This assumption is often inade- under consideration are designated by search terms: quate and, hence, retrieval models have been proposed, GUITAR, JAZZ, ROCK MUSIC. Second, these terms have which incorporate statistical methods for grading the to be grouped according to their logical dependencies. relevance of terms for a document search. There is a The final search query then looks like this: (GUITAR very general assumption underlying the use of statisti- and (JAZZ or ROCK MUSIC)). This query triggers a cal methodology for IR, which relates the frequency of search process in the underlying document collection, occurrence of terms to their relevance as content which can be paraphrased as follows: ‘Find me all descriptors. This can be expressed more precisely as 531 INFORMATION RETRIEVAL follows: determine the absolute frequency of each term ments (where k usually ranges between 20 and 50), j, tfij (excluding irrelevant, although highly frequent which he or she considers relevant (all remaining ones stop words such as ‘the’, ‘for’, ‘again’, etc.) for each are considered irrelevant, by convention). The terms of document i ∈ [1, n] in the entire collection (where n the relevant documents are ‘added’ to the vector repre- gives the total amount of documents in this collection). sentation of the original query, while those from the Then, determine the corresponding relative term fre- irrelevant documents are ‘subtracted’ from it. The quencies, documentwise and collectionwise. A term is entire search is then rerun with the modified query. considered relevant if its relative document frequency With two or three iterations, this usually leads to by far exceeds its relative collection frequency. In order improved search results within the feedback loop. to stress the specificity of a term, we may also add extra Statistical models of term association that use vec- weight to those terms that have a high occurrence fre- tor space representations and various association fac- quency in few documents and penalize those terms that tors (such as the cosine measure), are a quantitative are widespread and occur rather frequently throughout way of making sense relations explicit, although they the entire collection (and, hence, do not discriminate do not distinguish between hyponyms, antonyms, or well for searches). In essence, this is achieved by the synonyms, etc. A more traditional way of accounting so-called tf*idf measure (term frequency—inverse doc- for lexical semantics is the use of thesauri and classi- ument frequency). Its tf part equals tfij from above, and fication codes in document retrieval. A thesaurus is a its idf part is constituted by log(n–n ), where n yields the lexical repository in which terms are linked to other i size of the document collection and ni specifies the terms via a small set of informal semantic relations, number of documents in which term i occurs. such as synonymous or quasisynonymous terms, There are more advanced methods, e.g. the vector broader or narrower terms, related terms, etc. This space model, in which a document is represented by an might be useful for document retrieval in search n-dimensional document-term vector, n being the modes based on query expansion: using a broader (i.e. number of different terms (excluding stop words) in general) term such as INSTRUMENT explicitly linked to the entire document collection. Interestingly, this a thesaurus will match documents that not only con- approach assumes the dimensions of the vector space tain INSTRUMENT as a term but also, say, GUITAR, to be orthogonal, i.e. all terms are treated as being DRUMS, PIANO, or TRUMPET, i.e. narrower (more spe- totally independent of each other (which sounds cific) terms. Hence, it frees users from thinking of bizarre from a linguistic perspective). The ‘similarity’ conceptual variants of search terms (synonyms, more (another more technical notion to express relevance) specific or related terms) when they formulate a query. of a document and a query can then be measured by Classification codes are another vehicle for express- comparing the corresponding document and query ing semantic relations although, usually, only along the vectors using well-known metrics such as the cosine general-specific axis. In classifications, concepts are measure (the higher the cosine value, i.e. the smaller represented by some kind of alphanumerical string the cosine of the angle of two vectors, the more simi- (e.g. INSTRUMENT comes as MI0) and more specific lar both the items are). Accordingly, we may also concepts are represented using this string as a prefix so determine the similarity between any two documents that, e.g. GUITAR may appear as MI01 or MI0G. Text in the document collection based on a vector compar- categorization (classification) is then defined as the ison between both of them. This can be exploited for problem of assigning a suitable set of category codes to the retrieval process in the following way: when a each document. Similarly, indexing denotes a process query identifies a relevant document, then another by which a document is assigned a set of index terms very similar document (which exhibits a high cosine describing its main topics, with the index terms either value relative to the already retrieved document) being directly extracted from the source document or should also be part of the system output. This derived from some controlled vocabulary, e.g. a the- approach to classifying documents usually requires saurus. As far as the retrieval process is concerned, clustering procedures, which build on the lexical vec- with lists of category codes or index terms being avail- tor representation of documents. In general, IR sys- able, these document surrogates are searched rather tems that apply the vector space model generate as than the full text of documents. output a list of documents, ranked by decreasing rele- The above scenario assumes a person actively sub- vance in terms of how similar each document is rela- mitting an ad hoc query to an IR system and awaiting tive to the query, given their vector representations. interactive system response. Nevertheless, the same A powerful extension to this single-step ranking methodologies can also be used for different informa- approach is called relevance feedback. Based upon the tion supply models. One of these focuses on document results of the first retrieval round, the searcher identi- routing (or, document filtering). In this case, a query fies those documents among the top-ranked k docu- that is stable over a certain time interval (some 532  INFORMATION RETRIEVAL months, or so) is continuously evaluated in the system. the extraction of factoid knowledge from textual When new, incoming documents match this query, sources, for text mining there are no a priori templates they are delivered automatically to the person who set available, which specify the type of relevant informa- up the content router (filter). This mode helps people tion a system should look for. Rather, text miners try keep up to date with the most recent documents to capture ‘new’ and interesting (relevant) information matching their (stable) search interests. for which no a priori conceptualization exists yet. By From a linguistic point of view, an interesting and way of exploring textual data and interpreting it, a text very general conclusion can be drawn from standard IR miner supplies its user with unexpected and, thus, par- approaches. A content-based search problem for find- ticularly valuable information. As with classical docu- ing relevant documents is treated as a lexical problem ment-focused IR, the notion of relevance and the (sometimes, when noun phrases are incorporated, it is interest value is crucial for even more advanced fact- slightly extended as a phrasal problem). Relevancy, finding IR activities. usually, is mirrored through co-occurrence patterns of Another challenging task is concerned with factual nouns (as denotations of topics, issues, etc.)—either in question answering directly from large collections of sets of documents (Boolean retrieval) or in a vector on-line textual resources (Harabogiu and Moldovan space. Since all other kinds of text-specific information 2003). Unlike information extraction whose scope is is lost (even the natural order of words in a text), this limited to the set of a priori available templates, text- approach has been dubbed as a ‘bag of words’. based question answering is open-domain, i.e. not Although numerous attempts have been made to pro- constrained to preselected topics. Successful question mote further linguistic methodologies beyond stem- answering systems use pretty advanced natural lan- ming (e.g. syntactic processing of phrases), their guage processing techniques, e.g. to capture the application yielded discouraging results in many exper- semantics of (wh) questions and to perform lexico- iments. By and large, linguistic methods did not out- semantic unifications of questions and candidate perform the much simpler statistical ones and failed to answers (usually, text snippets) through the collection improve the performance of IR systems substantially. and assessment of evidence from multiple documents. While the methods discussed so far cover a tradi- Recently, the field has further expanded into the tional, fairly established view on the IR task, viz. the area of text summarization, although this term is a bit provision of relevant documents, current research tries of an overstatement considering the methods being to complement this document-centered view on IR by used (Mani 2001). Basically, a summary is an abrevi- focusing on relevant facts and data. A corresponding ated version of a source text, which contains the most information extraction task (Gaizauskas and Wilks relevant information from its source, with compres- 1998; and Jackson and Moulinier 2002) can then be sion ratios ranging from 5 to 25%. Single-document phrased as follows: given a set of documents, extract summarization can even be considered a special case all of the information considered relevant from these of multi-document summarization where the contents textual sources by filling a priori supplied sets of of a whole set of thematically homogeneous docu- domain-specific templates (reminiscent of the knowl- ments has to be transformed into one comprehensive edge representation construct of frames). Typically, summary (redundancy and overlap determination, res- these templates address names of people, locations or olution of inconsistencies, and temporal sequence organizations, as well as simple relations and complex errors then become crucial issues). events involving these entities. In order to achieve this Since in-depth abstraction from the source has goal, a much more sophisticated analytic machinery proved to be infeasible so far, the problem has been than the one used for classical IR tasks has to be pro- rephrased as a sentence extraction problem. An extract vided. Usually, it is based on regular expressions for is then formed from sentences, which are likely to con- named entity recognition, (cascades of) finite-state tain the gist of or crucial statements of the source doc- automata for partial syntactic analysis (parsing), tem- ument. Few types of surface indicators seem to be plate-filling rules which move linguistically analyzed sufficient. Given the prevailing text genres in IR text portions (e.g. noun or verb groups) into the canon- (newspaper articles and technical reports), it is no sur- ical template structures, while template merging rules prise that cue phrases indicating the salience of a sen- account for the final assembly of consistent and com- tence (such as ‘in summary’, ‘our experiments have plete data about the same entity or event as captured shown’, etc.) or the position in which a sentence from the entire document. Also in hybrid approaches, occurs (e.g. lead sentences in a journal article, the first statistical methods play their part, although not as and last paragraph of a document) are really good indi- dominantly as in the pure IR framework. cators for picking up relevant sentences for an extract, Text mining (Hearst 1999) is slightly different from while statistically relevant terms (e.g. based on some the information extraction task. Although the target is of the statistical measures mentioned for the classical 533 INFORMATION RETRIEVAL IR approach) are less reliable but still useful. Most for which only modest progress has been made so far. extraction systems vary in the degree they tune the par- With the WWW becoming the prime information ticulars of these parameters, manually or even automat- exchange platform for information seekers, additional ically (when considered as a machine learning task). linguistic challenges come up associated with the mul- The field of IR has a strong experimental backing. tilinguality of the Web. In this area, a distinction is Given soft notions such as relevance, salience, impor- made between cross-language retrieval (a native-lan- tance, which are crucial for almost all IR tasks, the guage query is processed on many other languages, accomplishment of how relevant the delivered infor- too, although the results are original documents in var- mation items really are needs quantitative assessment. ious foreign languages) and multilingual retrieval (in Among the various quality metrics discussed in the this case, hits are, usually roughly, translated to field, two particularly relevant ones have found con- empower the searcher to assess the relevance of a doc- siderable consensus, viz. precision and recall (cf. ument and, possibly, arrange for a detailed high-quali- Table 1). ty translation, if the document is really relevant). The f-measure (cf. Table 2) is a convenient way to combine the effects of recall and precision, allowing References the designer to stress the impact of either one of these (via the weights α and β ) depending on the experi- Baeza-Yates, Ricardo, and Berthier Ribeiro-Neto (eds.) 1999. mental setting. Modern information retrieval. Reading, MA: Addison- Wesley & Longman. While all the above-mentioned tasks deal with tex- Gaizauskas, Robert, and Yorick Wilks. 1998. Information tual data only, a real challenge to IR will be the increas- extraction: beyond document retrieval. Journal of ing amount of multimedia data. This includes audio Documentation 54(1). 70–105. (speech, sounds, music), graphics (tables, formulae, Harabagiu, Sanda, and Dan Moldovan. 2003. Question answer- graphs, images, animation), and video data (movies), ing. The Oxford handbook of computational linguistics, ed. by Ruslan Mitkov, 560–82. Oxford, UK: Oxford University Press. Hearst, Marti A. 1999. Untangling text data mining. TABLE 1 Precision and Recall Metrics Proceedings of the 37th annual meeting of the Association for Computational Linguistics, 3–10. College Park, MD, #relevant items retrieved Precision :  USA, 20–26 June 1999. San Francisco, CA: Morgan #items retrieved Kaufmann. #relevant items retrieved Jackson, Peter, and Isabelle Moulinier. 2002. Natural language Recall :  processing for online applications. Text retrieval, extraction #relevant items in the collection and categorization. Natural Language Processing, Vol. 5. Amsterdam, Philadelphia, PA: John Benjamins. Mani, Inderjeet. 2001. Automatic summarization, Natural lan- guage processing, Vol. 3. Amsterdam, Philadelphia, PA: John Benjamins. TABLE 2 F-Measure van Rijsbergen, C.J. 1979. Information retrieval, 2nd edition. ( Recall) ( Precision) London, UK; Boston, MA: Butterworths. F-measure :  Salton, Gerard, and Michael J. McGill. 1983. Introduction to ( Recall)( Precision) modern information retrieval. New York, NY: McGraw Hill. UDO HAHN Interpreting In interpreting, a spoken message or text in the origi- Three key variables in interpreting are time lag, nal (‘source’) language is converted into a spoken physical/professional setting, and language direction. message or text in another (‘target’ or ‘receptor’) lan- Looking first at time lag, the crucial difference is guage. Interpreting and written translation have certain between consecutive and simultaneous interpreting. similarities, but the on-line, more directly interactive In consecutive mode, the interpreter hears the source nature of most spoken text gives spoken interpreting text or a section of it (e.g. one speaker’s turn, as with specific features. Message 1 in Figure 1); the source speaker then stops 534  INTERPRETING Message 1 Language Y Language X Message 2 User of User of Language X Interpreter Language Y Figure 1 speaking, and the interpreter delivers a translated version ● Monitoring (closely scrutinizing) the emerging to the target listener. While listening, interpreters can use target text and, if necessary, repairing it (correct- strategies such as note-taking and mnemonics to remem- ing, e.g. inserting ‘I’m sorry, that should have ber source text data, and they have time to work out the been X’). text structure (e.g. whether the first sentence contains a ● Strategies can be used to make the task a little preamble or the main point). Good interpreters will pass easier. For example, on all information contained in the source message, but ● Several segments may need to come in before they are unlikely to produce a ‘literal’ translation. Just as one has enough information to start translating ‘normal’ listeners, after a certain delay, remember the (e.g. because the source and target languages meaning of a message rather than exact wording or have very different sentence structures, such as grammar, consecutive interpreters use memory of mean- English and Japanese): then, when the inter- ing to construct the target message; thus, ‘places like preter can finally deliver a target version, he or Thailand, Malaysia, Indonesia, and Vietnam’ might be she does so as quickly as possible to clear work- translated as ‘Southeast Asian countries’ (see Figure 1). ing memory. In simultaneous interpreting, interpreters produce a ● Exploit the source speaker’s pauses to deliver as target text while the source text is coming in: once much target text as possible without having to they have received a translatable source-text segment, process new input. they pass it on in the target language while listening to Nevertheless, simultaneous interpreting remains a the next source-text segment. The translatable segment specialized and taxing skill. Becoming a simultaneous is typically a meaning-based unit a couple of phrases interpreter requires extensive training and practice, long: e.g. ‘// the government maintains that // BSE, or and even experienced interpreters cannot interpret for ‘mad cow disease’, // is of no danger to the public. //’ so long at a time (typically, 20 minutes between Simultaneous interpreting thus requires two closely breaks: hence, conference interpreters often operate in related subskills: divided attention, i.e. the ability to two-person teams). Preparation, however, is crucial in pay attention to both the incoming and the outgoing easing the difficulty of both consecutive and simulta- message, and multitasking, i.e. the ability to perform neous interpreting: thus, before an assignment, inter- several tasks simultaneously. The key tasks to be done preters usually ask for documentation, drafts of simultaneously (see Figure 2) are as follows: speeches, etc., to be given to them in advance. ● Analyzing and comprehending the incoming A second key variable in interpreting is the physical source-language segment, fitting its information setting. The simultaneous interpreter is usually in a into an emerging picture of sentence and overall soundproofed booth, where he or she hears the source source-text meaning, discourse structure, speak- speaker through headphones and produces the target er attitude, etc., and holding the meaning in message into a microphone. Conference delegates, short-term memory. courtroom advocates, etc., can listen either to the inter- ● Planning and producing the target-language seg- preter through their own headphones (at conferences, ment according to the previous source-text seg- switch settings may give a choice of target languages) ment and to the emerging sentence and or to the source speaker directly. With just one target- whole-text structure. language listener, simultaneous interpreting may also 535 INTERPRETING Modelling whole input text Listening Segment 2 Speaking Segment 2 Listening to Segment 2 Speaking Segment 1 Monitoring whole output text PAST PRESENT FUTURE Figure 2 Message 1 Language X Language Y Language Z User of User of Language X X–Y Interpreter Y–Z Interpreter Language Z Figure 3 occur through chuchotage, where the interpreter sits assumptions behind one party’s behavior—e.g. telling close to the listener, hears the source message directly a bereaved American that an Indonesian’s smile indi- and whispers the target version into his or her ear. cates sympathy rather than amusement. In ‘liaison interpreting’, a consecutive interpreter As for language direction, interpreters usually prefer enables communication between two parties—as in to translate from their ‘B language’ (a nonnative lan- business or in ‘community’ interpreting (working with guage they know well) into their ‘A language’ (the lan- members of linguistic minorities at hospitals, with guage they know best—usually their native language). immigration authorities, etc.). The interpreter is phys- Research, however, indicates little difference between ically present, often positioned between the two par- A→B and B→A performance with skilled interpreters ties, and thus tends to be regarded as a fuller (in the former, less attention is needed for source-text participant in the interaction. Usually, there is only one comprehension, and in the latter, less attention is need- interpreter (see Figure 1), although sometimes—e.g. ed for target-text production). Moreover, in some lan- in international political talks—there may be two (one guage pairs, nearly all interpreters are native speakers for each language direction). of one language rather than the other (e.g. most In most settings—courtroom interpreting, for Bosnian–English interpreters are native speakers of example—accurate rendering of source text and strict Bosnian rather than English) and thus often find them- neutrality between parties are key professional ethics. selves having to work in both directions. In some settings (e.g. when mediating between mem- Sometimes, interpreters may not be available for a bers of different community groups, or in business certain language pair (e.g. at a conference with negotiations), however, interpreters may see their job English, Spanish, and Japanese as working lan- as facilitating communication in a wider sense. This guages, there may be no Spanish-Japanese inter- may involve, for instance, explaining the cultural preters). Then, ‘relay interpreting’ may take place, for 536  IRONY example, a Spanish speech is interpreted into English, Gile, Daniel. 1995. Basic concepts and models for interpreter and the English version is interpreted into Japanese; and translator training. Amsterdam: Benjamins. Jones, Roderick. 1998. Conference interpreting explained. see Figure 3. Manchester: St. Jerome. Lambert, Sylvie, and Barbara Moser-Mercer (eds.) 1994. Bridging the gap: empirical research in simultaneous inter- References pretation. Amsterdam: Benjamins. Carr, Silvana E. (ed.) 1997. The critical link: interpreters in the Pöchhacker, Franz, and Miriam Shlesinger (eds.) 2002. The community. Amsterdam and Philadelphia: Benjamins. interpreting studies reader. London: Routledge. Gambier, Yves, Daniel Gile, and Christopher Taylor. 1997. Wadensjö, Cecilia. 1998. Interpreting as interaction. London: Conference interpreting: current trends in research. Longman. Amsterdam: Benjamins. FRANCIS R. JONES Irony The word irony comes from the Greek eironeia (‘pre- theory notes that not all irony seems to be interpretable tense, dissimulation’), as does the history of its defini- as the echo of someone else’s words and that an admit- tion and analysis. Irony is seen as a trope (i.e. a figure tedly rarer, nonnegative, praising irony exists. An of speech) in ancient rhetorics, and this analysis has example of the latter is the utterance ‘Sorry to bother remained essentially unchallenged until recently. In you’ before the announcement of excellent news (e.g. the traditional definition, irony is seen as saying the winning the lottery). opposite of what one means. This definition is demon- All the theories of irony mentioned so far (except strably incorrect, because a speaker may be ironic but mention theory, which is presented as a direct access not mean the opposite of what he or she says, as in ‘It theory; see below) share the idea that the processing of seems to be a little windy’ (uttered in the middle of a irony is a two-step process in which one sense (usual- violent storm), in which the speaker is saying less than ly assumed to be the literal meaning) of the utterance what is meant. A recent and fruitful restatement of the is accessed and then a second sense of the utterance is irony-as-trope theory has been presented by H. Paul discovered (usually under contextual pressure). Thus, Grice, who sees irony as a linguistic structure with in a Gricean account of irony as implicature, the hear- only implied meaning, which puts a significant burden er of an utterance such as ‘That was smart’ (uttered as on the listener, who has to decode the implication. a description of clumsy behavior, such as spilling Broadening the definition to, for example, ‘saying one’s wine on someone’s clothing) will first process something while meaning something else’, runs the the utterance as meaning literally roughly ‘This behav- risk of obliterating the specific difference between ior was consonant with how smart people behave.’ irony and other forms of figurative or indirect speech. This interpretation will then be discarded in favor of Despite the problems with the idea of ‘oppositeness’, the implicature that the speaker means that the behav- approaches to irony as negation have been presented. ior was not consonant with how smart people behave. Speech-act approaches to irony see it as an insincere This account has been challenged recently by ‘direct speech act. Other approaches to irony include the access’ theories. ‘tinge’ theory, which sees irony as blending the two The direct access theories claim that the hearer does meanings (the stated and the implied ones) with the not process the literal meaning of an ironic utterance effect of attenuating the ironic one. first and access only later the figurative (ironic) mean- A very influential approach to irony is the mention ing. They claim instead that the literal meaning is theory (Sperber and Wilson 1981), which claims that accessed either not at all or only later. Direct access an utterance is ironic if it is recognized as the echoic interpretations of irony are directly at odds with the tra- mention of another utterance by a more or less clearly ditional interpretation of irony as an implicature. Some identified other speaker. Furthermore, the ironic state- results in psycholinguistics have been seen as support- ment is critical of the echoed utterance. Similar theo- ing this view (Gibbs 1994). The mention theory of ries based on the ideas of ‘pretense’ and ‘reminder’ irony is a direct access theory. Other researchers have have been presented as well. Criticism of the mention presented contrasting views that support the two-step 537 IRONY approach, although not always the claim that the literal have argued that the two can be distinguished (for meaning is processed first: claims that interpretations example, irony can be involuntary, whereas sarcasm are accessed in order of saliency or in parallel have cannot), others maintain that no clear boundary exists. been put forth. Another unresolved issue is the connection between Psycholinguistic studies of irony have focused on irony and humor. It is clear that the two overlap sig- children’s acquisition of irony, progressively lowering nificantly, but also that they differ: not all irony is the age at which they understand it to under 10 years. humorous, and not all humor is ironic. These studies focus on the neurobiology of the pro- cessing of irony, emphasizing the role of the right References hemisphere alongside the left one, in which most lan- guage processing takes place, and the order of activa- Attardo, Salvatore. 2000. Irony as relevant inappropriateness. Journal of Pragmatics 32(6). 793–826. tion of the various meanings in the ironic text. Booth, Wayne. 1974. A rhetoric of irony. Chicago: University Considerable attention has been paid to the option- of Chicago Press. al markers of irony, i.e. indications of the speaker’s Colston, Herbert L. 1997. Salting a wound or sugaring a pill: ironic intent through, primarily, intonation and body the pragmatic function of ironic criticism. Discourse language. Although several phonological (sound) fea- Processes 23. 25–45. Dews, Shelley, and Ellen Winner. 1999. Obligatory processing tures have been considered ‘markers’ of irony, it seems of literal and non-literal meanings in verbal irony. Journal of that none is exclusively a marker of irony. Pragmatics 31(12). 1579–99. Recently, the social and situational context of irony, Gibbs, Raymond W. 1994. The poetics of mind: figurative as well as its pragmatic ends, have begun being inves- thought, language, and understanding. Cambridge, New tigated in sociolinguistics and discourse/conversation York: Cambridge University Press. Giora, Rachel. 2003. On our mind. Salience, context, and figu- analysis, and also in psycholinguistics. Work on the rative language. Oxford: Oxford University Press. social functions of irony has found a broad range of Haiman, John. 1998. Talk is cheap: sarcasm, alienation, and the functions, including in- and out-group definition, eval- evolution of language. Oxford, New York: Oxford uation, aggression, politeness, verbal play, and many University Press. others. It is possible that the list may be open ended. Katz, Albert N. (ed.) 2000. The uses and processing of irony and sarcasm. Special issue of Metaphor and Symbol Work on the reception of irony by its audience is also 15(1/2). 1–107. beginning to appear. Kreutz, Roger, and Richard M. Roberts. 1995. Two cues for It should be noted that the term irony is also com- verbal irony: hyperbole and the ironic tone of voice. monly used to describe ‘situational’ irony (i.e. irony of Metaphor and symbolic activity 10. 21–31. facts and things entirely dissociated from their linguis- Sperber, Dan, and Deirdre Wilson. 1981. Irony and the use— mention distinction. Radical pragmatics, ed. by Peter Cole. tic expression), such as a fire station burning to the 295–318. New York, London: Academic Press. ground, Socratic irony, romantic irony, and even a type Toplak, Maggie, and Albert N. Katz. 2000. On the uses of sar- of religious experience (Kierkegaard). Although there castic irony. Journal of Pragmatics 32(10). 1467–88. may be connections between situational and verbal Winner, Ellen. 1988. The point of words: children’s under- irony, it does not seem that literary and religious uses standing of metaphor and irony. Cambridge, MA: Harvard University Press. can be fruitfully explained in terms of linguistic irony. SALVATORE ATTARDO Other definitional problems include the purported dis- tinction between irony and sarcasm. Although some See also Grice, H. Paul Israel One does not ordinarily speak of ‘miracles’ in linguis- True, Hebrew was never completely dead; it is more tics, and with good reason. However, if any event in lin- accurate to speak of the ‘revival’ of the Hebrew lan- guistics and language history deserves the name guage rather than its rebirth. Rabbis and learned men ‘miracle’, it is the rebirth of the Hebrew language in never ceased to use Hebrew for high purposes long modern Israel. Languages, once dead, do not normally after it had declined as the spoken language of the rise again from the dead. This the Hebrew language did. Jewish people; it was always the language of prayer 538  ISRAEL for observant Jews; it served to some extent as a lingua Romans, and Arabs, only to name the principal play- franca among Jews from different countries. But for ers who came on and off the Palestinian stage during over a millennium and a half until toward the end of its long history. The languages of all its occupiers con- the nineteenth century, Hebrew was virtually nobody’s tributed to a multilinguistic profusion and vibrancy first language. It was a language acquired in adoles- that is preserved in the linguistic pluralism of present- cence and adulthood by at most a minority of Jewish day Israel. males, by Jewish females almost never because of Hebrew was the common language of the Jewish religious strictures. Today, it is a language like any people—the everyday spoken language of the people other—that is to say, a language spoken by ordinary of the kingdom of Judah—from around 1000 BCE and people going about their everyday activities and thus is at least 3,000 years old. It was the language in acquired in the usual way: by learning the language in which most of the Jewish Bible (called in Christian infancy. usage the Old Testament) was written (the ‘New Hebrew is a member of the Semitic family of lan- Testament’ of Christianity was written in Greek, not guages, to which Arabic and Amharic and extinct lan- Hebrew). However, another Semitic language, guages such as Aramaic and Phoenician also belong. It Aramaic, was used for legal and official purposes by is written from right to left in what is essentially a con- priests and officers of the court. In 800 BCE, most of sonantal script. Semitic languages are organized the Jews of Palestine were forced into exile—the around the principle of the root: roots have a basic Babylonian Exile. Seventy years later, when Jews meaning expressed by means of, usually, three conso- were permitted to return to their ancestral land and nants. The Semitic root SLM elucidates the principle. most did, the linguistic situation in Palestine had Its basic, abstract meaning is ‘peace’, and from it are begun to shift dramatically. Aramaic had descended derived such words as Arabic SaLaaM—the greeting from the court to the marketplace, and it developed (in essence ‘peace be with you’, cf. Hebrew ShaLoM. rapidly among both Jews and non-Jews as the major Consider also iSLaM and (m)uSLiM and the proper vernacular language in the Middle East, a position it names from Jewish history such as (Hebrew) maintained until well into the Common Era when SoLoM(on) and SaLoMe or (Arabic) SuLeiM(an). The Arabic took its place. Jews for the most part became reader of Arabic or Hebrew recognizes that the basic bilingual in Hebrew and Aramaic, and from this time ‘meaning’ of the word or name is carried by the root forward the days of Hebrew as the spoken language of structure SLM. The common semantic intersection of everyday life were numbered. Jesus of Nazareth spoke these and other words formed with the three conso- Palestinian Aramaic as his native tongue. nants SLM is ‘peace’. When the Holy Land came under Greek dominion, Normally, Semitic words are written without vow- the Jewish upper classes increasingly became els. This is possible because the context suggests the ‘Hellenized’, meaning that they adopted Greek cultur- range of readings of a given word in print and because al practices and the Greek language. As Palestine there are rules of grammar that help the reader know passed from Greek into Roman hands, Latin also what vowel is to be supplied. In Hebrew, for example, became a language that upwardly mobile Jews had to the word spr consisting of three letters can be read learn. Thus, by the beginning of the Common Era, /safar/ ‘he counted’, /siper/ ‘he told’, /supar/ ‘it was bilingualism (Aramaic/Greek or Aramaic/Hebrew) told’, /sapar/ ‘barber’, /sefer/ ‘book’, and /sfar/ ‘fron- was common, trilingualism (Aramaic/Greek/Hebrew tier’. (In Hebrew, the same letter is used for /p/ as for or Aramaic/Greek/Latin) not uncommon, and the /f/.) ‘Full spelling’, in which diacritic marks are used Jerusalem Talmud—the Talmud is the body of author- to indicate the vowel, can be used in Hebrew but nor- itative Jewish tradition and commentary comprising mally is not. The reader simply has to know from con- the Mishnah and Gemara—states that ‘four languages text and native linguistic knowledge of the language are of value: Greek for song, Latin for war, Aramaic which reading is the correct one. In Hebrew, vowel for dirges, and Hebrew for speaking’. Monolingual marking is used principally in pedagogical materials, Hebrew speakers carried on only in remote villages in children’s books, Biblical texts, poetry, and dictionar- southern Judea. By the second century CE, Hebrew ies. had become so precarious as a spoken language that Israel, earlier called Palestine, has always been an we find rabbinical injunctions that fathers should area of intense historical and linguistic collisions with speak Hebrew with their sons, a certain sign that most aftershocks extending far beyond the tiny piece of soil fathers were not speaking Hebrew with their sons. that the country actually encompasses. It was the birth- Although Hebrew receded as spoken vernacular, it place of two major world religions, Judaism and retained its significance as the language of the Bible— Christianity, always a country pulled this way and that holy, sacred, divine—and both Aramaic and Hebrew and culturally enriched successively by Jews, Greeks, are joined in the Hebrew designation leshon hakodesh, 539 ISRAEL meaning ‘the language of sanctity’. To this day, countries of central and eastern Europe, where there Aramaic is the language in traditional Judaism of legal was at the time little anti-Jewish sentiment and where instruments such as a divorce decree (get) and the Jews were welcomed for economic reasons. There was marriage contract (ketuba). Bilingualism in Hebrew no sanctuary in the west. The language the fleeing and Aramaic is implicit in the transmission of tradi- Jews carried with them was medieval German, which tional Judaism. on the soil of eastern Europe and in isolation from the What is known as the Jewish Diaspora began with German language in Germany, developed into the lan- the destruction by the Romans of the Second Temple guage for which the preferred designation is Yiddish in 70 CE. As Jews spread out from their ancestral (earlier ‘Judeo-German’). homeland into Europe and the lands bordering the Yiddish is written in the Hebrew alphabet, like Mediterranean, like all emigrants to foreign lands, Ladino and virtually all Jewish languages, but is some they began to speak the language of the country in 80% German in vocabulary, 15% Hebrew, and 5% which they had settled during the early and medieval Slavic and other languages. The marginal vocabulary periods of our millennium. Jews in Spain spoke of Yiddish depends on where the Yiddish language is medieval Spanish—perhaps an identifiably Jewish spoken—American Yiddish has borrowings from variant of Spanish in some cases, but Spanish never- American English, French Yiddish from French. theless. Jews in France spoke Old French or, again, However, the basic language structure of Yiddish is possibly a Jewish dialect of Old French. Jews living Germanic through and through. along the Rhine and Danube rivers in Germany spoke Yiddish is one of the many languages that have Middle High German—the German of 1050 to 1350 always suffered from inferiority complexes. Because, CE—of one kind or another, probably always with a although it shared ancestry with German, it had devel- Jewish flavor and accent. oped in ways different from the standard German lan- When the Jews were expelled from Spain in 1492 guage, Yiddish was condemned by some as ‘bad they settled for the most part around the German’. Others argued, especially with the spread of Mediterranean Sea—in northern Africa, Italy, Greece, Zionism in the late nineteenth and early twentieth Turkey—but the language they carried with them was centuries, that Jews should return to their ancient Spanish. In time, their Spanish changed and developed ancestral language, Hebrew, that only this ancient into the language now usually called Ladino (the ancestral language could unite Jews from all parts of usage is not settled: one also finds ‘Judeo-Spanish’, the world in a common quest for security and a place ‘Judezmo’, ‘Dzhudezhmo’, and ‘Spaniolish’): it is of their own. Still others argued for assimilation: Jews Spanish with admixtures from Hebrew and the lan- living in Poland should speak Polish, Jews living in guages of the lands to which the Spanish emigrants America English, Jews living in Russia Russian, and had fled. Many of these Jews gave up their ancestral so on. Spanish altogether and acquired the language of the Every place that Jews have lived they have devel- country to which the winds had blown them. This oped ‘Jewish’ variants of the local language. Thus, we branch of the Jewish people is called the Sephardim have Yevanic (Judeo-Greek), Dzhidi (Judeo-Persian), from the medieval Hebrew word for ‘Spain’. Judeo-Georgian, Judeo-Tat, Judeo-Tunisian Arabic, A different set of language choices obtained among Judeo-Iraqi Arabic, and many others. In the medieval the Jews, who had settled during the Diaspora from the era, there was Laaz (Judeo-French) and Knaanic Holy Land in western Europe north of the Pyrenees, (Judeo-Slavic). These languages, to the extent that primarily in Germany. These Jews are called the they still are spoken, are generally on the list of endan- Ashkenazim, after the medieval Hebrew word for gered languages. Ladino and Yiddish stand out among ‘Germany’. Until the eleventh century CE, Jews had ‘Jewish languages’ because they are the two best lived in France, Germany, and England in relative known with the most extensive literatures and the peace and security. They were technically, if often largest numbers of speakers—at least until the World only notionally, under the protection of the Holy War II destroyed most of European Jewry. The number Roman Emperor. This changed terribly for the worse of speakers varies widely. In Israel today, there are with the onset of the Crusades in 1096. What began as estimated to be some 200,000 speakers of Yiddish, a war against the ‘Saracens’—‘Arabs’, ‘Turks’, later 100,000 speakers of Ladino, 250,000 speakers of ‘Moors’—to reclaim the holy sites of Christianity rap- Judeo-Moroccan Arabic, 60,000 speakers of Dzhidi, idly became a war against ‘unbelievers’ of every stripe, and 35 speakers of Yevanic. These Jewish ‘legacy’ lan- and hence Jews were expropriated and massacred and guages are in addition to languages still spoken by expelled until finally there was nothing left for them some Israelis such as Bulgarian, French, German, but wholesale emigration away from the troubles. In Hungarian, and English. Very few people would be practical terms, this meant fleeing to Poland and other monolingual in one of the Jewish heritage languages, 540  ISRAEL almost all would be fluent to some degreee in Hebrew. ing’). Aliyah brought a renewed interest in spoken Fluency in Hebrew has had the highest priority since Hebrew, and in the newly regained ancestral land, the creation of modern Israel in 1948. schoolteachers and families here and there began to Hebrew remained dormant through the centuries, its use spoken Hebrew. Eliezer Ben-Yehuda (1858–1922) use confined largely to the written form of the was one of the leaders in the movement, but it is a language, especially among the rabbinate and in the myth that he single-handedly brought about the revival synagogue. However, it never ceased to be a written of spoken Hebrew by refusing to speak anything else language. People whose cradle language had been any- with his family. The movement was slow in the begin- thing in the world but Hebrew composed substantial ning. Competition from traditional Jewish Diaspora works in the language, often elegantly and with origi- languages such as Yiddish and Ladino was stiff, not to nality. Every Jewish boy had to learn Hebrew in the mention French and German in the schools, and it has Hebrew alphabet—well or badly, as is true today—in been estimated that by 1904 there were only about 20 order to become bar mitzvah, signifying that he had Hebrew-speaking families in Palestine. reached his thirteenth birthday and had, therefore, The movement accelerated with the Second Aliyah attained maturity and had reached the age of religious (1904–1906). Waves of po’alim ‘workers’ arrived, and duty and responsibility. Females did not have a coming- they soon had established Hebrew-medium high of-age ceremony in traditional Judaism and thus were schools (gymnasia) and teachers’ seminaries to cater largely illiterate in Hebrew. ‘Literacy’ is, of course, not to the growing commitment in the Jewish population nearly the same as ‘fluency’: to be literate simply meant of Palestine to the resurrection of spoken Hebrew. The that religiously observant Jewish males could read movement fed upon itself. To speak Hebrew was to Hebrew. Jewish languages like Yiddish and Ladino, assert that one belonged to the worldwide community indeed virtually all Jewish variants of languages, were of Jews and was a proud member of that small but and are written in Hebrew characters. This goes back to growing band of Zionist pioneers who had returned to the widespread literacy of Jews in Hebrew—of Jewish the ancestral home. The larger the number of Hebrew- males at least—during the ‘Dark Ages’. speaking students coming out of the school system, The revival of Hebrew as a secular written language the greater the demand for entertainment and reading began during what is known as the Jewish material in Hebrew—plays, novels, newspapers. By Enlightenment (Haskalah) in the eighteenth century. 1916–1918, some 40% (34,000 of 85,000) Jews in Talented writers began to write in the language, play Palestine recorded Hebrew as their first language. with it, revivify it, although they often wrote in The problems of bringing an ancient language up to Yiddish as well. The revival of Hebrew as a spoken the complexity of modern life were enormous. Words language began in small steps in the late nineteenth had to be invented for chicken pox, coffee, opera, elec- century. One of the main reasons, ultimately, was anti- tricity, idealists, billiard cure, mailman, and so on. Semitism. Ill treatment of Jews—pogroms, discrimi- Phrases not found in the Bible had to be devised: all nation—was on the increase in eastern Europe and right, I’ll do it; make it brief; at your service. What was Russia, and even western Europe, where since the correct pronunciation of Hebrew? Sephardic and Napoleonic times the legal situation of Jews had Ashkenazic traditions differed, for example David vs. improved considerably, was not immune. The Dreyfus Dovid, Shalóm vs. Shólem ‘hello’. As a rule, the Case in France, in which a Jewish officer, Alfred Sephardic pronunciation prevailed. By the 1920s, a Dreyfus, was falsely accused of espionage and con- majority of the Jews living in Palestine spoke Hebrew, demned to a harsh imprisonment on Devil’s Island although this represented a tiny minority of the simply because he was Jewish, laid bare how thin the world’s Jewish population, and the only ones who veneer of civility was when it came to being Jewish spoke it fluently were the sabras, the Israelis who had even in supposedly enlightened western Europe. grown up with it. The darkening clouds over Jewish life in Europe led After World War II had savagely reduced the to the Hibbat Zion movement (‘Love of Zion’) and to world’s Jewish population from some 11,000,000 to the Zionist movement itself, founded by the Austrian around 5,000,000, Jews emigrated in large and Jewish journalist Theodor Herzl, which called for the growing numbers to Palestine, and in 1948 the new creation of a Jewish homeland. Jews began to leave country of Israel gained its independence from the eastern Europe and settle in largely agricultural settle- British. The question was: what should the language ments in Palestine that they had purchased or leased of the new country be? Other things being equal, from Arab or Turkish landowners. (Palestine was then Yiddish would have been the leading candidate since part of the Ottoman Turkish Empire.) most of the eastern European Jews who had survived The first wave of immigrants arrived between 1882 the Holocaust and managed to get to Palestine were and 1903 (called the First Aliyah, or first ‘homecom- speakers of Yiddish. Yiddish, however, suffered from 541 ISRAEL various disabilities, one of which was that it was References stigmatized as the ‘language of the ghetto’. It was Ben-Rafael, Eliezer. 1994. Language, identity and social divi- thought of as a victim’s language. No language sion: the case of Israel. Oxford: Clarendon Press. besides Hebrew could possibly do for the new Israel— Bunis, David M. 1975. Problems in Judezmo linguistics. New not English, not Yiddish, not Ladino. Hebrew linked York: The American Sephardi Federation. the Jewish past and the Israeli future as no other Chomsky, Willliam. 1957. Hebrew, the eternal language. Philadelphia: The Jewish Publication Society of America. language could. Hebrew was a sublime symbol of Fellman, Jack. 1973. The revival of a classical tongue: Eliezer hope, of aspiration—not only of Jewishness but of a ben Yehuda and the modern Hebrew language. The Hague: muscular strain of Jewishness that would never allow Mouton. another Holocaust to happen to its people: Never Fishman, Joshua A. (ed.) 1974. The sociology of language in Again! became the rallying cry of modern Jewish Israel. International Journal of the Sociology of Language 1. The Hauge: Mouton. pride and militancy. The Hebrew language is its sym- Fishman, Joshua A. (ed.) 1981. Never say die! A thousand years bol, its icon. of Yiddish in Jewish life and letters. The Hague: Mouton. As unique an event in the annals of linguistics as Fishman, Joshua A. (ed.) 1985. Readings in the sociology of the revival of the Hebrew language was, it is only part Jewish languages. Leiden: E.J. Brill. of a much larger story of linguistic legacy. If the Hetzron, Robert. 1987. Hebrew. The World’s major languages, ed. by Bernard Comrie. New York: Oxford University Press. revival of the Hebrew language in Israel was a miracle, Paper, Herbert H. (ed.) 1978. Jewish languages: themes and it was part of a greater miracle: the preservation of variations. Cambridge, MA: Association for Jewish Studies. Jewishness through centuries of discrimination, mas- Rabin, Haim. 1973. A short history of the Hebrew language. sacre, and oppression. All of the languages of Israel— Jerusalem: Jewish Agency. Hebrew, Aramaic, Yiddish, Ladino, and all the other Spolsky, Bernard, and Elana Shohamy. 1999. The languages of Israel: policy, ideology and practice. Clevedon, Buffalo, ‘Judeo-X’ variants—are part of the Jewish legacy, and Toronto, Sydney: Multilingual Matters Ltd. all have contributed to the preservation of Jewish eth- Weinreich, Max. 1980. History of the Yiddish language. nicity from the days when Moses led his people out of Chicago: University of Chicago Press. Egypt into freedom in The Holy Land. ROBERT D. KING Italian Italian (Italiano) is a Romance language spoken by as the official language for the whole Italian geo- more than 65 million people in Italy (including graphic region. The reason for this choice was that in Sardinia and Sicily), Vatican, Malta and Corsica, the fourteenth century, Dante Alighieri, Francesco Southern Switzerland (where it is one of the four offi- Petrarca, and Giovanni Boccaccio wrote their master- cial languages), San Marino, Istria, on the pieces using this Italian dialect; more precisely, the Northeastern shore of the Adriatic Sea, in the former Old Florentine dialect of Tuscan, and their works Italian colonies of Libya, Somalia, Eritrea and became genuine models of grammar. Their impact was Ethiopia, in Tunisia, Egypt, and Greece. It is also quite so strong that, even if later in the nineteenth century spread in the whole Mediterranean area, and is spoken the dialect of Rome started gaining considerable pres- by the large immigrant communities in the Americas, tige, it could never equal the Florentine’s one. especially in the United States and Argentina, in Apart from the Standard Italian, there are a number Canada, North Africa, and Australia. of distinct dialects spread along and across the Italian- The earliest known written materials in Italian date speaking territories. Some of them are very close to from the tenth century; more precisely, a document in the national language, such as those from Tuscany, the Archives of Montecassino, dated year 960, con- Northern Latium, and Umbria, while other dialects are tains a whole sentence, repeated four times, that is more remote, such as those in the Northern and practically Italian vernacular. The first literary work of Southern Italy, as well as those in the isles. length is the Ritmo Laurenziano (Laurentian Rhythm), This diversity is mainly due to the individuality of a cantilena in praise of a bishop by a Tuscan, of the various parts of the territory, which determined that the late twelfth century. distinctiveness of a certain region be preserved, both in Standard Italian is essentially based on the Tuscan the case of pre-Latin population, coming from different dialect, which was adopted first as the literary and then areas and speaking different languages, and in the case 542  ITALIAN of present-day populations living in different parts of It is possible in Italian to have double consonants. the country. Another factor that contributed to the indi- Usually, there is an opposition between single and vidualism of various geographical dialects was the bar- double consonants, as in: eco ‘echo’ vs. ecco ‘here is’, barian invasion in the High Middle Ages, which brought rupe ‘rock’ vs. ruppe ‘he broke’, where the double new elements into the already shaped language vari- consonants are clearly pronounced. eties, and, of course, the fact that the political unifica- Intonational stress in Italian generally falls on the tion of Italy occurred only late in the nineteenth century. next to last or third from last syllables, but it may range The main groups of Italian dialects are: the Northern freely over the last three syllables of the word. The dialects, the Central-Southern dialects, and the Italian orthography marks in writing the grave ( ` ) and Sardinian group—the last one considered by some lin- the acute ( ´ ) accents, which are usually used to distin- guists as an entirely different language. The so-called guish otherwise similar-looking one-syllable words, ‘Spezia-Rimini line’, a dividing line that runs East–West e.g. se ‘if’ vs. sè ‘oneself’, e ‘and’ vs. è ‘he is’, and to across the territory, separates the first two groups. indicate stress on the final vowel in polysyllabic words, The Northern dialects constitute two major sub- e.g. perchè ‘because’, così ‘so’. The circumflex accent groups: the Venetian (also called Venetic) dialects, ( ^ ) was reserved to poetry and it was used to mark containing the Venetian proper and the Trentino either the ellipsis of a syllable in polysyllabic words, or dialects, and the Gallo-Italian dialects, containing the the contraction of two adjacent vowels in a word. Piedmont, Liguria, Lombardy, Emilia, and Romagna The punctuation of Italian is rather simple. The dialects, all of which display a close affinity to French main differences from English are that the dash ( – ) is in their pronunciation and truncated terminations. commonly used to mark the exchange of dialogue, The Central-Southern dialect groups include the instead of inverted commas, or marks and additional subgroup of the Tuscan dialects, consisting of the comment, instead of brackets. Also, in the case of indi- Florentine, Arezzo, Cortona, Pisa, Livorno, Lucca and rect questions, Italian does not use the question mark. Siena dialects, the subgroup of central dialects, includ- The capitals are less used than in English. Italian does ing the Modern Roman, Lazio, Umbria, Southern not use capitals for the names of the months, for the Marche, Corsican, and Northern Sardinia dialects, the days of the week, for the names of the languages, for subgroup of the Southern dialects, further subdivided adjectives referring to a nation, or individuals belong- into the Neapolitan dialects and the Sicilian ones, and ing to a nation, e.g. il popolo italiano ‘the Italian peo- the subgroup of the Southern and Southeastern ple’, due italiani ‘two Italians’. dialects, containing the Abruzzi, Naples, Campania, As a direct offspring of the Latin language, Italian Lucania, Calabria, Otranto, and Sicily dialects. displays a Latin-based vocabulary. The greater part of The last group lists the most distinct dialects, i.e. the Italian vocabulary (or lexis) is formed by Latin words Ladin, Sardinian, and Istrian. Around 60% of the pop- that are either directly or indirectly inherited from Latin, ulation can speak the national language, but most of or borrowed from a certain Italian dialect. However, in them usually speak a local variety. And, given the the course of centuries, several foreign languages have communication problems between different dialects, influenced the Italian vocabulary. These languages are most of the population is now at least ‘bilingual’, French, Spanish, English, German, and Arabic. The although the proper term would be ‘bidialectal’. French influence is most obvious especially in the terri- The Italian alphabet is: a b c d e f g h i l m n o p q tory of the Gallo-Italian dialects, but it is not restricted to r s t u v x z; j, k, w, and y are only used in foreign it. French have influenced the Italian language over a words, but j can also be found in some proper names very long period, and the French loan words entered in old writings. both the everyday language—some basic terms as man- In general, all letters are pronounced in Italian, with giare ‘to eat’ and giallo ‘yello’ are borrowed from the exception of h. The letter c is pronounced /k/ when French—and the more technical jargons—approccio preceding a, o, and u, but /t/ when it is followed by ‘approach’, destriere ‘charger’. The Spanish influence e and i. The same rule applies to the letter g denoting imposed terms like toreador, bolero and, nowadays, its voiced counterpart. G is pronounced as // before goleador. The English influence was noticeable in the a, o, and u, but as /d/ before e and i. The consonant eighteenth and nineteenth centuries, but it was much cluster gg is also pronounced /d/ before e and i. Z and more prominent in the twentieth century. However, zz are generally pronounced either /t/ or /d/. Sc given the influence of the nationalist movement in the before e and i is pronounced //. politics during the Fascist period, many English loans The sound system of Italian is quite similar to that have been replaced with Italian counterparts, while oth- of Latin or Spanish. Italian has five vowels /3, , , , ers changed their original meaning. Germanic exerted its / and four semivowels: /j, e, o, w/, as well as 21 con- influence as a result of the Teutonic conquests, predom- sonants: /p, b, m f,v, t, d, n, s, z, ts, l, r, , t, inantly in the contact area of Lombardy. The Arabic d, k, , h, , /. influence was most important in Sicily and Southern 543 ITALIAN Italy, and it was a result of the Saracenic invasions. Some cles can be combined with some simple prepositions, of the most common Arabic words are albicocco ‘apri- resulting in contracted forms of the definite article. The cot’, ragazzo ‘boy’, and zecca ‘mint’. combination of the definite article with a simple prepo- In turn, Italian has exerted its influence on the lan- sition is labeled ‘the combined preposition’, e.g. il + a guages of Western Europe, especially on English, : al, gli + di : degli, la + con : colla, le + in : nelle. French, Spanish, German. Some of the English words In Italian, there are two groups of adjectives: those of Italian origin are: balcony, bandit, broccoli, casino, ending in the vowel –o, e.g. severo, and those ending gondola, incognito, inferno, lava, macaroni, malaria, in –e, e.g. grande. The adjective also agrees in gender opera, spaghetti, studio, umbrella, and volcano. Most and number with the noun it modifies, but it does not of the words imposed in different languages are in the have the same ending: e.g. la ragazza forte – le field of literature, art and, of course, music. In this ragazze forti ‘the strong girl/s’. Più ‘more’, meno respect, it must be mentioned that the contemporary ‘less’ and the suffix -issimo are used to form the com- musical terminology all over the world includes Italian parative and the superlative degrees: e.g. più bello words like allegro, andante, aria, concerto, crescendo, ‘more beautiful’, il più bello ‘the most beautiful’, bel- libretto, maestro, oratorio, piano, quartet, solo, lissimo ‘very beautiful’, meno bello ‘less beautiful’, sonata, soprano, tempo, and virtuoso almost. and il meno bello ‘the least beautiful’. The Italian grammar is similar to that of the other The forms of the pronouns differ according to the modern Romance languages. category of case. The Personal Pronoun for the first- The nouns have distinct inflected forms for gender person singular, for example, has different forms and number. Most Italian nouns end in a vowel, while according to the case: io ‘I’ in the Nominative (sub- those ending in a consonant are usually borrowings from ject) changes to me or mi ‘(to) me’ in the Dative (indi- a foreign source, e.g. il bar ‘pub’, lo sport ‘sport’, il film rect object) and Accusative (direct object), according ‘movie’. Generally, nouns ending in -o are masculine, to whether the pronoun is stressed or unstressed. As nouns ending in -a are feminine, while those ending in - for their use, subject pronouns are omitted in everyday e may be either masculine or feminine. The foreign bor- speech, and they appear when needed for emphasis or rowings ending in a consonant are generally considered contrast. The stressed form of direct object pronouns, to be masculine. The nouns change their vowel endings e.g. me ‘(to) me’, te ‘(to) you’, lui ‘(to) him’, lei ‘(to) from singular to plural. The masculine nouns have a plu- her’ is also used for emphasis or contrast; otherwise, ral ending -i, while the feminine nouns have a plural the unstressed forms are used, e.g. mi, ti, lo, la. ending either in –i, when their singular form ends in –e, The possessive adjectives and pronouns have differ- or in –e, when their singular form ends in –a. A number ent forms to distinguish whether the possessor is the of masculine nouns have two distinct plural forms, one first, second, or third person, e.g. mio ‘my/mine’, tuo for the masculine gender and one for feminine gender, ‘your/yours’, suo ‘his-her/his-hers’, whether the pos- each of them with its own, distinct meaning: e.g. il mem- sessor is singular or plural, e.g. mio ‘my/mine’, nostro bro ‘member’ – le membra ‘limbs’, i membri ‘mem- ‘our/ours’, and, moreover, whether the possessed entity bers’; il braccio ‘arm’ – le braccia ‘arms’ (of a body), i is masculine, e.g. mio, or feminine, e.g. mia, singular, bracci ‘arms’ (of an object). Some nouns have an irreg- e.g. mio, or plural, e.g. miei. Therefore, the possessive ular feminine form. In this case, it is quite common for pronoun form vostre ‘yours’, for example, displays the some speakers to use only a masculine form of the nouns following information: the possessor is a second-per- instead of the irregular feminine forms: il ginocchio son plural (as ‘you plural’), and the possessed entities ‘knee’ – le ginocchia or i ginocchi ‘knees’. are plural and feminine (as ‘things’ feminine). In Italian, articles and other modifiers may accom- The verbs are traditionally classified into three class- pany the noun and usually their forms agree with the es called ‘conjugations’, according to their ending in the form of the noun in the categories of gender and num- present infinitive form: the verbs ending in –are belong ber: e.g. il riso ‘the laughter’, la moglie ‘the wife’. to the first conjugation, e.g. comprare ‘to buy’, those The article forms are chosen according to the sound ending in –ere belong to the second conjugation, e.g. of the initial letters of the word it precedes. In the case credere ‘to believe’, and those ending in –ire belong to of invariable nouns, the article helps to identify the cat- the third conjugation, e.g. dormire ‘to sleep’. There are egories of gender and number of the respective nouns: seven moods in Italian, namely Indicative, Conditional, e.g. il re – i re ‘the king/s’, la gru – le gru ‘the crane/s’. Subjunctive, Imperative, Gerund, Participle, and Also, in the case of nouns of different gender ending in Infinitive, all of them having similar functions as in the same vowel –e, the article reveals the gender of the English. Some of these moods have additional tenses noun in question: e.g. il monte ‘mountain’ (m.) vs. la that express different time relations, such as ‘past his- mente ‘mind’ (f.). Unlike in other languages, the plural toric’, which is used for completed actions that are not forms of the indefinite article are rarely used. The arti- related to the present, and ‘imperfect’, which is used for 544  ITALY continuous action during which some other action one (plural): e.g. librone ‘big book’. To convey the idea occurs. Some of the tenses are compound, i.e. they are of a bad or ugly quality, the following suffixes are made up of an additional verb, called ‘auxiliary’, and a added: -accio, -accia, -acci, -acce, e.g. ragazzaccio specific form of the verb in question (the main verb). In ‘rough boy’. To form collective nouns, the following general, the transitive verbs (i.e. those that take a direct suffixes are used: -ame, -ume, -aglia, the latter two also object) take the auxiliary avere ‘to have’, and the intran- having a derogatory connotation, e.g. pollame ‘poul- sitive (i.e. those that do not take a direct object) and try’, forestierume ‘foreign scum’, antiglia ‘old junk’. reflexive verbs (i.e. the verbs with a reflexive pronoun) The suffix –one is used to denote specific agents, e.g. take the auxiliary essere ‘to be’. imbroglione ‘cheat’, beone ‘drinker’, while the suffix The adverbs are invariable parts of speech. Most of –io indicates repetitive tasks or actions, e.g. lavorio them are formed by adding the suffix -mente to the fem- ‘constant working’, mormorio ‘murmuring’. The suffix inine singular form of the corresponding adjectives, e.g. –ardo is used to form pejorative words, e.g. bastardo rapidamente ‘quickly’. While the comparative form of ‘bastard’, vecchiardo ‘nasty old man’. Some of the suf- the adverbs follows the rules of the adjectives (see fixes, like –accio may also be used as independent above), e.g. più gentilmente ‘kinder way’, the superla- words, e.g. proprio accio ‘really bad’. Some of the suf- tive form is always formed with molto plus the adverb fixes also combine with each other, e.g. –etto + -ino: e.g. molto gentilmente ‘very kind way’. The adverb pre- librettino ‘little book’, -one + -ino: cartoncino ‘small cedes the adjective or the adverb and follows the verb it card’, -etto + -accio: cagnettaccio ‘nasty little dog’. modifies, e.g. ha finito presto ‘he finished early’. The suffixes may also be added to verbs, e.g. lavorare In the case of numerals, cardinal numerals indicate ‘to work’ – lavoricchiare ‘to work a little’, dormire ‘to the quantity; they are invariable in form and (apart for sleep’ – dormicchiare ‘to snooze’, to adverbs, e.g. beno uno ‘one’) they all go with plural nouns, e.g. cinque ‘well’ - benino ‘quite well’, presto ‘early’ – prestino studentesse ‘five students’. The Italian system of writ- ‘quite early’, and to personal names: e.g. Gigi – ing uses points to separate thousands and commas to Gigino, Peppe – Peppino, Simona – Simonetta. separate decimals. When expressing the day of the month, except for the first day, the date is indicated by Referenes cardinal numerals, e.g. nove marzo ‘the 9th of March’. Ordinal numerals indicate a sequence; they agree in Aust, Derek, and Mike Zollo. 1997. Azione grammatica: a new gender and number with the noun to whom they refer. Italian grammar. London: Hodder & Stoughton. Clivio, Gianrenzo P., and Marcel Danesi. 2000. The sounds, The cardinal numeral uno ‘one’ is used like the indef- forms, and uses of Italian: an introduction to Italian linguis- inite article un, una, uno ‘a/an’ and agrees with the tics. Toronto: University of Toronto Press. noun to whom it refers in gender. Dardano, Maurizio, and Pietro Trifone. 1997. La nuova gram- Suffixes play an important role in the Italian vocab- matica della lingua italiana (The new grammar of the Italian ulary. Italian frequently uses affective suffixes, usually language). Bologna: Zanichelli. Haller, Hermann W. 1999. The other Italy: the literary canon in related to size or emotions. In order to indicate small- dialect. Toronto: University of Toronto Press. ness or express affection, the following suffixes are Ledgeway, Adam. 2000. Comparative syntax of the dialects of used: -ino/a/i/e, e.g. mamma ‘mother’ – mammina, - Southern Italy : a minimalist approach. Blackwell Publishers. etto/a/i/e, e.g. libretto ‘little book’, -ello/a/i/e, e.g. pae- Lepschy, Anna Laura, and Giulio Lepschy. 1988. The Italian sello ‘little village’, and -uccio, -uccia, -ucci, -ucce, language today. London: Hutchinson. Proudfoot, Anna, and Francesco Cardo. 1996. Modern Italian e.g. caruccio ‘quite expensive’. On the other hand, to grammar: a practical guide. London: Routledge. denote largeness, the suffixes used are -one/-ona, -oni/- RADU DANILIUC Italy Italy, officially known as the Italian Republic ters, counts 57,679,955 inhabitants and is divided into (Repubblica italiana), is a country situated in southern 20 regions (regioni). Europe. It consists of the Italian Peninsula, two larger The country’s official language is Italian, which is a isles (Sicily and Sardinia), and quite a number of Romance language, i.e. a language developed from the smaller ones. It has an area of 301,338 square kilome- spoken variant of Latin (other Romance languages are 545 ITALY French, Spanish, Portuguese, Romanian, etc.). Italian on every occasion. However, most Italians use However, the country is characterized by a large lin- alternatively their proper ‘dialect’ and the standard guistic diversity. This diversity includes various language, depending on the communicative situations dialects and minority languages. It is important to note they are in (e.g. familiar discussion vs. formal dis- that, in the case of Italy, the term ‘dialect’ is not course). With regard to this state of affairs, one could intended in the way it usually is: it indicates a local think of using the terms bilingualism or diglossia. It variety that has rather little in common with the stan- seems that, as far as Italy’s standard languages and dard language and with other Italian ‘dialects’ ‘dialects’ are concerned, the term diglossia is the most (Lepschy-Lepschy). Actually, these varieties should be appropriate, although Ferguson’s definition does not considered as an outcome of different developments match exactly the Italian situation. Anyway, nowa- from spoken Latin just like standard Italian, French, days, the very large majority of Italy’s inhabitants are Spanish, and other Romance languages. able to express themselves in standard Italian. Apart An explanation for this peculiar situation can be from everyday speech, ‘dialects’ can also be used in found in history: during centuries, the Italic Peninsula literary works and for the major ‘dialects’, there are was a patchwork of different states that, politically, translating dictionaries from and to standard Italian. economically, and culturally, more or less differed one As for the classification of the different ‘dialects’, from another. On the linguistic scale, the only form of important divisions can be made, but it should be standardization consisted in that the speech of major borne in mind that on the geographical plan, the centers was sometimes considered as a model to reach ‘dialects’ gradually blend into one another and that (Maiden). For quite a long time, Latin remained the most distinct borders represent, in fact, gray zones of language in religious, scientific, administrative, and transition. judicial matters. In literature, however, there was a Traditionally, three major groups are identified: growing interest in the use of the regional language; —northern ‘dialects’ the discussion, known as the Questione della Lingua —Tuscan ‘dialects’ (‘Language question’), about the nature of a literary —central and southern ‘dialects’. language for the Italian Peninsula ended up with the hegemony of the Florentine language, as used in the It is considered that an imaginary line between the works of Dante Alighieri (1265–1321), Francesco cities of La Spezia and Rimini delimitates the northern Petrarca (1304–1374), and Giovanni Boccaccio zone from the central and southern ‘dialects’, Tuscan (1313–1375), and described in Pietro Bembo’s gram- ‘dialects’ (also situated below the La Spezia-Rimini mar Prose della volgar lingua (1525). line) occupying a specific position. This line roughly It was only in 1861, after the movement known as delimits a number of more or less general features, but Risorgimento (‘Resurrection’), that the country was it should be clear that these are abstractions and that a politically unified. At that time, the writer Alessandro large variation exists within each of the ‘dialect’ Manzoni (1785–1873) had already been confronted groups. In any case, the ‘dialects’ closest to standard with the lack of a national language. For the third ver- Italian are the Tuscan ones. sion of his famous novel I promessi sposi (1840, ‘The In addition to the so-called ‘dialects’, one can find, Betrothed’), he had deliberately chosen the contempo- in some regions of Italy, other varieties different from rary speech of the cultivated Florentine classes, since standard Italian. These varieties (among which the literary Florentine missed vocabulary to allow him to most important are Sardinian, Friulian, and Ladin) write about everyday life. Although Manzoni’s novel developed from spoken Latin, just like standard Italian aimed and obtained national spread, it was rather the and Italian ‘dialects’. Their status, however, has been a literary variant that was diffused in the newly unified matter of discussion for quite a long time. Nowadays, country (Maiden), where, around 1861, only a very most scholars refer to them by using the term ‘lan- small part of the literate population had a thorough guages’ rather than ‘dialects’. command of the standard language. Despite the polit- Due to its geographical isolation, the language of ical unification, the linguistic unification took place the isle of Sardinia is considered as the most conser- rather slowly in its beginnings. However, it was boost- vative of Romance languages, i.e. closest to Latin. ed by a number of external factors, such as national Nevertheless, it should be stressed that there is no such education, the rise of mass media, general conscrip- thing as a standard Sardinian language: several vari- tion, migration, and urbanization (De Mauro). eties exist, and they are traditionally classified into Research executed by the Italian opinion investiga- four groups (speech of the Logudoro, the Campidano, tion institute Doxa in 1991 shows that 12.8% of the the Gallura, and of Sassari). As the spread of a nation- country’s population claims to use ‘dialect’ in all or al language after 1861 also reached Sardinia, standard nearly all situations, while 33.6% speaks standard Italian is used on most formal occasions. However, 546  ITALY Sardinian varieties are spoken by about 1.2 million spoken in Trentino-Alto Adige joins the Tyrolean and people nowadays (Klinkenberg), mostly in informal Bavarian type of speech (Mazzotta). contexts. Therefore, the term bilingualism (when French is an official language in the autonomous Sardinian is considered to be a language) and diglos- region of Val d’Aosta, in Italy’s most northwest corner. sia (when it is regarded as a dialect) is also used in the In this region, it also constitutes a compulsory subject case of Sardinia. in education. Yet another Romance language is spoken Friulian and Ladin are two languages spoken in by about 90,000 people (Klinkenberg): this language, northern Italy. Friulian is spoken in parts of the called Franco-Provençal, represents merely a group of autonomous region of Friuli-Venezia Giulia. The use different dialects without a standard. These dialects of Ladin is limited to a number of valleys in Alto historically belong to a zone covering a part of France, Adige (or South Tirol, part of the autonomous region Switzerland’s French-speaking region and Val of Trentino-Alto Adige, see map). Both languages are d’Aosta. In formal and official contexts, however, they closely related to Romansh, one of Switzerland’s four are almost completely surpassed by French and official languages, and have characteristics in common Italian. with northern Italian ‘dialects’. Yet, like Romansh, A similar situation can be found in Piedmont, they each represent one of a series of ‘dialects’. In the where small Franco-Provençal, Occitan, and Walser case of Friulian, the variety from Udine and environs minorities exist. Occitan is also a Romance language, is sometimes considered as a base for a standard, but mainly spoken in southern France, yet remaining a this assertion has also been criticized. Friulian is esti- minority language there as well. Occitan has to cope mated to be spoken on a regular basis by approxi- with problems related to standardization, but like mately 430,000 people (Picco). However, it is again Franco-Provençal it is recognized as a minority lan- standard Italian that is used in formal and official con- guage by national law and is promoted by various cul- texts. One might be tempted to use the terms diglossia tural associations. The same holds for the speech of and bilingualism, but none of them fully accounts for the Walser, which is a variety of German spoken by a the situation: diglossia would imply that Friulian is small group of people in Val d’Aosta and Piedmont. considered as a ‘dialect’ of standard Italian and the use Catalan, the official language of the region of of bilingualism would mean that both Friulian and Catalonia in Spain, is spoken by a minority of Italian are used in all circumstances. Attempts to use 15,000–20,000 people in the Sardinian town of Friulian at school and in mass media are not always Alghero, where it is also a subject in education successful (Francescato 1989); the lack of a Friulian (Klinkenberg). standard appears to be one of the major problems. Slovenian is spoken by approximately 20,000–30,000 As for Ladin, it has, like standard Italian and people in Friuli-Venezia Giulia, close to the Slovenian German, an official status in Trentino-Alto Adige and border; there are a number of schools that provide edu- is probably spoken by about 35,000 people (official cation in Slovenian (Francescato-IvašiJ KodriJ). figures and estimates). There is a written standard, and Serbo-Croatian is estimated to be spoken by a small Ladin can be taught and used at school. However, the comunity of 4,000–5,000 people in central Italy Ladin language is under pressure, as it is less presti- (Mazzotta). gious than standard Italian and German. In fact, these Albanian and Modern Greek are spoken by scat- two languages are far more spoken, not only seen in a tered communities in southern Italy; it is mostly elder- European context but also within the region itself. ly people who seem to hold on to their language. Moreover, German and Italian are languages that con- In a country where so many language varieties are vey a widespread culture. used, different phenomena due to language contact can Like German, a number of minority languages spo- be expected. In regions where one or more languages ken in Italy have in common that they are an official coexist with standard Italian, these languages (French, language in one or more countries or regions outside German, etc.) may influence standard Italian. Italy. Their status within the Italian Republic, howev- However, it is clear that the position of standard Italian er, is not the same for all. as the Republic’s official language remains first, as its German is an official language in Trentino-Alto knowledge is boosted through public education, mass Adige, together with standard Italian and Ladin. media, etc. Schools can be either in German or in Italian, but in As for the contact between standard Italian and both of them the second language is the region’s other Italian ‘dialects’, it is obvious that the spoken and—to official language; there are also plurilingual schools a lesser extent—written use of standard Italian bears with a fifty–fifty division of German and Italian, marks of ‘dialectal’ features, such as differences in where Ladin is also taught (Freddi). As far as the lin- pronunciation or in vocabulary. An example of the for- guistic features are concerned, the German variety mer is the Roman -NN- pronunciation, where standard 547 ITALY Italian has -ND- (annamo for andiamo ‘we are References going’); an example of the latter can be the use of the De Mauro, Tullio.1963. Storia linguistica dell’Italia unita. Bari: Venetian word campo for piazza ‘square’. By the way, Laterza. 2nd edition, 1970. ‘dialectal’ words can be accepted in standard Italian, Ferguson, C.A. 1959. Diglossia. Word 15. 325–40. like the word tortellini (‘tortellini’, i.e. a type of Francescato, Giuseppe. 1989. Friaulisch: Soziolinguistik. pasta), which has its origin in Emilia (Zolli). Sociolinguistica. Lexikon der Romanistischen Linguistik (LRL). Band/Volume III. Die einzelnen romanischen But the standard language also leaves its traces in Sprachen und Sprachgebiete von der Renaissance bis zur ‘dialectal’ speech. This mainly happens in the Gegenwart. Rumänisch, Dalmatisch/Istroromanisch, ‘dialects’s’ sound system and vocabulary (where a Friaulisch, Ladinisch, Bündnerromanisch. Les différentes local term is abandoned in favor of an Italian one; e.g. langues romanes et leurs régions d’implantation de la Piedmontese ['pejla] ‘pan’ is replaced by [pa'dela], Renaissance à nos jours. Le roumain, Dalmatico/ Istroromanzo, Friulano, Ladino, Le romanche, ed. by from standard Italian padella) (Sobrero 1997). The Günter Holtus, Michael Metzeltin, and Christian Schmitt. diffusion of the standard language also influences the Tübingen: Niemeyer. global use of the ‘dialect’. Referring to the contempo- Francescato, Giuseppe, and M. IvašiJ KodriJ. 1983. La comu- rary situation, it is possible to speak of an abandon- nità slovena in Italia. L’Italia plurilingue, ed. by Giovanni ment of ‘dialect’, which is, however, far less frequent Freddi. Bergamo: Minerva Italica. Freddi, Giovanni. 1983. Maggioranze, minoranze e plurilin- in rural areas than in urban zones, where several peo- guismo nella Provincia di Bolzano (Südtirol-Alto Adige). ple from different origins live together and where var- L’Italia plurilingue, ed. by Giovanni Freddi. Bergamo: ious cultures coexist. Minerva Italica. Recently, the term italiano popolare (‘popular Klinkenberg, Jean-Marie. 1994. Des langues romanes. Italian’) has been used to indicate a kind of substan- Louvain-la-Neuve: Duculot. Lepschy, Anna Laura, and Giulio Lepschy. 1977. The Italian dard Italian. This substandard Italian originates from language today. London: Hutchinson. an imperfect acquisition of the standard language, pos- Maiden, Martin. 1995. A linguistic history of Italian. London sibly—but not necessarily—due to a ‘dialectal’ back- and New York: Longman. ground. Some typical features of popular Italian are: Maiden, Martin, and Mair Parry (eds.) 1977. The dialects of Italy. London and New York: Routledge. English Standard Italian Popular Italian Mazzotta, Giuseppe. 1983. Profilo linguistico dell’Italia odier- it seems to me ... A me pare ... A me mi pare ... na. L’Italia plurilingue, ed. by Giovanni Freddi. Bergamo: (or Mi pare ...) Minerva Italica. to have avere averci Picco, Linda. 2001. Ricercje su la condizion sociolenghistiche (= avere + ci ‘there’) dal furlan. Ricerca sulla condizione sociolinguistica del friu- the place where il posto dove il posto che lano. Udine: Forum. we have been siamo stati siamo stati Rohlfs, Gerhard. 1949–1954. Historische Grammatik der (‘the place that we Italienischen Sprache und ihrer Mundarten. I Lautlehre, II have been’) Formenlehre und Syntax, III Syntax und Wortbildung. Bern: Francke. Popular Italian is not a language on its own, but a Sobrero, Alberto. 1997. Italianization of the dialects. The whole of characteristics that differ from standard dialects of Italy, ed. by Martin Maiden and Mair Parry. London and New York: Routledge. Italian. These characteristics can be widespread Zolli, Paolo. 1986. Le parole dialettali. Milano: Rizzoli. through the peninsula (like the examples above), but The data concerning Italy’s area and population are the figures sometimes they are limited to one or more (parts of) provided for 2000 by the Italian institute ISTAT regions. (www.istat.it). The campaign that aimed, shortly after Italy’s uni- The data from the opinion investigation institute Doxa can be obtained through www.doxa.it. fication, to promote ‘standard Italian’ as a national lan- ALEXANDER LOENGAROV guage, seems to have reached its goal. Local and regional types of speech often remain lively, which See also Bilingualism; Diglossia; Indo-European 4: leads to intense contact between the different codes. Romance; Italian; Switzerland 548  J Jakobson, Roman Roman Jakobson was one of the major linguists, liter- /b/ in bat is differentiated from /m/ in mat as oral vs. ary theorists, and semioticians of the twentieth century, nasal. He defined a small set of such features that and a leading proponent of functional structuralism. underlie the phonological systems of all the languages of the world. He also showed the relevance of these notions for grammatical meaning, which he decom- Structural Phonology, Child Language, and posed into semantic features. Aphasia Jakobson helped to found the Prague Linguistic Invariance and Relational Structure Circle, the cradle of (functional) structuralism, in 1926. For the Prague School, language serves for com- Invariance was, for Jakobson, the dominant topic and munication and thus its internal structure has to be methodological device underlying his research: any studied from the standpoint of the tasks it performs. sign is defined by its invariant properties. Across lan- Jakobson and Nikolai Trubetzkoy argued that phono- guages as well, he also insisted on the importance of logical systems are structural wholes, based on their universal invariants (language universals), establishing relational properties, and that language change has to the equivalence of diverse signs. Invariance for be seen in terms of systems. They investigated types of Jakobson was always invariance in relationship phonological systems and established the study of lin- (equivalence), and the primary type of relation in lan- guistic affinities (Sprachbünde)—how languages that guage is that of binary opposition, phonological (e.g. are geographically and culturally related may acquire nasal vs. oral), grammatical (e.g. plural vs. singular), traits in common. or lexical (e.g. near vs. far), etc. Opposition is based Jakobson also turned his attention to the acquisition on an asymmetry in the relationship between a marked of phonology by children and its dissolution in apha- (focused or weighted) and an unmarked (neutral) term: sics and demonstrated that the order of acquisition e.g. unmarked oral, singular, far vs. marked nasal, plu- goes from universal to nonuniversal, simple (optimal) ral, near. For Jakobson, markedness is a fundamental structures to more complex ones. He also showed that way by which we organize our linguistic, semiotic, aphasic loss follows a reverse order. and sociocultural systems in general. Theory of Distinctive Features and Functional Functions of Language View of Sound For Jakobson, the raison d’être for language is com- Jakobson showed that phonemes such as /b/ and /m/ in munication and language is suited to various commu- bat and mat are not the smallest constituents of lan- nicative goals, which in turn are correlated with the guage. Rather, distinctive features are. For example, speech event in which language is used. He defined 549 JAKOBSON, ROMAN six primordial functions in terms of a focus on one of Jakobson’s Legacy in Linguistics the facets of the speech event: (1) emotive function Many of Jakobson’s contributions to twentieth-centu- (focus on the speaker)—e.g. intonation showing ry thought have become a permanent part of American anger; (2) conative function (focus on the and European views of language. He contributed to addressee)—e.g. imperatives and vocatives; (3) refer- linguistics such concepts as (distinctive) feature, bina- ential function (focus on the context)—e.g. talk about ry opposition, structuralism, markedness, and univer- the real world; (4) poetic function (focus on the mes- sals. He also showed the importance for linguistics of sage)—e.g. poetry; (5) metalingual function (focus on child language acquisition, aphasia, poetry, the act of the code)—e.g. definitions of words; and (6) phatic communication (language usage), the meaning of function (focus on the contact)—e.g. ‘hello, do you grammatical categories, and the systematicity of lan- hear me?’ These functions may be dominant in a mes- guage change. He has had such a towering role in lin- sage or may be subsidiary: e.g. a referential message guistics that his work has defined the field itself and may also carry expressive information about the opin- many of his concepts and discoveries are now thought ion of the speaker. to be commonplace or self-evident. Jakobson was a leading proponent of structuralism, Metaphor and Metonymy: arguably one of the most influential trends not only in Similarity/Contiguity linguistics but in other humanistic areas as well. His Jakobson also analyzed the relation between commu- functional viewpoint has inspired functionalist nicative processes and properties of linguistic struc- approaches more generally and abetted developments ture. First, he distinguished the two fundamental such as discourse analysis, text linguistics, ethnogra- operations: selection (substitution) for encoding (pro- phy of communication, anthropological linguistics, duction) and combination for decoding (comprehen- sociolinguistics. Work in pragmatics has led linguists sion). Then, he contrasted two types of relations: to explore the boundaries of their discipline with the similarity (all types of equivalence) and contiguity neighboring fields of anthropology, sociology, mythol- (temporal and spatial neighborhood). Put simply, the ogy, philology, and philosophy in a truly Jakobsonian elements in a selection set are normally associated by interdisciplinary spirit. similarity, and those in combination by contiguity. For Jakobson stressed the need to search for general Jakobson, similarity/contiguity is a fundamental polar- laws governing all linguistic systems and thus helped ity of language, texts, culture, and human thought to foster the current interest in universal and law-gov- (cognition) in general. He proposed using it as a way erned, general (explanatory) properties of language. of categorizing types of aphasic disturbances and also His seminal work on child language and aphasia has of defining the poetic function: in poetry, where focus been signaled as the point of departure of a new era in on the message is dominant, equivalence (similarity) linguistics because it showed the relation between relations help to build the combination. Thus, rhyme, child language and phonology, and it launched psy- alliteration, parallelisms, metaphor, etc. help to struc- cho- and neurolinguistics and contributed to contem- ture the poetic text, whereas in prose contiguity is the porary work on the functional organization of the essential constructional principle. human brain. The widespread movement that presently investi- gates universal properties of language, both universal Grammar and Semantics grammar (Noam Chomsky) and language universals Jakobson also focused on the function of grammatical (Joseph Greenberg), is deeply indebted to Jakobson. categories. For example, he worked on those elements Many scholars agree with Jakobson that typology is whose general (invariant) meaning in the code can important for the study of language change, areal lin- only be specified by taking into account their use in guistics, and historical reconstruction. His seminal speech events, e.g. deictic pronouns designate speaker idea that changes must always be treated in view of the (I) and addressee (you). Thus, language encodes prag- system that undergoes them is the essential premise matic factors of the context of utterance. for textbooks in historical linguistics. Grammatical categories (both morphological and Many linguists have claimed that the distinctive fea- syntactic), for Jakobson, are obligatory, whereas par- ture was Jakobson’s greatest insight and, after the ticular lexical categories (e.g. words referring to phoneme, the most significant step forward in modern space) are optional. Thus, he provided a semantic and phonology. Generative phonology recognizes Jakobson operational approach to the relation between language as a founding father, markedness is now widely used in and cognition: grammatical categorizations provide linguistics and other disciplines, and componential the necessary patterns of thought. analysis—analyzing an item into smaller properties 550  JAPANESE and representing them as a combination—has been 1960 he became professor of general linguistics as well. adopted by many. In 1957, he was also named Institute Professor at MIT. With the breadth and depth of his knowledge, his He was President of the Linguistic Society of America brilliant imagination, his prodigious output, the origi- in 1956, and received the International Prize for nality of his accomplishments, and his international Philology and Linguistics in 1980 and the Hegel Prize influence on a variety of fields, Roman Jakobson has in 1982. Jakobson died in Cambridge, Massachusettes been recognized worldwide as one of the major cre- on July 18, 1982. ative minds of our century. References Biography A Tribute to Roman Jakobson 1896–1982. New York: Mouton. Roman Jakobson was born in Moscow, Russia on 1983. October 11, 1896. He was educated at Moscow Armstrong, J. Daniel, and C.H. van Schooneveld (eds.) 1977. Roman Jakobson: echoes of his scholarship. Lisse: Peter de University, in the Slavic section. He was awarded the Ridder Press. Buslaev Prize in 1916 for his work on North Russian Bradford, Richard. 1994. Roman Jakobson: life, language, art. folk epics, and the master’s was awarded in 1918. He London: Routledge. helped to found the two groups known now as Russian Holenstein, Elmar. 1976. Roman Jakobson’s approach to lan- Formalism in 1915–1916. He moved to Prague in 1920. guage. Bloomington: Indiana University Press. Jakobson, Roman. 1990. A complete bibliography of his writ- In 1926, he was cofounder of the Prague Linguistic ings, 1912–1982, ed. by Stephen Rudy. Berlin: Mouton. Circle and was its vice-president until 1939. He Jakobson, Roman, and Krystyna Pomorska. 1980. Dialogues. received his doctorate in 1930 from Prague University. Cambridge, MA: MIT Press. In 1933, he began teaching at the T.G. Masaryk _________. 1987. Language in literature, ed. by Krystyna University at Brno, and assumed the chair of Russian Pomorska and Stephen Rudy. Cambridge, MA: Harvard University Press. Philology and Old Czech literature at Brno in 1937. In _________. 1990. On language, ed. by Linda Waugh and 1939, he fled Nazi invasion and went to Scandinavia. Monique Monville-Burston. Cambridge, MA: Harvard He was visiting lecturer in Copenhagen and Oslo until University Press. 1940, and then in Uppsala until 1941. In 1941, he took _________. 1962. Selected writings I, phonological studies. a freighter to the United States, and taught at the Ecole The Hague: Mouton; 2nd expanded edition, 1971; II, Word and language, The Hague: Mouton, 1971; VII, Contributions Libre des Hautes Etudes in New York in 1942–1946. In to comparative mythology; studies in linguistics and philol- 1943, he was cofounder of the Linguistic Circle of New ogy, 1972–1982, Berlin: Mouton, 1985; VIII, Major works, York and its vice-president until 1949. From 1943 to 1972–1982, Berlin: Mouton, 1987. 1946, he was visiting professor of general linguistics at Waugh, Linda R., and Monique Monville-Burston. 1990. Columbia University and became T.G. Masaryk Introduction: the life, work and influence of Roman Jakobson. In Jakobson. Professor of Czechoslovak Studies in 1946. In 1949, he LINDA R. WAUGH was named Samuel Hazzard Cross Professor of Slavic Languages and Literatures at Harvard University, and in See also Phonology; Trubetzkoy, Nikolai Sergeyevich Japanese Japanese is the national and official language of Japan, Japanese is written with ideographic characters, spoken by approximately 126 million people on the called kanji, originally borrowed from China, and also Japanese islands, as well as by Japanese minorities in with two syllabaries, the hiragana and katakana, Hawaii, North and South America, and other parts of developed in Japan by simplifying the Kanji Chinese the world. The language is called Kokugo, ‘language characters. A kanji is connected to semantic content of the country’ in Japan, and Nihongo, literally ‘Japan- and has several different pronunciations, while hira- language’, when spoken by nonnative speakers. In gana/katakana represent syllables. Roughly, kanji are terms of number of speakers, Japanese ranks eighth on used for content words, e.g. ‘sun’ or ‘read’, hiragana the list of major languages of the world. for grammatical endings and grammatical function 551 JAPANESE words, e.g. -ta past tense or moshi ‘if’, and katakana relation to Dravidian languages (spoken in the south of for loan words, e.g. erebeetaa ‘elevator’. Additionally, India) is hypothesized, as Dravidian languages also Roman script is used for abbreviations, e.g. NHK, and share the grammatical features listed above for Altaic words cited from foreign languages. Traditionally, languages. Japanese is written vertically and from right to left, but The earliest Japanese writings date back to the today horizontal left to right writing is also used, eighth century CE. The language then developed from depending on text type and space. Old Japanese to Modern Japanese between the twelth The question of any language family membership and sixteenth centuries. Late Old Japanese and Middle of Japanese is still disputed. Japanese is not related to Japanese are considered today as the literary language Ainu, formerly spoken on Japan’s biggest northern- (bungo). The term ‘Classical Japanese’ is used either most island Hokkaido, the Kurile islands, and to correspond to bungo, or may designate any stage of Sakhalin (an island off the northeast coast of the the language between the eighth and fourteenth cen- Asiatic continent). Nor does Japanese belong to the turies. This epoch produced a rich literature of lasting Sino-Tibetan language family spoken on the Asiatic importance by court nobles, especially court ladies, continent, the most prominent representative of which both in prose and verse. is Chinese. The most widely discussed hypothesis After Japan was politically unified in the seventh today is an affiliation to Korean (spoken in Korea) and and eighth centuries, the greater Kyoto region in the to Altaic languages (Turkic, Mongolian, Manchu, and west of the main island of Honshu became the politi- Tungusic languages, spoken in continental Asia). cal and cultural center in the eighth century, which it Japanese shares a substantial number of grammati- remained, almost without a break, for about a thou- cal characteristics with both Korean and Altaic lan- sand years. The political center was then moved east- guages: word order in the sentence; suffixal inflection ward to Tokyo at the beginning of the seventeenth (with endings rather than with prefixed elements); century, and gained cultural authority until the end of postpositions (functioning like English prepositions, the eighteenth century. Today, the speech of the Tokyo but following the noun); the existence of ‘converbs’ region, adorned by influences from the Kyoto variety (verb forms that are used for clause linking instead of and interspersed with traits of the prevailing local conjunctions); question formation via particles at the dialect, especially accentual features, functions as the end of a sentence; vowel harmony (a restriction on standard national and official language. vowels occurring together in a word, of which there Japanese is rich in dialectal variation. Several dif- are traces in Old Japanese); and the absence of a gen- ferent classifications have been proposed, but two der system, articles, and relative pronouns. Attempts to major dialect divisions are generally recognized: (1) establish lexeme correspondences between Altaic lan- between the Eastern dialect group spoken in the north- guages and Japanese, and to reconstruct a common eastern half of the main island of Honshu up to the proto- (ancestral) language, however, have not led to a northern island of Hokkaido and the Western dialect significant amount of convincing phonological corre- group spoken in the southwestern part of Honshu, and spondences. The absence of complex consonant clus- (2) between the Western dialects and the dialects spo- ters (e.g. tk, pl, str), and the near absence of ken on the island of Kyushu, south of Honshu. The syllable-final consonants in the Japanese sound pat- East–West dialect boundary runs through Central tern, contrasts with Altaic languages, where these uses Honshu between the Niigata and Toyama prefectures do occur. Additionally, vowel harmony occurred only in the north, along the Japanese Alps, and between the briefly in the linguistic history of Japanese, and inflec- Shizuoka and Aichi prefectures in the south. The tional endings on verbs for persons involved in the boundary is constituted by the existence and pronunci- event being discussed are absent in Japanese. ation of several sounds, by the accent pattern, and by a The sound pattern, the open (vowel-final) syllables, number of inflectional forms of the verb. Other and the basic bisyllable pattern of words in Japanese dialects are differentiated by their sound and accent are reminiscent of a neighboring language family, the patterns, lexicon, and functionality of case markers. Austronesian languages (spoken in Taiwan, the A fourth linguistic area is formed by the islands of Philippines, Indonesia, Oceania, and Madagascar), in Okinawa, south of Kyushu. The language of Okinawa which these same properties are characteristic. But is sometimes considered to be another major dialect of again, a sufficient number of phonological correspon- Japanese, but it is classified by other scholars as a sep- dences between Japanese and Austronesian have not arate language, known as Ryukyuan or Luchuan. The been established, and sentence structure is disparate. It latter approach is supported by the mutual unintelligi- is therefore assumed by some scholars that Japanese bility of Japanese and Ryukyuan. Ryukyuan is esti- might be a hybrid language, amalgamating features of mated to have split from the Japanese stock around the Altaic and Austronesian languages. Besides this, a beginning of the Christian era and was the language of 552  JAPANESE an independent kingdom between the fifteenth and form, which is used with reference to a socially supe- seventeenth centuries CE, before the Ryukyu islands rior person; and a humble form, which is used with were incorporated into Japan. Ryukyu is divided into reference to the speaker or the speaker’s group: three larger dialects, which are mutually unintelligible Neutral Elevating Humble both among themselves and with Japanese dialects. ‘a respected person ‘(speaker) does X While Ryukyuan has a different sound pattern with does X’ for/to a respected fewer vowels than Japanese, sound correspondences person’ were easily established. Morphology (word structure) ‘read’ yomu o-yomi-ni o-yomi-suru ‘... naru and syntax (clause structure) are quite similar to ‘exist’ iru irassharu oru Japanese, but Ryukyuan preserves the lexical, mor- ‘come’ kuru irassharu mairu phological, and syntactic features found in Old ‘do’ suru nasaru itasu Japanese, having also introduced inflectional forms: ‘eat’ taberu meshiagaru itadaku Japanese (Tokyo) Ryukyuan (Naha) Respect forms also exist for a number of nouns, ‘nose’ hana hana especially kinship terms. The prefixes o- and go- are ‘wind’ kaze kazi ‘hand’ te ti’i honorific: ‘smoke’ kemuri kibusi Own thing/kin Respected person’s ‘bird’ tori tu’i thing/kin, address term Among the 1.3 million inhabitants of the Ryukyu tegami ‘my letter’ o-tegami ‘your/his/her/their letter’ islands, Ryukyuan is spoken mainly by the older gen- hon ‘my book’ go-hon erations. Those under 20 years of age speak only ‘your/his/her/their book’ Japanese. Thus, Ryukyuan is a severely endangered haha ‘my mother’ o-kaasan language. Although some private broadcasting servic- ‘your/his/her/their mother’, es provide news in Ryukyuan, its prospects for surviv- ‘mother!’ chichi ‘my father’ o-toosan ing are bad for political reasons. ‘your/his/her/their father’, Japanese grammar and lexicon present a number of ‘father!’ phenomena that are of theoretical interest to general lin- ane ‘my elder sister’ o-neesan guistics. Japanese has a pitch-accent pattern, which dif- ‘your/his/her/their fers from both stress-accent patterns (as in English)—in elder sister’, ‘sister!’ otooto ‘my younger brother’ o-toosan which accent is marked by sound pressure or loud- ‘your/his/her/their ness—and tone languages such as Chinese—in which younger brother’ each syllable bears one fixed pitch out of a pool of four pitch types. In Japanese, every word has a pitch ‘con- An elaborate honorific system is also present in tour’ with, at most, one pitch drop. The place of the drop Korean. from high to low pitch in a word distinguishes meaning: As in Altaic languages and Korean, Japanese is rich e.g. hashi-ga with the pitch pattern high–low–low (i.e. in onomatopoetics (sound-symbolic words), which are high pitch on ha) means ‘chopsticks’, while hashi-ga conventionalized mimetic expressions of natural with the pitch pattern low–high–low (i.e. high pitch on sounds, mental and emotional states, sensations, and shi) means ‘bridge’, and hashi-ga with the pitch pattern physical states and manners. These words are usually low–high–high (i.e. high pitch on shi and ga, no drop) used as adverbs. Examples are wan-wan (bow-wow), means ‘corner’. patan (with a bang), zaa-zaa (sound of downpour), A complex system of honorifics (expressions of kossori (stealthily), ira-ira (nervously), zara-zara respect/politeness) plays an important role in any (rough texture), hara-hara (falling down softly; fear- Japanese utterance. Two kinds of honorifics are inter- fully), and bara-bara (falling of heavy objects). woven, namely, the expression of politeness toward Most Japanese loanwords are borrowings from the interlocutor, and the expression of respect toward a Chinese and have a status comparable to Latinate person one is talking about. Appropriate use of hon- words in European languages. Other loanwords come orifics is determined by the social context of the utter- from European and other Asian languages, especially ance, and is socially obligatory. Politeness to the from English. Loan-word adaptation both to the interlocutor is expressed by an inflectional ending on Japanese sound pattern and word classes resulted in the verb (-masu); respect toward a person being men- compromise forms, which introduced new sounds and tioned is shown by the choice of special verbs or syllable-final consonants into Japanese. Native words derivational forms (word formation products) of verbs and loanwords with similar meanings may coexist and and nouns. There is a neutral form for every verb (used usually have specialized meanings: e.g. torikeshi with reference to equally ranked persons); an elevating (native), kaiyaku (Chinese loan), and kyanseru (from 553 JAPANESE English ‘cancel’) all mean ‘cancellation’, but the first and/or ending, or ‘unbounded’, no relevant beginning, example has a very general meaning, while the Chinese or ending). The -te iru form (a kind of continuous loan is used for formal occasions such as in contracts form) of a great number of verbs refers to a continu- and the English loan for reservations or appointments. ing/continuous event, e.g. yomu (read), miru (look), A phenomenon of Japanese word formation of great furu (fall), and kangaeru (consider). For example, theoretical importance occurs when from a clause con- yonde iru translates as, ‘s.o. is reading’. With other struction larger than one word, the last element may verbs such as shinu (die), sameru (wake up), and mit- undergo word formation without any other changes in sukaru (be found), the -te iru form expresses a result: the construction taking place. For example, the phrase shinde iru means, ‘s.o. is dead’. A third class lacks this sake-o nomita-sa ni (out of desire of drinking sake) is form entirely, while for still other verbs, it is the only formed from the clause sake-o nomitai (s.o. wants to form that can be used to refer to present tense. drink (nomitai) sake) with the help of the ending -sa. Japanese has several inflectional verb forms that This is remarkable in view of the theoretical assump- cannot be used as the predicate form, i.e. the kernel, of tion that word formation is a word-level process and an independent sentence, but are the predicate of a should not be accessible beyond the word level. dependent clause or in connection with an auxiliary On the morphological (word structure) level, verb, such as the form ending in -te; these are used for Japanese is agglutinative, i.e. various grammatical end- clause linking. Such verb forms are known as ‘con- ings are chained together to form a word. Japanese has verbs’, which are also typical of Altaic languages. word classes of nouns, verbs, and adjectives, although An interesting grammatical form of a verb is the so- the latter are a subclass not of nouns but of verbs, and called passive, formed with the ending -(r)are-. This they are inflected in the same way as verbs. form is used for events that are not controlled by the The function that a noun fills in a clause is signaled person to whom they are happening, e.g. shinobareru by elements loosely attached to the end of a noun, (something spontaneously comes to my mind) or ame- which are traditionally called ‘postpositions’. A post- ni furareta, literally ‘I was fallen (on) (furareta) by the position is a word that fills the function of a preposi- rain (ame-ni)’, meaning ‘I was adversely affected by tion (‘at’, ‘in’, ‘for’, etc.), but follows the noun instead the rain falling’. The ‘passive’ is also used for the of preceding it. In Japanese, postpositions fill the func- expression of ability, such as nerarenai (I cannot tions of case: e.g. Hanako-ga ‘Hanako’ (nominative), sleep), and as an honorific form, such as korareru (a Hanako-no ‘Hanako’s’ (genitive), Hanako-o ‘Hanako’ respected person comes). (accusative), Hanako-ni ‘to Hanako’, Hanako-to ‘with Japanese is also a classifier language, i.e. it makes Hanako’, etc. Postpositions are also characteristic of use of special elements in counting. Nouns in Japanese Altaic languages. cannot be counted directly, as in English: ‘one book’, The persons involved in an action are not usually ‘two books’, etc. Instead, the numeral element is com- explicitly expressed, neither on the verb nor by nouns bined with another element that expresses the proper- accompanying it. The bare verb constitutes a minimal ties of the noun being counted (such as shape, kind, or sentence, and the persons involved must be deduced function) or measures it. This classifier is then used in from the larger context: e.g. yomu is a complete clause conjunction with the noun. Thus, the element -ko is that may mean, ‘I read it/something’, as well as used for small round objects, -hon for cylindrical ‘you/he/she/it/we/you/they read it/something’. The objects, -ri and -nin for persons, -dai for machines, persons involved in the action can be understood from -ken for houses, -sai for age, -hai for cupfuls, and so a range of verb pairs sharing a common root (basic forth. A counting construction takes the form hon component). One verb of a pair signals that s.o./sth. issatsu (one book, literally ‘book one-bound entity’), other than the speaker or her group is acting (so-called hon nisatsu (two books), and so forth, with issatsu and ‘intransitive’), while the other signals that the speaker nisatsu being the classifiers for bound objects. is acting on s.o./sth. (so-called ‘transitive’): On the clause level, the Japanese word order is sub- Nonspeaker acts Speaker acts ject–object–predicate, as in Hondasan-wa Fujiisan-o agaru ‘rise’ ageru ‘raise’ mita (Mr./Ms. Honda saw Mr./Ms. Fujii). Modifying tomaru ‘stop’ tomeru ‘stop’ elements precede the word they modify, according to aku ‘open’ akeru ‘open’ the patterns kono ie (this house), akai ie (red house), okureru ‘be late’ okurasu ‘postpone’ watashi-no ie (my house), Hondasan-no ie (Mr./Ms. naku ‘cry’ nakasu ‘make cry’ sakeru ‘split’ saku ‘split’ Honda’s house), and Hondasan-ga mita ie (Mr./Ms. Honda saw (mita) house (ie)), which is the equivalent Furthermore, Japanese verbs are classified accord- of the English, ‘the house that Mr./Ms. Honda saw’. ing to their behavior in tense and aspect (presentation Relative pronouns (such as the English ‘which’) do not of an event as ‘bounded’, having a relevant beginning exist. Adverbs likewise precede verbs, according to the 554  JAPANESE SIGN LANGUAGE pattern yukkuri-to aruku (s.o. walks slowly, or literal- chaining of events. In the study of conversation, strate- ly slowly walks). A complicated topic discussed in the gies used for turn-taking, sustaining the audience’s field of Japanese clause structure is configurationality, attention, and confirming attention are studied, as well or whether or not the elements of a clause are hierar- as verbal signs used in such interaction, such as chically structured. As there are arguments both for absence of grammatical elements, reordering of com- and against Japanese being configurational, the ques- ponents of a clause, sentence-final particles (small tion remains open. words), echoing, and affirmative sounds and gestures. An aspect of clause structure for which Japanese has become famous is the existence of two structural References components of the sentence (‘phrases’) that are candi- dates for serving as the subject. These components are Hinds, John. 1986. Japanese. London: Croom Helm. Jacobsen, Wesley M. 1992. The transitive structure of events in identifiable by the markers at their ends, -wa and -ga, Japanese. Tokyo: Kuroshio. respectively. They have different conditions of usage Kishitani, Shoko. 1985. Die Person in der Satzaussage (Person and varying functions, and they may occur simultane- in sentential predication). Wiesbaden: Harrassowitz. ously in one sentence (known as the ‘double subject Kuno, Susumu. 1973. The structure of the Japanese language. construction’), as in zoo-wa hana-ga nagai (the ele- Cambridge: MIT Press. Kuroda, Shige-Yoshi. 1972. The categorical and the thetic judg- phant has a long trunk), or literally, ‘the elephant (zoo- ment: evidence from Japanese. Foundations of Language 9. wa, topic), the nose (hana-ga, subject) is long (nagai, 153–85. predicate)’. The component marked with -ga is the Lewin, Bruno. 1959, 1996. Abriß der japanischen Grammatik grammatical partner of the predicate and therefore the auf der Grundlage der klassischen Schriftsprache (An out- subject; the component marked with -wa is the topic, line of Japanese grammar based on the classical written lan- guage). Wiesbaden: Harrassowitz. i.e. the thing or person about which the speaker says Martin, Samuel E. 1975. A reference grammar of Japanese. something. The topic may have a looser relation to the New Haven: Yale University; Rutland: Tuttle, 1988. predicate, as in boku-wa unagi da (eel for me), or lit- Maynard, Senko K. 1998. Principles of Japanese discourse. A erally ‘I (boku-wa) it is (da) eel (unagi)’. The topic handbook. Cambridge: Cambridge University Press. construction can be better understood if translated by Miller, Roy Andrew. 1967. The Japanese language. Chicago: University of Chicago Press. the English ‘as for . . . ’; for example, ‘as for the ele- Sansom, George. 1928, 1968. An historical grammar of phant, its trunk is long’. Japanese shares the existence Japanese. Oxford: Clarendon Press. of such a topic construction with Korean. Shibatani, Masayoshi (ed.) 1976. Japanese generative grammar Japanese discourse, or larger bodies of spoken or (Syntax & Semantics 5). New York: Academic Press. written text, is thoroughly studied as well. Topics of ———. 1990. The languages of Japan. Cambridge: Cambridge University Press. interest in narrative discourse are the organization of a JOHANNA MATTISSEN text and the elements used for signaling its structure; the way the persons involved in a story are handled lin- See also Ainu; Altaic; Chinese and Japanese Tradi- guistically to maintain a perspective; and the formal tional Grammar; Korean Japanese Sign Language Japanese sign language (JSL, or NS after its Japanese dialects of JSL, respectively (see Smith 1990). JSL name, Nihon-Shuwa (alternatively, Nihon Syuwa), lit- shares an extremely high degree of basic lexical cog- erally ‘Japanese hand-talk’) is the dominant language nates with each of the other two members of the of the deaf community of Japan, and to a lesser extent Japanese sign language family. of the hard-of-hearing community. Owing to the estab- lishment of deaf schools during the time of the mili- Sociolinguistics tary occupations in World War II, JSL is also cognate with the (South) Korean and Taiwanese sign lan- The Tokyo dialect is the socially dominant variety of guages. In the latter case, the dialects of Taipei and Japanese sign language, owing in large part to efforts of Tainan are said to derive from the Osaka and Tokyo the Japan Association of the Deaf and broadcasts of sign 555 JAPANESE SIGN LANGUAGE language courses and sign language news on the NHK ferences between the two sign systems; hence, signing (Nihon-Housou-Kyoukai = Japanese Broadcasting here is also often on the Signed Japanese end of the Corporation, or public television) educational channel. spectrum. In both deaf schools and interpreting situa- Due to these influences, differences between dialects, tions, the reduced degree of comprehensibility of the although present, are declining. Lexical differences sign system used means that much of the content is continue to be considerable in some dialects, particular- understood only partially. ly among elderly signers. Although little research has been undertaken in this area, syntactic differences also Visual Cue Structure arise, affecting auxiliary verb and subject/object–verb agreement in western Japan. As proposed for other sign languages such as As is the case with perhaps all minority languages, American sign language (ASL), the structural features JSL is more or less influenced by the majority lan- of JSL consist of (1) hand shape, (2) handedness (one guage—in this case, by spoken (and written) Japanese. vs. two-handed signs), (3) location of the sign in sign- The extent of influence varies according to several fac- ing space (including relative to a passive base hand), tors. First, JSL varies with the education of the speak- (4) motion (both the hand and arm as a whole, and fin- er (and whether the speaker obtained his or her gers and parts of the hand to affect hand-shape education in the period between the world wars and up change), (5) palm and finger-tip orientation, and (6) to the early 1990s, when the Ministry of Education nonmanual signals. Of all the features entering into (Monbushou) took over control of deaf school admin- JSL signs, the most significant studies have concen- istration and promoted a strictly oral education). trated on hand-shape structure and nonmanual signals. Considerable difference exists, for example, between JSL comprises approximately 50 hand shapes, the the signing of older and younger people. Significant vast majority of which are presumably phonemically numbers of older signers use little or no finger- distinctive. Of these phonemic hand shapes, perhaps spelling, and younger signers tend to allow mouthing half a dozen or so form the basic building blocks of the patterns to play a much greater role in discourse. language. JSL lacks some hand shapes that are present Second, JSL also varies with the speech situation in other sign languages (e.g. ASL /t/), and conversely (more formal situations often result in signing closer possesses some that are absent elsewhere (the JSL to the signed Japanese than to JSL per se), and on the hand shape with an extended middle finger is the judged signing proficiency (and hearing status) of the prime example). In each of these cases, the given hand interlocutor. It would be better, therefore, to speak of a shape is taboo in the other (dominant as well as deaf) continuum from Japanese sign language to signed culture. Japanese. Signed Japanese uses mostly JSL signs, but In contrast to the half dozen basic hand shapes, with Japanese word order and without most of the which are widespread and carry a high functional load, grammatical features of JSL (and, it should be noted, some hand shapes have exceedingly limited distribu- without many of the grammatical features of spoken tion, at times restricted to a single sign and a single Japanese either). Two notable (and highly noticeable) meaning, giving them marginal status. Thus, for features of signed Japanese (as opposed to JSL) are the instance, the hand shape with all but ring finger extend- extensive use of the sign for exist (the inanimate form) ed (and ring finger totally bent) occurs only in the sign as a copular (or linking) verb (Japanese desu) and the for swallow (bird). Virtually the same hand shape, but ubiquitous use of a question particle sign at the end of with only the proximal joint of the ring finger bent, all questions (equivalent to Japanese ka); both signs occurs only in the sign for medicine. The hand shape have an extremely limited use with these meanings in with only the little finger bent is limited to signs with native signing (the latter marking perhaps only a focal- the meaning ‘eight’ (also ‘800’ and ‘8,000’), and the ized question). same hand shape with all the extended fingers bent is The differences between JSL and signed Japanese, limited to the meaning ‘80’. and the limited comprehensibility of the latter to many Quite a few JSL hand shapes have meanings associ- deaf people, has a deleterious impact, especially in the ated with them; for instance, of the single-extended- fields of education and interpreting. Since almost all finger type of hand shape, all except the extended index deaf school teachers can hear, with spoken Japanese as finger have distinctive associated meanings. Thus, an their primary language, and since few of those who extended thumb (with the exception of a few signs with sign have contact with JSL-signing deaf adults, the body contact) indicates a male person (or sometimes signing produced in deaf schools is almost totally gender-neutral person); the extended middle finger signed Japanese. Also, in the field of interpreting, means brother (and, at least in some dialects, an allo- much interpreting is performed by volunteers, with lit- morph for son, alternating with extended thumb); an tle linguistic training or real understanding of the dif- extended little finger means female person, and the 556  JAPANESE SIGN LANGUAGE extended ring finger is a dialectically limited alternate ‘falseness, lie’, and often occurs as an independent of the same sign with extended little finger. gesture. The tongue-in-cheek can also occur simulta- The set of JSL hand shapes is subject to historic neously with another sign to indicate that what is change. For instance, the I love you hand shape of ASL being stated by the hands is in some way false (e.g. to (with thumb, index, and little finger extended) origi- sign feigned absence from home, one would co-sign nally existed only as a partial assimilation of the first absent from home along with a /tongue in cheek/). morpheme to the second morpheme of the compound word parents. More recently, it has spread in the I love Sentence Structure you borrowing, in airplane (and related lexemes, all originally and indigenously with the /Y/ hand shape), JSL is primarily a topic-dominant language, with topic especially among the young, and at least in Tokyo fronting being accompanied by certain nonmanual sig- among some young signers in the sign for sparrow, nals. In the absence of topic fronting, the semantically normally signed with the middle finger extended as neutral sentence order is Subject–Object–Verb (SOV). well. Also, thumb extension, which has been said to be Of the other five possible combinations of S, V, and O, less prominent in JSL than in ASL, is more widely only *VSO is not possible, given the proper context and used in western Japan in variations of lexemes where topicalized element(s). Otherwise, generally the old it does not occur in eastern Japan, such as Hokkaido information is given first, and the new information last. (Japan’s northernmost island), bridge, and sing (all Nonmanual signals (e.g. raising an eyebrow, tuck- three ‘normally’ produced with just index and middle ing in the chin, nodding the head, blinking an eye, tilt- finger extended together). ing the head, etc.) play a major role in syntax. For example, the difference between a wh-question— (where is) Takashi?—and a yes/no question—(are Word Structure you) Takashi?—can potentially be indicated (e.g. in a It has been noted that JSL has no separate features dis- one-word utterance) by a slight difference in such sig- tinguishing traditional parts of speech, such as noun nals. Nonmanual markers are also used in the forma- from verb, or adjective from either. What is distin- tion of topicalization, relative clauses, conditionals, guishable are forms that can be made into predicates negative sentences, and so forth. Individual nonmanu- and forms that cannot. Perhaps better yet is the state- al signal elements may occur in a variety of syntactic ment that JSL lexemes can be grouped into form class- functions, alone or in combination with other non- es, based on whether or not they can be modulated for manual signal elements. One example of a multifunc- agreement, aspect, etc. tional, nonmanual signal is tilting the head (or cocking JSL verbs are divided into those that can be inflected the head), which is used for a range of modal mean- for person and/or number (agreement verbs), where ings including noncommital statements (possibility, inflection is shown by direction of sign motion and/or allegedness); strangeness; approximation (in either palm and finger orientation, and those that cannot (plain number or location); and so forth. Perhaps all these verbs). Other verbal inflections (for instance, aspect and meanings can be grouped in a unified way as deviation tense) have been inadequately studied, but clearly exist. from the expected, stated, or normal. Nonmanual sig- At the very least, a repeated form of a sign means a nals are an area where extensive research continues. repeated, continuing action. Alternating motions, pro- duced with one hand at a time, mean that the repetition Lexicon and Borrowings is distributed over several participants. In general, inflection is used for aspect rather than tense, and the The overwhelming majority of JSL lexemes are indige- two signs finish and middle, which sometimes occur nous. A certain percentage are shared with gestures of with verbs, are indicators of perfective and imperfective the majority hearing populace (e.g. money, girl(friend)). aspect rather than past and present (continuous) tense. Borrowings from other sign languages have played a Nouns and pronouns are inflected for number role in JSL, generally a small one, although in certain and—unlike spoken Japanese—gender, but not case. areas (e.g. computer technical vocabulary) the contribu- Thus, number is indicated in nouns by reduplication, tion of ASL stands out. Examples of nontechnical bor- and in pronouns by multiple or sweeping pointing ges- rowings include communication and analyze. In earlier tures; gender is indicated in nouns, pronouns, and borrowings, ASL also served as the base for the finger- verbs by the extended thumb for a male person and by spelling syllabary—the vowels and the basic ‘consonant the extended little finger for a female person. + /a/’ syllables unmarked by diacritics (e.g. ka, sa, na, Nonmanual signals, which play a major role at the ha, ma, ra, ya, wa) were borrowed from ASL vowels syntactical level, also occur at the lexical level—for and consonants. The exception is ta, with the thumb tip example, the tongue protruding in the cheek indicates provocatively protruding between closed index and 557 JAPANESE SIGN LANGUAGE middle fingers, which is indigenous because the ASL Phonetics, Phonology and Morpho-syntax) 98(1–3). form is taboo. Neologisms are sometimes literal trans- 103–20. Ichida, Yasuhiro. 1991. Shuwa no Kihon-Bunpou [The basic lations of ASL words (e.g. total-communication, with grammar of sign language]. Shuwa-Tsuuyaku no Kisou: /to/ -hand shape in place of ASL /t/), and the ASL Shuwa-Tsuuyaku-Shi wo mezasite [The basics of sign lan- method of initialism have been used (e.g. kadai ‘sub- guage interpreting: aiming at being a sign language inter- ject, theme’ from the /ka/ hand shape incorporated into preter], ed. by Kanda Kazuyuki, 138–50. Tokyo: Daiichi the form of mondai ‘problem, question’), although to a Houki. ———. 1994. Nihon-Shwua no Bunpou to Goi [Japanese sign very limited extent. language grammar and vocabulary]. Nihongogaku 13(2). 25–35. Discourse Features ———. 1998. Nihon-Shuwa no Bunpou [Japanese sign lan- guage grammar]. Gengo 27(4). 44–51. Discourse features such as selective use of signing Minoura Nobukatsu. 1998. Nihonshuwa-hinshi-shiron — space, index pointing, and role shift play a role in JSL Seirisshu-shogo to no taishou wo toosite [An essay on parts discourse. JSL has a lexical (as well as nonmanual) of speech in JSL: by way of comparison with the Salish lan- guages]. Nihonshuwa-gakkai Dai 24kai-taikai Yokoushuu topic marker in-the-case-of, but lacks a simple con- [Proceedings of the 24th conference of the Japan junctive (‘and’), although it possesses a disjunctive but Association of Sign Linguistics]. 46–9. and to-change-the-subject, as well as conjunctions of Morgan, Michael. 1998. JSL ni okeru atama no naname: dan- causation therefore and in-order-to. To a certain watekina kinou kara imi teki na konkyo he [Head tilt in JSL: extent, enumeration can play the role of conjunction, from discourse function to semantic basis]. Nihonshuwa- gakkai Dai 24kai-taikai Yokoushuu [Proceedings of the 24th but this is largely left up to discourse-cohesive features conference of the Japan Association of Sign Linguistics]. such as topic continuity, sequential use of signing 62–5. space, index-finger pointing (to the point in signing ———. 1999. Tracking topic in Japanese sign language dis- space assigned to a given referent or topic, or to a sign course: icon and index (summary of TISLR 7 report, in classifier already in place in that space), etc. Japanese). Japanese Journal of Sign Linguistics 15(1–2). 34–5. Nonmanual signals also play a very important role in Osugi, Yutaka. 1999. The semantic status of ‘INDEX+MALE’ discourse topic tracking. Shared (extralinguistic) in Nihon Syuwa (Japanese Sign Language). Japanese background knowledge and assumptions and linguistic Journal of Sign Linguistics 15(1–2). 1–3. indications of topic shift are very important in dis- Patschke, Cynthia. 1998. A comparative study of body lean in course flow, and repetition of the same sign or sign ASL and JSL contrastive focus. Japanese Journal of Sign Linguistics 14(2). 21–39. sequences are very important to ensure that signer and Smith, Wayne H. 1990. Evidence for auxiliaries in Taiwan sign addressee are on the same page. language. Theoretical issues in sign language research, Vol 1: Linguistics, ed. by Susan D. Fischer and Patricia Siple, References 211–28. Chicago, London: University of Chicago Press. MICHAEL MORGAN Fischer, Susan. 1996. The role of agreement and auxiliaries in sign languages. Lingua (Special issue on Sign Languages: See also American Sign Language Javanese Javanese is one of the Austronesian languages, belong- and Balinese, which are all spoken on or near the ing to the Western Malayo-Polynesian subgroup and island of Java. An ancestor language for Javanese, the Sundic family. The Austronesian languages exhib- Proto-Malayo-Javanic, has been reconstructed by it a high ratio of vowels to consonants. Most root Nothofer (1975). words consist of two syllables, and from these, gram- Javanese is spoken by over 40% of the people in matical variants are derived by means of affixes. Indonesia. It is the mother tongue of 62 million people Austronesian languages use reduplication to indicate in Java, the most populous island in Indonesia. It is spo- the plural and other grammatical concepts. All these ken mainly in central and eastern Java, but is also spoken features are manifest in Javanese. Other Sundic lan- in a thin strip along the north coast of west Java, except guages are Sundanese, Tenggerese, Osing, Madurese, for the area around Jakarta where a form of Malay is 558  JAVANESE spoken. The regional dialect of Solo and Yogyakarta, the There are at least six syntactic forms of word redu- historical centers of Javanese culture, is called Kejawen, plication: and is considered the standard form of Javanese. Whole word reduplication without any phonologi- Javanese has a literature dating back to the eighth cal change, e.g. mangan (eat)  mangan-mangan (eat century and its own Indian-based script, where each informally with other people). symbol corresponds to a syllable consisting of a con- Partial doubling, producing a noun from an adjec- sonant and a vowel. These characters may be viewed tive, e.g. lara (sick)  lelara (sickness), peteng (dark) on www.skypoint.com/~gimonca/huruf-jawa.html.  pepeteng (darkness). The reduplicated fragment is Old Javanese is the language of texts written in the a prefix consisting of the first phoneme of the root pre-Islamic era. Nowadays, Roman script is more word followed by e. commonly used, but Javanese can also be written in Partial doubling + an, e.g. tembung (word)  Arabic script. The literature of Javanese has been cat- tetembungan (wording, expression). Whole word alogued by Pigeaud (1967). reduplication of a verb with a phonological change, The Javanese vowels are a, e, i, o, and u, and there e.g. bali (return)  bola-bali (to and fro), mubeng (go are open(long) and closed(short) forms of each. There around)  mubang-mubeng (beat around the bush). is also a variant of the open a when it is the final syl- Lexical doubling: the root words are already dou- lable, pronounced half way between o and a, and a bled, since the single form does not exist. neutral (pepet) e, as in the English word open. The Morphological doubling—a new meaning is consonants are shown in Table 1 (Robson 1992). formed in contrast to the nondoubled one. Root words are typically disyllables of the form Robson (1992) lists the following semantic cate- (C1) V1 (C2) V2 (C3), where (C1), (C2), and (C3) are gories created by reduplication: optional consonant clusters. Allowable consonant Do something at leisure, e.g. mlaku (walk)  clusters include mb, nd, ndh, nj, and nng, which can all mlaku-mlaku (go for a stroll). occur in the initial position. There is a light stress on Do something repeatedly, e.g. njerit (shriek)  the second last syllable, or the final syllable when the jerit-jerit (go on shrieking). second last syllable contains a neutral e. This light Interrogative pronouns given indefinite meaning, stress does not occur when a suffix is added. e.g. sapa (who)  sapa-sapa or sapaa (anyone). Grammatical variants of a root word may be com- Mild exasperation, e.g. mentah-mentah iya dipan- posed by affixation, reduplication, or combination. gan (even though it is unripe he still eats it). Affixes, which may be prefixes, suffixes, or infixes, Plurality with diversity for both adjectives and are more common in Javanese and Tagalog than in nouns, e.g. gedhong dhuwur-dhuwur (buildings all Malay. Affixes may result in the production of either a more or less high). noun or a verb. Sometimes, the surface forms of affix- Doing something together, e.g. omong-omongan es that result in the formation of a noun are identical to (to chat together). those that result in the formation of a verb. The lists of To compete in, e.g. gelis-gelisan (to see who is Javanese affixes used for noun and verb formation, fastest at running). given by Suharno (1982), are shown in Tables 2 and 3, Plurality, e.g. wet–wet (trees). respectively. An important characteristic of Javanese is the Adjectives can take affixes, e.g. cukup (enough)  speech decorum of the language, where different lev- cukupan (more or less enough), dhuwur (high)  els or stylemes of speech are used depending on the kedhuwuren (too high). Adjectives can also be formed relative social status of the two speakers. This system from nouns with affixes, e.g. jamur (fungus)  has been in existence since the sixteenth century, and jamuren (moldy). is a legacy of the feudal system left behind by the old Hindu court tradition. The speech levels are not differ- TABLE 1 The Javanese Consonants ent languages, but manners of speaking that vary according to the relationship between the speaker and Unvoiced Voiced Nasal the addressee. Each level within the language has its Labial P B M own characteristic set of vocabulary. Dental T D N The three main levels of modern Javanese are krama, Retroflex TH DH Palatal C J NY madya, and ngoko—high, middle, and low, of which Velar K G NG krama and ngoko are most commonly used. Someone Liquids R L of high status speaking to someone of low status will Semivowels Y W use ngoko, while the other will use the (more formal) Sibilant S krama (pronounced kromo). The basic level ngoko is Aspirant H used between friends and equals. Ngoko means the 559 JAVANESE TABLE 2 Javanese Affixes Used in Verb Formation Affix Usage Examples moro-, mer- Prefix forming a verb from a simple word Dayoh (guest)  morodayoh, merdayoh noun. Not productive. (pay a visit). kapi- Prefix forming a verb from a simple word Dereng (strong wish)  kapidereng noun. Not productive. (overeager). kami- Prefix normally occurring in conjunction Seset (peel)  kamisesetan (suffer from with the suffix -an to form a verb from skin peel). another simple word verb. Not productive. kumo- Prefix forming a verb from a simple word Wani (dare)  kumowani (recklessly verb or a simple word noun. Not dare). productive. a- Prefix forming a verb from a simple word Rupo (appearance)  arupo (to have the noun. appearance of). ma- Prefix forming a verb from a simple word Guru (teacher)  maguru (to learn from noun. This sometimes involves a sound a teacher); sembah (respect)  change: a nasal consonant must be manembah (pay respect to); aju homorganic with the initial phoneme of the (progress)  maju (move forward). root word. Also applies to n-. n- Prefix forming a verb from a simple word Tules (write)  nules (write); becak verb or simple word noun. Nasalization is (pedicab)  mbecak (ride in a pedicab); found with nearly all transitive verbs. The gule (stew)  nggule (make stew); inep rules for nasalization of a root word are:  nginep (spend the night). initial p  m; b  mb; t  n; d  nd; th  n; dh  ndh; c  ny; j  nj; k  ng; g  ngg; r  ngr; l  ngl; s  ny; w  m or ngw. ka-, ke- Prefix forming a verb with a passive Junjong (lift)  kajunjong (lifted); Jupoq meaning from a simple-word verb. (take)  kejupoq (taken). di- Prefix forming a verb with a passive Tandor (plant)  ditandor (planted); tuku meaning from a simple-word verb. (buy)  dituku (bought). taq- Prefix related to aku (I) forming a verb Dol (sell)  taqdol (sold by me). with a passive meaning from a simple word verb. koq- Prefix related to kowe (you) forming a verb Gawe (make)  koqgawe (made by you). with a passive meaning from a simple word verb. -in- Infix forming a verb with a passive Sawang (watch)  sinawan (seen); meaning from a simple word verb. barong (accompany)  binarong (accompanied). -um- Infix forming a verb from a simple word laku (walk, gait)  mlaku (walk); verb or a simple word noun. Usually kumrisik (make a rustling sound). contracted to initial m-. -r- Infix forming a verb from a simple word Tutol (spot)  trutol (to be spotty). noun. -i Suffix forming a verb from a composed- Taker (measure)  nakeri (measure word verb, itself formed by one of the repeatedly); tugel (break)  ditugeli prefixes n-, di-, taq-, koq-. (broken into pieces); tembong (mention)  koqtembungi (you ask for a particular thing). -an, ka- -an Suffix or affix combination that may Lunggoh (sit)  lungguhan (sitting); lebu form a verb from a simple word verb. (enter)  kelebon (intruded). Continued. 560  JAVANESE TABLE 2 Continued Affix Usage Examples -en Suffix forming an imperative verb from a Jupoq (take)  jupuqen (Take it!); simple word verb. pangan (eat)  panganen (Eat it!). -o Suffix forming an imperative verb from a Turu (sleep)  turuo (Sleep!); gowo single word verb, or forming a (carried)  gawoo (even if carried). hypothetical verb from a verb formed using di-, tag, koq, -in-. -ake Suffix, always in combination with Maguro (learn from a teacher)  one of ma-, n-, ka-, di-, taq-, koq-, -in-, maguroqake (send someone to learn from forming a verb from a verb. a teacher). Njupoq (take)  njupoqake (take for someone); sinugoh (served with)  sinugohake (served to somebody). TABLE 3 Javanese Affixes Used in Noun Formation Affix Usage Examples pi-, pang- Prefix forming a noun from a simple word Anggo (use)  panganggo (clothing); verb or noun. utang (debt)  piutang (credit). ka-, ke- Prefix forming a noun from a simple word Weroh (see)  kawruh (knowledge); verb. In combination with the suffix -an it lurah (village mayor)  kelurahan forms a verb from a simple word noun. (village mayor’s office). -an Suffix forming a noun from a simple word Jaran (horse)  jaranan (hobby horse); verb or noun. It may be used in tegal (nonirrigated field)  tegalan (area combination with pi- to form a noun from of such fields); Tulung (help)  a verb or a noun. pitulungan (assistance). -e, -ne Suffix forming a definite noun from a Jaran (horse)  jarane (the horse); tuku simple word verb or noun. (buy)  tukune (the purchase). -ku Suffix related to aku (I). Kembang (flower)  kembangku (my flower). -mu Suffix related to kowe (you). Omah (house)  omahmu (your house). ngoko form of I, while krama means marriage. The Madya (the old servant to the girl): Kula mpun madya level consists of krama containing certain words nedha sekule (I have eaten the rice). shortened and with ngoko style affixes. It is often used Another form, basongan, is only used in the kratons among strangers. There are about 900 words exclusive (Sultan’s palaces) of Jogjakarta and Solo. The lan- to krama, and also a few hundred modesty words called guage of religion is called ‘Jawa Halus’ (refined krama inggil (inggil means high). These words can be Javanese) and many words are based on Sanskrit or mixed into either ngoko or krama as required. Krama Kawi, but a diminishing number of people are able to inggil words are used when one speaks about the per- use this form of the language. The number of levels son, actions, or possessions of someone to whom may vary according to regional dialect, and between respect is due—either addressed or referred to. These urban and rural areas (Geertz 1960). A sample of words cannot be used of oneself. Examples of the use of words that differ at four different levels is shown in the three levels of Javanese speech are given by Robson. Table 4. Ngoko (girl to her younger sister): Aku wis mangan The Javanese personal pronouns are shown in segane (I have eaten the rice). Table 5. Krama (girl to her uncle): Kula sampun nedha For we, ngoko uses awake dhewe, while both sekulipun (I have eaten the rice). ngoko and krama can use the Indonesian loanword Krama with krama inggil (girl to her uncle about kita. The second- and third-person pronouns are her father: Bapak sampun dhahar sekulipun (Father rarely used, and are generally replaced by kinship has eaten the rice). terms, titles, or proper names. For example, a woman Ngoko with krama inggil (girl to her sister about may be addressed as Bu (literally, mother). A pronoun her father): Bapak wis dhahar segane (Father has eaten may be omitted altogether if the referent’s identity is the rice). understood. 561 JAVANESE TABLE 4 Words that Differ at Four Different Levels English Ngoko Krama Madya Krama Inggil Allow Kareben Kajengipun Kajenge Kersanipun Obedient Gugu Gega Dhaharatur Ngestokaken dhawun Speak Celathu Wicanten Canten Ngendika Wear Enggo Engge Ngge Agem You Kowe Sampeyan Samangdika Panjenengan TABLE 5 The Javanese Personal Pronouns English Ngoko Madya Krama Krama Inggil I Aku __ Kula Dalem You Kowe Samang Sampeyan Panjenengan He, She Dheweke __ Piyambakipun Panjenengane, panjenenganipun TABLE 6 Aspect Markers in Javanese preceding the verb. The list given by Robson is given Ngoko Krama Meaning in Table 6. The normal word order within the sentence is Aja Sampun Don’t Arep, bakal Badhe Will subject–predicate. There is no copulative verb, e.g. Durung Dereng Not yet klambiku reget (my shirt is dirty). No changes are found Isih Taksih Still in nouns or verbs for number, case, or gender. Lagi Saweg In the process of doing Meh — Almost References Meksa — Even so, still Ora Mboten Not Geertz, C. 1960. The religion of Java. New York: Free Press. Padha Sami Also; indicates the plurality Pigeaud, Theodore G. 1967. Literature of Java: catalogue of the subject performing raisonne of Javanese manuscripts in the library of the the action University of Leiden, Vol. 1. Leiden: Leiden University Press. Sok — On occasion, ever Nothofer, B. 1975. The reconstruction of proto-Malayo- Tansah — Always, constantly Javanic. The Hague: Martinus Nijhoff. Wis Sampus Already Robson, Stuart. 1992. Javanese grammar for students. Monash papers on South East Asia 26, Monash University, Clayton, Victoria, Australia. Suharno, Ignatius. 1982. A descriptive study of Javanese, Verbs are not inflected to denote tenses, but Pacific Linguistics, Series D, No. 45. instead auxiliary words are used as aspect markers MICHAEL PHILIP OAKES Jespersen, Otto Otto Jespersen was a markedly independent scholar. he developed what came to be called an antalphabetic He did not adopt the terminology or the viewpoints of system for describing speech sounds using letters and others, he belonged to no ‘school’ of linguistics, and he numbers to represent the articulating organ in the created none, but over his long and active academic mouth (lips, tongue, etc.), the exact place of articula- career he created an astonishing body of work in a wide tion in the mouth, and the degree of openness of the variety of linguistic subfields, virtually all of which vocal tract in the production of a sound. The result constitutes a rich and continuing source of information analyzed sounds into their components in a manner for later scholars and an inspiration for many. more akin to modern distinctive feature analysis than Jespersen first became known outside his native to traditional phonetic transcription. For purposes of Denmark for work in articulatory phonetics. In 1889, transcribing Danish and its dialects, Jespersen invented 562  JESPERSEN, OTTO Dania, a dialect alphabet still in use. He also published the time in both Europe and America in the work of a detailed book on phonetics in Danish in 1897; it was Ferdinand de Saussure and Franz Boas, with their soon translated into German, making it more widely emphasis on synchronic study without recourse to his- accessible to scholars. Little was new here, but torical information, Jespersen always maintained that Jespersen’s emphasis on first-hand observation and the the science of language was essentially historical scope of his examples made the texts invaluable in because ‘a language or a word is . . . a result of previ- promoting the empirical study of speech sounds. ous development and at the same time . . . the starting- In a campaign to reform Danish foreign language point for subsequent development’ (Language: its teaching and reacting against traditional pedagogical nature, development and origin, 1921:7). methods that stressed rote learning, grammatical rules, The MEG was Jespersen’s life work, his magnum and translation, Jespersen became an advocate of what opus, and he saw it as the grammatical equivalent of the he considered a more ‘natural’ approach, emphasizing New English dictionary on historical principles, later the spoken language, the use of meaningful and inter- known as the Oxford English dictionary (OED). The esting texts for written language, and the learning of volume on Sounds and spellings and another on grammar by observation and induction. How to teach Morphology were quite traditional and are of little a foreign language (1904) was influential not only in interest today, but the five Syntax volumes remain an his own country but in England and particularly in invaluable resource, with an abundance of citations and North America. Indeed, Leonard Bloomfield in his great insights into the nature of English syntax, cited by most important book, Language (1933), presented twentieth-century linguists as diverse as Noam Jespersen’s ‘direct’ method, and elements of this later Chomsky, Eugene A. Nida, and W. Nelson Francis. made their way into the so-called ‘Army method’ In the MEG and in shorter books that presented developed by Bloomfield and other American linguists similar material on syntax (e.g. The philosophy of for foreign language teaching in the years immediate- grammar, 1924; Essentials of English grammar, ly preceding and then during World War II. The focus 1933), Jespersen introduced the notion of ranks, a on spoken language, on meaningful material in func- hierarchy of levels of subordination and dependency tional context, and a more inductive approach to the among the words of phrases (e.g. in a phrase such as acquisition of grammar all characterized both the audi- very young child, the highest rank is held by child, olingual method used in the United States during the which Jespersen would call the primary; young has the 1950s and 1960s and its successor, sometimes referred next highest rank, the secondary, and very is the low- to as the communicative approach. est, the tertiary). Junction referred to subordinate con- An early idea and a fundamental theme that structions, nexus to predicative structures; so the Jespersen pursued for many years was the notion of barking dog displayed junction, the dog is barking progress in language. Based on aspects of the history nexus. Because Jespersen always held to a close con- of the English language, he maintained that as a lan- nection between linguistic form and content, his gram- guage changes, original inflectional endings will dis- mar also made use of notional categories, universal appear and be replaced by principles of fixed word categories of meaning. order. Challenging prevailing nineteenth-century Although work in syntax was somewhat uncommon beliefs that languages decay over time, Jespersen con- in the linguistics of Jespersen’s day, many of his ideas sidered such changes as progress. The modern forms, are precursors to aspects of modern linguistic theories. he argued, are shorter, requiring less production effort; The relation of mental categories and linguistic cate- they are fewer, requiring less memory; they are more gories is certainly central to contemporary approaches, regular morphologically and syntactically; they avoid and the ideas Jespersen sought to develop with ranks, redundancy; and their regular word order assures junction, and nexus are fundamental to modern theo- understanding. Jespersen admitted that his conclusions ries such as valence grammar, dependency grammar, were based on information from very few languages, and some types of functionalgrammar. but nevertheless he claimed that a progressive ‘tenden- Jespersen’s most technical work on syntax was the cy towards grammatical simplification is a universal small book Analytic syntax (1937), in which he devel- fact of linguistic history’ (Language: its nature, devel- oped a notational system for representing ranks, junc- opment and origin, 1921:366). Further research has tion, nexus, and other syntactic constructions. The supported neither the older theory of decay nor system was formal and complex, and at the time, few Jespersen’s theory of progress. responded to it. But as syntax and formalism became In 1909, Jespersen began publication of what would more and more a part of twentieth-century linguistics, become his most extensive study of the English lan- the book was reprinted several times in the later twen- guage, A modern English grammar on historical prin- tieth century, a tribute to Jespersen’s lifelong interest ciples (MEG). Contrary to an emphasis emerging at in syntax and to his lasting influence. 563 JESPERSEN, OTTO Unlike linguists of more modern times, Jespersen’s grammatical structure, the epitome of progress in lan- theoretical ideas often appeared not in highly technical guage. Jespersen called his language Novial (nov = monographs aimed at his peers, but rather in books ‘new’ + i ‘international’ + a ‘auxiliary’ + l ‘language’). directed toward the educated reading public. Books such But the movement was fractured and no consensus as Language: its nature, development and origin (1921), developed. Jespersen’s Novial disappeared, sharing The philosophy of grammar (1924), and Mankind, the fate of other created languages. nature, and individual from a linguistic point of view There is scarcely a linguistic topic of interest today (1925) were described in popular journals as ‘fascinat- that does not appear at some point in Otto Jespersen’s ing’, and many members of the public came to know voluminous works, which number more than 800 about linguistics by reading Otto Jespersen. The most scholarly items in print. While some of his lines of important idea that Jespersen brought to both the public investigation seem dated and some of his theories have and to other linguists was the rejection of traditional been challenged by more extensive data drawn from a grammar that set forth standards of correctness. In its wider variety of languages than he used in his work, it place, he proposed a living grammar, based on direct is nevertheless often fruitful and inspiring to go back observation of contemporary spoken usage, a grammar to his writings on foreign language teaching, language that was founded on the past but continually changing, a use, and, especially, English syntax. grammar that was systematic but not without the irregu- larities we should expect in anything human. Biography Jespersen tried to relate current usage to the users of the language. In Language: its nature, development Otto Jespersen was born on July 16, 1860 in Randers, and origin, six chapters are devoted to ‘The child’. Jutland, Denmark. He entered the University of Here, Jespersen explored at length issues in the devel- Copenhagen in 1877 and studied law; in 1881, he opment of language in children, treating sounds, turned to the study of languages. For seven years, while words, and grammar, as well as possible influences a student, he worked as a shorthand recorder in the that children’s language acquisition might have on lan- Danish House of Parliament. He received a master’s guage change. This pioneering effort to relate the degree in 1887 with major in French, secondary con- nature and the universals of human language and of centrations in English and in Latin (the latter obligatory language change to children’s language acquisition for language majors). He traveled for a year, meeting anticipated major themes in linguistics of the late linguists in England (Henry Sweet, James Murray), twentieth and early twenty-first centuries. Germany (Karl Brugmann, August Leskien, Eduard The same book contained another chapter well in Sievers), and France (Paul Passy, Jules Gillíeron). He advance of its times, a chapter titled ‘The woman’, in returned to Copenhagen, and taught English and French which Jespersen presented and analyzed various in private schools while working on the doctorate. He accounts of differences between women’s and men’s defended his doctoral dissertation ‘Studies on English speech. Modern gender studies of language use have case’ in 1891. From 1891 to 1893, he served as an certainly supported Jespersen’s premise that women unpaid Privatdocent (instructor) at the University of and men use language differently, often for reasons Copenhagen, teaching Old English and Chaucer. In having to do with societal differences. But Jespersen’s 1893, he was appointed Professor of English Language ideas were reflective of his times and place, and his and Literature, University of Copenhagen. In 1899, he chapter on ‘The woman’ is also a source of statements was elected to membership in the Royal Danish that modern scientific scholarship soundly rejects, for Academy of Sciences and Letters. He visited the United example, ‘Men will certainly with great justice object States and lectured at Congress of Arts and Sciences in that there is a danger of the language becoming lan- St. Louis in 1904, and received the French Volney Prize guid and insipid if we are always to content ourselves for Growth and structure of the English language in with women’s expressions . . .’ (p. 247). 1905. He was Dean of Faculty of Arts, University of Jespersen’s dedication to the international auxiliary Copenhagen from 1904 to 1906. In 1909–1910, he language movement is sometimes treated as an odd again visited the United States, lecturing at the avocation, but this interest was a rational extension of University of California and Columbia University. In some of his most fundamental linguistic work, espe- 1920–1921, he was Rektor (Vice Chancellor), cially his advocacy of foreign language learning for University of Copenhagen. Jespersen retired from the communication, his concern with universals of lan- University of Copenhagen in 1925, and was elected to guage, his theories on syntax, and his principles of honorary membership in the Linguistic Society of progress in language change. He maintained that an America in 1926. In 1936, he was President, Fourth artificial language constructed on scientific principles International Congress of Linguists, Copenhagen. He should be easy to learn and efficient to use, simple in received honorary degrees from Columbia University in 564  JONES, SIR WILLIAM 1910, St. Andrews University, Scotland in 1925, and ———. 1921. Language: its nature, development and origin. The Sorbonne, University of Paris, in 1927. He died on New York: Holt; London: Allen & Unwin, 1922. ———. 1924. The philosophy of grammar. London: Allen & April 30, 1943 in Reskilde, Denmark. Unwin and New York: Holt. ———. 1925. Mankind, nation and individual from a linguistic point of view. London: Allen & Unwin. References ———. 1928. An international language. London: Allen & Aitchison, Jean. 1991. Language change: progress or decay?, 2nd Unwin; New York: Norton, 1929. edition. Cambridge, New York: Cambridge University Press. ———. 1933. Essentials of English grammar. London: Allen & Baker, C.L. 1995. English syntax, 2nd edition. Cambridge, Unwin and New York: Holt. MA: MIT Press. ———. 1933. Linguistica: selected papers in English, French Falk, Julia S. 1992. Otto Jespersen, Leonard Bloomfield, and and German. Copenhagen: Levin & Munksgaard. American structural linguistics. Language 68. ———. 1937. Analytic syntax. Copenhagen: Levin & Haislund, Niels. 1966. Otto Jespersen. Portraits of linguists: a Munksgaard and London: Allen & Unwin. biographical source book for the history of western linguis- ———. 1941. Efficiency in linguistic change. Copenhagen: tics, 1746–1963, Vol. 2, ed. by Thomas A. Sebeok. Munksgaard. Bloomington and London: Indiana University Press. Juul, Arne, and Hans F. Nielsen (eds.) 1989. Otto Jespersen: Jespersen, Otto. 1889. The articulations of speech sounds rep- facets of his life and work. Amsterdam and Philadelphia: resented by means of analphabetic symbols. Marburg in John Benjamins. Hessen: Elwert. Juul, Arne, Hans F. Nielsen, and Jørgen Erik Nielsen ———. 1894. Progress in language, with special reference to (eds.)1995. A linguist’s life: an English translation of Otto English. London: Sonnenschein and New York: Macmillan. Jespersen’s autobiography with notes, photos and a bibliog- ———. 1904. How to teach a foreign language. London: raphy. Odense, Denmark: Odense University Press. Sonnenschein and New York: Macmillan. Large, Andrew. 1985. The artificial language movement. ———. 1905. Growth and structure of the English language. Oxford, UK and New York: Basil Blackwell. Leipzig: Teubner and New York: Stechart. Lightfoot, David. 1999. The development of language: acquisi- ———. 1909. A modern English grammar on historical princi- tion, change, and evolution. Oxford, UK and Malden, MA: ples, 7 vols. Heidelberg: Winter, 1914, 1927, 1931; Blackwell Publishers. Copenhagen: Munksgaard, 1940, 1942, 1949. JULIA S. FALK Jones, Sir William Sir William Jones, also known as ‘Oriental’ Jones, was the independent Whig aristocrats. Tutoring financed a seminal linguist of the English Enlightenment, pio- his studies at Oxford, where he began studying the neering the study of comparative linguistics. Through same year. his singular knowledge of western and oriental lan- At the Spencer family seat at Althorp, he met Count guages, he discovered the root relationships among Carol Revicsky, an enthusiast of Arabic and Persian, them that would later be defined as the Indo-European who was Polish ambassador to Great Britain and whose language family. His devotion to and authoritative homeland lay in the shadow of the Ottoman Empire. publications on Sanskrit and Arabic languages togeth- Jones avidly added knowledge of these languages and er with Hindu and Moslem religions and cultures cultures to the western classical ones he already knew. made him a pioneer in introducing oriental culture to Arabic and Persian were of great importance in the ori- Europe. His admiration and respect for them defied ental world, the first as the religious tongue of the and gradually began to reverse centuries of western region and the latter as its international one. religious and cultural ignorance and ridicule of them. In 1770, when he was 24, Jones translated from the Jones studied with great success at Harrow College Persian, the Histoire de Nader Chah. He wrote in and Oxford University, where he held the Bennet [sic] French, the western world’s international language at fellowship. His intellectual prowess allowed him to the time, publishing the book in 1773 as The history of elevate himself in society. His knowledge of the clas- the life of Nader Shah, King of Persia, a tyrannical sics at Harrow so impressed Jonathan Shipley, Bishop Persian ruler in India, who lived from 1688 to 1747. In of St. Asaph (Wales), that he recommended the young 1771, Jones published A grammar of the Persian lan- man to become the tutor in 1764 to the children of the guage, following this with pioneer translations of vastly wealthy Earl of Spencer, a leading member of poetry from oriental languages. 565 JONES, SIR WILLIAM To establish his economic and professional inde- from his own financial resources. As the first president pendence, he left the Spencer household in 1770 to of the society, his annual presidential speeches became study law in London at the Middle Temple. He became significant communications of his ideas and findings. a lawyer and circuit magistrate in the districts of The speech of 1786 (Third Anniversary Discourse) Oxford and Wales five years later and a commissioner traced the common elements of western and oriental for bankrupts. Now a noted young scholar, Jones was languages, the basis whereby others later identified the elected in 1773 to the Literary Club of Dr. Samuel Indo-European family of languages. It also contributed Johnson, the renowned sage of the English language. to the development of comparative linguistics. Later Convening regularly at the Turk’s Head Inn, club discourses established the basis for establishment of members included the historian Edward Gibbon, econ- comparative religious studies. omist Adam Smith, painter Joshua Reynolds, and Jones also began a series of translations of epic statesman Edmund Burke. A sympathizer of the move- works from oriental literature. The introduction of ment for independence of the English colonies in these classics into the English-language world and America, Jones was also a friend of Benjamin Europe would alter not only the perception of the Franklin, having known him from the Shipley family. weight and accomplishment of eastern civilizations Jones visited him three times in Paris after the but also influence directions and developments in American Declaration of Independence in 1776, modern European culture itself. where Franklin had become the young American During 1784–1785 he translated nine poem-hymns to republic’s representative. Hindu deities. In 1789, he translated Gitagovinda (‘The Jones’s writings now advanced into the fields of law song of Govinda’), a lyric poem that was one of the last and politics. He especially concentrated on estate and Sanskrit devotional texts (bakhti) written. The increasing commercial law, now a legal adviser to the Spencers use of vernacular languages in the twelfth century began and other aristocrats, with a work on Athenian inheri- to obliterate the use of Sanskrit, somewhat as in Western tance law in 1779 and An essay on bailments in 1781. Europe vernacular languages, such as Florentine Italian, In addition, he wrote pamphlets and smaller works began to obscure the use of Latin. In the same year, he related to numerous issues regarding individual and translated the romantic verse play, the Shakuntala (also political rights, proposing a plan for reconciliation Sakuntala). This was the masterpiece of the greatest between Britain and America based on common busi- writer of classical Sanskrit, Kalidasa, a fifth-century res- ness interests. ident of the Gupta court. In 1792, he oversaw the first To refine and deepen his oriental studies and to printing in Sanskrit of Kalidasa’s Ritusamhara, and advance his professional position, Jones had long translated from Arabic Al Sirajiyyah. sought a judgeship on the bench of the English courts Given the vast range, quantity, and originality of the in India that especially oversaw operations of the East work of Sir William Jones, it seems incongruous that one India Company. In 1783, he succeeded in being can say he left his work incomplete at the time of his appointed an associate judge of the Supreme Court of death in 1794. Nevertheless, he left many projects still to Judicature for Bengal (now Bangladesh) in Calcutta. complete and, even more, a future promise to satisfy. The same year, he was knighted by King George III, In recent decades, there has been much controversy and he married Anna Maria Shipley, daughter of regarding the focus and treatment of eastern cultures Bishop Shipley. They had known and been fond of that Jones introduced. ‘Orientalism’ has been denigrat- each other for many years, but he vowed only to ed as colonialist, manipulative, and self-serving. marry once he had established his financial independ- Nevertheless, one of the central issues of modern histo- ence. En route to the Orient, he wrote down a plan for ry has been the emergence of relations between East his research, which he styled ‘Objects of enquiry dur- and West as vital global concerns. It is indisputable that ing my residence in Asia’. It outlined how he would Sir William Jones played a key initial role in making the survey Hindu and Moslem languages, literatures, reli- West aware of the East and of the need to respect the gion, business, technology, and social and physical Orient’s vast cultural accomplishments and to recognize sciences. That he actually accomplished the study of the manifold characteristics that East and West share. this array of topics in the remaining 11 years of his life vaulted him to the realm of outstanding seminal Biography scholars. Founding the Asiatick Society in Calcutta in 1784, Sir William Jones was born in London in 1746. He this association became the vehicle for propagating his was a student of Harrow College in 1753–1764, and of voluminous research. He published Asiatick miscel- University College, Oxford in 1764–1768. He was pri- lany in 1785 and, from 1788 until his death Asiatick vate tutor to the children of the first Earl of Spencer in researches. He supported his research and publishing 1765–1770. He studied law in Middle Temple, London 566  JUBA AND NUBI ARABIC in 1770–1773. Jones was elected Fellow of the Royal York University, April 21, 1994]. New York: New York Society, London in 1772, and was elected as a member University Press. Franklin, Michael J. 1995. Sir William Jones. Writers of Wales of the Literary Club of Samuel Johnson, London in series. Cardiff: University of Wales Press. 1773. He also practiced as a lawyer in London; as a ———. 1995. Sir William Jones: selected poetical and prose circuit judge in Oxford and South Wales districts; and works. Cardiff: University of Wales Press. as commissioner of bankrupts in 1775–1783. Jones Jones, Sir William. 1785. Asiatick miscellany. was knighted in 1783 and married Anna Maria Shipley ———. 1788–1794. Asiatick researches. ———. 1771. Dissertation sur la littérature orientale. the same year. He was appointed to the Supreme Court ———. 1781. Essay on the law of bailments. of Judicature for Bengal, Calcutta in 1783–1794. He ———. 1789. Gita Govinda by Jayadeva, translation from founded the Asiatick Society of Bengal, Calcutta in Sanskrit. 1784. Sir William Jones died as a result of liver ———. 1771. Grammar of the Persian language. inflammation, and was buried in Calcutta in 1794. ———. 1770. Histoire de Nader Chah, translated from Persian; translated into English as The history of the life of Nader Shah, King of Persia. 1773. References ———. 1789. Shakuntala by Kalidasa, translation from Cannon, Garland. 1990. The life and mind of Oriental Jones: Sanskrit. Sir William Jones, the father of modern linguistics. ———. 1807. The collected works of Sir William Jones, 13 Cambridge: Cambridge University Press. vols. London: J. Stockdale; reprinted, London: Curzon; and Cannon, Garland, and Kevin R. Brine (eds.) 1995. Objects of New York: New York University Press, 1993. inquiry: the life, contributions, and influences of Sir William Said, Edward. 1979. Orientalism. New York: Random Jones (1746–1794) [Papers from Sir William Jones House. Symposium on the occasion of bicentenary of birth, New EDWARD A. RIEDINGER Juba and Nubi Arabic Juba and Nubi Arabic are two contemporary related and form of this Arabic pidgin remains unknown. Its Arabic-based Pidgins–Creoles descending from an target language seems to have been a mixture and an Arabic military pidgin that developed in southern approximate form of Egyptian and northern/western Sudan following the annexation of Sudan by the Sudanese Colloquial Arabic. It is not certain whether a Turkish–Egyptian Government in 1820. While the single variety spread all around southern Sudan or northern and central parts of Sudan were already whether various varieties emerged simultaneously in largely Arabized at the end of the eighteenth century, the different camps. But the available historical and the southern part of Sudan laid behind the southern contemporary data testify to a very similar restructur- borders of Arabic expansion due to physical barriers. ing process and indicate that this Arabic-based pidgin Thus, the conquest of southern Sudan between 1839 stabilized in a very short span of time. and 1841 instituted the first trade contacts between the Following various military events at the end of the north and the southern hinterlands. The subsequent nineteenth century, the southern Sudanese Arabic pid- establishment of military and trade camps and the gin first known as Bimbashi Arabic (from Osmanli, development of a large-scale slave trade between 1854 Bimbashi ‘officer’) developed in three further vari- and 1889 led to major social upheavals and to the eties: Turku in Chad, Nubi in Kenya and Uganda, and emergence of an Arabic pidgin in this previously non- Juba Arabic in southern Sudan. Turku (from Arabic, Arabic-speaking area. Many members of the local Turuk, i.e. ‘Turks’) was brought from Bahr al Ghazal southern population (sometimes up to 25%) were cap- to Chad at the end of the nineteenth century by some tured as slaves or employed in the camps, whose pop- Sudanese soldiers from Rabeh’s army. It remained a ulation could number up to 10,000 people. In these pidgin and became one of the trade languages of Chad camps, the southern population soon outnumbered the and neighboring countries in the first decades of the Arabic-speaking population. Coming from heteroge- twentieth century. No present-day Turku is recorded; neous linguistic and ethnic backgrounds, this southern also, some kind of Arabic functions as interethnic lin- detribalized population quickly adopted a pidginized gua franca in this area. Nubi (or Ki-Nubi) was brought form of Arabic as its lingua franca. The exact origin to Kenya–Uganda in 1888 by the retreating Sudanese 567 JUBA AND NUBI ARABIC soldiers of Amin Pacha’s army. Those Arabized and passes a wide range of individual variations. In spite of Islamized soldiers stayed and formed a new ethnic a short time of common history and a century of sepa- groups till locally known as Nubi and speaking an rate development, Nubi and Juba Arabic are structural- Arabic-based Creole, ki-Nubi, as their mother tongue. ly closed and mutually intelligible, which may indicate Cut off from further Arabic influence, Ki-Nubi devel- that their common ancestor stabilized before their split. oped in a dominant African/Swahili environment and The lexicon derives mainly from Arabic roots, but with is not mutually intelligible with standard Arabic (both important phonological restructuring and the addition Colloquial or Classical Arabic). Most East African of a number of borrowed words from African vernacu- Nubis are urban dwellers and are at least bilingual. In lars. The morphology of Arabic has been lost (such as southern Sudan and more specifically in Equatoria, verbal flexions, derived verbal and nominal forms). Bimbashi Arabic continued to develop during the first Both languages used invariable verbal and nominal decades of the twentieth century first in the military stems and independent grammatical markers for garrisons and urban centers and then spread as the expressing grammatical categories such as tense, main lingua franca in the heterogeneous rural areas aspect, persons, definiteness, comparative, etc. Like with minimal contact with standard Arabic. Today, it is for most Pidgins and Creole languages, the attention of becoming the mother tongue of a growing number of linguists focused on the supposed origin of these lin- children in southern urban areas due to mixed neigh- guistic features. In this respect, Arabic-based boring and interethnic marriages. It thus functions Pidgins–Creoles appear to share many similarities with both as a first and second language. It became known other non-Arabic-based Pidgins–Creoles especially as Juba Arabic from Juba, the capital of Equatoria. regarding the verbal system. Their restructuring has Since 1956 and more intensively since the 1980s, Juba gone further than any other Arabic Colloquial varieties, Arabic speakers have been more exposed to Colloquial including the most peripheral ones, and interferences Sudanese Arabic and Classical Arabic through mass from neighboring African languages have been pointed media, schooling, urbanization, and migration to out at various levels. The three above-mentioned northern Sudan. Therefore, many Juba Arabic speak- Arabic-based Pidgins–Creoles might not have been the ers can shift from a more Creolized variety to a more only Arabic-based existing ones but their history testi- Colloquial one according to the setting. However, the fied to a radical context of emergence compared to antagonistic relationship between the North and the other situations of contact. The maintenance of Nubi South fosters the identity function of Juba Arabic, and Juba Arabic as specific Pidgin/Creole varieties also which is perceived as expressing and symbolizing an indicates the crucial role played by identity factors. ‘African identity’. Ki-Nubi and Juba Arabic are mainly spoken lan- guages. They are not taught and have neither been stan- References dardized nor normalized through an official script. Heine, Bernd. 1982. The Nubi language of Kibera: an Arabic Ki-Nubi is an ethnic language for approximately Creole. Berlin: Dietrich Reimer. 15,000 Kenyan speakers and an unknown number of Kaye, Alan, and Mauro Tosco. 1993. Early East African Pidgin Ugandans. Juba Arabic, on the other hand, is both a Arabic. SUGIA 14. 269–305. Mahmud, Ushari. 1982. Arabic in the Southern Sudan: history vehicle for the majority of the southerners and a ver- and spread of a Pigin–Creole. Khartoum: FAL. nacular for some urban dwellers. Its formal contexts of Owens, Jonathan. 1985. The origin of East African Nubi. use include limited radio broadcastings, theatrical per- Anthropological Linguistics 27(3). 229–71. formances, songs, Christian religious preaching, etc. ———. 1997. Arabic-based Pidgins and Creole. Contact lan- Some prayer books are written in Juba Arabic using guages, a wider perspective, ed. by Sarah G. Thomason. Amsterdam, Philadelphia: John Benjamins Publishing. Latin script, while individuals educated in Literary 125–72. Arabic may use Arabic script. Both Nubi and Juba Tosco, Mauro, and Jonathan Owens. 1993. Turku: a descriptive Arabic include regional varieties. But Nubi appeared and comparative study. SUGIA 14. 177–268. more homogenized than Juba Arabic, which encom- CATHERINE MILLER 568  K Kayardild and the Tangkic Languages The Tangkic languages are spoken in Queensland, them, albeit in a distinctive form containing many Australia, in the Wellesley Islands and adjoining main- words adapted from Tangkic languages. land. The family comprises three main languages: Apart from scanty word lists from the nineteenth Lardil, Kayardild, and Yukulta. Additional and now and early twentieth centuries, all materials on the extinct varieties are Yangkaal and Nguburindi, Tangkic languages have been recorded since the early although the limited materials we have on these show 1960s. The practical orthographies (written language Yangkaal to be a sister dialect of Kayardild and systems) used to write these languages were devel- Nguburindi a sister dialect of Yukulta. Within Tangkic, oped in this period. These orthographies use digraphs the family bisects into ‘northern Tangkic’, comprising for a variety of sounds; e.g. r before a stop or nasal let- Lardil alone, and ‘southern Tangkic’, comprising ter denotes retroflexion. For example, rd or rn indi- Yukulta and Kayardild. cates that the d and n are pronounced with the tip of The Tangkic languages have no close relatives, the tongue curled backward. Similarly, an h after a although they are related, at a distant level, to other stop (th) or nasal letter (nh) indicates that the blade of Australian languages and share most grammatical sim- the tongue is placed between the teeth for the pronun- ilarities with languages along the Roper River, well to ciation of the relevant sounds. As far as their sound the west. system is concerned, the Tangkic languages are typical Speakers of the Tangkic languages were traditional- Australian languages. ly hunter–gatherers, with a strong emphasis on marine Lardil differs from the other Tangkic languages in resources, building stone walls around the coasts to having lost the final syllable in words greater than two catch fish and hunting for the abundant turtle and syllables: cf. Kayardild kandukandu, ‘red’, but Lardil dugong found in the area. Apart from the Kayardild, kandukan. Furthermore, Lardil dropped off any who were relatively isolated, the other Tangkic groups remaining word-final ng, k, m, or b: cf. Kayardild were linked together in a complex network of tribal kurkangka, ‘bulrush’, but Lardil kurka (rather than interrelationships with people to the west and south. kurkang). Because various grammatical suffixes ‘pro- From the 1860s for the Yukulta, the 1920s for the tect’ the original word stem from such truncation, Lardil, and the 1940s for the Kayardild, they came into Lardil has developed a systematic distinction of word increasingly intensive contact with pastoralists on the forms, depending on their function and position in the mainland and missionaries on the islands, which dis- sentence. The Tangkic languages are typical of rupted many aspects of traditional life, although there Australian languages in using a rich system of case has been a recent resurgence in many traditional activ- suffixes, i.e. word endings that make distinctions sim- ities. However, no Tangkic language is now spoken ilar to English the house vs. to the house. This allows fluently by people below the age of 40 and none has for great freedom of word order. Beyond this, their more than a dozen speakers left: English has replaced case systems are perhaps the most remarkable in the 569 KAYARDILD AND THE TANGKIC LANGUAGES world for several interrelated reasons. First—and by grammatical suffixes) with special forms drawn from no means uniquely in Australia—they exhibit double a set of approximately 150 items with abstract mean- case marking, because one noun phrase embedded in ing and a bizarre sound structure. In fact, the sound another inflects both for its own case (e.g. the posses- structure of Damin is unique among the world’s lan- sive) and that of the main noun: cf. Kayardild thabuju- guages in using five distinct airstream types: some karra, ‘brother’s’, and wangal-nguni, ‘with the sounds are produced with the usual mechanism of boomerang’ (literally ‘boomerang-with’), combine to pressing air outward from the lungs (pulmonic egres- thabuju-karra-nguni wangal-nguni, ‘with brother’s sive), others are produced with inward breath (pul- boomerang’ (literally ‘brother’s-with boomerang- monic ingressive), and others are click sounds (velaric with’). Second, Kayardild and Lardil mark e.g. tense, ingressive) and sounds produced by ejecting air via not only on the verb but simultaneously via modal tongue movements (labiovelar lingual egressive) and case suffixes on nouns. Thus, Ksayardild dangka-a by movements of the Adam’s apple (glottalic egres- burldi-jarra yarbuth-ina thabuju-karra-nguni-na wan- sive). Of these, the pulmonic ingressive and labiovelar gal-nguni-na, ‘the man hit the bird with brother’s lingual egressives are found nowhere else in the boomerang’, shows a past-tense suffix -jarra on the world’s languages, at least not with a similarly promi- verb burldi, ‘hit’, and also a modal case suffix -(i)na nent function. As far as meaning is concerned, Damin on the nouns yarbuth, ‘bird’, thabuju-karra-nguni, solves the problem of compressing the entire everyday ‘with brother’s’, and wangal-nguni, ‘with the language vocabulary into 150 terms by four methods: boomerang’. Third, in Kayardild, case suffixes can be ● Highly abstract words, such as a single form for used to connect sentences with each other. In this ‘act harmfully/damagingly upon’ to replace the usage, the relevant suffix occurs on all words of the everyday words for ‘eat’, ‘chop’, ‘bite’, ‘shoot’, subordinate (dependent) clause. Much of how this and ‘cut’. strange system evolved has now been reconstructed ● A variety of semantic extensions are used to tie with the help of data from Yukulta, the Tangkic lan- together many distinct senses for the relevant guage that conserves the most characteristics of the word; e.g. ‘stomach’ is also used to mean ‘large common ancestor language. intestine’ (by contiguity), ‘guts, excrement’ (by Another twist to case in the Tangkic languages is association), ‘defecate’ (product to process), and added by a further set of verbal case inflections, ‘emerge’ (by metaphor). semantically and structurally part of the set of normal ● The same word can be used as a noun or a verb, case inflections, but with the peculiarity that they con- as shown by the extension from ‘excrement’ to vert their hosts from nouns into verbs. Kayardild ‘defecate’, above; inflectional suffixes show ngada waa-jarra wangarr-ina ngijin-maru-tharra which is intended. thabuju-maru-tharra, ‘I sang a song for my brother’, ● The Damin words are supplemented by hand for example, contains thabuju-maru-tharra, ‘for my signs to distinguish subtypes, e.g. the abstract brother’, which literally means ‘brother-put-(past term for ‘fish’ means ‘bluefish’, which crunches tense)’. Thus, the verbal dative case suffix -maru- con- coral with its teeth, if the speaker points to the verts the noun thabuju into a verb, which then takes the teeth. Because these speech–sign combinations regular verbal inflection for past tense: -tharra. are conventionalized and taught together, Damin Like many other Australian languages, the Tangkic should probably be viewed as a mixed speech languages have a highly developed set of derivatives and sign language. from compass terms. To locate an entity, one normally says things such as ‘the east uncle’ or ‘the shark com- ing from the east’; some Kayardild examples of deriv- atives based on the root ri-, ‘east’, are riinda, ‘coming References from the east’, rilungka, ‘eastward’, riliida, ‘heading Dench, Alan, and Nicholas D. Evans. 1988. Multiple case- ever eastwards’, riyananganda, ‘to the east of’, marking in Australian languages. Australian Journal of ringurrnga, ‘east across a geographical boundary’, Linguistics 8. 1–47. Evans, Nicholas D. 1992a. Kayardild dictionary and ethnothe- riyanyinda, ‘at the eastern extremity of’, rilumir- saurus. Melbourne: University of Melbourne Department of damirda, ‘sea-grass territory to the east’, rilurayaan- Linguistics. da, ‘from one’s previous night’s camp in the east’, Evans, Nicholas D. 1992b. ‘Multiple semiotic systems, hyper- rilijulutha, ‘move to the east’, and riinmali, ‘hey you polysemy and the reconstruction of semantic change in coming from the east’. Australian languages. Diachrony with synchrony: language history and cognition, ed. by Günter Kellermann and Lardil and Yangkaal people had a special auxiliary Michael D. Morrissey, 475–508. Frankfurt: Peter Lang. language, known as Damin, taught to second-degree Evans, Nicholas D. 1995. A grammar of Kayardild, with histori- initiates, which involved substituting all roots (but not cal-comparative notes on Tangkic. Berlin: Mouton de Gruyter. 570  KHMER AND MON-KHMER LANGUAGES Hale, Kenneth L. 1973. Deep-surface canonical disparities in Keen, Sandra. 1983. Yukulta. Handbook of Australian lan- relation to analysis and change: an Australian example. guages, Vol. 3, ed. by R.M.W. Dixon and Barry J. Blake, Current Trends in Linguistics 8: Linguistics in Oceania, ed. 190–304. Canberra: Australian National University Press. by Thomas A. Sebeok, 401–58. The Hague: Mouton. McConvell, Patrick. 1981. How Lardil became accusative. Hale, Kenneth L. 1982. The logic of Damin kinship terminolo- Lingua 55. 141–79. gy. Languages of kinship in Aboriginal Australia, ed. by McKnight, David. 1999. People, countries and the Rainbow Jeffrey Heath, Alan Rumsey, and Francesca Merlan, 31–37. Serpent. New York, Oxford: Oxford University Press. Sydney: Oceania Linguistic Monographs. Ngakulmungan, Kangka Leman. 1997. Lardil dictionary. Hale, Kenneth, and David Nash. 1997. Damin and Lardil phono- Gununa, Queensland: Mornington Shire Council. tactics. Boundary Rider: essays in honours of Geoffrey Plank, Frans (ed.) 1995. Double case: agreement by O’Grady, ed. by Darrell Tryon and Michael Walsh, 247–59. Suffixaufnahme. Oxford: Oxford University Press. Canberra: Pacific Linguistics. NICHOLAS EVANS Khmer and Mon-Khmer Languages The largest of the branches making up the and creakiness, and the aspiration of consonants is Austroasiatic language family, the Mon-Khmer lan- recognized as phonemic as well. An additional charac- guage group comprises over a hundred languages spo- teristic that makes Mon-Khmer languages unique is ken throughout Southeast Asia, particularly in the that they possess very large, complex vowel systems. countries of Kampuchea, Vietnam, and Malaysia. For example, the most complex vowel system among While primarily concentrated on the mainland of these languages is that of Bru, a member of the Katuic Southeast Asia, Mon-Khmer languages can be found branch of Mon-Khmer, which has 68 vowels in its as far west as the Republic of India and on various inventory. Conversely, the inventory of consonants in islands surrounding the Asian mainland. The distribu- Mon-Khmer languages is relatively small, due in part tion of Mon-Khmer languages throughout Southeast to the absence of voiced sounds found in many other Asia suggests that they were among the first to be spo- languages of the world. ken in the region. Although it is hard to be certain without written evidence, some scholars have specu- Khmer lated that the branches of Mon-Khmer separated at approximately the same time during the second mil- One of the oldest languages in Southeast Asia, Khmer lennium BC. The most commonly accepted subgroup- is the official language of Kampuchea (formerly ing of the Mon-Khmer languages proposes a Cambodia). In addition to the six million classification of nine branches: Khmer, Monic, Pearic, Kampucheans who speak Khmer, political turbulence Bahnaric, Katuic, Khmuic, Palaungic, Khasi, and Viet- in Kampuchea, especially during the 1970s, resulted in Mu’o’ng. many people fleeing the country and creating sizable Mon-Khmer languages possess several characteris- populations of Khmer speakers outside Kampuchea, tics distinguishing them from other languages spoken primarily in the neighboring countries of Thailand, in Southeast Asia. Among the most significant of these Vietnam, and Laos. Khmer speakers also reside in is that words in Mon-Khmer languages tend to be France and the United States, bringing the total num- sesquisyllabic, i.e. a word consists of an initial ber of Khmer speakers to over seven million. Modern unstressed syllable followed by a stressed syllable Standard Khmer is used throughout Kampuchea; with full vocalism, in contrast to the high frequency of however, regional differences are typically slight monosyllabic and disyllabic words in other languages enough that there are few obstacles in cross-dialectal in the region. Another characteristic making Mon- communication. Khmer languages unique to the region is that they gen- As a result of thousands of years of contact erally do not recognize tone as phonemic, i.e. creating between various populations in Southeast Asia, many a difference in meaning, in contrast to tonal languages of the languages spoken in the region share much of in the region such as Thai, Lao, and Chinese. Mon- the same vocabulary. During the Angkor period from Khmer languages do, however, make a phonemic dis- the ninth to the fifteenth centuries CE, Khmer had a tinction between such vowel qualities as breathiness powerful influence on other languages in the region, 571 KHMER AND MON-KHMER LANGUAGES an influence that extended to unrelated languages tional morphology, Khmer uses derivational morphol- such as Lao and Thai, as well as to other Mon-Khmer ogy to create new words. With respect to affixing, languages, such as those of the Pearic branch, which Khmer does not use suffixing, but uses prefixing and includes Chong and Pear, and the Bahnaric branch, infixing, particularly as a means of deriving nouns which includes Phnong and Chrau. Khmer and Thai, from verbs. Khmer also uses compounding extensive- which is primarily spoken in neighboring Thailand, ly and reduplication is common, particularly with share so much with respect to vocabulary that early vowel alternation in the prefixed reduplicant to create scholars were led to believe that the two languages a class of words known as iconic expressives. were genetically related. It is now assumed, however, The basic word order of Khmer is subject– that the similarity between the two languages is due to verb–object, although topicalization, or the movement mutual borrowing and reborrowing between them. of an object to the beginning of a sentence, is a com- This borrowing has extended beyond lexical borrow- mon phenomenon in Khmer. Questions retain the ing to include syntactic borrowing as well. The same word order as statements in Khmer. Adjectives Khmer lexicon also includes borrowings from and other elements dependent on the noun generally Vietnamese. follow the nouns they modify, an exception being that Khmer has also borrowed extensively from lan- numerals typically precede the noun. Negation is pre- guages outside the region of Southeast Asia. During verbal with an intensifier typically following the verb. the Angkor period, Khmer borrowed much of its insti- tutional vocabulary from Sanskrit and some of these Mon borrowings can now be heard in colloquial speech, while others are typically only found in literary works. Prior to the sixteenth century CE, the Mon ruled over Since the introduction of the Buddhist religion to parts of Myanmar and Thailand, but today the Mon Kampuchea in the fifteenth century, Khmer has also live primarily in small villages in Myanmar (formerly borrowed heavily from Pali. During the political dom- Burma) and in Thailand. In Myanmar, they generally ination of Kampuchea by the French from the mid- live along the Tenasserim coast between Thaton and nineteenth to the mid-twentieth century, Khmer Tavoy, and also live in the vicinity of Pegu. Although borrowed many French words, some of which can still the Mon have made many efforts to attain recognition be heard, particularly in Kampuchea’s urban centers. as a state in Myanmar, these attempts have met with Khmer has also borrowed a small number of Chinese little success. As a result, over the past several cen- words. Since Khmer words are generally monosyllab- turies many Mon have fled to Thailand, where they ic and sesquisyllabic, polysyllabicity is typically live primarily in the Chao Phraya basin between indicative of borrowings, most often from Pali, Uthiathani and Bangkok. Although they do not live in Sanskrit, or French. the city of Bangkok itself, the Mon have established a Like other Mon-Khmer languages, the inventory of cultural center in Thailand on the southern outskirts of sounds in Khmer is interesting, mostly owing to its the city in the town of Prapradaeng. Despite this suc- large vowel system. While not as vocalically rich as cess, the southern portion of Myanmar is still consid- Bru, most standard dialects of Khmer have between 25 ered the homeland of the Mon. and 28 vowels, with a typical inventory comprising 27 The language of the Mon has a story parallel to the vowels. The Northern Khmer dialect has a system story of its speakers. A member of the Monic branch comprising at least 31 vowels. The consonantal inven- of Mon-Khmer languages, Mon was once a major lan- tory of Khmer is somewhat smaller than its vocalic guage in Southeast Asia; however, it is now spoken by inventory, partially due to the absence of voiced fewer than a million people. Although for centuries sounds such as g and z that speakers of other lan- there has been pressure on the Mon in both Myanmar guages would expect to find. Phonologically, conso- and Thailand to assimilate to the dominant languages nant clusters may occur at the beginning of a Khmer and culture in the countries—Burmese and Thai, word, but not at the end. Additionally, there exist sev- respectively—the pressure to assimilate has increased eral consonants never found word-finally in Khmer, since World War II due to modernization and urban- including r, s, b, and d. ization. In Thailand, where the Mon have been living Like Chinese and Vietnamese, Khmer is an isolat- in exile for centuries, there is intense pressure for all ing language, which means that it tends toward a one- ethnic minorities to assimilate—the Thai government to-one correspondence between morpheme and word. makes no distinction between ethnic groups and Thai. Khmer never uses inflectional morphology, but relies Additionally, the Thai language and culture are the on separate morphemes or context to show such things only local language and culture recognized by the gov- as gender, number, and tense. Although lacking inflec- ernment schools. Not surprisingly, the Mon exiles 572  KHMER AND MON-KHMER LANGUAGES have made no attempts to gain autonomy in their Despite their differences, Vietnamese and other adopted country of Thailand. Mon-Khmer languages share a number of characteris- tics. Vietnamese is an isolating language with no inflectional morphology; however, like other Mon- Khasi Khmer languages, Vietnamese uses derivational mor- The westernmost language in the Mon-Khmer group, phology as a means of creating new words. Besides Khasi is spoken by approximately 500,000 people liv- prefixing, Vietnamese makes extensive use of com- ing in the Khasi and Jaintia Hills of Meghalaya; as pounding and also uses reduplication, particularly in such, it is the only Mon-Khmer language found in the onomatopoetic expressions. Republic of India. Khasi is also spoken by about Among the Austroasiatic languages, three have 85,000 people in Bangladesh. Because of its geo- long traditions of writing: Khmer, Mon, and graphical isolation from other Mon-Khmer languages, Vietnamese. The earliest Khmer writings in which the Khasi is the sole member of the Khasi branch of the date is known were written in 611 CE, and other extant Mon-Khmer group. Its status as a singular language in writings are estimated to be from the second or third India, as well as its strong matrilineal society, has centuries CE. Khmer is written in a phonologically attracted a great deal of interest in the Khasi people, based writing system that derives from the Pallava particularly among ethnographers and other scholars. script, a system that was used in South India at approx- As a result, the Khasi people have managed to main- imately the same time and descended from the Bhrami tain some prestige in the region. script. Having evolved since its introduction, the script used for writing in Khmer is now called the Khmer let- ters. Mon is also written with an Indian-derived alpha- Vietnamese bet and extant texts include some dating back to the Vietnamese is the official language of Vietnam and eleventh century CE. In addition, epigraphs have been comprises approximately 80 million speakers, primari- found that date back to approximately 600 CE. ly in Vietnam, but also in neighboring countries, as Although Vietnamese also has a long written tradition, well as in France and the United States. Although the history of this writing is considerably different Vietnamese is now generally recognized as being a from the histories of writing in Khmer and Mon. member of the Mon-Khmer family, its genetic affilia- Written Vietnamese begins with Chinese characters, tion was the subject of debate among linguists for some then with a writing system derived from Chinese char- time. The controversy arose due to the status of acters, and finally a Roman script introduced by Vietnamese as a tonal language, a trait also possessed Catholic missionaries in the seventeenth century. by Chinese and Thai, but a trait not shared by the Although scholars have speculated that the majority of Mon-Khmer languages. Although tone is Vietnamese were using the Chinese script to record not the only characteristic that Chinese and Vietnamese their own language as early as the eighth century CE, share, another being the great number of lexical items the earliest extant writings of Vietnamese in this script the two hold in common, linguists ruled out any close date back to the thirteenth century. genetic affiliation between the two languages long ago, instead attributing similarities, particularly with respect to vocabulary, to the long domination of the References Vietnamese by the Chinese. Given the choice between Haudricourt, André-Georges. 1954. De l’origine des tons in Vietnamese as a Mon-Khmer language that had some- viêtnamien. Journal Asiatique 242. 68–82. how adopted tone, or a Tai language that had borrowed Huffman, Franklin E. 1970. Cambodian system of writing and much of its vocabulary from Mon-Khmer languages, beginning reader with drills and glossary. New Haven: Yale many linguists adopted the latter view until the con- University Press. Jacob, Judith M. 1968. Introduction to Cambodian. Oxford: vincing argument of Haudricourt (1954). A French Oxford University Press. botanist, Haudricourt argued that the phenomenon of Parkin, Robert. 1991. A guide to Austroasiatic speakers and tonal genesis in Vietnamese was an evolutionary their languages. Honolulu: University of Hawaii Press. process in which the language acquired three tones by Thomas, David D. 1964. A survey of Austroasiatic and Mon- the sixth century through language contact with speak- Khmer comparative studies. Mon-Khmer studies, Vol. I. Saigon: The Linguistic Circle of Saigon. ers and through internal changes, and doubled this Thomas, David D., and Robert K. Headley, Jr. 1970. More on number by the twelfth century to the six tones current- Mon-Khmer subgroupings. Lingua 25. 398–418. ly used in the Hanoi dialect of Vietnamese. Haudricourt LAMONT ANTIEAU went on to argue that the tones in Chinese and Thai developed in much the same way. See also Austroasiatic; Southeast Asia; Vietnamese 573 KHOISAN Khoisan According to a widely accepted genetic classification some (e.g. Honken 1977; Ehret 1986) but rejected by by Joseph Greenberg (1963), Khoisan constitutes one others (e.g. Westphal 1971). A third group of scholars of the four language phyla (or families) to be found on (e.g. Köhler 1981; Sands 1998; Traill 1986; the African continent. Apart from two languages Güldemann and Vossen 2000) do not explicitly reject (Sandawe, Hadza) spoken in north-central Tanzania, Greenberg’s hypothesis but argue that so far there has all languages are found in southern Africa, mostly in been no proof of genetic relationship; given the evi- Botswana and Namibia, and to a minor extent also in dence available, this appears to be the most reasonable Angola, Zambia, and South Africa. While there must position to be assumed. have been more than 100 Khoisan languages prior to A substantial number of lexical and grammatical the European conquest of southern Africa, less than 30 similarities across various groupings of Khoisan have have survived. The total number of Khoisan speakers been identified, but in most cases it remains unclear probably does not exceed 200,000, of whom around whether these similarities are due to genetic relation- two thirds belong to one language community: ship or language contact. Only for one of the Khoisan Nama/Damara (cf. Güldemann and Vossen 2000). families, Khoe, there exists a sound comparative Earlier work on Khoisan had assumed, largely analysis (Vossen 1997). based on nonlinguistic criteria taken from physical and Table 1 gives an overview of Khoisan languages cultural anthropology, that there exists a basic division and their genetic classification (for more details and a between ‘Hottentot’ and ‘Bushman’ languages. The slightly different treatment, see Güldemann and former were said to consist of the Khoekhoe branch of Vossen 2000). Some of the languages are presumably the Khoe family (see Table 1). The latter includes the extinct; others form dialect clusters. The !Xun family, remainder of Khoisan, with the exception of the for example, constitutes a complex dialect cluster con- Tanzanian languages Sandawe and Hadza, which were nected by chains of mutual intelligibility, even if many not associated with this general entity at that time (see East !Xun speakers do not understand North !Xun e.g. Meinhof 1912). This division is reflected in the speakers. In most cases, it remains unclear whether label ‘Khoisan’, proposed by Schultze (1928), which there is a language or a dialect boundary. For example, is composed of two lexical items from the Kxoe, Buga, and Ani are treated here as different lan- Nama/Damara language: khoi (or khoe) ‘person’, guages of Kalahari Khoe; yet, there are also reasons to standing for the ‘Hottentot’, and san ‘forager’, stand- consider them as dialects of one and the same lan- ing for the ‘Bushman’ languages. guage (see Table 1). Subsequent research established that there exists While it remains unclear whether indeed the some kind of linguistic relationship uniting the Khoisan languages form one language family, there Khoisan languages of southern Africa (cf. Köhler are a few typological characteristics that set these lan- 1975). Greenberg (1963) was the first to argue that guages off from the neighboring Bantu languages in Khoisan forms a genetic unit, and that this unit not particular and from many other languages in general: only includes the ‘click languages’ of southern Africa (1) Most Khoisan languages distinguish between at but also two East African languages. He proposed the least four different types of click; languages of following classification of Khoisan: the !Ui-Taa and the 苷Hõã families even have (1) South African Khoisan five. In addition to dental (), alveolar (!), a. Northern palatal (苷), and lateral () clicks, there is a bila- b. Central bial click (>) in these two families. c. Southern (2) Furthermore, there are a number of distinctions (2) Sandawe made in vowels; !Xun, for example, has distinct (3) Hatsa (= Hadza) sets of oral, nasal, glottalized, and pharyngeal- ized vowels. A number of Khoisanists are reluctant to adopt (3) Accordingly, the number of phonemes in these Greenberg’s (1963) hypothesis that the Khoisan lan- languages is comparatively high; some of these guages really are genetically related to one another. languages, such as !Xõõ and !Xun, have more Greenberg’s phylum hypothesis has been defended by than one hundred distinct phonemes and thus 574  KHOISAN TABLE 1 Khoisan Languages DC = dialect cluser, C = central, E = east, N = north, S = south, W = west, + = presumably extinct language. Family Branch Group Languages Where Spoken Number of Speakers !Xun North !Xun (DC) S Angola, N Less than 5,000 (Northern) Namibia East !Xun (Ju‘hoan; NE Namibia, Less than 5,000 DC) NW Botswana Khoe Khoekhoe North Namara/Damara, Namibia Over 120,000 (Central) Hai‘om, 苷Aakhoe (DC) + South !Ora; +Cape varieties (DC) S Namibia, Presumably extinct South Africa Kalahari West Kxoe, Buga, Ani NE Namibia, Over 10,000 Khoe (DC); Naro (DC); Gana, Botswana, SE Gui, 苷Haba (DC) Angola East Shua, Ts’ixa, Danisi, Botswana Less than 5,000 Xaise, +Deti; Kua-Tsua (DC) !Ui-Taa !Ui + Xam, +‘Auni, South Africa, Presumably extinct (Southern) + 苷Khomani, +Xegwi, etc. E Namibia, W Botswana Taa !Xõõ (DC) W Botswana A few thousand 苷Hõã 苷Hõã WC Botswana A few hundred + Kwadi Kwadi SW Angola Presumably extinct Sandawe Sandawe C Tanzania A few ten thousands Hadza Hadza C Tanzania A few hundred belong to the phonetically most complex lan- have verbal derivational morphemes, invariably suffix- guages of the world. es or enclitics on the verb. (4) Although Khoisan languages have a fairly rich Most Khoisan languages have subject–verb–object inventory of grammatical distinctions, their (SVO) as their basic word order, the noun precedes its morphological structure is fairly isolating modifiers, but in attributive possessive constructions analytic. the head follows its modifiers. Hadza has both verb-ini- tial (VSO) and verb-medial (SVO) order, while all A widespread morphological characteristic of Khoe languages and Sandawe have essentially verb- Khoisan languages is the presence of a noun class sys- final (SOV) syntax with modifiers preceding the head. tem. There is, however, a remarkable difference While clause subordination is common in Khoe, coor- between the Khoe languages on the one hand, and the dination involving structures commonly found in verb- !Xun and Taa languages on the other. While the latter serializing languages appears to be widespread in other do not make a morphological distinction based on nat- Khoisan languages. A syntactic characteristic of a ural sex, marking distinctions such as human vs. non- number of Khoisan languages is the presence of a mul- human or animate vs. inanimate instead, Khoe tipurpose oblique case marker used to introduce languages have three gender categories, which are adjuncts, i.e. clausal participants that are not part of the masculine, feminine, and common. Khoe languages valency of the verb (Güldemann and Vossen 2000:110). are further characterized by a set of portmanteau mor- While one language (Nama/Damara) has been phemes simultaneously expressing person, gender, described in some detail, the majority of Khoisan lan- and number (pgn-markers); since there are three cate- guages are virtually unknown, apart from a few gram- gories of each person, gender, and number (singular, matical details and/or a wordlist (see Güldemann and dual, plural), there are close to 27 different pgn-mark- Vossen 2000:103). With the exception of Nama/ ers, as well as personal pronouns, in these languages. Damara and a handful of other languages, all lan- Sex-based gender systems are also found in Kwadi, guages are seriously endangered, many of them on the Sandawe, and Hadza. verge of extinction. Widespread in the verbal system (except for Khoe) are suppletive verbs, where there are different forms for singular and plural. Suppletive verbs show number References agreement with the object in the case of transitive Baumann, Hermann (ed.). Die Völker Afrikas und ihre tradi- verbs and with the subject in the case of intransitive tionellen Kulturen. Part I, Allgemeiner Teil und südliches verbs. Furthermore, a number of Khoisan languages Afrika. (Studien zur Kulturkunde, 34.) Wiesbaden: Steiner. 575 KHOISAN Bleek, Dorothea F. 1956. A Bushman dictionary. (American Meinhof, Carl. 1912. Die Sprachen der Hamiten. Hamburg: Oriental Series, 41.) New Haven: American Oriental Friederichsen. Society. Sands, Bonny. 1998. Eastern and Southern African Khoisan: Ehret, Christopher. 1986. Proposals on Khoisan reconstruction. evaluating claims of distant linguistic relationships. SUGIA (Sprache und Geschichte in Afrika) 7(2) 105–30 (Quellen zur Khoisan-Forschung, 14.) Cologne: Köppe. Greenberg, Joseph H. 1963. The languages of Africa. The Schultze, Leonhardt. 1928. Zur Kenntnis des Körpers der Hague: Mouton. Hottentotten und Buschmänner. Jena: Fischer. Güldemann, Tom, and Rainer Vossen. 2000. Khoisan. In Heine Sebeok, T.A. (ed.) 1971. Current trends in linguistics, Vol. 7: and Nurse, pp. 99–122. Linguistics in sub-Saharan Africa. The Hague: Mouton. Heine, Bernd, and Derek Nurse (eds.) 2000. African languages: Traill, Anthony (ed.) 1977. Khoisan Linguistic Studies, Vol. 3. an introduction. Cambridge: Cambridge University Press. (African Studies Institute Communications, Vol. 6.) Honken, Henry. 1977. Submerged features and Proto-Khoisan. Johannesburg: University of the Witwatersrand. In Traill, pp. 145–69. ———. 1986. Do the Khoi have a place in the San? New data Köhler, Oswin. 1975. Geschichte und Probleme der Gliederung on Khoisan linguistic relationships. SUGIA (Sprache und der Sprachen Afrikas. Von den Anfängen bis zur Gegenwart. Geschichte in Afrika) 7(1) 407–30. In Baumann, pp. 135–373. Vossen, Rainer. 1997. Die Khoe-Sprachen: Ein Beitrag zur ———. 1981. Les langues khoisan, In Manessy. Erforschung der Sprachgeschichte Afrikas. (Quellen zur Manessy, Gabriel. 1981. Les langues de l’Afrique subsahari- Khoisan-Forschung, 12.) Cologne: Köppe. enne. (Les Langues dans le monde ancien et moderne, ed by Westphal, Ernst O. J. 1971. The click languages of southern and Jean Perrot.) Paris: Centre National de la Recherche eastern Africa. In Sebeok, pp. 367–420. Scientifique. BERND HEINE Kinship Terms The inventory of names for family members found in younger siblings), whether the kinship tie is maternal or a language can tell us a great deal about the culture of paternal, whether the relation is by blood or by marriage, the speakers of that language. How these names are and whether the relation is lineal (father, daughter) or used, both within and outside the family, is also close- collateral (aunt, cousin). In English, a combination of ly related to culture. For these reasons, the study of the components 2 generations, male, and lineal kinship terms has been important to anthropological would indicate the person called grandmother. Different linguistics. It has also played a role in theories of word contrasts show up in different languages. meaning. In cognitive linguistics, kinship terms can Anthropologists are interested in which contrasts are provide us with examples of ‘unmarked’ and ‘marked’ maintained and which ones are merged or overlap in categories, that is, categories that fit neatly into a sym- different societies. In Arapaho, a man would use the metrical structure as well as categories that are asym- term neyóo to refer to his brother-in-law, while a metrical. This allows us to see the interplay between woman would use the term neiθébi for her brother-in- particular social structures and linguistic universals in law. Neiθébi actually means something like ‘sibling-in- forming systems of meanings within a language. law of a different sex’, and it would be used by a man Kinship terminology forms a well-defined subset of to refer to his sister-in-law. In the Iroquoian languages, vocabulary within a language, and it is a subset that there are different words for older and younger sisters tends to be more structured, symmetrically and hierar- and brothers; lineal and collateral relatives are not sep- chically, than other areas of the lexicon. These charac- arated as in English: in Seneca, noʔye¯h refers to one’s teristics make kinship terms good candidates for a type mother and her sisters, and haʔnih refers to one’s father of meaning analysis known as ‘componential analysis’. and his brothers. In English, cousins are not distin- This type of analysis seeks to break down word mean- guished by sex, but in Czech, they are. ings into their basic components, typically using a bina- The Omaha Indians had a system that has been ry classification system. For example, many kinship described as ‘skewing’ and ‘merging’ certain categories. terms designate the sex of the person referred to, so that Generational distinctions are skewed depending on nephew could be labeled as male, while niece would one’s sex, so that a woman’s brother and her father are be labeled as male. Besides the sex of the referent, equivalent. In other words, the brother of a female is other components of kinship terminology are the sex of shifted to an older generation. This leads to the merging the speaker, generation (x generations older or younger), of other relations as well: the maternal uncle would be birth order (used e.g. to distinguish among older or equivalent to the grandfather, and a man’s sister would 576  KITUBA be equivalent to his daughter. It has been proposed that to denote relative status and a person’s attitude of a system like this one can account for the facts of Proto- humility toward an older addressee. Indo-European kinship terminology also (cf. Latin avus Family terms frequently extend metaphorically even ‘grandfather’ and avunculus ‘maternal uncle’). into nonhuman realms. Gods and goddesses are called A look at Proto-Indo-European kinship terms by the terms for father and mother. The Navajo have the reveals certain features of the social organization of same concept that some English speakers do of ‘Mother these ancient people. If a woman’s brother takes on the Earth’, which they further extend to agricultural fields status of her father, the men of the family likely held and corn. Words for brother and sister have been used in higher status positions of dominance or protection various Indo-European and other languages to denote a over the women within their own generation. Not only spiritual kinship as well as a physical one. Whether our was the culture therefore patriarchal, it was also behavior toward the earth or toward our fellow human patrilocal, meaning that when a woman married, she beings can ever parallel behavior expected toward our went to live with her husband’s family. This is reflect- closest family members remains to be seen. ed in the greater frequency and number of terms for a woman’s in-laws than a man’s in the Proto-Indo- References European language. Thus, the kinship terms provide Casson, Ronald. 1981. The semantics of kin term usage: trans- us with clues about the role of women in this society. ferred and indirect metaphorical meanings. Language, cul- We can ask whether the categories that are important ture, and cognition, ed. by R. Casson, 230–44. New York: Macmillan. in distinguishing kinship terms affect the way in which Friedrich, Paul. 1965. Proto-Indo-European kinship. Ethnology people perceive the world. Does the lack of a particular 5. 1–36. distinction in the vocabulary of a language mean that Goodenough, Ward. 1956. Componential analysis and the study people who speak that language are unable to mentally of meaning. Language 32. 195–216. make that distinction? Most would reject this extreme Goodenough, Ward. 1965. Yankee kinship terminology: a prob- lem in componential analysis. Formal semantic analysis, ed. version of linguistic relativity. That is, using the same by E. Hammel, 259–87. American Anthropologist 67. word for both mother and mother’s sisters does not Greenberg, Joseph H. 1980. Universals of kinship terminology: mean that the speaker cannot distinguish who is his/her their nature and the problem of their explanation. On lin- biological mother. It does, however, imply that one guistic anthropology: essays in honor of Harry Hoijer, ed. by behaves in certain ways toward both the mother and her Jacques Maquet. Los Angeles: Undena for UCLA Anthropology Department. sisters, perhaps having the same obligations toward Greenberg, Joseph H. 1966. Language universals. The Hague: them. In this system, as found in the Arapaho language, Mouton. 9–32. a man would be obliged to treat his sons and his broth- Heath, Jeffrey, Francesca Merlin, and Alan Rumsey (eds.) 1982. er’s sons the same way, whenever possible. The languages of kinship in aboriginal Australia. Sydney: As far as behavior is concerned, the extension of University of Sydney. Lounsbury, Floyd. 1964. The formal analysis of Crow- and kinship terms to refer to people outside the family is Omaha-type kinship terminologies. Explorations in Cultural also symbolic of a certain expected behavior. In earli- Anthropology, ed. by Ward H, 351–93. Goodenough. New er times, African Americans would sometimes refer to York: McGraw-Hill. any respected elders in the community as aunt or Lounsbury, Floyd. 1964. The structural analysis of kinship uncle. Turkish villagers, when speaking to a person of semantics. Proceedings of the IXth International Congress of Linguists, ed. by H. Lunt, 1073–93. The Hague: Mouton. their own generation, can use either the term for moth- Nogle, Lawrence E. 1974. Method and theory in the semantics er’s brother (dayi) to show respect or the term for and cognition of kinship terminology. The Hague: Mouton. father’s sister’s son (aga) to express decreased solidar- Schneider, David M. 1980. American kinship: a cultural ity. Chinese speakers can use terms for elder uncle, account, 2nd edition. Chicago: University of Chicago Press. younger uncle, or younger aunt to nonfamily members ELLEN JOHNSON Kituba Kituba, a.k.a. Kikongo-Kituba, is a contact-based all the way to the Atlantic Ocean—and in the southern Bantu language spoken in the western part of the part of the Republic of the Congo. It is known by sev- Democratic Republic of Congo—from the Kwilu River eral other names, including Kikongo ya leta ‘Kikongo 577 KITUBA of the colonial agent (leta)’, Kikongo ya bula-matadi Overall, Kituba differs from Kikongo languages ‘Kikongo of the rock-breaker’ (in reference to rock- and other Bantu languages in the following structural blasting during the construction of the railroad from the respects: (1) Kituba does not use tone or pitch varia- Atlantic coast to Kinshasa), and Mono kutuba ‘I tions to distinguish meaning to the same extent as the speak’. Outside the Bakongo area, west of Kinshasa, it other languages; (2) it has a less complex word struc- is better known by the name Kikongo, in reference to ture; and (3) it has lost e.g. the agreement system typ- the fact that most of its vocabulary comes from one of ical of Bantu languages, in which the inflectional the Kikongo language varieties associated with the pattern of the subject noun matches that of the verb. Bakongo people. In the Republic of the Congo, where A large proportion of Kituba words have the same some of the Bakongo people live, it is called Ikele ve ‘it intonational system as Swahili. In the case of verbs, isn’t’, in reference to the nonnative way of negating the the position of the intonational stress changes, depend- verb kele ‘be’. Only an accident of history has thus ing on whether or not it has a tense-aspect suffix. favored the name Kituba in linguistics. Thus, móno kwénda is ‘I go’ (in the narrative tense), Although the emergence of Kituba is undeniably but móno kwendáka is ‘I went’ (in the past tense). associated with the exploitation colonization of the However, there is a significant proportion of words Congo Basin by Belgium (roughly, east of the Congo that seem to follow the basic high-and-low tone sys- River) and by France (west of the River), its roots lie tem of Lingala, e.g. munoko ‘mouth’ (with low tones in the usage of (Kikongo-) Kimanyanga as a trade lan- only), kímbambala ‘old (machine)’ (with a high tone guage during the precolonial period. Manyanga (situ- on the first syllable only), and makilá ‘blood’ (with a ated a few dozen miles west of present-day Kinshasa) high tone on the last syllable only). Kituba thus incor- was already an important trade center before the porates both tone distinctions and pitch accent, which exploitation colonization of the Congo Basin—which is not typical of Bantu languages. started soon after Africa was apportioned to European Unlike in the Bantu canon, Kituba verbs do not colonial powers in the Berlin Treaty in 1885. During carry pronominal or agreement prefixes, nor do they the Kongo Kingdom and its subsequent trade colo- carry any other preverbal marker for tense or negation. nization by Europe, several caravans from long dis- Negation is expressed sentence-finally with an invari- tances in the interior went through Manyanga. ant marker vé, as in kwísa vé ‘don’t come’ (literally, Kimanyanga, its Kikongo variety, served as the lingua ‘come not’). The pronouns derive from the emphatic franca. The Belgian and French colonizers adopted it ‘independent’ pronouns of Kimanyanga (viz. móno to communicate not only with the Bakongo people in ‘I/me’, ngé ‘you, singular’, yándi ‘he/him/she/her/it’, the territory between Kinshasa and the Atlantic coast béto ‘we/us’, béno ‘you, plural’, báu/bô ‘they/them’), but also with other indigenous populations farther in and they occupy the same positions as nouns around the interior. They took it as far east as the Kwilu River. the verb, as in Yándi/Pételo mon-áka ngé/María vé They also brought colonial auxiliaries from West ‘He/She/Peter didn’t see you/Mary’ (literally, ‘X saw Africa and laborers from as far east as Zanzibar to Y not’). Aspect and mood are expressed by independ- build the railroad between the Atlantic coast and ent words, as in yándi ké(l)e kwénda ‘he/she is going’ Kinshasa. (Labor migrations were a common phenom- (progressive aspect), yándi lénda kwénda ‘he/she enon during the exploitation colonization of Africa by may/can go’ (potential mood), and yándi ata kwénda Europeans during the late nineteenth and early twenti- ‘he/she will go’ (future aspect). eth centuries.) The language contact that ensued from In complex noun phrases, modifiers are connected these population movements and contacts produced to the head noun by na if they are pronominal and by Kituba from Kimanyanga. There are at least three ya in other cases, as in mukandá na móno ‘my book’ major dialects of Kituba to date: (1) the Western (literally, ‘book of me’), mukandá ya Pételo ‘Peter’s dialect, spoken in the former Kongo Kingdom south of book’ (‘book of Peter’), mukandá ya ngé tang-á(k)a the Congo River (i.e. in today’s Democratic Republic ‘the book that you read’ (‘book of you read’), and of Congo); (2) the Northern dialect, spoken north of bíma ya kú-dia ‘food items’ (‘thing of eat’). the Western dialect, in the Republic of the Congo; and As a nonnative lingua franca in rural areas, Kituba is (3) the Eastern dialect, spoken from the Kwango to the often spoken with different accents that reflect its Kwilu rivers in the Democratic Republic of Congo. As speakers’ linguistic backgrounds. The ‘accent’ may be explained below, they reflect specific influences of the identified mostly in word choice (for instance, when a more indigenous languages on Kituba. This has been speaker imports a term from his or her ethnic language) facilitated by the fact that it is spoken as a second lan- and in some grammatical features. For instance, in the guage in the rural areas, although more and more chil- western dialect, the structure of the progressive con- dren have been acquiring it at school because it serves struction uses the connective na, as in móno kéle na kú- as the medium of education. dia ‘I am eating’ (literally ‘I am LOC eat’). The 578  KOREAN counterpart of the construction in the eastern dialect is borrowings. Relative to Kimanyanga, the main vocab- móno ké(l)e (kú-)dia, without the connective. Within ulary source, Kituba seems to have autonomized to an the eastern dialect, the pattern for commands varies. extent seldom reached by most creoles of the Up the Kwilu River, kwis-éte ‘come, please’ is the nor- Caribbean and the Indian Ocean. Kituba speakers mal imperative, whereas elsewhere the bare imperative hardly think they speak a dialect of ethnic Kikongo, kwísa ‘come’ will do and will not offend the addressee. although they use the latter name in reference to their In the same variety, the connective tí is used in yándi language in the eastern dialect region. It is interesting kwisáka tí bô ‘he/she came with them’, whereas else- that Kituba is hardly ever included in the classification where in the eastern dialect the same general purpose of Bantu languages, which reflects an old linguistic connective na is used. tradition of not including acknowledged contact-based Whereas varieties spoken in rural areas often reflect varieties in classifications of language families. Yet influence from the more indigenous ethnic languages, Lingala, which has a similar history, is typically those spoken by the educated reflect French influence. included among the Bantu languages. The rural varieties are looked down upon, but their educated counterparts are at best criticized by purists; References they are well tolerated. The source language for the Fehderau, Harold. 1966. The origin and development of Kituba. vocabulary is no longer considered as the norm from Ithaca, NY: Cornell University dissertation. which the other varieties putatively deviate. Rural and Mufwene, Salikoko S. 1997. Kitúba. Contact languages: a educated varieties just deviate from the urban non- wider perspective, ed. by Sarah G. Thomason, 173–208. Amsterdam: John Benjamins. Frenchified norm. In fact, French, the official language Samarin, William. 1989. The black man’s burden: African colo- of both the Democratic Republic of Congo and the nial labor on the Congo and Ubangi rivers, 1880–1900. Republic of the Congo, is not even part of the Kituba Boulder, CO: Westview Press. continuum, although it is the source of many lexical SALIKOKO MUFWENE Korean Although the Korean language shares many character- With regard to the historical division of Korean, istics with the Japanese and Altaic languages, its Lee (1972) calls the period before the Goryeo dynasty genetic relationship is unknown. Its syntactic structure ‘Old Korean’, and the Goryeo dynasty period ‘Early most closely resembles Japanese, but its phonological Middle Korean’, the period from the end of the fif- system differs markedly. Morphologically, it displays teenth century in the Joseon (or Choso ˘n, Yi) dynasty a so-called agglutinating characteristic, i.e. words are and up to the Japanese invasion in the late sixteenth formed from roots and affixes. century ‘Late Middle Korean’, the period up to the late The oldest written records about Korean are in third- nineteenth century ‘Modern Korean’, and after the century Chinese, eighth-century Japanese ‘Nihon- nineteenth century ‘Contemporary Korean’. During shoki’ (Chronicles of Japan, 720) historical records, in the transition from Middle Korean to Contemporary which are written Korean place names, and in refer- Korean, some sound changes and various grammatical ences to ‘Samkwuk-saki’ (Chronicles of the three king- changes occurred. doms, 1145) and ‘Samkwuk-yusa’ (Memorabilia of the three kingdoms, 1285) of the Goryeo (or Kolye in the Area of Use and Variants Yale Romanization system) dynasty. Korean is only fully recognizable in historical records after 1443, Korean is the national language of the Democratic when the Hangeul (or Hangu ˘l, Hankul) writing system People’s Republic of Korea in the northern part of the was created, allowing full written representation of the Korean Peninsula, and the Republic of Korea in the Korean language. Until then, the language was written south. Koreans are an almost completely homogenous mainly with Chinese characters and supplementary ethnic group. North Korea has a population of 22.27 characters called Gugyeol (or Kugyo ˘l, Kwukyel), but million people, while South Korea’s population is these were only able to present a partial picture of the 47.27 million people. Additionally, over 1.8 million language at the time. ethnic Koreans live in China and more than 400,000 579 KOREAN reside within the borders of the former Soviet Union. and pressed into existence in the form of a book about As a result of Stalin’s forced settlements, Koreans also Hankul. In Hankul, the written representation form of inhabit Central Asian countries such as Kazakhstan consonants is based on the forms of the point of artic- and Uzbekistan. Between 1910 and 1945, Korea was ulation. From the following year, 1447, they were used in a state of annexation by Japan and forced migration in the publication of the official hymn of the Joseon to Japan occurred during this era. There are an esti- dynasty ‘Yongpi-echenka’ and the rhyme dictionary mated 700,000 Koreans living in Japan. Korean emi- ‘Tongkwuk-cengwun’, and were subsequently used in gration has also extended its reach to countries like the the publications of many books. Other than book pub- United States and Australia. Most Korean speakers lications, the official script at the time was written with outside of the Korean Peninsula live in bilingual soci- Chinese characters; Hankul was mostly used for pri- eties. Now, there are more than 75 million speakers of vate communication, but at the end of the nineteenth Korean worldwide. century and the start of the modern age, Hankul came The national languages of South Korea and North into general use. Today, the literacy rate in both North Korea are basically one and the same, but in South and South Korea is nearly 100%. Korea the standard language is based on the Seoul Each Korean letter represents one sound, that is, a dialect, while in North Korea the standard language is letter of an alphabet, which are combined and arranged based on the Pyeongyang dialect and is called to form characters. Each combination of letters, or a mwunhwae ‘the language of culture’. Because of this, character, represents one syllable. Therefore, Hankul there are slight variations in vocabulary, pronuncia- simultaneously has a single sound, alphabetic charac- tion, and orthography. teristics like the Roman alphabet and syllabic charac- teristics like Chinese characters or Japanese kana: Dialect k a Korean dialects can be roughly divided into the fol- kam m lowing six categories. Following the Korean peninsula from north to south, they are: the northeast dialect in Hamgyeong and Ryanggang Provinces, the northwest Hankul has several methods of transcription: it can dialect in Pyeongan and Jagang Provinces, the central be Romanized according to the McCune–Reischauer dialect in Hwanghae, Gangwon, Gyeonggi and method, the Yale method, or official methods desig- Chungcheong Provinces, the Southeast dialect of nated by the governments of North and South Korea. Gyeongsang Province, the southwest dialect in Jeolla The Yale method is basically used in this writing. The Province, and Jeju dialect in Jeju Province. The Jeju Hankul, its Romanized transcriptions, and main pro- dialect retains the Middle Korean vowel ‘3’, which is nunciations are shown in Table 1. now lost in Contemporary Korean, and the southwest dialect makes no distinction between the vowels ‘ey’ and ‘ay’ or between ‘uy’ and ‘e’. Also, the fortis ‘ss’ is Phonetics and Phonology not present in the southeast dialect. In the northwest Words in contemporary Korean are formed from at dialect, the affricate [tʃ] is pronounced [ts]. Middle least one syllable. For most words, the basic vocabu- Korean had a pitch accent system to distinguish word lary stem is one or two syllables long. Syllables gen- meanings, but in Contemporary Korean only the erally fit this pattern: consonant + semivowel + vowel northeast and southeast dialects retain a pitch accent, + consonant. A vowel always forms the core of the syl- while it has completely disappeared from others such labic, and one consonant may come on each side of the as the central dialect. vowel. For basic vocabulary, a relatively large number of words end with a consonant. Writing System These are the eight vowel phonemes present in the Seoul dialect: The Korean writing system, which is today called Hankul in South Korea and wulikul in North Korea, was created under the guidance of King Sejong, the fourth king of the Joseon dynasty. Hankul first appeared on the world stage in written form in 1446 in Hunminjeongeum (or Hunminjo ˘ngu˘m, Hwunmincengum), a woodblock publication containing details of the reason for its cre- Of the eight vowels, the back vowel wu and o are ation, construction basics, and usage examples. In other rounded vowels, while the others are unrounded. The words, Hankul is unique in the fact that it was carved shape of the lips when pronouncing i and u are nearly 580  KOREAN TABLE 1 Consonants Vowels the same. Also, in the present-day Seoul dialect, to the disappearance of the distinction between the distinction of pronunciation has all but disappeared subjective form of the first-person pronoun nay-ka ‘I’ between the mid vowel ey and low vowel ay. The lack and the subjective form of the second-person pronoun of distinction between these two vowels has even led ney-ka ‘you’. In the present-day Seoul dialect, the first 581 KOREAN person ‘I’ is pronounced ney-ka and the second person mapping of the spoken language; a certain differentia- ‘you’ is pronounced ni-ka, or ne-ka by the younger tion of vocabulary and grammar is allowed. generation. In addition to the above-mentioned The syntactic structure of a Korean sentence is very vowels, the central vowel allophone e [ ] was used by similar to that of Japanese. The word order is members of previous generations, but it has also all but S(ubject)–O(bject)–V(erb), and modifiers are located disappeared. oy [ø] has been diphthongized to [we]. It in front of the word to be modified (example provided is believed that the disappearance of these vowels has with rough literal and free translations): progressed dramatically since the 1960s. A distinction existed between long and short vowels na-nun ecekkey alumtawu-n kkwum-ul kkwu-e-ss-ta. I-topic yesterday beautiful dream-object dream-past. in the first syllable of words, which were used to distin- ‘I dreamed a beautiful dream last night’. guish meaning such as [nun] (eye) and [nun] (snow), but this long vowel is also rapidly disappearing from the Sentence types can be divided into predicate sen- Seoul dialect, with most middle-aged and younger tences integrated by a predicate (verb or adjective), speakers pronouncing the long vowel as a short vowel. and nonpredicate sentences that are not integrated by a There are two semivowels that form a rising diph- predicate. As a general rule, spoken language has a thong: j and w. j comes before all vowels except i and higher occurrence of nonpredicate sentences than writ- u, and w comes before all vowels except wu, o, and u. ten language. Unlike languages such as English, the Consonants can be categorized based on both the subject is a component of the sentence that is only point and manner of articulation, as in Table 2. present when required, and the existence of a predicate A distinguishing trait of stops and affricates is that does not necessarily guarantee that a subject will also they have three different series. /p, t, c, k/ are called be present. Even in written language in predicate sen- lenis and are nonaspirated, /ph, th, ch, kh/ are voiceless tences such as text in novels, it is normal for there to aspirated, and /pp, tt, cc, kk/ are called fortis and are be no subject in half the cases. Therefore, not only sen- voiceless sounds with tensing of the glottis. The frica- tences but also clauses may be considered a complete tives include the aspirated alveolar sound /s/ and the clause if even just a predicate is present. fortis sound /ss/, as well as the larynx sound /h/. At the Predicates take various converb forms to create sub- beginning of words, /h/ becomes a fricative [h], but in ordinate clauses such as manner clauses, reason claus- the middle of words it becomes a voiced [1] or is es, conditional clauses, and opposite clauses. The dropped. At the beginning of words, the lenis /p, t, c, k/ elements that can be contained within a subordinate become the voiceless [p][t][tʃ][k], and between other clause, such as topic words, subjects, or adverbials, are voiced sounds they become voiced [b][d][d][]. The largely predetermined by their functions. Generally, liquid sound /l/ is a flap [4] at the beginning of syllables the element that makes the proposition is situated in or before /h/, while after /l/ or at the final position in the the middle of the sentence, and the modal element syllable it becomes a lateral [l] or a retroflex lateral [5]. forms a hierarchy that surrounds it. Also, as another One trait of the stops is that when they come at the example of fixed hierarchy structure, whether or not a end of a syllable, the labial [p], lingual [t], and velar subordinate clause contains other subordinate clauses [k], instead of being pronounced explosively, are depends on its function. Multiple clause structures like stopped and formed implosively. the one below are possible, but one cannot switch the When consonants appear next to each other, various relationship between the clauses: kinds of consonant assimilation can occur, such as nasalization. [[[[ thipi-lul po-myense] pap-ul mek-u-myen ] kenkang- ey an coh-u-nikka ] pap-puthe mek-e-la ]. Grammar [[[[ while watching TV] eating ] is bad for your health ] so eat first]. ‘Watching TV while eating is bad for your Korean has two modes of existence, as spoken or health, so eat first’. [[[[ manner clause] conditional written language. The written language is not simply a clause] reason clause] main clause] TABLE 2 Labial Alveolar Alveolo-palatal Velar Larynx Stops p,ph,pp t,th,tt k,kh,kk Fricatives s,ss h Affricates c,ch,cc Nasals m n ng Liquid l 582  KOREAN In languages like English, the person and number of nals use the dative form -ey as in san-ey (to the moun- the subject dictate the predicate’s form, but in Korean, tain). For the inanimate form, locative-elative forms also the form the predicate takes is determined by what exist: san-eyse (at the mountain, from the mountain). degree the speaker respects the listener—whether the Special vocative -iye/ -ye and comitative -kwa/ -wa writ- listener is being treated politely or nonpolitely. This ten language forms also exist. Additionally, -uy is a gen- treatment level is called the speech level. For example, itive, which means ‘N1’s N2’ or ‘N2 of N1’ when used the verb ha-ta ‘to do’ originally took these six conju- as ‘N1-uy N2’, but only when the genetive is specified; gational forms spanning four different speech levels: a rudimentary genitive meaning is also understood from ha-pnita, hay-yo; ha-o; ha-ney; and ha-nta, hay. But the basic construction ‘N1 N2’ without -uy: wuli nala over the past 50 years, the modern Seoul dialect has kkoch ‘we’ ‘nation’ ‘flower’ > ‘our nation’s flower’. rapidly evolved to use only the polite forms ha-pnita There are other markers in addition to those just and hay-yo, and the nonpolite form hay. Of the polite mentioned, such as the topic or contrast marker -un/ - forms, ha-pnita is formal and hay-yo is informal, but nun, or markers that set limitations on meaning such as the latter is used far more frequently in daily conver- -man ‘only’ or -to ‘also, as well’. These markers come sation. The nonpolite hay form is used when address- after nominals. Also, it is possible to string two or ing children or subordinates, or in situations where the more markers, as in ku salam-hanthey-man-un ‘only to speaker is the same age or older than the listener and that person’. they have a close relationship. The ha-o and ha-ney The morphemes that present themselves with nom- forms are rarely used except in very limited forms of a inals are thought to be dependent words and are called kind of pseudospeech, such as television dramas. Ha- particles or postpositions by some scholars, or endings nta is used as the basic speech level form for written or clitics by others. Regardless, these case and limita- language in news, academic, and fiction texts. tion markers are attached outside of the nominative Furthermore, speech levels in spoken language can form, and they are not independent words, but they are undergo dramatic level shifts, even within the same also different from the endings that occur within words conversation with the same speakers and listeners. in languages such as Russian or German. In addition to the speech level paradigm, which is As for word classes, in addition to verbal and nom- listener-oriented, there is the honorific/nonhonorific inal, there are also adverb, interjection, and adnomi- paradigm, which is referent-oriented. The honorific nal. Adnominals modify nominals, such as on ‘the form is made by combining the verbal with the suffix whole’, in on seysang ‘the whole world’ or say ‘new’, - - si , as in ha-nta (nonhonorific) and ha-si-nta (honorif- in say sensayngnim ‘new teacher’, but they themselves ic). The honorific form is always used when the speak- do not carry inflection or declension. er expresses an event in reference to an older or superior person(s); usage of the nonhonorific form Vocabulary would be considered rude. Noun groups are called nominals (cheyen in Korean). Depending on origin, Korean vocabulary can be divid- Nouns, pronouns, and numerals belong to this group. ed into three lexical strata: (1) native words, (2) Sino- The nominal declension system is formed by aggluti- Korean words, or (3) foreign loanwords. Native words nating dependent morphemes such as salam (person: are words considered to have their origin in the Korean nominative), salam-i (the person [did something]: sub- language, such as hana ‘one’, nwun ‘eye’, salam ‘per- jective), salam-uy (person’s: genitive), salam-ul (a per- son’, salang ‘love’, or ha-ta ‘to do’. Sino-Korean son: accusative), salam-hanthey ([someone gave words are words that have their root in classical something] to a person: dative), salam-hantheyse (from Chinese like il ‘one’, san ‘mountain’, or chayk ‘book’. a person: ablative), salam-ulo (by a person: instrumen- Foreign loanwords are borrowed words from English tal), salam-a (oh, person!: vocative), and salam-hago or other languages, such as khemphyuthe ‘computer’ (with a person: comitative). Of these, the subjective - or khephi ‘coffee’. The Sino-Korean category also i/ka, accusative -ul/ -lul, instrumental -ulo/ -lo, and voca- includes Japanese words made from Chinese charac- tive -a/ -ya have multiple forms; the former is used with ters, such as chelhak ‘philosophy’ or yepse ‘postcard’. nominals that have consonant stems, while the latter is Newspapers and research papers are usually more than used with nominals that have vowel stems. For the sub- 50% Sino-Korean word content, while at the other end jective and dative cases, as in sensayngnim-kkeyse (the of the spectrum, literary works can have more than teacher being the subject) and sensayngnim-kkey (to the 70% native word content. Generally speaking, basic teacher), the honorific form differs from the non- Korean vocabulary contains more native words than honorific form. Also, the dative case for animate nomi- the other two types. Numerals have both native word nals is salam-hanthey in spoken language but and Sino-Korean word forms, and each is used salam-eykey in written language, while inanimate nomi- depending on the co-occurring words. 583 KOREAN Onomatopoeia is even more plentiful in Korean Kono, Rokuro. 1979-1980. Kono Rokuro Chosakushu than in Japanese, with over 8,000 examples found in a (Selected writings of Kono Rokuro). Tokyo: Heibonsha. Kwuklip-kwuke-yenkwuwen (National Academy of Korean typical dictionary. Also, even for a single adjective like Language). 1999. Phyocwun Kwuke-taysacen (Standard ‘red’, vowel/consonant alternation results in a variety Korean dictionary). Seoul: Twusantonga. of forms like ppalkah-ta, ppelkeh-ta, palkah-ta, and Lee, Hyun Bok. 1989. Korean grammar, Oxford: Oxford pelkeh-ta, where the pronunciation shift serves to dis- University. tinguish between various nuances through a well- Lee, Ki-moon. 1972;1998. Sincengphan Kwukesa Kaysel (Revised introduction to the history of Korean language). developed system of sound symbolism. Korean also Seoul: Thayhaksa. has a rich abundance of adjectives. Martin, S.E., Lee,Y.H, Chang, S-U. 1968. New Korean English dictionary. Seoul: Minjungseorim. References Martin, S.E., Martin, S.E. 1992. A reference grammar of Korean. Tokyo: Charles E.Tuttle. Chang, Suk-Jin. 1996. Korean. Amsterdam and Philadelphia: Nam, Ki-Shim, and Ko Yong-kun. 1985;1993. Phyocwun John Benjamins. Kwuke Mwunpeplon (Standard Korean grammar). Seoul: Hankwuk Cengsin-mwunhwa Yenkwuwen (The Academy of Thapchul-phansa. Korean Studies). 1991. Hankwuk Mincok-mwunhwa Tay- Noma, Hideki. 2002. Hankwuke Ehwiwa Mwunpep-uy paykkwa Sacen (Encyclopedia of Korean Culture). Seong- Sangkwan-kwuco (Korean: correlative construction of nam, Korea: Hankwuk Cengsin-mwunhwa Korejskogo. vocabulary and grammar). Seoul: Thayhaksa. Kholodovich, A.A. 1954. Ocherk grammatiki korejskogo jazy- Sohn, Ho-min. 1994. Korean. London: Routledge. ka (Introduction to Korean Grammar). Moskva: Izdatel’stvo Suh, Cheong Soo. 1996. Kwuke-mwunpep (Korean grammar). literatury na inostrannykh jazykakh. Seoul: Hanyang University Press. King, J.R.P. 1987. An introduction to Soviet Korean. Language HIDEKI NOMA research, Vol. 23, No. 2. Seoul: Language Research Institute: Seoul National University. See also Altaic; Japanese Krio (and West African Pidgin English) Krio is an English-lexicon creole spoken natively by with (in this order) both Gullah and Jamaican Maroon about half a million people in the Freetown area of speech, as well as with preexisting African varieties of Sierra Leone, but is used as an auxiliary language by Pidgin English. Although the question is not uncontro- virtually all the country’s inhabitants. The foundations versial, most observers agree that Krio is in essence a of the language were laid in 1787, when a settlement New World creole. was founded in Sierra Leone intended to receive freed Compared to other Atlantic English-lexicon creoles, Blacks from Britain. These were later joined by two Krio is typologically rather distant from its lexifier, major groups of immigrants. The largest of these was something that is visible not least in its phonology. a contingent of about 1,200 American ex-slaves who Krio has seven oral and five nasal vowels, but it has had fought on the British side against the Americans, been suggested that earlier forms of the language oper- and who arrived by way of Nova Scotia. A couple of ated with a five-vowel system. In the consonantal sys- years later, 550 Jamaican Maroons (slave rebels) were tem, the lack of /z/ and /h/ is noteworthy. Sporadic transported to the new colony. As the groups merged, cases of depalatalization of /∫/ may suggest that this Krio emerged as the language of the new population. too was lacking in older Krio. Although /v/ is phone- After the abolishment of the slave trade in 1807, the mic, it has become /b/ in the noninitial position. British navy began intercepting slave vessels bound Furthermore, there is reason to assume that earlier for the Americas, landing the recaptives (in all, almost forms of Krio had less liberal phonotactics than 100,000) in and around Freetown. Within a short time, English, with a marked preference for CV syllables. they outnumbered the original settlers. Given their On the other hand, contact with African languages has ethnolinguistic heterogeneity, the Recaptives also introduced coarticulated stops. adopted Krio as their language, and by the second half Krio is a tone language opposing high and low of the nineteeneth century, all Freetown ethnicities had tones in a small number of minimal pairs. merged to one single Creole population. Lexically, Krio is marked in part by its continuing Given the settlement history of the Freetown penin- contact with local languages (mainly Temne, Sherbro, sula, it is not surprising that Krio shows similarities and Mende), but even more by influence from the 584  KRIOL (ROPER RIVER CREOLE) languages of the Recaptives. Thanks to these, Krio It then spread inland, until it came to be spoken in features hundreds of items of Yoruba origin. most of southwestern Cameroon. Cameroon is offi- Characteristic features in Krio that appear to be of cially divided into an Anglophone and a Francophone Transatlantic origin include the completive dɔn, the zone. While the WAPE of the former is essentially an preposed pluralizer dεm, serial comparatives involving extension of Nigerian WAPE, the latter area presents pas, existential and locative copula lib, habitual blant, some unique features due to the absence of English as equative copula na, progressive de, locative copula de, a prestige variety. It is more conservative (thus closer 2pl una, relativizer we, and habitual kin. However, a to its Krio roots), and there is a noticeable impact of smaller, but far from negligible, number of items can be French in the lexicon. Cameroonian WAPE may be demonstrated to represent local African developments. spoken by as many as 70% of the country’s popula- These include the multipurpose preposition fɔ, the pro- tion, including virtually everybody in its southwestern gressive lib fɔ, the modal fit ‘be able to’, and the pos- parts. It appears to have acquired a larger proportion of sessive construction of the type ‘John his book’. native speakers than Nigerian WAPE. While Krio has no official status, there does exist a In Ghana, WAPE is less widespread than else- sizeable body of literature, consisting both of original where, and it does not seem to have any native speak- works and translations. In particular, Sierra Leonean ers. It is nevertheless used emblematically in several radio makes use of it in its news broadcasts. sectors of Ghanaian society, in particular between The familiarity of Sierra Leoneans with western men and in milieux such as schools, universities, and culture and their ability to speak ‘English’ favored in the police corps. It was apparently imported in the their employment as middlemen in the British late nineteenth or early twentieth century by Sierra colonies. This was of crucial importance for the lin- Leoneans in British service. Related varieties have guistic makeup of much of West Africa, where vari- also been attested in Togo and Côte d’Ivoire, although eties (in part) derived from Krio are referred to as West the prestige of French has prevented its further spread African Pidgin English (WAPE). in these countries. Krio was first implanted on the island of Fernando Since the nineteenth century, finally, a variety of Poo (now part of Equatorial Guinea) by the British, Krio called Aku is spoken by a community numbering who founded a naval base there in the early nineteenth some 6,000 in the Gambia. century. The native-speaking community is only a couple of thousand strong, but the local variety of Krio has developed into the island’s lingua franca. Its posi- References tion was later strengthened by massive immigration of Dwyer, David. 1967. An introduction to West African Pidgin Nigerians in the mid-twentieth century. English. East Lansing: Produced for the United States Peace Sierra Leone Creoles were engaged in the British Corps by the African Studies Center, Michigan State conquest of Nigeria, not only because of their being University. Faraclas, Nicholas. 1996. Nigerian Pidgin. London, and New Anglophile but also because of their Recaptive ances- York: Routledge. try having retained links with local Nigerian cultures. Féral, Carole de. 1980. Le pidgin-english camerounais: Essai de At least a third of the Nigerian population is now définition linguistique et sociolinguistique. Nice: Publication believed to know WAPE, making it (if L2 speakers are de l’Institut d’Études et de Recherches Inerethniques et included) the most widely spoken pidgin/creole lan- Interculturelles. Fyle, Clifford, and Eldred Jones. 1980. A Krio-English dic- guage in the world. Nativization is attested from the tionary. New York: Oxford University Press/Sierra Leone early 1960s, and there are a sizeable number of University Press. Nigerians speaking WAPE as their first language. MIKAEL PARKVALL In Cameroon, WAPE was first implanted by mis- sionaries expelled from Fernando Poo by the Spanish. See also Nigeria Kriol (Roper River Creole) Kriol is one of two large nontraditional indigenous lan- Strait Islands). Kriol is spoken by between 15,000 and guages of Australia (the other being Kep Yok (Cape York 20,000 people in an area covering in excess of 350,000 Creole), spoken in Northern Queensland and the Torres square kilometers (150,000 square miles). Kriol is spoken 585 KRIOL (ROPER RIVER CREOLE) throughout the Kimberley region of Western Australia, in made polysyllabic by epenthesis. Some correspon- the Northern Territory as far south as the Barkly, and dences between SAE and Kriol include: extensively in Western Queensland. The spread of pidgins (1) SAE Kriol through this area during the late nineteenth and early twentieth centuries was due to the cattle and livestock fellow [fεloυ] bala/pela [bala]/[pεla] river [rivə] riba [riba] industry, but creolization occurred more recently, possibly that [ðæt] det/jat [dεt]/[at] in mission settlements. There is some evidence that these visit [vIzIt] bizit/bijit [bIzIt]/[bIIt] pidgins actually had their origins in the Sydney area with another [ən3ðə] naja [naa] the earliest European settlements of the late eighteenth been [bi:n] bin [bIn] century. Kriol is now the primary means of communica- all [ɔl] ol [ol] sneak [sni:k] jinik [InIk] tion among aboriginal people in the above-mentioned area, whose main social mobility is in the direction of broader aboriginal contacts, and not economically Morphology ‘upward’ into the dominant white Australian society. Pronoun forms come from English morphemes, but While early investigations (Kaldor and Malcolm the paradigm (given in (2)) is typical of traditional 1979; Sandefur 1979) identified Kriol as a postcreole languages. There is a contrast between dual (two peo- continuum with Standard Australian English (SAE), ple) and plural (more than two), and a contrast Sandefur (1982) makes an excellent case for consider- between whether the term ‘we’ includes the person ing Kriol and SAE to be distinct languages. Apparent being spoken to or not. There is no gender contrast, mesolects (‘lighter’ vs. ‘heavier’ creoles) are actually and hence no distinction between ‘he’ and ‘she’. There interlanguages, which Kriol speakers learning SAE are no separate subject and object forms, unlike both use in their dealings with non-Aboriginal people. SAE and most traditional languages. Attitudes toward Kriol differ widely throughout the region. Most older speakers do not know the term (2) person singular dual plural ‘Kriol’ at all, and identify their language as either 1 addressee mi yunmi wi ‘Pidgin’ or ‘English’. Of those speakers who identify it included as either ‘Kriol’ or ‘English’, attitudes vary from disdain addressee - mindubala mibala to pride. The attitude of educators had formerly been to excluded 2 yu yundubala yubala/yumob treat Kriol as a corrupt or substandard form of English, 3 im dubala ol/olabat/dei and to attempt to eradicate it. However, most govern- ments and education systems now recognize Kriol as a language in its own right. In some areas, bilingual While Standard Australian English has a plethora of Kriol–English programs exist, although these are con- prepositions (e.g. of, at, for, with, etc.), traditional stantly under threat from both local communities and languages lack similar separate words, instead using a the (ever-changing) centralized governments. Within narrow range of suffixes to indicate these relationships. some aboriginal communities, the recognition of Kriol Kriol does not have such suffixes, and it does not have is seen as a threat to the maintenance of traditional abo- the broad range of prepositions either. Instead, it uti- riginal languages, all of which are severely endangered. lizes two prepositions, each covering many functions: Kriol is based on the grammar and vocabulary of la or langa are used to identify locations. la riba can SAE, although its phonology, morphology, and to mean ‘in the river’ or ‘at the river’. blanga marks pos- some extent its syntax show enormous influence from session or association, as in dis ngawu blanga Jingili, traditional aboriginal languages. meaning ‘this (is the) home of the Jingili people’. Kriol has a particle na that is used to indicate emphasis. Im bin bogi-bogi means ‘she or he was Phonology swimming’ (literally: ‘she or he been swim-swim’; the Most traditional languages lack sounds like f, v, s, z, verb bogi is ‘reduplicated’ to indicate an action in and th, and this property is carried over into many vari- progress or a repeated action). Im na bin bogi-bogi, on eties of Kriol (many varieties have s but not the other the other hand, means ‘she or he was swimming’ or ‘It sounds). While most traditional languages have three- was her who was swimming’. vowel (a/i/u) systems, Kriol has a five-vowel If verbs take an object, they are marked with a suf- (a/e/i/o/u) system (which is found among some tradi- fix -im in Kriol. -im is derived from English ‘him’. Sili tional languages in Northern Australia). Although baga im bin dringk-im means ‘the silly bugger drank it’ many traditional languages distinguish long from short (literally ‘silly bugger it been drink-it’), and ol bin pik- vowels, most dialects of Kriol do not. Initial and final im-ap manggo means ‘they picked up mangos’ (literal- consonant clusters are typically either reduced or ly ‘they been pick-it-up mango’). Notice that the im ‘it’ 586  KRIOL (ROPER RIVER CREOLE) in the first sentence can be left out, since the -im mark- References ing on the verb already indicates that the verb is transi- Harris, John W. 1986. Northern Territory Pidgins and the origin tive, i.e. the sentence is about the fact that the silly of Kriol. Canberra: Pacific Linguistics C 89. bugger drank something in particular. Kriol does not Kaldor, Susan, and Ian Malcolm. 1979. The language of the make a distinction between definite ‘the silly bugger’ school and the language of the Western Australian aborigi- and indefinite ‘a silly bugger’. The context must con- nal schoolchild—implications for education. Aborigines of the West: their past and present, ed. by Ronald M. Berndt vey what is meant. Kriol does not distinguish between and Catherine H. Berndt, 406–37. Nedlands: University of singular ‘mango’ and plural ‘mangos’ either. Here, Western Australia Press. again, the listener has to decide which interpretation is Koch, Harold. 2000. The role of Australian Aboriginal lan- more appropriate in the situation. guages in the formation of Australian Pidgin grammar: tran- Here is a brief Kriol text from Newcastle Waters in sitive verbs and adjectives. Processes of language contact: case studies from Australia and the Pacific, ed. by Jeff the Barkly Tableland, which illustrates many of the Siegel. Montreal: Fidas. properties discussed above: Sandefur, John. 1979. An Australian Creole in the Northern Territory: a description of Ngukurr-Bamyili dialects, Part 1. ‘Wel, dat dem pipul bin gu solja, wantu fait langa naja Work Papers of SIL-AAB, series B, Vol. 3. pleis, blanga naja pipul, dei bin blanga kool blad. Dey ——— 1981. Kriol—an Aboriginal language. Hemisphere 25. bin jinik ap an fait langa alabala waj deya wel bin fait. 252–6. Sambala dei bin lusim pipul dei bin get supia tru.’ ——— 1982. Kriol and the question of decreolisation. International Journal of the Sociology of Language Well, some people went off to make war [go soldiers], 36. 5–13. to fight in another country, another people’s country, ——— 1991. A sketch of the grammar of Kriol. Language in they went in cold blood. They snuck up and fought with Australia, ed. by Suzanne Romaine, 204–12. Cambridge: all the people there and there was a battle. Some of Cambridge University Press . them lost people when they got speared through. ROB PENSALFINI 587  L Labov, William Although William Labov took many years to follow ‘labovian framework’, with no capitalization or any his true calling in linguistics, he made good use of the attribution as to the year or the research involved. time and the life experiences that he acquired in the Implicit in the variationist approach is the concept of nonacademic world. Following ideas propagated by the ‘speech community’. A speech community tradi- his mentor, Uriel Weinreich, he brought to the field of tionally referred to a group of people who were linguistics the vision that data could be naturalistical- delimited solely by linguistic criteria. The definition ly collected and empirically tested and analyzed. In of a speech community in the variationist approach, the 1960s and 1970s, this was considered an extreme however, puts much less emphasis on language as a contrast from the ideal speaker/hearer dyad data used necessary and sufficient condition for group member- by Noam Chomsky and his disciples. Labov elaborat- ship. Instead, it suggests that the three key require- ed and finessed a style of linguistics that was practi- ments for membership in a linguistic/social group are cally unheard of at the time––sociolinguistics. Using the existence of a social community, that patterns of the concepts of language variation, the linguistic vari- language use are shared by its members, and that able, and the interactions of social factors with both of norms for the social evaluation of speech are shared these, Labov literally created a new style of linguistics. by community members. Using data from Martha’s Vineyard, New York City, This shared set of social attitudes posited for a and Philadelphia, he proved that language change can speech community were proposed in what has come be studied and documented as it occurs. to be known as the classic variationist research, After working as a chemist for 11 years, Labov Labov’s (1966) study of New York City speech. In went back to school, and created a coup d’état with this work, he established a range of speech styles that his master’s thesis work. He conducted research on included word lists, minimal pairs, reading passages, the small island of Martha’s Vineyard, Maine. His interview style, and casual style. He posited that the work on the pronunciations of the vowels of the linguistic variable that was under scrutiny (r-fulness native islanders was groundbreaking because he or r-lessness, for example) would reveal patterned found that, although the variation of the vowels variation by speakers, which could be correlated with seemed random, it was in fact related to occupation, the independent variable of stylistic range. His geography, and ethnicity. This was one of the first hypothesis was that, since the amount of attention studies that incorporated social factors into linguistic paid to speech is highest (more closely monitored, in variation, a style of investigation that began to be Labov’s terms) while reading lists of words, and the referred to as ‘variationist’ research. This study and least amount of attention is paid to pronunciation in a further works by Labov are so well accepted as the casual setting, there would be a range of variation on linchpins of variationist research that the concepts going from the most ‘standard-like’ pronunciation in contained in them are often simply referred to as the the most highly monitored speech (the word lists) to 589 LABOV, WILLIAM the least ‘standard-like’ pronunciation in the least It occurs both throughout time and in the present, monitored speech (the casual style). Although the and can be shown by studying language change data collection methods were similar for both the in ‘real time’, as he did in his Martha’s Vineyard Martha’s Vineyard and the New York City studies, study (Labov 1972), or by studying language change they showed diametrically different results. In in ‘apparent time’, as he did with his study of the Martha’s Vineyard, the locals were using the ‘marked’ evolution of New York City vowels (Labov 1966, vowel pronunciation as a way of showing pride in 1972). their heritage, while in New York, the ‘lower class’ In an attempt to quash criticism from those who pronunciation (r-lessness) gave way to the more pres- claim that the variationist approach is too subjective, tigious variety of the pronounced ‘r’ when social fac- statistical programs that take into account and corre- tors came into play. late social and linguistic variables have been written. In Sociolinguistic patterns, Labov (1972) set out to This results in quantitatively analyzed data collection prove yet again that language should not be separated and analysis and allows for cross-linguistic and cross- from the culture and society in which it finds itself. cultural comparisons. Labov has recently written a Labov was certainly not the first linguist to advocate program called Plotnik for plotting, analyzing, and this concept. Older works by Meillet (1921) and comparing vowel systems. In addition, in recent years, Fischer (1958) also recognized the fact that social fac- he has also tackled the issue of forensic linguistics by tors and language use are inextricably intertwined and giving expert testimony in court cases, promoted the suggested that one could not be adequately described cause of accepting African American Vernacular without accounting for the influence of the other. English as an accepted and normal language in and of However, Labov carried this idea a step further by itself, promulgated the idea that linguists should give applying it not just to synchronic processes of lan- something back to the community from which they guage change, but to diachronic processes as well. He take their data, and overseen the creation of the used data from a series of studies done on ‘change in ‘Phonological Map of North America’. progress’ to show that various functions of language The most important concept to come out of his can determine and force language change, that gram- research was the idea of the linguistic variable. mar rules can (to a certain extent) be affected by Although dialectologists have long acknowledged changes in society, and that linguistic evolution or the fact that different people speak in different ways, change is not dysfunctional, and therefore ‘(not a) no one had taken into account the fact that the same massive testimony to original sin’ (p. 323). person speaks differently on different occasions. It Labov (1972) also examined the philosophy and was commonly thought that this intraspeaker varia- views of various linguists about accepting and using tion was random, and therefore unquantifiable. With social factors as one explanation of linguistic change. the advent of variationist sociolinguistics, it can now He suggested the following dichotomy: a ‘social’ be seen that the variation is systematic, and is the group that believes that one can see change in result of the interaction between linguistic variables progress, follows the ‘wave model’ of linguistic evo- and social factors. Rather than variables (or linguis- lution, and believes that social factors can and do tic elements) being in free variation (or random), affect language, vs. an ‘asocial’ group that disregards they suggest that variables have been and are condi- the effects of linguistic diversity and contact between tioned by certain social and cultural factors. languages, works with a Stammbaum or family tree Chambers (1995) puts this quite strongly by noting model of linguistic evolution, and does not believe that correlating linguistic variation as the dependent that social factors can help explain linguistic change. variable with independent variables such as linguis- Labov places himself squarely in the first group, and tic environment, style, or social categories is the pri- argues persuasively that social variation does have a mary empirical task of sociolinguistics. The place in not only synchronic but also diachronic linguistic fallout from this belief has resulted in an description of change. These musings on philosophy explosion of work dealing with the effects of gender and its interconnection with language are carried on language, social class on language, race and eth- even further in Labov (1994), in which he compares nicity on language, education level on language, and linguistic change to geographical change, and notes geographical location on language. The interplay of that both can be either gradual or catastrophic, and these factors and their effect on speech can now be that both may put linguists in a better position to cor- used to view language change in progress. In the relate the social and linguistic factors and respond past, language change had to be tediously recon- better to Meillet’s suggestion that social change is the structed using ancient texts and sheer speculation. only possible alternative to explain linguistic change. One of the most exciting facets of the variationist For Labov, variation is a reflex of change in language. approach to linguistics is the fact that it can be used 590  LADEFOGED, PETER as a key to the past. Language change is invariant; all References languages change. The modern-day methods pio- Bailey, Charles-James N.1973. Variation and linguistic theory. neered by Labov and others can be used as a tem- Arlington, VA: Center for Applied Linguistics. plate from which to examine previous changes that Chambers, Jack. 1995. Sociolinguistic theory. Oxford: Blackwell languages have undergone as well as to document Press. the changes that speakers impose on their languages Fasold, Ralph. 1990. The sociolinguistics of language. Cambridge: Blackwell. every day. Fischer, John L. 1958. Social influences on the choice of a lin- guistic variant. Word 14. 47–56. Labov, William. 1963. The social motivation of a sound change, Biography Word 19. William Labor was born in Rutherford, New ———. 1966. The social stratification of English in New York City. Jerseyon, December 4, 1927. He recived his B.A. ———. 1969. Contraction, deletion, and inherent variability of (1948) from Harvard College; M.A. (1963) from the English copula. Language 45. 715–62. Columbia University for a thesis on social aspects of ———. 1972. Language in the Inner City: studies in the Black sound change on Martha’s Vineyard, tutored by Uriel English vernacular. Philadelphia, University of Pennsylvania Weinreich; and Ph.D. from Columbia for work on Press. ———. 1972 Sociolinguistic patterns. Philadelphia, PA: class differences in the dialects of New York City. He University of Pennsylvania Press. was Assistant Professor, Columbia University in: ———. 1975. What is a linguistic fact. Lisse, The Peter de 1964–1970, and Associate Professor, Professor, Ridder Press. University of Pennsylvania (1971–present). He was a ———. 1994. Principles of linguistic change: Vol. 1, Internal Guggenheim Fellow (1970–1971 and 1987–1988), factors. Oxford: Basil Blackwell. Meillet, Antoine. 1921. Linguistique historiique et linguistique President, Linguistic Society of America (1979), and generale. Paris: La Societe Linguistique de Paris Fellow, American Association for the Advancement Saussure, Ferdinand de. 1916. Cours de linguistique generale. of Science (1997–present). He received honorary Paris: Payot. Ph.D.s form University of Uppsala (1985), Wardhaugh, Ronald. 1986. An introduction to sociolinguistics. University of Liege (1990), University of York Oxford: Blackwell. Weinreich, Uriel. 1968. Languages in contact: findings and (1998), and Haifa University (1999). He was Sapir problems. The Hague: Mouton. Professor, Linguistic Institute in 1986, and is since Wolfram, Walt, and Natalie Schilling-Estes. 1998. American 1988 to date Editor of Language Variation and English. Oxford: Blackwell. Change (1988–present). MEGAN E. MELANÇON Ladefoged, Peter Like his teacher David Abercrombie, Peter Ladefoged who was himself the son of a poet, Lascelles continues the tradition of British phonetics and lin- Abercrombie. I think it is this early exposure to exper- guistics, which includes the names of such scholars as imental sciences that propelled Ladefoged to concen- Henry Sweet, Daniel Jones, and John Rupert Firth (all trate on experimental phonetics. Ladefoged’s student three of whom are featured in the photograph opposite at UCLA, John J. Ohala of the University of the title page of Asher and Henderson (1981), one of California, Berkeley, has called the maestro the the major sources in phonetic historiography, and with ‘world’s leading phonetician’, with which I am in firm the exception of Firth, specifically credited for their agreement. Born in England and educated in Scotland, pioneering research by Ladefoged in A course in pho- the long-time UCLA professor of phonetics and direc- netics (2001:ix)). Ladefoged began his career at the tor of the UCLA Phonetics Laboratory (currently age of 17 studying physics, chemistry, and mathemat- UCLA Research Linguist and Professor of Phonetics ics at Cambridge University, but after a stint in the Emeritus) is practically a household name in many lin- army, he decided to study poetry. By chance, he hap- guistics departments in the United States and around pened to run into a man named David Abercrombie, the world. Countless numbers of students (including who had just started up a Phonetics Department, and mine, ever since the first edition in 1975) have been 591 LADEFOGED, PETER introduced to phonetics from the four editions of his A acoustic rather than articulatory properties of sounds. course, not to mention his numerous other books and Also, he has shown that vowels are better described in articles (see under References for typical examples). terms of auditory rather than acoustic or articulatory Many more readers will be exposed to phonetics from parameters (Johnson et al. 1993). another survey work of his that appeared recently, In addition to his fieldwork, Ladefoged has been Vowels and consonants: An introduction to the sounds influential in linguistic terminology and theory. of languages (Oxford: Blackwell, 2001), which is less Ladefoged (1964) e.g. coined the term ‘approximant’, technical than A course in that it omits many detailed which has gained widespread acceptance in the pro- phonetic concepts and is couched in simpler language; fession as synonymous with semivowel and friction- however, there is, not surprisingly, some overlap less continuant. It is so named because one articulator between the two, such as with the vowels of English approximates the position of another. He has demon- ((pp. 26-30) and A course (p. 29, p. 81)) and the con- strated a number of interesting phonetic laws or ten- trasting stops in Malayalam ((p. 149)) and A course dencies; e.g. that there is no clear one-to-one ((p. 141)). At first, however, it may seem more com- correlation between syllables and chest-pulse move- plex than A course since it starts with acoustic phonet- ments, or that apical trills are more common than lam- ics (it is Chapter 8 of 11 chapters in A course) –– a inal trills. However, it was as the President of the Ladefogian specialty ever since the publication of his International Phonetic Association that he seized the The nature of vowel quality (1962) and his Elements of opportunity to propound the Kiel Convention of 1989, acoustic phonetics (1962). which was organized to revise the International One of Ladefoged’s most outstanding contributions Phonetic Alphabet (IPA). As a researcher of rare to phonetic science has been the publication of his speech sounds himself, such as bilabial trills, voiceless phonetic fieldwork and the insights gleaned therefrom. implosives, and velar laterals, he was instrumental in He has conducted field research in Africa, India, making contributions to the adopted IPA revision. It is China, Papua New Guinea, Australia, etc., recording safe to report that he was the one who first noticed a hundreds of languages over the years abroad or at the flapped labial stop in an ideophonic word in Margi (a UCLA Phonetics Laboratory. These recordings, which Chadic language of Nigeria). Ladefoged should also are housed in the UCLA phonetics archives, provide be credited for inventing new ways of looking at the requisite data for the phonetic theory to which speech sounds. In fact, he is largely responsible for Ladefoged has contributed so much (see e.g. insisting that the IPA chart be regarded as a one-page Ladefoged 1965). As a pioneer in experimental pho- theory of phonetics. netics (see e.g. Ladefoged 1967), including still and Distinctive feature theory is yet another area where cine x-ray photography (see as far back as Ladefoged Ladefoged’s work has been peremptory. Critical of the 1958), palatography, spectrograms, waveforms, and features proposed in Chomsky and Halle (1968), aerodynamic data techniques, Ladefoged’s career has Ladefoged (1971) proposed a new system, which rec- been devoted to the phonetic phenomena of the ognized the importance of an auditory basis for some world’s languages (see e.g. Gordon et al. 2000), and as features. As a result of this, it is now accepted that some a result of this type of work, there now exists a scien- sounds are to be grouped together on the basis of audi- tific basis for distinguishing linguistic from nonlin- tory phenomena. In view of his numerous accomplish- guistic sounds. Thus, through the theoretical and ments described above, there can be little doubt that he experimental sides of his research (and that of many of has earned his noteworthy international reputation. his colleagues), we now know enough of the sounds of the world’s languages for there to be a book with that Biography very title (Ladefoged and Maddieson (1996)). For Ladefoged and his student John J. Ohala, pho- Peter Ladefoged was born in Sutton, Surrey, England netics is but part of phonology, the goal of which is to on September 17 1925. He attended Haileybury understand all aspects of sound patterning in lan- College, Hertford (1938–1943). He received his M.A. guages (see Ohala 1994:3053). It is precisely this kind (1951) and Ph.D. (1959) from the University of of emphasis that prompts Ladefoged and Maddieson Edinburgh for a dissertation on the acoustics of vowel (1996:7) to assert that the typical five-vowel system quality, with Daniel Jones, recently retired from the for many languages is the more distinct set /i e a o u/ chair of phonetics at University College, London, rather than their rounded (or more marked) counter- serving as a consultant to the project. He was Assistant parts /y, ø, a, γ, ɯ/ (/a/ is neutral and is included in Lecturer in Phonetics, University of Edinburgh in both sets). Further, Ladefoged has demonstrated 1953–1955 and Lecturer in Phonetics, University of repeatedly that some phonological statements may be Edinburgh, 1955–1959, University of Ibadan, better formulated and understood by reference to the 1959–1960, and University of Edinburgh, 1960–1961. 592  LAKOFF, GEORGE He was a Field Fellow in the West African Languages Johnson, Keith, Peter Ladefoged, and Mona Lindau. 1993. Survey in 1961–1962. He was also Assistant Professor Individual differences in vowel production. Journal of the Acoustical Society of America. 94. 701–14. of Phonetics, UCLA, 1962—1963; Associate Ladefoged, Peter. 1958. Use of palatography. Journal of Speech Professor of Phonetics, UCLA, 1963–1965; and and Hearing Disorders 22. 764–74. Professor of Phonetics, UCLA, 1965–1991. He: was ––––––. 1962. Elements of acoustic phonetics (2nd edition with Chair, Department. of Linguistics, UCLA in added chapter on computational phonetics, 1996). Chicago: 1977–1980, and Fellow, Acoustical Society of University of Chicago Press. ––––––. 1962. The nature of vowel quality. Monograph supple- America and American Speech and Hearing ment to Revista do Labóratorio de Fonética Experimental de Association. He was a recepient of the Distinguished Faculdade de letras da Universidade de Coimbra. Teaching Award and Research Lecturer, UCLA, in ––––––. 1964. A phonetic study of West African languages. 1972 and 1989, respectively. Ladefoged was Cambridge: Cambridge University Press. President, Linguistic Society of America in 1978; ––––––. 1965. The nature of general phonetic theories, Georgetown University Monographs on Language and President, International Phonetic Association in Linguistics. 18. 27–42. 1987–1991; Fellow of the American Academy of Arts ––––––. 1967. Three areas of experimental phonetics. Oxford: and Sciences, 1990; and UCLA College of Letters and Oxford University Press. Science Faculty Research Lecturer in 1991. He ––––––. 1971. Preliminaries to linguistic phonetics. Chicago: received a Gold Medal at the XII International University of Chicago Press. ––––––. 1975. A course in phonetics, 4th edition, 2001. Fort Congress of Phonetic Sciences in 1991. He was also Worth: Harcourt. Corresponding Fellow of the British Academy in ––––––. 2001. Vowels and consonants: an introduction to the 1992. He also received an honorary D. Litt. from the sounds of languages. Oxford: Blackwell. University of Edinburgh in 1993, and a Silver Medal Ladefoged, Peter, and Ian Maddieson. 1996. The sounds of the from the Acoustical Society of America in 1994. World’s languages. Oxford: Blackwell. Ladefoged, Peter, and D. Everett. 1996. The Status of Phonetic Rarities. Language. 72. 794–800. Ladefoged, Peter, and Matthew Gordon and Pamela Munro. References 2000. Some phonetic structures of Chickasaw. Asher, R. E. and J. A. Henderson (eds.) 1981. Towards a histo- Anthropological Linguistics 42. 366–400. ry of phonetics. Edinburgh: Edinburgh University Press. Laver, John. 1994. Principles of phonetics, Cambridge: Chomsky, Noam and Morris Halle. 1968. The sound pattern of Cambridge University Press. English. New York: Harper & Row. Ohala, John J. 1994. Phonetics, ed. by R. E. Asher, The ency- Fromkin, Victoria A. 1994. Peter Ladefoged (1925- ). The clopedia of language and linguistics, Vol. 6, 3051–3. encyclopedia of language and linguistics, Vol. 4, ed. by R. E. Oxford: Perganon Press. Asher, 1891–2. Oxford: Pergamon Press. ALAN S. KAYE Lakoff, George George Lakoff had his first contact with Linguistics (ii) the universal-base hypothesis: initial representa- while he was an undergraduate student at MIT. There, tions of derivations are cross-linguistically identical he worked on the structure of discourse, and tried to put logical representations, and (iii) the derivation of a sen- a narrative structure into a transformational framework. tence is a direct transformational mapping from seman- In those days, one of the basic tenets of transforma- tics to surface structure. tional grammar was the independence of syntax from Despite its initial success, Generative Semantics semantics. However, beginning in 1963, Lakoff started started to fade away in the mid-1970s. One of the main to find evidence that this central claim was not totally reasons was the growing body of evidence about the correct. Together with other linguists such as James D. embodiment of mind from the fields of cognitive sci- McCawley, Paul Postal, and John Robert Ross, he ence and neuroscience. Work by Charles Fillmore in developed the theory called Generative Semantics. Its frame semantics, Paul Kay in color categorization, main hypothesis is that there is no principled difference Leonard Talmy in universal spatial cognition, and between syntactic and semantic processes. Along with Eleanor Rosch in basic categories contradicted one of this, there are three other main assumptions: (i) nonex- the major tenets of this theory, namely the belief in istence of a purely syntactic level of ‘deep structure’, formal logic and model-theoretic semantics. This new 593 LAKOFF, GEORGE research showed that it was not possible to define career as a linguist and cognitive scientist, George meaning in terms of reference and truth conditions, in Lakoff has been a founder of three different intellectu- terms of the objective structure of the external world, al movements: Generative Semantics, Cognitive with no reference to the brain or mind. Therefore, for- Linguistics, and the Neural Theory of Language. mal logic could not be an adequate theory of natural language semantics, and neither could be transforma- tional grammar of syntax. As a consequence, Lakoff Biography stopped doing Generative Semantics and set about George Lakoff was Born in Bayonne, New Jersey on, developing a field of Cognitive Linguistics. May 24, 1941. The graduated from MIT in 1962 in Cognitive Linguistics views linguistic knowledge English Literature and Mathematics, and studied as part of general cognition and thinking; linguistic Linguistics with Roman Jakobson, Morris Halle, and behavior is not separated from other general cognitive Noam Chomsky. He received his doctorate in abilities that allow mental processes of reasoning, Linguistics in 1966 from Indiana University for his memory, attention or learning, but is understood as an work on the irregularity in syntax and transformation- integrated part of it. One of Lakoff’s major contribu- al grammar. He was a Research Fellow and Lecturer in tions to the theoretical foundations of this framework Linguistics at Harvard (1965–1969). Then, he worked is the theory of conceptual metaphor. as associated Professor of Linguistics at the University In the late 1970s early 1980s, Lakoff, then working of Michigan (1969–1972). He moved to the University in collaboration with Mark Johnson, discovered that of California at Berkeley, where he is Professor of the classical view of metaphor as a mere figure of Linguistics since 1972. He has been a fellow of the speech based on inherent similarity to indicate a com- Institute for Advanced Study in the Behavioural parison was empirically false. Instead, they showed Sciences at Stanford (1971–1972), and a visiting pro- that metaphor is part of the cognitive unconscious. It is fessor at the École des Hautes Études en Sciences a cognitive mechanism that allows us to experientially Sociales in Paris (1995). He served on the Governing ground conceptual systems and to reason about Board of the Cognitive Science Society (1990–1996), abstract domains (e.g. love) in terms of more concrete as President of the International Cognitive Linguistics domains (e.g. journeys). In other words, metaphor is a Association (1990–1994), and as a member of the cross-domain conceptual mapping. Lakoff has applied Science Board of the Santa Fe Institute (1995–pres- the metaphor theory to the analysis of the conceptual ent). Lakoff was a founding member of two recently structure of several disparate disciplines such as poet- formed political think tanks: The Rockridge Institute ry, visual arts, politics and social issues, philosophy, (1998), where he is a Senior Fellow, and the and mathematics. His work proves that a system of Frameworks Institute (1999), where he is a member of metaphorical mappings allows the abstract ideas of the Advisory Board. He was a recipient of the 2001 any discipline to be conceptualized in terms of embod- Hatfield Scholar Award for socially relevant research ied experience. by a social scientist, Mark O. Hatfield School of In the 1990s, together with other researchers such Government at Portland State University. as Jerome Feldman, Lakoff founded the Neural Theory of Language, which seeks to answer the ques- tion of how the physical brain can give rise to concepts References and language. The basic foundations of this research Gibbs Jr., Raymod W., and Gerard J. Steen (ed.). 1999. Metaphor program, previously known as the L0 Group, are in cognitive linguistics. Amsterdam and Philadelphia: John drawn from many different disciplines: neurochem- Benjamins. Johnson, Mark. 1987. The body in the mind. The bodily basis of istry, neuroscience, neural computation, psychology, language, reason and thought. Chicago: Chicago University and cognitive linguistics. The result is a completely Press. different understanding of language from that in tradi- McCawley, James. 1976. Grammar and meaning. New York: tional Chomskyan theories. In this framework, lan- Academic Press. guage is characterized as: ‘embodied’ ––structured by Lakoff, George 1963. Toward generative semantics. MIT Mechanical Translation Project Report. our bodily and social experience of the world, as well ———. 1970. Irregularity in syntax. Austin, TX: Holt, Rinehart as by the general properties of neural systems and the and Winston. neural structure of systems such as vision and motor ––––––. 1970. Global rules. Language. control; ‘content-dependent’ ––physical and biological ––––––. 1972. Hedges: a study in meaning criteria and the logic constraints; ‘cognition-dependent’ ––other aspects of of fuzzy concepts. Papers from the Eight Regional Meeting of the Chicago Linguistic Society. human cognition; ‘evolutionary’ ––use of earlier evo- ––––––. 1974. Berkeley studies in syntax and semantics, Vol. I, lutionary development; and ‘learned’ ––on the basis of ed. by Charles Fillmore and Robin Lakoff. innate structure that is not specifically linguistic. In his Lackoff, George, and Mark Johnson. 1980. Metaphors we live by. 594  LANGACKER, RONALD ––––––. 1987. Women, fire, and dangerous things: what cate- Lakoff, George, and Rafael Núñez. 2000. Where mathematics gories reveal about the mind. comes from: how the embodied mind brings mathematics into Lackof, George, and Mark Turner. 1989. More than cool rea- being. son: a guide to poetic metaphor. Ortony, Andrew, (ed.), 1986. Metaphor and thought. ––––––. 1993. The contemporary theory of metaphor. Metaphor Cambridge: Cambridge University Press, 2nd edition. 1993 and thought, 2nd edition. ed. by Andrew Ortony Regier, Terry. 1996. The human semantic potential. Spatial lan- ———. 1996. Moral politics: what conservatives know that lib- guage and constrained connectionism. Cambridege, MA: erals don’t. MIT Press. ––––––. L0 –– The first five years of an automated language Ruiz de Mendoza, Francisco José. 1997. An interview with acquisition project. AI Review, Special Issue on Integration George Lakoff. Cuadernos de Filología Inglesa 6(2). of Nautral Language and Vision Processing. Ungerer, F., and H.-I. Schmid. 1997. An introduction to cogni- Lakoff, George, and Jerome Feldman, David Bailey, Srini tive linguistics. London: Longman. Narayanan, Terry Regier, and Andreas Stolcke. 1996. IRAIDE IBARRETXE-ANTUNANO Lakoff, George, and Mark Johnson. 1999. Philosophy in the flesh: the embodied mind and its challenge to western See also Fillmore, Charles; McCawley, James D.; thought. Metaphor Langacker, Ronald Ronald Langacker had his first contact with Linguistics resides in ‘conceptualization’ and ‘mental experi- at the University of Illinois, where he received his train- ence’. This includes both fixed and novel conceptions, ing and pursued his early research in generative syntac- sensory and motor experience, and a full grasping of tic theory. His dissertation was a transformational physical, social, cultural, and linguistic context. As a description of aspects of French syntax. The first ten consequence, the boundaries between the traditional years of his professional career, 1966–1976, were devot- distinction between linguistic and encyclopedic ed to the study of comparative grammar and historical knowledge are blurred. In conceptualist semantics, reconstruction of the Uto-Aztecan language family of the meaning of a lexical item is not restricted to the Native American languages. meaning of the expression designating it, but is open In 1976, he decided that current theoretical to all the knowledge one may have about that entity approaches to language such as Generative linguistics itself. no longer offered satisfying answers for understanding Every expression and every symbolic unit has two language. Consequently, he started developing a new equally significant sides for linguistic meaning: the linguistic framework called Cognitive Grammar, ini- conceptual ‘content’ it evokes and the particular ‘con- tially known as Space Grammar, which presents a rad- strual’ imposed on that content. The content of an ically different conceptualization of language and its expression comprises a set of cognitive domains, each structure. of which corresponds to a different aspect of the In contrast to modular approaches, this new theory expression’s semantic value. Therefore, the starting proposes that language is an integral part of human point for semantic description is not a set of semantic cognition and that grammatical structure can be char- features or conceptual primitives, but an appropriate acterized in terms of cognitive processing. Cognitive array of integrated conceptions organized at different grammar assumes that (i) grammar is not autonomous levels of conceptualization. Equally important for an with respect to semantics, (ii) grammar can be fully expression is to know how it is construed. Langacker described in terms of ‘symbolic units’ i.e. form-mean- proposed several different cognitive operations for the ing pairings, and (iii) grammar together with the lexi- structuring of knowledge, different dimensions of con- con (vocabulary), morphology (word-structure and strual that a speaker may take into account when word-formation rules), and syntax form a continuum describing a certain situation. Among the most signif- of symbolic relationships. icant, we can mention the following: (i) ‘specificity’, This view of grammar as symbolic in nature relies the perception of objects and situations at different on a very specific understanding of linguistic seman- levels of detail or granularity; (ii) ‘scope’, all the pos- tics. In this approach, meaning is not identified with sible aspects that a given expression may evoke in its truth conditions (as it is in classical logic), but it characterization; (iii) ‘background’ assumptions and 595 LANGACKER, RONALD expectations, that is, the distinction between given and Science at UC San Diego (1983–2003). He is also new information; (iv) ‘perspective’, a speaker’s view- Visiting Professor and Fellow at a number of institu- point on the situation, e.g. orientation, vantage point, tions all over the world such as University of Chicago directionality; and (v) ‘prominence’, which includes (1969), University of Illinois (1969), Belgrade very influential concepts such as profile/base organi- University (1971), Summer Institute of Linguistics zation, trajector/landmark alignment. (1976), Australian National University (1991), In Cognitive Grammar, an expression’s grammati- University of New Mexico (1995), Graduate School of cal category depends on the nature of its profile. For Linguistics, Amsterdam (1996), Graduate School in Langacker, there are ‘things’ that profile nominal Language and Communication, Odense University expressions such as nouns (dog, table...), pronouns (I, (1998), Chulalongkorn University, Bangkok (2001), you...), determiners (the, this...) and even noun phras- and University of California, Santa Barbara (Edward es (the large table...), and ‘relationships’ that profile Sapir Professor, 2001). He served on both the Advisory relational expressions; some are atemporal relations Board (1989–1997) and the Governing Board such as adjectives (green, big...), adverbs (well, quick- (1999–present) of the International Cognitive ly...), and prepositions (behind, of...), and some others Linguistics Association and as President of the same are processes or temporal relations such as verbs (take, association (1997–1999). He is a recipient of several sleep...). Grammar consists of patterns that combine awards such as the National Endowment for the simpler symbolic structures with symbolic structures Humanities (senior fellowship, 1973–1974), of greater complexity, the latter called ‘grammatical Guggenheim Fellowship (1978–1979), the Medal of constructions’. the University of Helsinki (1996), and an honorary Another important feature of this model is that it is doctorate from the University of Lodz, Poland (2003). a usage-based approach. That is to say, it is based on Langacker was co-editor of the monograph series how speakers actually speak and understand language, Cognitive Linguistics Research and a member of which, in turn, is composed of networks of units that numerous editorial and advisory boards. become conventionalized or entrenched through usage. The first general description of Cognitive Grammar References was presented in Langacker (1982), but the fundamen- Croft, William, and Alan D. Cruse. Cognitive linguistics. tal piece of work was published in two companion vol- Cambridge: Cambridge University Press, in press. umes: one in 1987––an introduction to the main Cuyckens, Hubert, and Dirk Geeraerts. Handbook of cognitive theoretical tenets and tools––and the other in 1991a––a linguistics. Oxford: Oxford University Press, forthcoming. descriptive application mainly based on English data. Langacker, Ronald. 1969. On pronominalization and the chain Through the years, Langacker continued to refine of command. Modern studies in English, ed. by Sanford A. Schane and David A. Reibel, 160–89. Englewood Cliffs, NJ: and further articulate this theory. Cognitive Grammar Prentice-Hall. has been applied to a wider array of linguistic phe- –––––––. 1968. Language and its structure. New York: Harcourt nomena, for example, locatives, agreement, passive Brace and World, 2nd edition 1973. voice, tense and aspect, complementation, grammati- ––––––. 1972. Fundamentals of linguistic analysis. New York: calization, discourse, and to a wider variety of lan- Harcourt Brace Jovanovish. ––––––. 1976. Non-distinct arguments in Uto-Aztecan. guages and language families such as Cora, German, Berkeley and Los Angeles: University of California Press. French, Greek, Samoan, Slavic, Spanish, just to name ––––––. 1977. Studies in Uto-Aztecan grammar, Vol. 1, An a few. For his work, Ronald Langacker is considered overview of Uto-Aztecan grammar. Dallas: Summer one of the leading figures in Cognitive Linguistics, a Institute of Linguistics and University of Texas at productive and influential theorist within this frame- Arlington. ––––––. (ed.) 1979. Studies in Uto-Aztecan grammar, Vol. 2, work and in linguistic theory in general. Modern Aztec grammatical sketches. Dallas: Summer Institute of Linguistics and University of Texas at Arlington. ––––––. (ed.) 1982. Studies in Uto-Aztecan grammar, Vol. 3, Biography Uto-Aztecan grammatical sketches. Dallas: Summer Ronald Langacker was born in Fond du Lac, Institute of Linguistics and University of Texas at Arlington. Wisconsin, on December 27, 1942. He graduated from ––––––. (ed.) 1984. Studies in Uto-Aztecan grammar, Vol. 4, the University of Illinois in 1963 in French. He Southern Uto-Aztecan grammatical sketches. Dallas: Summer received his doctorate in Linguistics in 1966 from the Institute of Linguistics and University of Texas at Arlington. same university for his work on generative syntactic ––––––. 1987. Foundations of cognitive grammar, Vol. 1, theory. He moved to the University of California at San Theoretical prerequisites. Stanford: Stanford University Press. Diego in 1966, where he has been teaching ever since. ––––––. 1991a. Foundations of cognitive grammar, Vol. 2, Currently, he is Professor Emeritus and Research Descriptive application. Stanford: Stanford University Professor. He is a member of the Program in Cognitive Press. 596  LANGUAGE CHANGE ––––––. 1991b. Concept, image and symbol: the cognitive Lee, David. 2002. Cognitive linguistics. Oxford: Oxford basis of grammar. Berlin and New York: Mouton de University Press. Gruyter. Taylor, John. 2003. Cognitive grammar. Oxford: Oxford ––––––. 1999. Grammar and conceptualisation. Berlin and New University Press. York: Mouton de Gruyter. Ungerer, F., and H.-I. Schmid. 1997. An introduction to cogni- ––––––. 2001. ‘Discourse in cognitive grammar’. Cognitive tive linguistics. London: Longman. Linguistics 12. 143-88. IRAIDE IBARRETXE-ANTUNANO Language Change One of the most fundamental linguistic universals is that The seeds of language change are found everywhere all languages change over time. This fact is quite obvi- in the form of linguistic variation. An interesting ous in the case of languages having a documented his- sociopolitical phenomenon is the popular reaction, tory over many centuries, such as English. An English almost universally negative, to linguistic variants or speaker attempting to learn Old English will quickly perceived changes. It is very often the case that reduced find the task equivalent to learning a foreign language. pronunciations are denounced as sloppy, new word for- For example, in Old English the word knight (original- mations are rejected as slang or substandard, and words ly spelled cniht) was pronounced [kniçt]; hence, it con- being borrowed from another language are treated as a tains both a foreign cluster (word-initial [kn]) and a form of contamination. Such judgments have a long foreign sound ([ç]). Old English was also very different tradition, and no doubt when some people first began in its word structure. The genitive plural form of knight pronouncing knight with [n] at the start instead of the (as in the knights’ horses) was cnihta [kniçta], and there “correct” [kn], some eyebrows were raised; but over were specific dative forms (e.g. plural cnihtum [kniç- time, a linguistic community adopts one variant over tum] “to the knights”) that no longer exist. There has another. Usually, this adoption is done at the uncon- also been semantic change, since in Old English the scious level, but sometimes linguistic markers become word knight meant “boy, youth, servant”. prestigious or stigmatized in a fairly conscious way; But even without historical documents, we are sur- e.g. whether you pronounce the first vowel in “drama” rounded by various types of evidence showing that as [æ] or [ɑ], the last consonant in “garage” as [d] or English has undergone significant change. Since [], etc. While such variation might seem haphazard or spelling conventions are often maintained in spite of random, both the variation and the resulting language sound changes, written forms can reflect earlier pro- change have highly systematic properties, and the same nunciations. We still write knight even though the k and patterns of language change are found universally as a the gh have not been pronounced for more than 300 natural consequence of our common cognitive and years. Another type of evidence for language change is physiological endowment. found in the comparison of English with other Language change can have both external and inter- Germanic languages. Accordingly, the similarity of nal motivations. External motivations involve factors English knight and German Knecht [kneçt] can be independent of the language system itself, such as explained by the fact that both words come from the borrowing. Languages in contact can borrow exten- same source; namely, an earlier West Germanic lan- sively from each other, as is evident in English, where guage that later split into individual dialects that even- the effects of borrowing are most obvious in the case tually developed into English and German. In this case, of vocabulary, but in fact all subsystems (pronuncia- German is more conservative than English in preserv- tion, grammar, meaning, etc.) can undergo change as ing the pronunciation of both [kn] and [ç], and the a result of contact. Language contact is also of inter- meaning of the word (“farm laborer, servant”) is closer est since it seems to be a catalyst of change. to that of Old English. Another type of evidence for Languages with less contact tend to be conservative language change is found in idioms and compounds, (e.g. Icelandic) and retain many archaic features. which can sometimes preserve older features of the Languages with extensive contact tend to be innova- language; e.g. the quick and the dead and quicksilver, tive (e.g. English), i.e. they develop away from the where quick reflects the Old English meaning “living”. ancestral form. Another external type of change 597 LANGUAGE CHANGE involves the impact of writing on pronunciation (e.g. Phonological Change the word often pronounced with a [t] is a spelling- Sound changes are typically triggered by a condition- induced pronunciation), although this impact is usual- ing environment. A very common type of change is ly much less than people normally assume. However, “regressive assimilation”, where a segment becomes even in the absence of such external factors, lan- more similar or identical to one that follows it ([kt] > guages undergo change. A primary factor behind such [tt] as in Latin noctis, Italian notte “night”––the arrow internally motivated change is found in the language > indicates the direction of change). Changes can also acquisition process itself. A grammar is not some- be conditioned by properties of larger structures such thing that is handed down from one generation to the as the syllable; e.g. segments are often deleted at the next, but each generation––or more correctly, each end of a syllable (Spanish cabo “end” shows the loss child––must construct a grammar anew. In the of [t] in syllable-final position; cf. Latin caput). process, the child’s grammar can differ subtly from However, some changes, such as “deaffrication” ([ts] > the adult target grammar, and through the accumula- [s] or [tʃ] > [ʃ]), do not necessarily require a particu- tion of such differences over many generations, lan- lar conditioning environment, but rather they occur in guages can change significantly. all environments. Other changes, such as “diphthon- Explanations of particular changes are often gization” (e.g. [u] > [aw], [i] > [aj]), appear to find based on reference to relative preferences relating to their genesis in the phonological system itself. Vowel linguistic structure. Such an approach works partic- systems that include seven or more vowels are highly ularly well in the case of change in pronunciation prone to diphthongization, while smaller vowel sys- and word structure. For example, the elimination of tems are not. Although most changes can be described word variants (thrive/throve > thrive/thrived; old in articulatory terms (e.g. manner or place assimila- /elder > old/older) is attributed to a preference for tion, where fewer articulatory movements are “one meaning, one form”, and the common loss of a required), some types of change show the importance final consonant in syllables is motivated by the pref- of auditory factors (e.g. [x] > [f] as in laugh, where the erence for syllables consisting of only a consonant gh was originally pronounced [x]). and a vowel (CV). This type of explanation is Sound changes are usually regular in the sense that strengthened in cases where the preferences find the particular change applies to all words displaying independent support in areas like language acquisi- the necessary phonetic conditions regardless of its part tion (e.g. CV syllables are the first to be acquired by of speech. However, in some cases a particular change children), and implicational universals (all lan- can generalize from one word to another over time. guages have CV syllables, and some languages have Such “diffusion” is evident in the case of stress shift- only CV syllables). ing in English nouns; e.g. stress has been shifted in From a historical perspective, it is often possible to rebel (formerly rebél), but not in mistake. Nouns such identify drifts; that is, the tendency for a particular as address and adult are subject to significant variation language or language family to display interrelated or (both addréss and áddress). similar changes over a long period of time. Such drifts Some changes involve interrelated shifts of entire may be attributed to the assumption that the same series of segments or sets of natural classes. One of the preferences are at work, or that a particular condition- most famous sound shifts, called Grimm’s Law after ing factor is present over an extended period; e.g. it is its nineteenth-century discoverer Jakob Grimm, often claimed that the fixing of the stress accent on involves a set of sound changes that occurred in early the root syllable in early Germanic led to the weaken- Proto-Germanic (the assumed common language from ing and loss of unstressed syllables that is evident in which the known Germanic languages developed). In all the modern Germanic languages. Accordingly, lan- its traditional formulation, Proto-Indo-European guages can drift in the same direction, even after they voiceless plosives shifted to voiceless fricatives, have become separated. For example, “umlaut” (a voiced plosives to voiceless plosives, and voiced aspi- change in which vowels become more alike, which is rated plosives to voiced plosives; e.g. Latin pater the source of alternations such as foot/feet) occurred maintains the original voiceless plosive [p], whereas in both the West and North Germanic languages. the English cognate father with the voiceless fricative Although umlaut itself cannot be attributed to Proto- [f] shows the effects of Grimm’s Law. Germanic, it is apparent that favorable conditioning factors were inherited in both the western and north- ern branches. Morphological Change The study of language change is the subject of his- torical or diachronic linguistics. The following is a Languages are often classified according to the type brief overview of the main subfields. and complexity of the word structure they have. 598  LANGUAGE CHANGE “Analytic” languages have very simple words, while Syntactic Change “synthetic” languages are morphologically com- Changes in sentence structure are also readily apparent plex––have complex words. Thus, English, a fairly in the history of English. The word order in a sentence analytic language, does not have a future tense marker such as ond heo hine in þæt mynster onfeng (literally or case inflections, whereas a synthetic language such “and she him into the monastery received”) was entire- as Latin has both. Over time, languages can lose com- ly acceptable in Old English, but is not in Modern plexity, primarily through sound change; e.g. the case English. Indeed, a variety of evidence supports the system of Old English was lost in this way. At the same reconstruction of the Subject–Object–Verb word order time, new morphological complexity can develop for Proto-Germanic. One of the oldest Germanic sen- through morphologization or grammaticalization. tences, written on the Golden Horn of Gallehus about “Morphologization” refers to the case where an origi- 1,700 years ago, displays Subject–Object–Verb order: nal sound change becomes a morphological regularity; ek hlewagastiR holtijaR horna tawido (literally “I e.g. the source of irregular plurals such as foot/feet can Hlewagistir of Holt horn made”) “I Hlewagistir of be found in an earlier sound change (umlaut). In one Holt made the horn.” type of “grammaticalization”, an original word can Attempts to explain the shift from Old English become a grammatical item such as a suffix; e.g. the S(ubject)–O(bject)–V(erb) to Modern English Subject– future tense morpheme -ò in Italian finirò “I will fin- Verb–Object are usually grounded in typological stud- ish” derives from an original auxiliary verb, habeo, as ies that established two basic word order types, in the Latin main verb–auxiliary construction finire “dependent–head” (also called “prespecifying” or habeo. Forms undergoing grammaticalization are typi- Object–Verb languages) vs. “head–dependent” (“post- cally subject to extensive phonological and semantic specifying” or Verb–Object languages). Dependent– change. head languages such as Korean display the following Leveling and analogy are important kinds of mor- types of orders: objects precede verbs, main verbs pre- phological change. “Leveling” refers to the elimina- cede auxiliaries, relative clauses precede their nouns, tion of word variants or “allomorphs” within a and noun phrases precede their adpositions (or, in other paradigm. German has allomorphy in verbal para- words, dependent–head languages have postpositions). digms of the type ich helfe “I help,” es hilft “it helps” Head–dependent languages such as Maori display the (i.e. helf- vs hilf-). Old English had the same type of opposite word order patterns (objects follow verbs, allomorphy, but it was leveled out so that Modern main verbs follow auxiliaries, etc.). Much syntactic English has only help- as in I help, it helps. work assumes the primacy of the order of the verb and “Analogy” is the extension of a regularity from one the object. If the OV order changes to VO, then the context to another, especially from members of one head–dependent order can be generalized to the other paradigm to those of another. For example, in earli- patterns. From this perspective, it is essential to find er English there were numerous “strong” verbs of reasons for the OV > VO shift. One very old theory that the type drive/drove/driven, sing/sang/sung, etc. has stood the test of time is based on Wackernagel’s Over time, many of the strong verbs have joined the Law, which observes that certain particles called “sen- larger -ed verb class through analogy. Thus, molten, tence clitics” have a tendency to be placed in clause- the original past participle of melt, was replaced by second position. If these clitics are verbal elements, melted through analogy with -ed verbs such as then the clause-second position can be reinterpreted as belt/belted, work/worked, etc. Sometimes the older the position of the verb, and (S)VO can result as the forms are maintained in the language with a special- default order. ized meaning (e.g. molten steel) or they may be com- pletely lost (e.g. throve, the original past tense of thrive). Semantic Change “Folk etymology” is an interesting kind of mor- phological change involving a false or incorrect Change of meaning also displays a systematic aspect, analysis. The word bridegroom traces back to an orig- and patterns of change can be clearly identified. Words inal compound consisting of the words bry¯d “bride” for “see” often take on a new meaning of “know, and guma “man”. Over time, the independent word understand”; e.g. both vision (borrowed from French) guma was lost, and when the original compound was and wise derive from the same Indo-European root. no longer understood, folk etymology occurred, and Words involving a sense of touch often take on a new the word was reanalyzed as “bride” + “groom”. Often, meaning related to the sense of taste (e.g. sharp). At the result of folk etymology is semantically incoher- the same time, determining the motivations for semantic ent; e.g. the dialectal reanalysis of asparagus as spar- change has proved a much more elusive task than is the row grass. case for other types of change. Nevertheless, general 599 LANGUAGE CHANGE categories of semantic change can be determined. In through internal sources (e.g. compounding as in share- semantic “broadening”, the meaning of a word ware) or borrowing (peyote, from American Spanish). becomes more general or inclusive, as in the word barn, which derives from an Old English compound bere “barley” + ærn “house, store” that was originally References used to refer to a place where barley was stored. The Campbell, Lyle. 1999 Historical linguistics: an introduction. opposite, and more common, type of change is called Cambridge, MA: MIT Press. “narrowing”. A word such as deer displays narrowing, Croft, William. 2000 Explaining language change. An evolu- tionary approach. New York: Longman. since Old English de¯or meant “animal”. The Modern Harris, Alice C., and Lyle Campbell. 1995 Historical syntax in German cognate Tier “animal” retains the older mean- cross-linguistic perspective. Cambridge: Cambridge ing. Social and technological factors can also play a University Press. role in semantic change; e.g. write originally meant Hock, Hans Henrich.1992 Principles of historical linguistics, “to scratch”, since early writing involved scratching 2nd edition. Amsterdam: Mouton de Gruyter. Hopper, Paul J., and Elizabeth Closs Traugott. 1993 letters onto wood or other materials. Grammaticalization. Cambridge: Cambridge University Press. Joseph, Brian D., and Richard D. Janda, (eds.) 2003 The hand- Lexical Change book of historical linguistics. Malden, MA: Blackwell. Labov, William. 1994 Principles of linguistic change, vols. 1 Change in vocabulary primarily involves the loss or and 2. Oxford and Cambridge, MA: Blackwell. addition of words. In many societies, taboo is an impor- Lass, Roger. 1997. Historical linguistics and language change. tant factor in lexical loss; e.g. the Indo-European word Cambridge: Cambridge University Press. Lehmann, Winfred. 1992 Historical linguistics, 3rd edition. for “bear” was lost in earlier Germanic as the result of London–New York: Routledge. taboo. More generally, social or technological changes McMahon, April M. S. 1994 Understanding language change. play an important role in motivating the creation of new Cambridge: Cambridge University Press. words. As the need arises, new words develop either ROBERT W. MURRAY Language: Contact—Overview Broadly interpreted, language contact encompasses Early work in language contact and society such diverse experiences as casual contact with a for- looked at the typology of multilingual societies, eign language during travel, study of a second lan- describing each language found according to its sta- guage as a school subject, and the acquisition of two tus and function. Often in multilingual societies, dif- or more languages in early childhood. Because of the ferent languages are used in the same way as power and prestige of their language, many English different varieties of the same language in monolin- speakers never need learn another language and may gual societies; domain analysis attempts to classify regard monolingualism as the normal state of affairs. these functional differences by looking at the class For the majority of the world’s population, however, of situations (or domain) in which each language is some degree of bilingualism (or multilingualism) is used. In stable contact situations, multilingualism the norm; even small-scale nonliterate societies can be often endures for long periods, but when contact remarkably diverse in their linguistic composition. In accompanies social change, multilingualism is often the modern era, contact with other languages has transitory, as one group shifts to the language of the increased through such phenomena as modern trans- other. A corollary of language shift is language portation, the Internet, mass communications, the death, a disturbingly common phenomenon in the globalization of commerce, and the growth of English modern world. as a world language. The study of language contact Languages in contact may be seen as languages in intersects with many areas of linguistics, and can be competition, and their status must be regulated by a lan- approached from the perspective of the society, the guage policy, whether overtly formulated or residing individual, or the language. covertly in the society’s linguistic culture (Schiffman 600  LANGUAGE: CONTACT—OVERVIEW 1996). Language status planning must balance the trying to establish constraints on where switches may symbolic value of a single indigenous national language occur. against the competing interests of linguistic minorities, In addition to purposeful code-switching, bilin- especially of sizeable groups that speak languages with guals may incorporate features of one of their lan- important literary traditions. Ex-colonial countries may guages while speaking the other. This phenomenon, also have to take into account the prestige of an ex-colo- known in various contexts as borrowing, transfer, or nial language and its value of an ex-colonial language as interference, may be due to incomplete mastery of a link to the outside world as well as the cost of devel- one of the languages and in this way it is similar to oping an indigenous language that has not been used in certain interlanguage phenomena seen in second-lan- administrative, educational, technical, and other “high- guage acquisition. However, it also takes place when er” domains. Complementing language status policy is speakers are fluent in both languages. Transfer may educational language policy, whose goal for speakers of affect either language and may involve any linguistic a minority language may be to assimilate them to a feature: pronunciation, word structure, grammar, dominant language or to help them maintain their lan- vocabulary, or semantics. The nature and extent of guage; for speakers of the dominant group, policy may transfer are partly determined by the sociolinguistic aim to provide some degree of competence in another characteristics of the contact situation (Thomason and language, often a minority language of the country or an Kaufman 1988). When transfer features become per- international language of wider communication, such as manent fixtures of the host language, we speak of English. contact-induced change. Most languages have bor- Turning the focus to language contact and the bi- or rowed vocabulary from other languages, but the trans- multilingual individual immediately raises the issue of fer of other features is now recognized to be more the definition of bilingual. How well does one have to common than it was traditionally thought to be. In know a second language before being considered historical linguistics, a language that has influenced bilingual? Certainly, the person who can use two lan- another during the course of language shift is known guages with equal ease in all situations is a rarity, per- as a substratum, superstratum, or adstratum depend- haps even an aberration. When one’s time is divided ing upon whether its speakers were socially inferior, between two (or more) languages, differences in com- superior, or equal to the speakers of the language to petence are almost certain to arise, depending on the which they were shifting. If languages of different exposure to each. More importantly, because lan- families in an entire region come to resemble one guages in a bi- or multilingual society are generally another because of contact-induced change, we speak differentiated by domain, the normal expectation is for of a linguistic area or Sprachbund. the individual’s competence to be spread unevenly Finally, language contact may give birth to new across the society’s languages. languages in the form of pidgins and creoles. There are many paths to individual bilingualism: Pidgins arise in restricted contexts for limited com- the languages may be learned sequentially or more or munication needs (such as trading in a bazaar) where less simultaneously, in a variety of differing contexts, contact groups have no language in common. They and with varying degrees of community support. are not the primary language of any group, and in There is a long tradition of interest in the organization their initial stages they are stripped down to the of language in the bilingual brain, bilingual language barest essentials requisite for the communication of acquisition, and cognitive differences between bilin- content. Over time, they may expand until they guals and monolinguals. The last issue is particularly become indistinguishable from normal languages. acute in relation to educational policy in bilingual Creoles are primary languages that arise from the societies, where the outcomes of various types of nativization of pidgins or develop directly, usually in bilingual programes may be under scrutiny. multilingual contact situations in which children Bilinguals may alternate between languages, or acquiring the emergent language are present at an code-switch for a wide variety of reasons: for purely early stage and play a significant role in shaping it. linguistic reasons, such as the “triggering” effect of a Although some see creole genesis as a distinct phe- loanword; for situational reasons, such as to mark nomenon (e.g. Bickerton 1981), for others the struc- transitions from one domain to another; for discourse ture of pidgins and creoles owes much to the reasons, such as to emphasize a point made, to clari- structure of the contact languages. Although each fy, or to mark a change in topic; and for social rea- context has its individual characteristics, similar sons, such as to signal a particular identity, to express transfer phenomena can be seen in stable bilingual- solidarity or distance. Much work has been done on ism, language shift and death, second-language the structural aspects of code-switching, for example, learning, pidginization, and creolization. 601 LANGUAGE: CONTACT—OVERVIEW References Paulston, Christina Bratt. 1994. Linguistic minorities in mul- tilingual settings. Amsterdam, Philadelphia: John Appel, René, and Pieter Muysken. 1987. Language contact and Benjamins. bilingualism. London Baltimore: Edward Arnold. Romaine, Suzanne. 1995. Bilingualism, 2nd edition. Oxford: Baker, Colin. 1996. Foundations of bilingual education and Blackwell. bilingualism, 2nd edition. Clevedon, UK, Philadelphia: Rot, Sándor. 1991. Language contact. Frankfurt am Main, New Multilingual Matters. York: Peter Lang. Bickerton, Derek. 1981. Roots of language. Ann Arbor, MI: Schiffman, Harold F. 1996. Linguistic culture and language Karoma Publishers. policy. London, New York: Routledge. Cooper, Robert Leon. 1989. Language planning and social Sebba, Mark. 1997. Contact languages: pidgins and creoles. change. Cambridge, New York: Cambridge University Press. New York: St. Martin’s Press. Fabbro, Franco. 1999. The neurolinguistics of bilingualism: an Thomason, Sara G. 2001. Language contact. Washington: introduction. Hove: Psychology Press. Georgetown University Press. Hoffman, Charlotte. 1991. An introduction to bilingualism. Thomason, Sarah G., and Terrence Kaufman. 1988. Language London, New York: Longman. contact, creolization and genetic linguistics. Berkeley: Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices: University of California Press. the extinction of the world’s languages. Oxford, New York: Oxford University Press. IAN R. SMITH Language Death We consider a language to be dead when it is no longer of the two languages. Language shift is characteristic spoken. Languages may die suddenly, with the extinc- of low-prestige languages, typically spoken by popu- tion of their speakers. More commonly, however, lan- lations that are disadvantaged economically, political- guages die gradually, as their speakers shift to another ly, and demographically. Examples include indigenous language. linguistic minorities in Europe, such as Basque in In rare cases, sudden language death results from France and the Celtic languages in Britain (Welsh and natural disasters, such as when a volcanic eruption on Scots Gaelic), and immigrant communities in North Sumbawa Island wiped out all the speakers of America and Western Europe. Tamboran in 1815. More often, sudden language death Although multilingualism is a necessary condition is the result of human intervention. For example, it is for language shift, it is not sufficient, because speakers estimated that up to 90% of the indigenous population may choose to remain fully bilingual and to use both of North America was eradicated between 1600 and languages. Language shift results from situations of 1800 by diseases carried by the European settlers and unstable multilingualism, in which speakers abandon their animals. The Tasmanians of Australia and the the minority language for the dominant language, Beothuks of Newfoundland both became extinct by either involuntarily, through political coercion from the nineteenth century as a result of conquest and the state, or voluntarily, to improve their economic genocide by the Europeans. In such cases, the lan- prospects. The process of language shift takes two or guages of these people died with them. three generations to complete: whereas the first gener- Gradual language death may occur for a number of ation is monolingual in the minority language or reasons. A language may fall out of daily use, becom- speaks the dominant language only as a second lan- ing restricted to ceremonial and formal functions. For guage, subsequent generations are bilingual, with the example, the Egyptian language Coptic is used only as minority language gradually becoming the second lan- a liturgical language by the Coptic Christians of guage. When language shift is complete, speakers are Egypt, whose everyday language is Arabic. Typically, monolingual in the dominant language. however, gradual language death takes the form of lan- Language shift has consequences for the structure of guage shift, in which speakers of a minority language the dying language. (Note that many of these conse- assimilate to a dominant language. Language shift is quences are distinguished from normal processes of promoted or inhibited not by properties of the lan- language contact and change in degree rather than in guages themselves, but rather by a number of social kind and in their relative rapidity.) First, there is mas- factors, such as the size and economic status of the sive and asymmetrical lexical and grammatical inter- minority language population and the relative prestige ference from the dominant language. For example, 602  LANGUAGE PLANNING more words are borrowed into the minority language economic reasons, subsequent generations often than into the dominant language. Borrowings are inte- express regret at the loss of the dying language. grated less and less into the sound system and grammar Indeed, language death is a tragedy not only because as speakers’ command of the language declines, every language is a unique and valuable communica- reflecting a more general pattern in the dying language, tive system, but also because language is so often cor- in which all linguistic systems are reduced and varia- related with sociocultural and ethnic identity. tion between speakers increases. Grammatical com- Language death has been halted or even reversed in a plexities are reduced, as are the sound systems, number of instances, such as the revival of Hebrew as especially in sound distinctions that do not exist in the a spoken language in Israel and the maintenance of dominant language. For example, in dying varieties of Irish Gaelic in Ireland. Although such success stories Scots Gaelic, inflected pronouns such as rium “with (relatively speaking) have been aided by a combina- me” are replaced by a form with separate particles, ri tion of political and educational provisions, the major mi. Finally, the complexity of sentence structure factor in preventing language death seems to be the decreases: word order becomes fixed, complex single- degree to which the dying language is emblematic of a word constructions are replaced by multiword con- social identity that people wish to preserve. structions, and subordinate and relative clauses are lost. Language shift also has consequences for patterns of References language use. Because minority languages tend not to be institutionally supported, the decline in usage, the Brenzinger, Matthias (ed.) 1992. Language death: factual and increase in variation, and the changes in linguistic struc- theoretical explorations, with special reference to East Africa. Berlin and New York: Mouton de Gruyter. ture fail to elicit a puristic reaction from the education- Craig, Colette Grinevald.1997. Language contact and language al establishment. The domains of use of the language degeneration. The handbook of sociolinguistics, ed. by also become progressively restricted, gradually becom- Florian Coulmas. Oxford: Blackwell. ing confined to the home or the local community. Crystal, David. 2000. Language death. Cambridge: Cambridge Because different domains are associated with different University Press. Dorian, Nancy C. 1981. Language death: the life cycle of a linguistic registers, or styles, this restriction leads to a Scottish Gaelic dialect. Philadelphia: University of loss of social norms governing the choice of styles and Pennsylvania Press. results in monostylistic speech. As language shift pro- Dorian, Nancy C. (ed.) 1989. Investigating obsolescence: studies ceeds, fewer speakers know the language well and thus in language contraction and death. Cambridge: Cambridge offer an inadequate linguistic model for younger speak- University Press. Dressler, Wolfgang. 1988. Language death. Linguistics: the ers. Although these semispeakers may identify them- Cambridge survey, Vol. IV: language: the socio-cultural con- selves as native speakers of the dying language and may text, ed. by Frederick J. Newmeyer. Cambridge: Cambridge in fact have excellent competence in the communicative University Press. skills of the language, they are aware that their gram- Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices: matical and lexical competence is deficient. the extinction of the world’s languages. Oxford and New York: Oxford University Press. There is some debate in the linguistics community Wurm, Stephen A., Theo, and Baumann. 1996. Atlas of the about whether and how language shift should be halt- world’s languages in danger of disappearing. Paris: ed or reversed. Although minority language speakers UNESCO Publishing and Canberra: Pacific Linguistics. may voluntarily shift to the dominant language for JAMES A. WALKER Language Planning Language planning is the deliberate, future-oriented, A language policy is a body of ideas, laws, regula- systematic change of language form or use, most visi- tions, rules, and practices intended to achieve the bly undertaken by government, in some communities intended planned language change. Language policy of speakers. Language planning is directed by, or leads may be realized in very formal (overt) language plan- to, the promulgation of a language policy by the gov- ning documents and pronouncements (e.g. constitu- ernment or another authoritative body or person. tions, legislation, and policy statements) or in informal 603 LANGUAGE PLANNING statements of intent (i.e. in the discourse of language, it is also closely associated with disciplines such as politics, and society), or it may be unstated (covert). sociolinguistics and education. Language planning sub- Formal policy statements may be symbolic or substan- sumes literacy planning, but language planning goals, tive in their intent. Although the field differentiates including those for literacy planning, are often under- between language policy (the plan) and language plan- taken on their own––not as part of some broader policy. ning (the plan implementation), the two terms are The practice of overt (explicit) or covert (implicit) often used interchangeably in the literature. language policy and planning may be one of four Language planning is an outgrowth of the positivis- types: status planning (about society), corpus plan- tic economic and social science paradigms that domi- ning (about language), language-in-education (acqui- nated the three post-World War II decades. Initially sition) planning (about learning), and prestige called ‘language engineering’, the discipline emerged planning (about image). Each of these four types of as an approach to developing programs for solving lan- language planning can be realized from two guage problems in newly independent developing approaches: a policy approach, with an emphasis on nations. However, by the 1970s, it was apparent that form, or a cultivation approach, with an emphasis on language problems were not unique to developing language development and use. These eight language nations, but were widely applicable to macro (i.e. state- planning perspectives are best understood through level) language problems and situations. In the 1990s, the goals that planners set out to achieve. Status and policy and planning principles have increasingly been corpus planning have been discussed extensively in applied to microlanguage situations (e.g. to language the literature, with language-in-education planning problems in companies and organizations). Language often categorized as a part of status planning. Little planning is a major field within applied linguistics, but attention has been paid to prestige planning. Approaches Policy Planning Cultivation Planning (on form) (on function) Types (overt – covert) Goals Goals Status planning Status standardization Revival (about society) • Officialization • Restoration • Nationalization • Revitalization • Proscription • Reversal Maintenance Interlingual communication • International • Intranational Spread Corpus planning Corpus standardization Lexical modernization (about language) • Graphization Stylistic modernization • Grammatication Renovation • Lexication • Purification Auxiliary code standard. • Reform • Graphization • Stylistic simplification • Grammatication • Terminological unification • Lexication Internationalization Language-in-education Access policy Reacquisition (Acquisition) planning Personnel policy Maintenance (about learning) Curriculum policy Foreign language / Methods & materials policy Second language Resourcing policy Shift Community policy Evaluation policy Prestige planning Language promotion Intellectualization (about image) • Official government • Language of science • Institutional • Language of professions • Pressure group • Language of high culture • Individual • Language of diplomacy Based on Hornberger (1994) 604  LANGUAGE PLANNING Status Planning Goals Prestige Planning Goals Status planning concerns social goals that are external Prestige planning is directed at the image a language to the language and that a society must make about the needs to develop to promote and intellectualize environment in which a language is used. Detailed sta- that language. It is conducted by individuals or tus planning is a relatively neglected activity. The for- groups that have or take the responsibility to create mal goals of status planning may include the image of the languages, e.g. in language promo- officialization, nationalization, and proscription. tion via government institutions or pressure groups. Functional goals may be the revival, restoration, revi- It may aim at enhancing the functional image talization, maintenance, and spread of languages or the and status of the language in key language domains, promotion of interlingual communication (internation- e.g. in intellectualization, the development of a lan- al or intranational). guage of science, a language of professions, or a lan- guage of high culture. Prestige planning represents a separate range of activities from status, corpus, and Corpus Planning Goals language-in-education planning. The latter are Corpus planning is directed at language and repre- productive activities, insofar as they effect a materi- sents an attempt to codify, standardize, or modify and al change. Prestige planning is receptive and influ- elaborate a language. These goals relate to the lin- ences how the three other planning activities are guistic form of the languages, e.g. the standardization acted on and received. Prestige or efficiency of orga- of the writing system, orthography, grammar, and nizational impact levels influences the success of a vocabulary, or to the cultivation of linguistic func- language plan and the uses to which languages are tions, i.e. lexical or stylistic modernization, renova- put. tion (purification, reform, stylistic simplification, or terminological unification), and internationalization. Language renovation may involve language purifica- References tion, i.e. the removal of foreign (lexical) influences or Baldauf Jr., Richard. B., and Allan Luke (eds.) 1990. Language the adherence to the classical forms and lexicon of a planning in Australasia and the South Pacific. Clevedon: language. Written forms of a language, by definition, Multilingual Matters. are more standardized and purified than the corre- Clyne, Michael (ed.) 1997. Undoing and redoing corpus plan- sponding oral forms. Although the application of ning. The Hague: Mouton de Gruyter. Cooper, Robert L. 1989. Language planning and social change. technical linguistic skills is central to meeting corpus Cambridge: Cambridge University Press. planning goals, corpus planning also involves making Corbarrubias, Juan, and Joshua A. Fishman (eds.) 1983. choices or selecting alternatives that have a social Progress in language planning. Berlin: Mouton. aspect and that must be resolved for such planning to Corson, David. 1999. Language planning in schools. Mahwah, be successful. NJ: Lawrence Erlbaum. Eastman, Carol. 1983. Language planning: an introduction. San Francisco: Chandler and Sharp. Language-in-Education (Acquisition) Hornberger, Nancy H. 1994. Literacy and language planning. Language and Education 8. 75–86. Planning Goals Kaplan, Robert B., and Richard B. Baldauf Jr. 1997. Language Language-in-education planning is user related and planning from practice to theory. Clevedon: Multilingual Matters. aims to develop language education programs and Kaplan, Robert B., and Richard B. Baldauf Jr. 2003. Language teach a language for various purposes. It may relate to and language-in-education planning in the Pacific basin. the form of a language-learning program, e.g. in poli- Dordrecht: Kluwer Academic. cies concerning access, personnel, curriculum, meth- Rubin, Joan, and Björn H. Jernudd (eds.) 1971. Can language ods and materials, and resourcing policy and/or be planned? Honolulu: East-West Centre and University of Hawaii Press. evaluation, or to the cultivation of language teaching Schiffman, Harold F. 1996. Linguistic culture and language functions, e.g. reacquisition of a former language, policy. London and New York: Routledge. maintenance of an existing one, or foreign or second- Spolsky, Bernard. 2004. Language policy. Cambridge: language teaching. Although language-in-education Cambridge University Press. planning should co-occur with the other planning Wright, Sue. 2004. Language policy and planning: from nationalism to globalism. New York: Palgrave types, this often does not happen, and language-in- Macmillan. education planning, through schooling, can become RICHARD B. BALDAUF, JR the sole language-change agent. 605 LANGUAGE SOCIALIZATION Language Socialization Language socialization is the process through which instruct a younger child in what to say by modeling new or novice speakers learn about the language and each utterance for the child to repeat. The prompting culture. Language socialization includes both social- routine is marked by characteristic linguistic features. ization through language and socialization to use lan- Such routines are usually, but not always, initiated by guage. an imperative verb form meaning ‘say’ or ‘do’, fol- Historically, theories of socialization have varied in lowed by the utterance to be repeated. Very often, the degree to which the individual is seen as active in these utterances have a distinct voice quality and into- their own socialization. Nineteenth-century theories national contour, which mark them as components of followed Hobbes’ notion of the individual as aggres- the prompting activity. sive, selfish, and asocial by nature and saw socializa- Second, new or novice speakers are taught what to tion as the process of reshaping these natural impulses say (or not to say) on various occasions. Language into prosocial feelings and desires. Freudian theory in plays a major role in the acquisition of activity/event the early twentieth century emphasized conflict knowledge with many formal and functional features between human nature (the id) and society (the super- of discourse carrying sociocultural information, ego). Functionalism viewed the individual as more including phonological and morphosyntactic construc- passive and more socially directed. tions, the lexicon, speech-act types, conversational A return to an active role for individuals in con- sequencing, genres, interruptions, overlaps, gaps, and structing social order was found in Mead’s theory of turn length. These structures are socially organized symbolic interactionalism, which proposed that indi- and carry information concerning social order. They viduals and society construct one another through are also culturally organized and express local concep- social interaction. This perspective influenced the tions and theories about the world. As new or novice work of phenomenologists, ethnomethodologists, and speakers acquire tacit knowledge and competence in social interactionalists. Ethnomethodologists also the use of these cues, they acquire knowledge and incorporate close examination of the interactional pro- competence in the social organization of activities and cedures or methods used to construct a sense of shared events. Thus, through exposure to and participation in context or shared realities, and social interactionalists language-mediated interactions, new or novice speak- consider socialization to be an interactional display in ers acquire a knowledge of principles of social order which a novice language learner is shown expected and systems of belief (ethnotheories). For example, ways of thinking, feeling, and acting. Thus, social among certain social groups, some activities and interactions are considered to be sociocultural envi- events may have highly predictable discourse struc- ronments through which language learners come to tures (e.g. greetings, jokes, ritual insults, teasing, beg- internalize and gain performance competence in socio- ging, clarification sequences, trick-or-treat routines), culturally defined contexts. whereas others may have more variable discourse Language intersects with socialization in three dis- organization (e.g. gossip, negotiations, giving advice, tinct domains: (1) language is used to instruct the explanations, instructions). learner about what to do, think, feel, etc. (e.g. ‘smile’); Third, new or novice speakers are instructed in cer- (2) learners are taught what to say (or not) on various tain features of interaction that vary systematically and occasions (e.g. ‘say “Thank you”’); and (3) learners are correlated with individual or group variables. For are instructed in certain features of interaction that example, the acquisition of sociocultural knowledge vary systematically and are correlated with individual by children is influenced by their level of cognitive, or group variables (e.g. ‘good little girls don’t shout’). social, and linguistic development. In other words, the First, language is used to instruct the new or novice communicative competence expected of the three- speakers about what to do, think, feel, etc. When care- year-old child is necessarily different from that expect- givers and other native speakers use language to and in ed of a teenager or adult. Similarly, different social the presence of new or novice speakers, they are pro- situations require different levels of politeness depend- viding information or cues concerning what members ing on the status of the speakers (e.g. higher, same, or are doing. Typically, such instruction takes the form of lower than the speaker), the environment (e.g. church, explicit prompting by the caregiver or an other mem- home, school, etc.), or the intent of the speaker (e.g. ber of a group. For example, an older person will emphasize group solidarity). 606  LATERALIZATION AND HANDEDNESS In addition, new or novice speakers are active activities/events for a child, and using a simplified lexi- socializers of others in their environment. Even con and grammar to do so are cross-culturally variable. infants and small children have a hand in socializing These practices are not, for example, characteristic of other members of the family into such roles as care- adult–child or child–child interactions in traditional giver, parent, and sibling. As such, second children Western Samoan households. This evidence of overt enter a different social environment than do first chil- instruction is also of interest in the ongoing debate in dren; often, first children ‘break in’ adults as care- the literature regarding the degree to which language givers. As older children, they may further socialize acquisition is influenced by innate factors (e.g. parents into modes of acting and communicating Universal Grammar). associated with their school and peer-group experi- ences. This is illustrated by children’s recent role as References socializers of computer literacy within their respec- tive households. Berko-Gleason, Jean. 1997. The development of language. Although many theories of language acquisition Needham Heights, MA: Allyn & Bacon. Coates, Jennifer. 1993. Women, men and language, 2nd edition assume that children develop social and cognitive skills Harlow, Essex: Longman. through participating in structured cooperative interac- Fletcher, Paul, and Brian MacWhinney (eds.) 1996. The hand- tions with more mature members of society, it is impor- book of child language Oxford. UK: Blackwell. tant to note that all societies do not rely on the very Gallaway, Clare, and Brian Richards (eds.) 1994 Input and same set of language-socializing procedures. The cross- interaction in language acquisition. Cambridge, UK: Cambridge University Press. cultural research of researchers such as Schiefflin and Schiefflin, Bambi, and Elinor Ochs (eds.) 1986 Language Ochs (1986) has shown that although prompting a child socialization across cultures. New York: Cambridge on what to say appears widespread, expanding chil- University Press. dren’s utterances, using leading questions, announcing KAREN WOODMAN Lateralization and Handedness The term (brain) lateralization derives from the Latin clinical discoveries of Pierre Paul Broca (in 1865) and word latus (= side of the body) and refers to the fact Carl Wernicke (in 1875) regarding the special role of that the two hemispheres of the human brain are not the left hemisphere in language were made public. exactly alike. The concept of lateralization has earlier Broca took a firm position about the left hemispheric given rise to the misleading idea that one hemisphere dominance for articulate speech, the comprehension of is actually ‘dominant’ over the other — namely, the spoken language, reading and writing, but was careful hemisphere in which the major capacity for language to add that language was not the exclusive function of resides dominating the contralateral hemisphere. This the left hemisphere. He further believed the right classic concept of ‘cerebral dominance’ has been hemisphere to mediate speech in left-handed people. replaced by the concept of ‘hemispheric specializa- It was not before the twentieth century that scien- tion’, with both hemispheres having different general tists began to wonder whether Broca might have been cognitive operational modes and subserving efficient wrong about right-hemisphere language dominance in subdivision of complex functions, rather than showing left-handed people. Current evidence from a number any superiority of one hemisphere over the other. of converging sources, notably the high incidence of The earliest known theory of brain lateralization is the language disturbance known as aphasia after left- about 2,400 years old and stated that different hemi- but not right-hemisphere damage, indicates that the spheres of the brain control the two functions of sen- left hemisphere is dominant for comprehension and sation and cognition. Although this early theory articulation of language in close to 99 % of adult right- involved speculations about the two hemispheres handed people. Sixty-three percent of left-handed and being functionally different, up to the 1860s the major- ambidextrous people also have left-hemisphere lan- ity of the scientific community maintained that the guage, 24 % have bilateral language abilities, and only hemispheres were structurally and functionally identi- 13 % of left-handed and ambidextrous people have cal. This viewpoint changed dramatically, when the predominantly right-hemisphere language. 607 LATERALIZATION AND HANDEDNESS Recent studies showed that some lateralization of been examined through intracarotid infusion of anes- function is already present in newborns and is thus thetics (e.g. sodium amytal), which is a technique to established long before puberty. Moreover, certain allow one to inactivate reversibly and for a brief peri- anatomical asymmetries in the temporal lobe of the od one or the other hemisphere (WADA-test). A vari- brain are already present in the fetus. This is in con- ety of neurophysiological methods that measure trast to early theories, which favored the idea that the cortical electric or magnetic field activity (electroen- two hemispheres were equipotential at birth and for cephalography, EEG; magnetencephalography, MEG) the first two years of life and that a left hemisphere or cerebral blood flow and metabolic rate (positron superiority for linguistic functions develops with age. emission tomography, PET; functional magnetic reso- There is also no extensive agreement as to why func- nance imaging, fMRI) were also applied to study tions need to be lateralized. A possible interpretation hemispheric specialization. Despite the problems that of the cause of lateralization is that it reflects the evo- each of these methods have, the best strategy for lution of systems, which have diversified because of studying the functioning of the cerebral hemispheres is their functional incompatibility. One major example to use combinations of multiple methods. If the same could be provided by linguistic and spatial skills, hemisphere difference is found with multiple methods, which are kept segregated in the human hemispheres, it is unlikely to reflect methodological differences. perhaps because optimal performance in one of them Efforts to characterize hemispheric differences in is detrimental to optimal performance of the other. terms of oversimplified dichotomies such as the left Another suggestion is that the dominance of the left hemisphere to be the analytic, verbal part of the brain hemisphere over the right hand and skilled movement and the right the perceptual, creative, or emotional part preceded its dominance over language. have failed. An interesting concept that has emerged Because parts of the unsolved problem of function- from various studies is that functional lateralization is al brain lateralization may stem from the nature of the not rigidly contingent upon the stimulus itself but method used, it is worth considering the primary meth- rather upon the way the stimulus is processed. Both ods used to study hemispheric specialization. A com- hemispheres play active roles during language pro- mon approach to study functional brain lateralization cessing, with each side contributing in a complemen- is the investigation of patients with unilateral hemi- tary, not exclusive fashion. The left hemisphere, for spheric lesions. However, the most compelling evi- instance, shows a slight advantage on recognizing dence of functional brain lateralization has been detail, whereas the right hemisphere concentrates on obtained from the results of studies on patients who the broad background picture. This would explain why have either undergone hemispherectomy (removal of left-hemisphere language areas are so good at a pre- the cerebral cortex of one hemisphere) or commisuro- cise representation of words (phonology) and word tomy, a transection of the corpus callosum (the bundle sequences (morpho-syntax), while the right brain of nerve fibers connecting the two hemispheres of the seems to supply a wider sense of context, prosody, brain). Commisurotomy permits a direct test of each pragmatics, and meaning. Moreover, the cortical areas hemisphere almost working in isolation from the other specialized for language are not solely concerned with in these so-called split-brain patients. Studies with words. Studies of congenitally deaf individuals have these patients revealed a clear dominance of the left shown that the brain areas devoted to sign language hemisphere for language, particularly the syntactic are the same as those that organize spoken and heard aspects of language and language output. However, the communication. Such regions are therefore special- right hemisphere appears capable of understanding a ized for symbolic representation and communication limited amount of words and rudimentary phrases but rather than for spoken language as such. Currently, the is only able to produce speech in an extremely limited relationship between lateralization and language rep- context. It is not able to recognize when one word is resents a continuing research challenge in neurolin- superordinate to another or judging antonyms and it guistics, but a great deal of research focuses on the cannot use word order to assist in deciphering mean- brain’s integrating rather than lateralized abilities. ing. However, it can indicate when a sentence ends with a semantically inappropriate word and it can make grammatical judgments. A further approach to References study functional lateralization is represented by stud- Bishop, Dorothy V. M. 1990 Handedness and developmental ies in normal subjects, using the simple technique of disorder. London: Mac Keith Press. Corballis, Michael C., and Stephen E. G. Lea, (eds), 1999. The restricting sensory input initially to one or the other descent of mind: psychological perspectives on hominid hemisphere. This has been done with visual (tachisto- evolution. New York: Oxford University Press. scopic presentation) and auditory (dichotic listening) Davidson, Richard, J., and Kenneth Hugdahl (eds), 1995. Brain linguistic material. Language lateralization has also asymmetry, Cambridge, MA: MIT Press. 608  LATIN Geschwind, Norman, and Albert M. Galaburda. 1987. Cerebral Kertesz, Andrew (ed) 1994. Localization and neuroimaging in lateralization: biological mechanisms associations and neuropsychology. San Diego: Academic Press. pathology. Cambridge, MA: MIT Press. Lenneberg, Eric H. 1967. Biological foundations of language. Hellige, Joseph B. 1993. Hemispheric asymmetry: what’s New York: Wiley, right and what’s left. Cambridge, MA: Harvard University Molfese, Dennis L., and Sidney J. Segalowitz, (eds) 1988. Brain Press. lateralization in children: developmental implications, New Hiscock, Merrill, Eran Zaidel, Marcel Kinsbourne, Francisco York: Guilford Press. Aboitiz, Andres Ide, and Joseph B. Hellige. 1998. Springer, Sally P., and Georg Deutsch. 1981. Left brain, right Lateralization of language and communication. Handbook brain. Oxford: W.H. Freemann, , New York: W. H. Freeman of neurolinguistics, ed by Brigitte Stemmer and Harry A. and Company, 5th edition, 1998. Whitaker. San Diego: Academic Press. Trevarthen, Colwyn B., (ed) 1990. Brain circuits and functions Iaccino, James F., 1993. Left brain–right brain differences: of the mind: essays in honor of Roger W. Sperry. New York: inquiries, evidence, and new approaches. Hillsdale: Lawrence Cambridge University Press. Erlbaum Associates. SABINE WEISS Latin The Latin language (Lingua latina) is a member of the literary Latin language. The vocabulary from Late and Italic subfamily of the Indo-European family of lan- Medieval Latin has been influenced by the Germanic guages; it was brought into the Italian Peninsula languages in the army terminology, whereas the around the tenth century BCE by a wave of immi- Arabic influence was most prolific in the case of the grants from the North. As for its ties to non-Italic Indo- scientific language — e.g. algebra. European languages, it is closely related to Sanskrit The history of the Latin language may be divided and Greek and to the Germanic and Celtic subfamilies. into four periods, similar to the periods of Latin liter- In Italy, Latin was initially the dialect of the minor dis- ature. The Early Period (240–70 BCE) corresponds to trict of Latium on the left bank of the lower Tiber the time when Latin became the dominant language in River, and then it spread in the entire Italian Peninsula, Italy. The Golden Age, or The Classical Period (70 acquiring a number of features from the indigenous BCE–14 CE), marks the age when the Latin language languages it replaced. Latin is supposed to have origi- developed into an artistic medium of expression and nally been a member of a group of closely related attained its greatest richness and flexibility, when it dialects known as Latin-Faliscan, and its success was became standardized in grammar and vocabulary. the consequence of the successful Roman Empire. Its During this period, when it also became the language name comes from Latini, a group of related tribes who of the Mediterranean area, three types of Latin could settled in the territory of Latium. be distinguished: Classical written Latin, Classical Early Latin inscriptions exist from around the sev- Oratorical Latin, and Colloquial, or Vernacular, Latin. enth century BCE. The oldest example of Latin con- The Silver Age (14–130 CE) is the period when Latin sists of a five-word inscription on a fibula dating from became the lingua franca of the western part of the sixth century BCE. The oldest texts clearly in Europe. During all this time, the language follows a Roman Latin date mostly from the third century BCE, continuous process of development, the most notice- and among them there are rich texts reflecting an ear- able changes being in the area of vocabulary and syn- lier, pastoral society. The affluence of fairly continu- tax. Finally, the Late Period (200–600 CE) is the time ous and linguistically consistent material that the when invading barbarian tribes brought many foreign emerging Latin literature offers has been indispensa- forms and idioms into the Latin language, to the extent ble in the reconstruction of the ancestor language, that this new form of Latin was called the Lingua Proto-Indo-European. Romana and was distinguished from the Lingua Latin underwent changes due to several foreign Latina, the classical language of the educated people. influences, the most important being those of the In the Medieval Period, Latin served as the interna- Celtic dialects in Northern Italy, the Etruscan language tional written medium of communication, as well as in central Italy, and Greek in Southern Italy, the last the language of science, philosophy, and theology. one being probably the most important, at least for the Approximately two thirds of all medieval European 609 LATIN literature was written in Latin, and the higher society belong to only one of these declensions. The first used it in everyday life. Its vocabulary absorbed many declension consisted mostly of feminine nouns. The words from the local languages to meet the changed second declension included masculine and inanimate intellectual and social conditions. The Medieval (or neuter nouns. The third declension consisted of nouns Low) Latin borrowed new words from various of all genders, but with similar endings. The fourth sources, so that new, distinct meanings emerged. In the declension contained masculine and neuter nouns fifteenth and sixteenth centuries, the New (or Modern) (with one exception), whereas the fifth declension had Latin emerged. The writers of the Renaissance pro- only a few words, all feminine. Latin nouns were duced a new and brilliant Latin literature that was a inflected for gender, number, and case. Latin had six very close replica of that of the Latin classical writers. cases: the nominative (or subject case: frater ‘broth- Latin was the diplomatic language of communication er’), the genitive (or possessive case: fratris ‘of the until the eighteenth century, the teaching language in brother’), the dative (or the giving case: fratri ‘to the the European universities until the late nineteenth cen- brother’), the accusative (or the direct object case: tury, and the language of the Roman Catholic liturgy fratrem ‘the brother’), the vocative case (frater ‘broth- until the twentieth century. Today, Latin is officially er!’), and the ablative case (or the agent case: fratre spoken only in Vatican. However, together with Greek, ‘by the brother’). it is the main source for creating scientific terms all All Latin adjectives are in form really nouns; that is, over the world. they have approximately the same endings as nouns. The Latin (or Roman) alphabet was created in the They fall into two classes: the first class contains the seventh century BCE on the basis of the Etruscan adjectives based on the first and second nominal alphabet. The original Latin alphabet consisted of 21 declensions, and the second class contains the adjec- letters–– A B C D E F Z H I K L M N O P Q R S T V tives based on the third declension. Apart from the X––, but later Z was replaced by G. The letters Y and adjectives proper, there is the class of pronominal Z were borrowed from the Greek alphabet after the adjectives, which follow the pronominal type of conquest of Greece in the first century BCE. During declension. The adjectives have three grades of com- the Middle Ages, J, U, and W were added. The final parison: positive (the basic form of adjective: beatus alphabet consisted of 26 letters, as follows: A B C D E ‘happy’), comparative (beatior ‘happier’), and F G H I J K L M N O P Q R S T U V W X Y Z. It must superlative (beatissimus ‘happiest’), the last two being be mentioned that in antiquity, the small letter (minus- formed by adding specific suffixes to the stem. cule) did not exist, although there were several types Adjectives agree with the noun they modify in the cat- of the capital scripts, such as the elegant book capitals, egories of gender, number, and case. with a rounded shape, the lapidary capitals, used on Pronouns in Latin are used to substitute nouns, monuments, and the rustic capitals, which were less adjectives, and numerals. The pronouns are of the fol- elaborate, but easier to write. The minuscules devel- lowing main types: the personal pronouns (e.g. ego ‘I’, oped later through transformations of the ancient cap- tu ‘you’), the reflexive (e.g. se ‘self’), the possessive ital letters. (e.g. meus ‘my’), the indefinite (e.g. aliquis ‘some- Classical Latin had a typically Indo-European one’), the relative (e.g. qui ‘who’), the interrogative sound system, with five long vowels, five short vow- (e.g. quis ‘who?’), and the demonstrative pronouns els, and plain consonants. Its major characteristic fea- (e.g. hic ‘this’, ille ‘that’). ture is the large number of diphthongs, which later The numerals are of Indo-European origin. The car- disappeared from the language. dinal numeral unus ‘one’ was declined like a first- or Early Latin had intonational stress on the first syl- second-declension adjective, because it was acting in lable of a word, as opposed to the Latin of the Latin as a pronominal adjective, with distinct feminine Republican and Imperial periods, when the accent fell and neuter forms. Apart from 20, the numerals do not on either the last or second to the last syllables of the inflect after tres ‘three’, except when they are turned word. into regular adjectives. The cardinal numerals (e.g. Latin word structure is the result of gradual simpli- unus ‘one’) are usually used with nouns in the nomi- fication of the Indo-European original structure. Latin native case. All ordinal numerals (e.g. primus ‘the has five inflected (or variable) parts of speech: the first’) are declined like regular adjectives of the first or noun, the adjective, the pronoun, the numeral, and the second declension. verb; and four uninflected (or invariable) parts of Most of the adverbs are basically adjectival forms speech: the adverb, the preposition, the conjunction, that derive from the ablative singular forms of the and the particle. respective adjectives. They are used to determine a The nouns are divided into five classes, called verbal action or a quality of an adjective. There are declensions, according to their endings. A noun can two types of adverbs: primitive and derivative. The 610  LATIN first type consists of proper adverbs referring to cir- a verbal noun in the neuter singular form, or more pre- cumstances of place, time, and manner (e.g. jam cisely an adjective in the neuter singular serving as an ‘already’), whereas the second type consists of abstract noun. adverbs that are derived from nouns or adjectives, by Classical Latin had a free word order, in the sense means of special suffixes (e.g. longe ‘far’). Adverbs do that all combinations are possible. However, sub- not grammatically agree with the verb they match; i.e. ject–object–verb seems to be the preferred order in they have one fixed form only. Classical Latin. The verbs have an entirely different set of endings In parallel with Classical Latin, Latin developed a from the nouns. They are divided into four classes, spoken vernacular, which was spread by the Roman called conjugations. The verb conjugations express the army throughout the empire. This sermo plebeius or following distinctions: Lingua Romana, i.e. Vulgar Latin, later developed into the Romance languages. However, it is quite dif- (1) Singular vs. plural: amo ‘I love’ vs. amamus ficult to study this variety, because almost all written ‘we love’. materials in Latin used the Classical Latin forms, and (2) First vs. second vs. third person, both in singular thus there is little documentation of Lingua romana. and plural: amo ‘I love’ vs. amas ‘you love’ vs. One of the most useful works about Vulgar Latin is amat ‘s/he loves’; amamus ‘we love’ vs. amatis the ‘Appendix of Probus’, a list of correct and incor- ‘you love’ vs. amant ‘they love’ rect word forms from around the third century. One (3) Imperfective tense (present, imperfect, future) may say that Ancient Spoken Latin is mainly charac- vs. perfective tense (perfect, pluperfect, and terized by freedom of syntax, by the presence of future-perfect): amo ‘I love’, amabam ‘I was numerous interjections, and by the frequent use of loving’, amabo ‘I shall love’ vs. amavi ‘I loved’, Greek words. amaveram ‘I had loved’, amavero ‘I shall have On the other hand, in the development of Latin, loved’. other characteristics of Vulgar Latin emerged. Thus, (4) Mood, which includes indicative, subjunctive, vowels lost the distinction between long and short: imperative, infinitive, gerund, gerundive, and most of them became short in all Popular Latin participle: amo, amem, ama, amare, amandum, dialects. Many diphthongs disappeared, as well as a amandus, amans, respectively. large number of final vocals and consonants –– some- (5) Active vs. passive voice: amo ‘I love’ vs. amor times all the final syllables –– and in general the pro- ‘I am loved’. nunciation experienced simplification. So-called ‘inchoative’ verbs denote the beginning The Latin noun declension was simplified over of an action, state, or occurrence. There are three types time, and the number of cases decreased. However, a of inchoative verbs––: primitive (e.g. nosco ‘to get to new part of speech appeared: the article emerged in know’), verbal (e.g. pallesco ‘to become pale’), and Popular Latin, the indefinite article from the cardinal nominal (e.g. nigresco ‘to become black’)–– accord- numeral unus, and the definite one from the demon- ing to the way in which they are formed: inchoative strative pronouns ille and iste. The place of the definite verbs proper, inchoative verbs derived from other article was before a noun in the western dialects, but it verbs, and inchoative verbs derived from nouns. followed the nouns in Dacia. The indefinite article ‘Impersonal’ verbs denote the action of an unspeci- invariably preceded the noun. Subject–verb–object fied agent and are used with no expressed subject (e.g. became the established word order. pluit ‘it rains’). They do not have forms for the imper- During the long-lasting and wide use of Latin for ative mood, and the compound tenses are formed with scholarly and literary purposes, its influence on all the the participle neuter form. European languages has been enormous. Not only did Participles are adjectives with regular noun/adjec- the Romance languages assimilate the Latin vocabu- tive endings added to a verb stem. It must be mentioned lary, but also all other languages that, directly or indi- that any Latin adjective or participle can be translated rectly, came into contact with Latin. The legal, as an adjective, but if there is no convenient noun to scientific, and technological jargons have thousands of attach itself to, it is translated as a noun. In Latin, there words of Latin origin. For instance, English, a are three types of participles: the present participle (e.g. Germanic language, inherited words from Latin both amans), the past passive participle (e.g. amatum), and directly and through other languages (especially the future active participle (e.g. amaturus). French), so that the English vocabulary consists of The gerundive (e.g. amandus) is a specialized ver- almost 66% Latin words. English uses words of Latin bal adjective implying ‘oughtness’; it was used either origin, such as cheese, cup, kitchen, and street in as an attribute with the noun in indirect cases or sim- everyday life, not to mention the technological terms, ply as an adjective of obligation, whereas the gerund is perhaps the most common one being computer. 611 LATIN References Pinkster, H. 1990. Latin syntax and semantics. London: Routledge. Allen, W.S. 1970 Vox latina, Cambridge: Cambridge University Sharpley, G.D.A.1999. Essential Latin, London: Routledge. Press. Wright, R.1982. Late Latin and early romance. Liverpool: Harris, M., and Vincent, N. (eds.), 1988 The romance lan- Cairns. guages. London: Croom Helm. Palmer, L.R. 1954. The Latin language, London: Faber and Faber. RADU DANILIUC Lehmann, Winfred Philipp Winfred P. Lehmann has defined a role for the com- directing the Texas effort to create a German–English parative-historical study of Indo-European languages translation system, funded for many years by the US as part of the general study of languages and language Air Force, and later by Siemens Corporation, which history. marketed a system capable of doing useful first draft With the coming of World War II, Lehmann, like translations of technical material such as that produced many of his contemporaries, put his linguistic skills to in quantity by the Common Market. Besides its work in the service of his country, participating in Machine Translation (METAL) project, the University code-breaking efforts and language instruction. His of Texas’ Linguistics Research Center carried out work as officer in charge of the Japanese language related research on other, often ancient, languages school (1942–1946) resulted in the publication of an and still does (http://www.utexas.edu/cola/depts/lrc/ early practical grammar of Japanese. index.html). Following his first academic position (Assistant The impact of Lehmann’s work brought him inter- Professor, Department of German, Washington national recognition. Among other things, he was University, St. Louis, 1946–1949), his move to the awarded the Brothers Grimm Prize during a summer University of Texas at Austin (1949–), and a Fulbright term as Professor at the University of Marburg (1974), Research Fellowship to Norway (1950–1951), a sud- and, in recognition of his scholarly and diplomatic den need took the Lehmanns from a Fulbright in Egypt abilities, he was chosen to head the linguistic delega- to Ankara and the pressing task of creating materials to tion to China when linguists were first invited after the teach English to Turkish speakers as Director of the long diplomatic break (see Lehmann 1975). Because Georgetown English Language Program there of his breadth and depth of contributions to the fields (1955–1956). Like his work with Japanese, this task of linguistics, modern languages, and computational involved the production of materials for a language linguistics, Lehmann is unique among linguists to very different from English. It also underlined the crit- have served as president of three very diverse scholar- ical need to strengthen language teaching in America. ly bodies: the Linguistic Society of America, the His move to the University of Texas at Austin soon Modern Language Association, and the Association found him building the German Department, as its for Computational Linguistics. Chair (1953–1964), into a department ranked first in Lehmann’s book, Historical linguistics (in its sev- the nation, and creating a Linguistics Department eral editions), has been widely translated (see (Chair 1964–72) ranked third. He then created a new Lehmann 1962, 1973, 1967, 1969, 1992). This work, center to teach a wide range of non-European lan- originally published as a textbook, encoded the ever guages and brought numerous colleagues to Texas. To new directions of a rapidly changing field, as did develop new language curricula, Lehmann took monographs such as Proto-Indo-European phonology, advantage of the National Defense Education Act to Proto-Indo-European syntax, and Theoretical bases of create programs at the University of Texas (Austin) Indo-European linguistics. for the study of South Asian, African, and Middle Before Lehmann’s Proto-Indo-European phonolo- Eastern languages. gy (1952), Indo-Europeanists reconstructed the sounds In the 1960s, the field of computational linguistics of the protolanguage as isolated sounds. Lehmann’s began to develop. By the mid-1960s, Lehmann was study of Indo-European phonology as a system of 612  LESKIEN, AUGUST interdependent sounds underscored the need to recog- Biography nize sounds as part of a larger system. Winfred P. Lehmann was born on June 23, 1916, in Lehmann’s Proto-Indo-European syntax (1974), Surprise, Nebraska. Growing up in post-World War I reacting to Chomsky’s early syntactic work and Wisconsin, his undergraduate education at North- Greenberg’s work on implicational universals, sug- western College (Watertown, WI, B.A. 1936) rein- gested that general syntactic structural systems of lan- forced his earlier focus on Greek, Latin, and German. guage might limit the accumulation of structures that As a graduate student at the University of Wisconsin often resulted from comparative-historical reconstruc- (Madison: M.A. 1938; Ph.D. 1941), he pursued the tion. When linguists were still discovering much about study of older Indo-European languages such as the range of grammatical structures in the world’s lan- Sanskrit and Old Irish. In his Old Irish class, taught by guages, this also called attention to the importance of Professor Miles Dillon of Trinity University (Dublin), older Indo-European language data and reminded he not only learned the language of poets but also met Indo-Europeanists that their data were important for Ruth Preston Miller, who became his wife and collab- more general questions of language. orator on Old Irish (Lehmann and Lehmann 1975). His recent books, Theoretical bases of Indo- His publications, awards, and various listings European linguistics (1993) and Pre-Indo-European in Who’s Who’s are summarized online (http:// (2002), further explore the relation of Indo-European www.utexas.edu/cola/depts/lrc/general/facultyhomes/ to typologically diverse structures among the world’s lehmann.html). languages and go on to explore the role of linguistic paleontology in dealing with prehistory. In essence, References Lehmann (2002) pushes the active language type to a period before Proto-Indo-European on the basis of Lehmann, Winfred P. 1952. Proto-Indo-European phonology. structures that one might agree are only relics in Proto- Austin: University of Texas Press. ––––––. 1962. Historical linguistics: an introduction. New Indo-European, an otherwise nominative–accusative York: Holt, 2nd edition, 1973. Japanese edition: Rekishi language. gengo gaku josetsu. Tokyo: Kenkyusha, 1967. German edi- Lehmann’s later work is a unique attempt to tion: Einführung in die historische Linguistik. Heidelberg: describe the language structures of prehistoric speak- Carl Winter, 1969. Spanish edition: Introduction a la lin- ers whom Lehmann calls ‘pre-Indo-European’ as guistica historica. Madrid: Editorial Gredos, 1969. ––––––. 1973. A structural principle of language and its impli- opposed to ‘Proto-Indo-European’. He brings older cations. Language 49.47–66. Indo-European linguistic data here to bear on the prob- ––––––. 1974. Proto-Indo-European syntax. Austin: University lem of what language was like before Proto-Indo- of Texas Press. European. Lehmann’s concern with the intellectual Lehmann, Winfred P., and Ruth P. M. Lehmann. 1975. An intro- and academic context of Indo-European is thus duction to Old Irish. NewYork: Modern Language Association. reflected, not only in his administrative impact on the ––––––(ed.) 1975. Language and linguistics in the People’s academic institutions of which he was a part but also Republic of China. Austin: University of Texas Press. in the changing focus of his publications. ––––––. 1992. Historical linguistics: an introduction, 3rd edi- Lehmann’s work has remained firmly rooted in the tion. London: Routledge. practicalities of language and culture, be they lan- ––––––. 1993. Theoretical bases of Indo-European linguistics. London: Routledge. guage teaching, translating, or the question of histori- ––––––. 2002. Pre-Indo-European. (= Journal of Indo- cal stages of human language. He has challenged the European Studies Monograph, 41.) Washington, DC: status quo on many fronts and impacted the course of Institute for the Study of Man. national and international curricula. CAROL F. JUSTUS Leskien, August Leskien, together with Karl Brugmann, Berthold In 1876, Leskien’s award-winning monograph Die Delbrück, and Hermann Osthoff, formed the so-called Declination im Slavisch-Litauischen und Germanischen ‘ursprüngliche Gruppe’ (original group) of the (The Declension in Slavic-Lithuanian (i.e. Balto-Slavic) neogrammarian movement that emerged around 1875. and Germanic) was published, which was to become a 613 LESKIEN, AUGUST fundamental contribution to the neogrammarian Bulgarian language, especially the codices approach. Within this treatise, he presented a critical (Supraslensis, Zographensis, and Marianus). As a analysis of both August Schleicher’s genealogical theo- result of this work, Leskien published two books on ry –– stating that the evolution and relationship of lan- Old Bulgarian (Old Church Slavonic) that are still guages could be captured in a ‘family tree’ model –– valid today: Handbuch der altbulgarischen and Johannes Schmidt’s wave theory, according to (altkirchenslavischen) Sprache. Grammatik, Texte, which specific changes within a language spread in a Glossar (1871; A Handbook of Old Bulgarian (Old similar pattern as the waves produced by a stone Church Slavonic). Grammar, Texts, Glossary) and his dropped into the water. Leskien rejected Schleicher’s Grammatik der altbulgarischen (altkirchenslavischen) idea of language being comparable to a natural organ- Sprache (1909; A Grammar of Old Bulgarian (Old ism (and thus linguistics being a natural science) and Church Slavonic)). He preferred the term Old criticized Schmidt’s wave theory for considering the Bulgarian rather than Old Church Slavonic, proving evolution of language as separate from its speakers, that Old Church Slavonic was mainly based on their history, migrations, etc. Leskien′s statement that Bulgaro-Macedonian dialects. Like August Schleicher sound laws (i.e. changes in the sound system of a lan- before him, Leskien thus rejected Franz von guage) have no exception was soon considered to be the Miklosich’s ‘Pannonian’ theory, which claimed that main principle of the neogrammarian movement. the oldest Slavic literary language was of pannonian- Analogical change (i.e. assimilation of patterns) as the slovene origin. Leskien’s reviews of Miklosich’s second important mechanism in language change is works on ‘Old Slovene’ are not to be seen merely as a only to be admitted as a valid explanation, after every criticism of Miklosich’s linguistic method, neverthe- conceivable other means was exhausted and a positive less, Leskien, underestimating the value of the first condition for analogy could be stated. complete Slavic etymological dictionary, vigorously Other features still considered typical of Leskien criticized Mikolsich’s Etymologisches Wörterbuch der and the other neogrammarians (although this charac- slawischen Sprachen (1886; Slavic etymological dic- terization is far from being exhaustive) are the preoc- tionary) and refused to review it. In his two books on cupation with historical linguistics, the atomistic Old Bulgarian, as well as in the first volume of his description of linguistic details, and the neglect of con- famous Grammatik der serbo-kroatischen Sprache tent analysis in favor of form analysis. A work of con- (1914; Serbo-Croatian grammar), Leskien, following siderable influence on his own and on the linguistic Schleicher, proposed a classification of the Slavic verb conception of the neogrammarians in general was according to its present stem. Leskien contributed rel- William D. Whitney’s book The life and growth of atively little to syntax; although his work on syntax is language (1875), which Leskien translated into documented as truly significant by his academic lec- German (1876; Leben und Wachstum der Sprache). tures held during the last years of his life, his two Whitney fully endorsed the neogrammarian position books on Old Bulgarian contain almost nothing and with regard to sound law and analogy, and the influ- the manuscript of the second volume of his Serbo- ence of his thinking is traceable in Leskien, when he Croatian grammar, which is supposed to have dealt insisted that language cannot be considered separate with syntax, got lost under the bombing of Leipzig in from its speakers. In the following years, Leskien pub- World War II. Together with Vatroslav Jagic, Leskien lished important works on the Baltic languages. His in 1876 founded the journal Archiv für slavische study Der Ablaut der Wurzelsilben im Litauischen Philologie. Besides his major works, Leskien pub- (1884; The ablaut in Lithuanian root syllables) is con- lished numerous articles on the phonology and mor- sidered to be the first monograph devoted to a special phology of the Slavic languages, as well as on problem of Lithuanian historical phonetics. Equally graphematic questions and Old Bulgarian literature. important is his work Die Bildung der Nomina im Although Leskien devoted most of his scholarly activ- Litauischen (1891; The formation of Lithuanian ities to the linguistic investigation of languages, he nouns), comprising valuable material for the following was also deeply interested in all cultural aspects of generations of researchers. His Litauisches Lesebuch language. His numerous publications on fairy tales, mit Grammatik und Wörterbuch (1919; Lithuanian folk songs, and related literary topics (mostly dealing reader with grammar and dictionary) was posthu- with Baltic, Slavic, and Balkan languages) were wide- mously published. Leskien is considered to be the pio- ly accepted as equal in quality to his linguistic works. neer and founder of Baltic studies in Germany. As the Undoubtedly, Leskien’s research laid the foundations first chair for Slavic philology in Germany, Leskien of modern Slavic and Baltic studies and left its imprint was one of the forerunners of systematic research into on historical and comparative linguistics in general. Slavic linguistics. From the mid-1870s he devoted Both as a scholar and as a university teacher, he influ- himself to the study of Old Bulgarian; his preliminary enced not only many European but also American lin- work focused on the oldest monuments of the Old guists. Among those who studied under Leskien, or at 614  LESSER ANTILLEAN FRENCH CREOLE least attended his lectures, were Jan Baudouin de Saxonian King and Title ‘Geheimer Hofrat’ (Court Courtenay, Aleksandar Belic´, Leonard Bloomfield, Councillor). Leskien died in Leipzig on September 20, Erich Berneker, Olaf Broch, Ben’o Conev, Filipp F. 1916. Fortunatov, Jooseppi J. Mikkola, Holger Pedersen, Ferdinand de Saussure, Aleksandur Teodorov-Balan, References Lucien Tesnière, Nikolai S. Trubetzkoy, Karl Verner, Gustav Weigand, and others. Many of his works, espe- Eckert Rainer. (1981). Zu Leskiens baltischen Forschungen. cially his Handbook and his Grammar of Old Zeitschrift für Slawistik 26. 199–204. Eichler, Ernst. (1981) August Leskiens Wirken für die Slawistik. Bulgarian as well as his Serbo-Croatian Grammar Zeitschrift für Slawistik 26. 168–91. have become classics and are still valid and stimulating. Eichler, Ernst, and Gerhart Schröter (1983) Nachträge zur Bibliographie der Veröffentlichungen August Leskiens, Zeitschrift für Slawistik 28. 280–83. Biography Jagic, Vatroslav. (1917) August Leskien. Almanach der kaiser- lichen Akademie der Wissenschaften in Wien 67. 481–99. August Leskien was Born in Kiel, Germany on July 8, Jankowsky, Kurt R. 1972. The neogrammarians. The Hague and 1840. He completed high school in 1860 and began his Paris: Mouton. university studies of Classical philology at Kiel Leskien, August. 1871. Handbuch der altbulgarischen University under Georg Curtius; in 1862, he followed his (altkirchenslavischen) Sprache. Grammatik, Texte, Glossar; teacher to Leipzig, Germany, where he obtained his doc- 2nd revised edition 1886, 10th edition 1990. ———. 1876. Die Declination im Slavisch-Litauischen und torate in 1864. In 1864–1866, he worked as a teacher of Germanischen. Latin and Greek at the ‘Thomasschule’ in Leipzig; he ———. 1882. Litauische Volkslieder und Märchen. Aus dem continued his academic studies of Comparative Indo- Preussischen und dem Russischen Litauen, gesammelt von European Linguistics with special emphasis on Baltic A. Leskien und K. Brugmann. and Slavic languages under August Schleicher at Jena, ———. 1884. Der Ablaut der Wurzelsilben im Litauischen. ———. 1891. Die Bildung der Nomina im Litauischen. Germany, and received his final academic degree venia ———. 1909. Grammatik der altbulgarischen (altkirchenslavi- legendi (habilitation) in 1867. In 1867–1869, he was uni- chen) Sprache. versity lecturer of Comparative Linguistics at Göttingen, ———. 1914. Grammatik der serbo-kroatischen Sprache. 1. Germany, and in 1869, succeeded Schleicher at Jena as Teil: Lautlehre, Stammbildung, Formenlehre. extraordinarius (associate professor) of Comparative ———. 1915. Balkanmärchen. Aus Albanien, Bulgarien, Serbien und Kroatien. Linguistics and Sanskrit. In 1870, he was appointed ———. 1919. Litauisches Lesebuch mit Grammatik und extraordinarius of Slavic Philology at Leipzig and as Wörterbuch. ordinarius (full professor) at Leipzig University in 1876. Morpurgo Davies, Anna. 1996. La linguistica dell’Ottocento. In 1875, he was Fellow of the Royal Saxonian Society of Bologna: il Mulino; as Nineteenth-century linguistics. Sciences; a member of the Jablonowski Society of London and New York: Longman, 1998. Schröter, Gerhart. (1981). Bibliographie der Veröffentlichungen Sciences in Leipzig and Foreign Member of the Imperial August Leskiens. Zeitschrift für Slawistik 26. 271–78. St. Petersburg Academy of Sciences in 1876; a member Streitberg, Wilhelm. 1919 [1921]. August Leskien. of the Bulgarian Learned Society (since 1911 Bulgarian Indogermanisches Jahrbuch 7. 138–43. Academy of Sciences) in 1884; Foreign Member of the Whitney, William D. 1875. The life and growth of language; as Imperial Academy of Sciences in Vienna in 1898; Leben und Wachstum der Sprache, transl. by August Leskien, 1876. Foreign Member of the Prussian Academy of Sciences ALEXANDER SITZMANN in 1900; a Member of the Swedish Academy of Sciences in Uppsala in 1914; and Foreign Member of the See also Baltic Languages; Brugmann, Karl; Histor- Academies in Belgrade, Munich, and Zagreb. He ical Linguistics; History of Linguistics: Overview; received an honorary Doctorate at the University of Old Church Slavonic; Serbo-Croatian and South Oslo; Serbian Order ‘Sava’ Second Degree, Order of the Slavonic Languages Lesser Antillean French Creole Lesser Antillean French Creole is spoken in the related to other Atlantic French Creoles such as Caribbean on the islands of Guadeloupe, Martinique, Guyanese, Haitian, and Louisianais, but such relation- Dominica, and Saint Lucia. It has been argued to be ships being controversial, we will not take into account 615 LESSER ANTILLEAN FRENCH CREOLE these other creoles in the present study. Within this four- At the phonological level, the vocalic inventory is island cluster, at least two major distinctions differenti- as follows: /i/, /e/, /ε/, /a/, /o/, /u/ with a little variation ate the varieties spoken in Guadeloupe and Martinique in the acrolectal varieties. Lesser Antillean contains from those spoken in Dominica and Saint Lucia. First, some consonants that do not exist in French: /h/, / N /, Guadeloupean and Martiniquais coexist with the French /tΣ/, and /dZ/. Palatalization is also quite common as language, whereas the varieties spoken in the in [ky], [gy], [ty], [dy], and [sy] (Gadelii 1997). Commonwealth Antilles on the islands of Dominica and While the influence of French on the lexicon of Saint Lucia coexist with English. Second, Lesser Lesser Antillean French Creole is obvious, a number Antillean emerged in Guadeloupe and Martinique of parallelisms between Guadeloupean and African (where a plantation colony was developed) and was languages, documented by Cérol (1997), are worth subsequently ‘exported’ to Dominica and Saint Lucia noting at the lexical level. They fall into distinct cate- by migrants. Given that this article is only concerned gories: with the study of Lesser Antillean French Creole, only Nouns: Kikongo musulongo ‘the name of a bakon- the varieties spoken in Guadeloupe and Martinique will go group’ > Guadeloupean mousoulongo ‘savage’. therefore be examined. In 1946, Guadeloupe and Kikongo bebele ‘piece of meat’ > Guadeloupean Martinique became French ‘Départements d’Outre- bébélé ‘dish made out of pieces of meat’. Mer’, a status that they still have today despite separatist African words abound in the realms of the fauna attempts during the 1960s and 1970s. and flora: Kikongo kyokyo meaning ‘a bird’ > Lesser Antillean French Creole developed in the Guadeloupean kyo ‘a species of bird’. Kikongo malan- seventeenth century at the onset of the French colo- ga ‘yam’ > Guadeloupean malanga ‘species of yam’. nization of Guadeloupe with African slaves in 1635, Cérol (1997) also provides numerous examples of the objective being the development of a plantation lexical parallelisms in the domains of anatomy, food, colony based on the cultivation of sugar cane. The habitat, music, and dance. influx of Black slaves lasted up to the 1860s. Ideophones: Kikongo, kya-kya-kya ‘onomatopea Geographically speaking, African slaves in the for laughter’ > Guadeloupean kya-kya-kya ‘laughter or Caribbean came from Kwa-speaking areas stretching to laugh’. from Sierra Leone to the Ivory Coast. They also came Grammatical morphemes: Kimbudu ‘for, to give’ > from Togo, Benin, Nigeria, Ghana, Gabon, Congo, Guadeloupean ba ‘for, to give’. Zaïre, and Angola (Gadelii 1997:2). Parkvall (1999) Adverbs: Kikongo phwele ‘crowd, a number of, a argued that Lesser Antillean originated in present-day lot of ‘ > Guadeloupean pwele ‘a lot of’. St.Kitts with two main substrata: Kwa and Western Besides direct lexical inheritances from the French Bantu languages. Singler (1995) takes a similar stand, lexifier and the African substrates, Lesser Antillean but suggests that the Bantu substratum was more has naturally produced many innovations of its own important than the Kwa one based on demographic such as palaviré ‘a couple of slaps’, pépa ‘cheap data. Lefebvre (1986) has defended that the Gbe clus- shoes’, or sanpalé ‘deaf’. Furthermore, the language is ter (including Ewe and Fon) in present-day Togo and extremely innovative in the domain of derivational Benin constitute the main substratum in the French morphology, as it borrows French affixes to make up colonies. The other languages having contributed to new words: makrèl ‘indiscreet woman’ + -é > makrélé the genesis of Lesser Antillean French Creole are not ‘to mind other people’s business’ or pwofité ‘to abuse’ only varieties of seventeenth-century French but also +* -asyon > pwofitasyon ‘abuse, tyranny’ (Cérol the languages of the Caribs, the indigenous group liv- 1997). ing in Guadeloupe prior to the arrival of the French In the realm of morphosyntax, the definite article la colonists. While some scholars do not fully acknowl- or a always occurs in a postnominal position and repre- edge the input of Carib languages to the genesis of sents one of the areas where Guadeloupean displays a Lesser Antillean due to the extermination of the tribes behavior different from Martiniquais. In Guadeloupean, by the end of the seventeenth-century (Cérol 1997), the definite marker is consistently la as in vwati-la ‘the others like Wittman and Fournier (1994) think other- car’. In Martiniquais however, while la occurs after a wise. According to these authors, the language of the consonant (or a semi-consonant) as in tab-la ‘the table’, Karina tribe that the French colonists first encountered its counterpart a must appear in a postvocalic position, in the 1620s was heavily influenced by Tupi and may as in loto-a ‘the car’ (Bernabé 1987). Damoiseau (1999: have given way to a French Carib pidgin (Gadelii 33) further documented in Martiniquais other variants 1997:37). They further suggested that Arawakan lan- that differ according to the environment. For instance, guages spoken by indigenous Ameridians before the lan appears after a consonant cluster as in madanm lan Europeans arrived also influenced French-related cre- ‘the woman’, and an appears after a nasal vowel as in oles in the Caribbean. pwason an ‘the fish’. From a cognitive perspective, it is 616  LÉVI-STRAUSS, CLAUDE worth observing that definite nouns are prime candi- yield different interpretations. As in a number of cre- dates for plural marking, otherwise the marker does not oles, one can observe that adverbials like jou tala ‘that appear, e.g. Sé tab la ‘the tables’ (Damoiseau 1999:34). day’ have an anchoring effect on the event and are As for the indefinite article, it is always found in a instrumental in deriving the relevant temporal and prenominal position as in an tab ‘a table’. aspectual interpretation. Such areas of investigation In the realm of question formation, Lesser Antillean display the true complexity of creole grammars. has recourse to two strategies: it may front an interrog- ative word as in es ou ka travay ‘do you work’ or use References raising intonation ‘ou ka travay?’ ‘you work?’ Bernabé, Jean. 1983. Fondal-Natal, Grammaire Basilectale Negation is expressed by the morpheme pa, which Approchée des Créoles Guadeloupéen et Martiniquais: always appears in a preverbal position: Patrik pa lavé Approche Sociolittéraire, Sociolinguistique et Syntaxique, loto a ‘Patrick did not wash the car’ (Damoiseau 1999). vols. 1,2,3. Paris: L’Harmattan. As far as tense, mood, and aspect markers are con- Bernabé, Jean. 1987. Grammaire Créole, Fondas Kréyol-la: cerned, just like most of its congeners, Lesser Antillean Eléments de base des Créoles de la Zone Américano- Caraïbe. Paris: L’Harmattan. makes use of preverbal unbound markers which occur Cérol, Marie-Josée. 1997. Langue et Identité en Guadeloupe: in isolation or in combination to yield diverse tense, Une Perspective Afrocentrique. Condé: Jasor. aspect, and mood interpretations. It is important to note Confiant, R. Césaire. 1993. Une Traversée Paradoxale du Siècle. that the Indo-European concepts of present, past, Paris: Stock. future... are not directly applicable to the study of creole Damoiseau, Robert. 1999. Eléments de Grammaire Comparée Français-Créole Martiniquais. Petit-Bourg: Ibis Rouge. languages but are nevertheless used as useful descrip- Gadelii, Karl Erland. 1997. Lesser Antillean French Creole and tive tools. The following examples from Damoiseau universal grammar. Göteborg: Gothenberg Monographs in (1999) stage a sample of markers that are representative Linguistics. of the Martiniquais variety of Lesser Antillean. Lefebvre, Claire. 1986. Relexification in Creole genesis revisit- ed: the case of Haitian Creole. Substrata versus universals in Iv ka travay ‘Yves works’. The present tense may be Creole genesis, ed. by Pieter Muysken and Norval Smith, expressed with the marker ka. 279–300. Amsterdam and Philadelphia: John Benjamins. Iv ø travay anpil ‘Yves has worked a lot.’ The present per- Parkvall, Mikael. 1999. A note on the peopling of English St Kitts. fect involves no marker. St Kitts and the Atlantic Creoles: the texts of Samuel Augustus Iv té ka travay lè nou rivé ‘Yves was working when we Mathews in perspective, ed. by Philip Baker and Adrienne Bruyn, 63–74. London: University of Westminster Press. arrived.’ The combination té ka expresses past progres- Singler, John. 1995. The demographics of Creole genesis in the sive while té in isolation may express past or pluperfect. Caribbean: a comparison of Martinique and Haiti. The early Iv té ø travay isi a lè i té jenn ‘Yves worked here when stages of creolization, ed. by Jacques Arends, 203–232. he was young.’ Amsterdam: John Benjamins. Jou tala Iv té ø travay anpil: i té las ‘That day Yves had Wittman, Henri, and Robert Fournier. 1994. Le Créole Haïtien, worked a lot, he was very tired.’ Langue Kwa Relexifiée: Vérification d’une Hypothèse “P&P” ou Elaboration d’Astuces Computationnelles? This last sentence shows that depending on the con- Plurilinguismes 8. 115–139. text, the same marker + verb combination ‘té + V’ may MARLYSE BAPTISTA Lévi-Strauss, Claude One of the most original and prolific thinkers of the discover a rationalist universal theory of mind by iden- twentieth century, the French anthropologist Claude tifying unconscious and abstract structures underlying Lévi-Strauss is associated with the development of a the diversity of human cultures. He derived its method school of thought in cultural anthropology known as from structural linguistics, deeply convinced that ‘lin- structuralism. Challenging the dominating empirical, guistics was the only social science which could truly positivistic and functionalist tradition in social sci- claim to be a science and which had achieved both the ences, he viewed cultures as systems and elaborated the formulation of an empirical method and an understand- principles of the structuralist method, which sought to ing of the nature of the data submitted to its analysis,’ 617 LÉVI-STRAUSS, CLAUDE and therefore should be applied as the main model for elements of meaning; like phonemes, they acquire the anthropological study of diverse cultural categories. meaning only if they are integrated into systems. His famous autobiographical work Tristes tropiques By extending the defining contrastive features of (1955) provides an insight into early influences during phonemes to the analysis of the meaning of cultural cat- his formative years of life that greatly contributed to egories, Lévi-Strauss suggests that the underlying prin- his later conceptualization of the structuralist doctrine ciples of human behavior and culture also take the form in anthropology. In addition to his probably innate of oppositions, such as nature vs. culture, male vs. inclination for formal logic, reinforced by academic female, earth vs. sky, hot vs. cold, etc. However, unlike training in philosophy, he was also fascinated by geol- Saussure, who held that meaning was not a preexisting ogy, Freudian psychoanalysis and Marxist ideas, concept but arbitrarily attached to the form and variable which all showed him that true reality was never that across languages, Lévi-Strauss maintains that all human which was the most manifest and that the process of minds are ultimately the same, the underlying organiza- understanding involved reducing one type of reality to tion of all cultures is the same, generated through the another. Later, when he had come to New York as a universal set of semantic oppositions, the apparent dif- refugee during World War II, he began to search for a ferences being due simply to varying arrangement of suitable method to analyze the data on social and fam- these oppositions. Because of the absolute universality ily life of Amerindian cultures collected during his of these principles, what can be discovered by analyzing own expeditions to central Brazil, as well as the accu- cultural categories and products in one culture is merely mulated ethnographic data on North American Indians a formal transformation of those in completely different published by the Smithsonian Institution. During this cultures. Accordingly, the focus of anthropological period, he met the Russian linguist, Roman Jakobson, research for Lévi-Strauss is the notion of social struc- who introduced him to the work of Ferdinand de ture, which represents a model of actual social relation- Saussure and the basic principles of structural linguis- ships and serves to explain them. This structure or model tics, thus exerting a decisive influence on his thinking. is a system of elements, none of which can undergo Trying to extend the formal structural features of lan- change without affecting others. The model has the pos- guage described by structural linguistics to cultural cat- sibility of predicting change. For every given model, egories, Lévi-Strauss began to view culture like a there is a system of transformations that form a set, a grammar, a formal code consisting of symbols and their group of models of the same type, which make it possi- logical relationships, while the job of the anthropologist ble to predict how the model will react if one or more of was to decipher this code and to reveal the logical work- its elements are submitted to change. ings of the human mind. In search for analogies In his first major work, he applied his theoretical con- between language and cultures, he concludes that just as cepts to one of the central themes of anthropology, the structural linguistics shifts from the study of conscious study of kinship. Contrary to the then prevalent view of linguistic phenomena, observed at the level of surface kinship patterns as explained only by relations of blood structures (speech), to the study of their deep uncon- and descent, under the influence of the French anthropol- scious infrastructure (language), so the anthropological ogist Marcel Mauss and his exchange theory as central to analysis should place priority on the underlying princi- social structure, Lévi-Strauss bases his theory of kinship ples (deep structures) that lie beneath observable every- on relationships of marital alliance and the exchange of day human behavior and cultural phenomena, such as women resulting from rules of incest avoidance. Incest kinship rules (surface structure), in order to explain taboos and marriage rules determine whom one should these phenomena. Structural linguistics also introduces marry and whom one should not, implying reciprocity the concept of the system of phonemic units that are put and the need for exchange while generating marriage together according to certain rules. The basis of its structures that determine the distribution of women analysis are not these units as independent entities, through marriage to different groups. The exchange of rather the relationships between them, based on func- spouses, he says, is the universal basis of kinship systems tioning binary opposites that constitute the structure of in the same way the exchange of goods is the basis of the the system. By analogy, Lévi-Strauss suggests, the true economic system and the exchange of words is the basis nature of social and cultural phenomena is not in these of the communication system. Elementary kinship struc- phenomena themselves, but in the relationships tures are those in which there exists a prescribed, positive between them. No element has any significance in rule for marriage to someone of a particular kinship cat- itself, but only in relation to all elements in the system, egory, for example, to a cross cousin (father’s sisters’ and so that it is the entire system with its axes of contrasts mother’s brothers’ children). Complex kinship structures and oppositions that defines each and every element are those that have negative marriage rules, specifying within it. Like phonemes, kinship terms and patterns, which persons one may not marry. marriage rules, similar prescribed attitudes between cer- According to Lévi-Strauss, the elementary structures tain types of relatives, and other elements of culture are of kinship of small-scale societies consist of a system of 618  LÉVI-STRAUSS, CLAUDE relationships, such as the mother’s brother/sister’s son In his most important work of four volumes of relationship called avunculate, which involves relation- Mythologiques or Introduction to the science of mytholo- ships between mother, father, son and mother’s brother gy, he developed the line of enquiry to its climax by ana- characterized by consanguinity, affinity, and descent. If lyzing 813 myths taken from 200 South and North there is a positive relationship (+) between husband and American Indians and presenting them as the purest man- wife, there will be a negative relationship (-) between the ifestation of the “Savage Mind”. He stressed that myths wife and her brother, and vice versa. If there is a positive do not have a practical function, but are aimed at solving relationship between father and son, there will be a neg- problems as an end in itself. Comparing myth to lan- ative relationship between that son and his mother’s guage, he borrows from structural linguistics the con- brother. Lévi-Strauss shows how these relationships vary cepts of syntagmatic and paradigmatic relations in different societies depending on whether the rule of (paradigmatic relationships constitute a set in which one descent is patrilineal or matrilineal, using the simple for- item can be substituted for another, and syntagmatic rela- mulaic expression A:B::C:D, as in mother’s brother (A) tionships are ones of contiguity and linearity) and : sister’s son (B) :: father (C) : son (D). Proceeding in this describes the workings of myths in terms of relations of way, he reduces diverse historical and geographical rules opposition, symmetry, substitution, and permutation. of kinship to only three elementary structures, construct- Myths function as systems of transformations, so that a ed through only two types of exchange of women, which single myth can never be understood in isolation, only in themselves depend on a single differential characteristic its relation to other myths. The events narrated in any sin- ––- harmonic or disharmonic regime considered, while gle myth form a syntagmatic chain, while the personages the whole set of prescriptions and prohibitions can be and events can be studied as members of paradigmatic deduced from the relationship between the rule of sets occurring in a corpus of myths. In his first volume of descent and the rule of residence. Mythologiques, The raw and the cooked, he begins by Lévi-Strauss fully illustrates the structuralist ideas in analyzing a myth of Bororo people and ends up examin- his analysis of the phenomenon of totemism, the prac- ing a whole collection of myths from South America, tice of associating people and social groupings, such as showing as their common structure the discovery of fire clans, with animals, plants, or other objects, related to and cooking, i.e. the transition from nature to culture, certain beliefs and rituals. Arguing against the evolu- symbolized through invariant oppositions such as tionary view on totemism as a form of primitive reli- raw:cooked (the cooking of culture transforming the raw gion, as well as against the functionalist explanation of of nature), or fresh:rotten (the natural process of purifica- the phenomenon in terms of usefulness of animals, tion or a return to nature). Within this common structure Lévi-Strauss defines totemism primarily as a system of of nature vs. culture, other analyzed myths show a series classification of social groups based on the analogy of transformations, by substituting e.g. the origin of fire with distinctions between species in the natural world. with the origin of water, honey or tobacco. In this way, all Totemic classifications, thus, reflect a system of opposi- the Amerindian myths studied by him are to be consid- tions based on natural objects to organize a set of cul- ered as variant versions of one another, linked together by tural ones. He opposed the widespread view on transformations, just “as the instrumental parts of a musi- presumed mental inferiority of natives and significant cal work, to be studied as one studies a symphony”. differences in the thought processes of so-called “prim- The main criticisms of Lévi-Strauss’s structuralism itive” or non-Western peoples and the modern scientif- have been concerned with its too formal logic based on ic Western culture. He shows that traditional literal adoption of linguistic methods, with its negation non-Western peoples with the “Savage Mind” have an of history and human initiative, as well as with overem- inductive “science of the concrete”, making generaliza- phasis on universals of pan-human unconscious struc- tions from primary sensible data, which “is as rigorous tures and on the intellectual aspect of culture on account as that of modern science”. Close to sensory intuition, of its emotive and practical sides. Despite these various the “science of the concrete” differs from Western sci- critiques, structuralism represented a dominant intellec- ence in that it operates by working directly through per- tual framework in the1960s and 1970s, so that the ception and imagination, classifying things in terms of impressive and voluminous work of Lévi-Strauss has their everyday, sensible features, rather than the under- influenced not only further developments in anthropol- lying abstract features and deductive formal principles ogy, particularly in the fields of symbolic anthropology that the Western scientist uses. In spite of these differ- (focused on the interpretation of meaning) and cogni- ences, Lévi-Strauss claims that the basic thought tive anthropology (related to mental representations of processes in terms of distinctive features and semantic cultural practices), but also a wide range of intellectuals oppositions remain the same in all individuals, “primi- of varied specializations, labeled post-structuralists, tive” or “civilized”. These oppositions for Lévi-Strauss such as the philosopher Jacques Derrida, the intellectu- are “good to think with” and they distinguish humans al historian Michel Foucault, the psychoanalyst Jacques from animals, culture from nature. Lacan, and the literary critic Roland Barthes. 619 LÉVI-STRAUSS, CLAUDE Biography Hénaff, Marcel. 1991. Claude Lévi-Strauss. Paris: Pierre Belfond. LaPointe, Francoise, and Claire LaPointe. 1977. Claude Lévi- Claude Lévi-Strauss was born in Brussels, Belgium, Strauss and his critics: an international bibliography of crit- on November 28 1908. He completed his B.A. in law icism (1950–1976) followed by a bibliography of the and philosophy from the University of Paris in 1931. writings of Claude Lévi-Strauss. New York: Garland. Leach, Edmund. 1970. Claude Lévi-Strauss. New York: He held a teaching post in a provincial secondary Viking. School, 1932–1934. He moved to São Paulo, Brazil, Lévi-Strauss, Claude. 1949. Les Structures élémentaires de la where he taught sociology at the University of São parenté; Paris: Puf; The elementary structures of kinship, Paulo, and made several ethnographic expeditions to transl. by James H. Bell, John R. von Sturmer and Rodney central Brazil, 1935–1939. He returned to France, Needham. Boston: Beacon Press, 1969. ———. 1955. Tristes tropiques, Paris: Librairie Plon, 1955; where he was mobilized by the French army in Tristes tropiques, transl. by John and Doreen Weightman. 1939–1940. He emigrated to New York in 1941, where London: Jonathan Cape Ltd, 1973. he was visiting professor at the New School for Social ———. 1958. Anthropologie structurale. Paris: Librairie Plon; Research, New York, 1942–1945; he was also Cultural Structural anthropology, transl. by Claire Jacobson and B.G. Counsellor at the French Embassy to the United States Schoepf. New York: Basic Books, Inc., 1963. ———. 1962. La Pensée Sauvage, Paris: Librairie Plon; The of America, 1946–1947. He presented a doctoral the- Savage Mind, transl. by George Weidenfeld. London: sis on ‘The elementary structures of kinship’ at the Weidenfeld and Nicolson, 1966. Sorbonne, Paris in 1948. He was Deputy Director of ———. 1962. Le Totémisme aujourd’hui, Paris: Librairie Plon the Musée de l’Homme, Paris, 1949–1950; Director of Totemism, transl. by Rodney Needham. London: Merlin Studies, École Pratique des Hautes Études, Paris, 1964. ———. 1964. Le Cru et le cuit, Paris: Librairie Plon; The raw 1950–1974; and Professor and Chair in Anthropology, and the cooked: Introduction to a science of Mythology, Vol. Collège de France, Paris, 1959–1982. He received the 1, transl. by John and Doreen Weightman. New York: Harper Prix Paul Pelliot in 1949, the Huxley Memorial and Row, 1969. Medalin 1965, and the Erasmus Prize in 1973; he was ———. 1973. Anthropologie structurale deux, Paris: Librairie also a Member of Académie Française in 1973. After Plon; Structural anthropology Vol. II, translated by Monique Layton. Chicago: University of Chicago Press, 1976. retirement in 1982, Lévi-Strauss currently resides and ———. 1975. La Voie des masques, Paris: Librairie Plon; The writes in Paris, France. way of the masks, transl. by Sylvia Modelski. Seattle: University of Washington Press, 1982. Mehlman, Jeffrey. 1974. A structural study of autobiography: References Proust, Leiris, Sartre, Lévi-Strauss. Ithaca, NY: Cornell Barbossa de Almeida, Mauro. 1990. “Mathematical metaphors University Press. in the work of Lévi-Strauss.” Current Anthropology 31. Pace, David. 1983. Claude Lévi-Strauss: the bearer of ashes. Boon, James. 1971. From symbolism to structuralism: Lévi- Boston, MA: Routledge. Strauss in a literary tradition. Oxford: Basil Blackwell. Rubel, Paula, and Abraham Rosman. 1996. Structuralism and Charbonnier, Georges. 1961. Entretiens avec Claude Lévi-Strauss. poststructuralism. in Encyclopedia of cultural anthropology, Paris: Plon-Julliard; Conversations with Claude Lévi-Strauss, ed. by David Levinson and Melvin Ember. New York: Henry transl. by John and Doreen Weightman, London: Cape, 1969. Holt. Hayes, Eugene, and Tanya Hayes (eds.) 1970. Claude Lévi- Sperber, Dan. 1985. On anthropological knowledge. Cambridge: Strauss: the anthropologist as hero. Cambridge, MA: MIT Cambridge University Press. Press. ANITA SUJOLDZIC Lexical Borrowing Lexical borrowing occurs when a language ‘borrows’ and the term ‘borrowing’ is so well established that we a word from another language and incorporates the will just accept the term. new word into its own lexicon. The term ‘borrowing’ Lexical borrowing comes about when two lan- is in some ways an unfortunate metaphor, since the guages are in contact with each other, and there are donor language is never asked whether it wants to lend speakers displaying some degree of bilingualism in the word or not, and the ‘borrowing’ language never both borrowing and source languages. Sarah repays the word. However, other potential metaphors Thomason and Terrence Kaufman (1988) propose a such as ‘giving’ or ‘stealing’ are equally problematic, ‘borrowing scale’ in which lexical borrowing occurs 620  LEXICAL BORROWING with only ‘casual contact’, while at the upper end of Some words are more ‘borrowable’ than others. the scale ‘very strong cultural pressure’ leads to ‘heavy Basic vocabulary (including body parts, natural geo- structural borrowing’. Lexical borrowing often occurs graphic phenomena, weather terms, small numerals, with only minimal contact. The English word, kanga- and the like) are rarely borrowed. Nouns are more bor- roo, was borrowed into English from the Australian rowable than other parts of speech. Einar Haugen Aboriginal language Guugu Yimithirr when Captain (1950) reports the following percentages of loans from Cook and his crew were making repairs to their ship English into American Norwegian: 75.5% nouns, on the northeast coast of Australia, and none of the 18.4% verbs, 3.4% adjectives, and only 1.2% adverbs crew ever learned to speak Guugu Yimithirr. Most or prepositions with 1.5% of others. probably, the word was brought into English after one Borrowing languages can be faced with the prob- of the crew pointed at an animal and was told by one lem of how to fit the borrowed words into their own of the local Aboriginals that the animal concerned was linguistic systems. Two basic strategies are found: a ‘kangaroo’. In fact, the Guugu Yimithirr word gang- adaptation and adoption. Sometimes, different speak- guru refers specifically to the large black kangaroo. ers of the same language choose different strategies, as More intensive cultural contact than occurred when different speakers of English with varying between Captain Cook’s crew and the Guugu Yimithirr degrees of familiarity with French use different pro- can lead to much more extensive borrowing. The nunciations for words like ‘croissant’. Norman Conquest of England in 1066 led to English Adaptation can be either phonological or gram- adopting literally tens of thousands of words from matical. Many languages do not allow consonant French, so that now almost half the words found in an clusters or final consonants, but the words they bor- average English dictionary will have their origins in row from other languages may have them. Japanese French. Sometimes, the same word has been borrowed generally resolves this problem by adding vowels more than once, at different times and with different between or after the consonants concerned, as it has meanings: the English word ‘chief’ came from Old with words like besaboru from English ‘baseball’. In French chef meaning ‘head’ or ‘boss’. The English some instances, English homophones have created word ‘chef’ was borrowed much later from the modern completely different Japanese words as with setur- French expression chef de cuisine or ‘head of the oku, a baseball ‘strike’, and seturoki, an industrial kitchen’. ‘strike’. For borrowing to occur, the borrowing language Grammatical adaptation is required when a noun is generally has to be gaining something, or it would borrowed from a language without grammatical gen- probably make do with the words that it already has. der, for example, into a language that has gender. Borrowing is often thought to occur for either reasons Most nouns borrowed into French are made into mas- of necessity or reasons of prestige. culine nouns, as were all the Americanisms given Necessity can explain the borrowing of ‘kangaroo’ above. Verbs that are borrowed into French also from Guugu Yimithirr, ‘banana’ from Wolof, or require adaptation. The infinitive forms of French ‘kayak’ from Eskimo. English has borrowed countless verbs all have a suffix of one of the forms -er, -ir, or - terms for flora, fauna, and other things that were not re. Virtually all borrowed verbs end up with the -er traditionally found in Britain. suffix, as, for example, chatter ‘to chat in an internet Prestige is often involved in situations where one chat room’. Sometimes, grammatical adaptation has language is thought by its speakers to have more pres- remarkable consequences. Swahili is a language with tige than the other. This motivation explains all of the a complex set of noun class prefixes. The prefix ki- French words that came into English after the Norman marks one set of singular nouns while the prefix vi- conquest. Up until the end of World War II, French had marks their plural counterparts. When the English borrowed relatively few words from English, but since term “keep left” was borrowed as kiplefiti, the initial then French has borrowed countless words from ki- was interpreted as the singular noun class prefix (American) English, referring to many aspects of pop ki-; hence, traffic roundabouts (plural) are referred to culture and technology. Words like sweat ‘sweatshirt’, as viplefiti. pull ‘pullover’, jean ‘jeans’ are used to refer to items Whether a language chooses to adapt or adopt often of clothing popularized by Americans. Sometimes, a depends on the degree of familiarity its speakers have culture can feel threatened by the arrival of large num- with the donor language. Taba (a minority language bers of borrowings. The Académie Française has from eastern Indonesia) borrowed the Malay word jadi fought a valiant but losing battle against all of these ‘thus’ replacing initial Malay // with the closest Americanisms, often proposing to replace them with native equivalent /d/. In more recent times, Taba more French equivalents. Few of their proposals have speakers have borrowed many more words (like baju ever had any success. ‘shirt’) with the Malay sounds preserved so that now, 621 LEXICAL BORROWING // can actually be considered to be a native Taba McMahon, April M.S. 1994. Understanding language change phoneme. Cambridge: Cambridge University Press. Thomason, Sarah Grey, and Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley and Los Angeles: University of California Press. References Trask, R.L. 1996. Historical linguistics. London: Arnold. Weinreich, U. 1968. Languages in contact: findings and prob- Haugen, Einar. 1950. The analysis of linguistic borrowing lems. The Hague: Mouton. Language 26. 210–31. Winter, Werner. 1973. Areal linguistics: some general consider- Hock, Hans Heinrich. 1986. Principles of historical linguistics. ations. Current trends in linguistics, Vol. 11: diachronic, Berlin: Mouton de Gruyter. areal and typological linguistics, ed. by Thomas A. Sebeok. Hock, Hans Henrich, and Brian D. Joseph. 1996. Language his- The Hague: Mouton. tory, language change and language relationship: an intro- JOHN BOWDEN duction to historical and comparative linguistics. Berlin, New York: Mouton de Gruyter. See also: Language: Contact—Overview Lexicalization Lexicalization is a term that is used by linguists for extension of meaning that took place in the senses of referring to several things. The most common uses of the word school. Originally, a school is ‘an institution this term are the following two: (a) When we create a of learning’. However, this meaning has been broad- word to express some meaning, we are ‘lexicalizing’ ened to that of ‘any group of people mentally engaged that meaning by means of the new term (e.g. the word upon shared activities or sharing views of style or opin- software). (b) Another use of the term ‘lexicalization’ ions’, giving as a result the senses of ‘group of artists’ is found in historical linguistics, where it stands for a or ‘group of people sharing opinions’. Another exam- process in which a sequence of words is reduced to a ple of generalization is the word moon. Originally, it single word (e.g. nevertheless). In both uses of the referred to the Earth’s satellite, but it is now applied to term, there is a new lexical item that is introduced to any kind of satellite. express something for which there was no single lexi- Metaphor can be very productive for meaning cal item before. Thus, it is possible to find cases of lex- extension. Metaphor (from Greek metapherein, ‘carry icalization in which the word that acquires a new over’) is a process based on perceived similarity. We meaning was already available in the dictionary, may use a word to refer to a certain notion because of although it was related to a different concept. their similarity of meaning, which does not need to be Semantic lexicalization has many causes. The most very close. The important point here is that the simi- typical one is the extension of the meaning of existing larity should be perceived by the user, whatever it is terms by means of specialization and generalization based on. Eventually, the metaphorical sense will processes, as well as through metaphor and metonymy. become lexicalized. The interpretation of a homonym In a process of specialization, the word’s original such as school in the sense of ‘group of fish’ can be meaning is always narrowed down to a smaller set of related to the senses of school as ‘group of learners’ special referents. An example of specialization is the and may thus be motivated by the relation of similari- English word corn, which was originally a term used ty that we perceive between a group of learners fol- for ‘all kinds of grain’. Later, it specialized to the most lowing a class and a group of fish swimming together typical referent in different countries, so that in the end, and following a leader. it acquired the meanings of ‘wheat’ in England, ‘oats’ In metonymy, on the other hand, the extension of in Scotland, and ‘maize’ in the United States. meaning is based on a relationship of contiguity, i.e. Something similar happened to the English word between the whole of something, e.g. school as an queen, which also went through a specialization ‘institution of learning’, and a part of it, e.g. the les- process. Its original meaning was any ‘wife or woman’, sons. The word school can metonymically stand for but now it is restricted to only one type of wife, that of each of its components, i.e. the building itself, the les- the king. As for generalization, it is the process oppo- sons, the pupils, the staff, the headmaster, etc. The part- site to specialization. We find generalization in the for-whole relationship is one of many possibilities in 622  LEXICON: OVERVIEW metonymy, and there are also many others, such as the (‘because of’), lo (‘the’) and tanto (‘so much’) would producer for product (I’ll have a Pepsi-Cola), material give something like‘because of (the) so much’. It is for thing (Take the iron), effect for cause (He had a interesting to notice here that their English counter- long face), etc. parts are also lexicalized expressions, but in this case All these are cases of lexicalization of new mean- they have become single words: however is made up ings using already existing words. In some other cases, of how and ever, and therefore of there and fore (fore new words are introduced into the language, by means no longer exists as a single word, but it has given us of borrowing (plateau from French, gestalt from the present-day English word before). Other examples German, etc.) or creation of new terms (telephone, are notwithstanding, nevertheless, and alright. We also software, etc.). These processes also lexicalize mean- have lexicalized compounds among nouns, adjectives, ings. In any case, both the extension of meaning and and verbs, e.g. blackbird, wheelchair, colour-blind, to the incorporation of new words are processes that can vacuum-clean, etc. also be examined from a diachronic (historical) point of view, because the new items enter the language at a References certain moment of its history and take some time to be firmly incorporated into the system. Bauer, Laurie. 1983. English word-formation. Cambridge: An interesting case in this respect is what happens Cambridge University Press. Cruse, Allan.1991. Lexical semantics. Cambridge: Cambridge when groups of words that function consistently University Press. together become only one word in the course of time. Dirven, René, and Marjoljin Verspoor. 1998. A cognitive explo- In some cases, they are still used as separate words, ration of language and linguistics, [Chapter 2: Morphology.] even though the expression has to be understood as a Amsterdam and Philadelphia: John Benjamins. whole. This is what happens, for instance, with the Leech, George. 1974. Semantics, Harmondsworth: Penguin. Lyons, John. 1977. Semantics. Cambridge: Cambridge Spanish expressions sin embargo, ‘however’, or por lo University Press. tanto, ‘therefore’, which have little to do with their Ullmann, Stephen. 1957. The principles of semantics. Oxford: components: sin (‘without’), embargo (‘embarrass- Blackwell. ment’) would give ‘without embarrassment’, and por CARLOS INCHAURRALDE Lexicon: Overview A lexicon (or ‘mental dictionary’) contains all the specify how particular nouns, verbs, and adjectives words that a person knows; it is a model of how words combine with arguments. For example, the verb give are organized and stored in the brain. The lexicon takes three arguments (e.g. the arguments for Rod gave specifies how a word is spelled, how it is pronounced, the cat to Jessica are Rod, the cat, and Jessica). The what it means, and its part of speech (e.g. noun, verb, verb’s lexeme specifies information about semantic adjective, preposition). roles associated with the verb (e.g. the entry for give The lexicon figures prominently in grammatical indicates that this verb takes an agent and a patient). theories, although the exact content of a ‘lexical entry’ The entry also indicates the mapping between seman- (also called a lexeme) differs among theories. tic roles and syntax (e.g. the lexeme for give indicates Structural theories propose that verb lexemes, for that the agent is the subject and the patient is the object instance, include information about the types of object when the sentence is in the active voice). that a verb might take. According to this approach, the Words often contain multiple morphemes, i.e. entry for kiss indicates that it takes a noun phrase (as meaningful word parts (e.g. walking consists of walk in Mary kissed the kitten). The subject noun phrase is and -ing, firefly consists of fire and fly). A debate has not specified in the lexical entry because its existence arisen over whether the mental lexicon contains whole is determined by more general rules. In contrast, lexi- words or only morphemes. Some theorists argue that cal theories assume that syntactic information is stored each variant of a word (e.g. brush, brushed, and brush- with individual lexical entries rather than as more gen- ing) has its own lexeme. However, this approach is eral syntactic rules. Under this approach, lexemes problematic for highly agglutinative languages such as 623 LEXICON: OVERVIEW Finnish and Turkish because, in these languages, many grammarians assume that these syntactic categories prefixes and suffixes can be attached to words. can be subdivided. For example, nouns can be divided Consequently, the words in these languages can have into common nouns (nouns that can be preceded by a thousands of variants and would require many lexemes ‘determiner’ such as a or the) and proper nouns to represent all possible variants of each word. (names, places, events, etc.). Common nouns can be Therefore, some theorists have adopted an alternative further divided into ‘count nouns’ (e.g. pebble) and approach to the one-lexeme-per-variant view. They ‘mass nouns’ (e.g. water)––count nouns can be used in propose that the lexicon contains solely morphemes the plural, but mass nouns ordinarily cannot. Thus, and that variants do not have their own lexeme. lexical entries contain information about minor cate- Instead, words are decomposed into their morphemic gories (such as whether the entry is a count noun) as parts (during recognition) or composed (during pro- well as major ones (such as nouns and verbs). duction). Which view is correct appears to depend on Third, semantic relations among open-class words the kinds of morphemes that make up the word. play an important role in determining the structure of Highly regular morphemes (called inflectional mor- the lexicon. Five of the most common relations are phemes) do not significantly alter the root morpheme synonymy, antonymy, incompatibility, hyponymy, and (e.g. the morpheme -s indicates plurality but otherwise meronymy. Synonymy is traditionally defined in terms does not alter the meaning of the word). However, less of substitution. That is, two words are said to be ‘syn- regular morphemes (called derivational morphemes) onyms’ if they can be interchanged without affecting significantly alter the root morpheme (e.g. the mor- the truth value of the statements in which they appear pheme -er changes learn from a verb to a noun). (e.g. sofa and couch). It should be noted that syn- Multimorphemic words with derivational morphemes onyms can rarely be exchanged in every possible con- appear to have their own lexemes, whereas words with text; thus, synonymy is also thought of as a case of inflectional morphemes do not appear to be stored as extreme semantic similarity. Words are said to be whole units. ‘antonyms’ if they are lexical opposites (e.g. hot and The representation of compounds in the lexicon cold). ‘Incompatibility’ refers to a relation in which depends on whether the meaning of the compound is two words cannot be used simultaneously (e.g. an ‘transparent’ or ‘opaque’. Compounds such as button- object can be either a circle or a square, but never hole and snowman are said to be semantically trans- both). ‘Hyponymy’ refers to the inclusion of one class parent because their meaning can be derived from the in another (e.g. apple is a type of fruit). ‘Meronymy’ individual morphemes. Other compounds, such as but- is a semantic relation that exists between words that tercup and hippocampus, are semantically opaque denote parts and wholes (e.g. a stem is part of an because their meaning is not derivable from their apple). parts. Semantically opaque compounds (e.g. table- By measuring how quickly and accurately people spoon) most likely have their own lexical entries in can recognize or name a word, psycholinguists have addition to entries for their morphemic constituents identified several variables that influence access to the (e.g. table and spoon). However, semantically trans- mental lexicon. One of the most robust findings is that parent compounds do not have their own lexical high-frequency words take less time to access than entries and instead are decomposed into their parts low-frequency words. However, this finding appears to prior to recognition. hold only for open-class words. A second finding is The lexicon is organized in a number of ways (three that concrete words (e.g. apple or cat) are more readi- are described here), and this organization affects how ly accessed than abstract words (e.g. freedom or words are accessed during active recognition or pro- democracy). A third finding is that the ease of access- duction. First, the lexicon can be divided into ‘closed- ing a word is affected by whether a semantically sim- class’ words and ‘open-class’ words. Closed-class ilar word has been recently presented in the words (also called ‘function’ words) comprise articles, experiment. That is, a word (e.g. doctor) is accessed prepositions, pronouns, and conjunctions. This class more easily if it has been preceded by a semantically of words is so called because the membership changes associated word (e.g. nurse) than if it has been pre- little over time. In contrast, new open-class words can ceded by an unrelated word. A fourth finding indicates easily be added. Open-class words comprise nouns, that lexical access is affected by how a word sounds. verbs, adjectives, and adverbs and are often thought of Researchers have shown that when people have diffi- as the ‘content’ words of a language. culty retrieving a word (a situation known as the ‘tip- Second, the lexicon is organized according to syn- of-the-tongue phenomenon’), they tend to retrieve tactic category. Lexical entries contain information words with similar sounds rather than words with sim- about their syntactic category (for example, cat is a ilar meanings. Taken together, these findings demon- noun; pet can be a noun or a verb). In addition, many strate that the organization of the lexicon is very 624  LINGUA FRANCA flexible and that grammar, meaning, and pronuncia- Jackson, Howard. 1988. Words and their meaning. London, tion can all play a role in retrieving a word from the UK: Longman. Marconi, Diego. 1997. Lexical competence. Cambridge, MA: lexicon. MIT Press. Pustejovsky, James. 1996. The generative lexicon. Cambridge, MA: MIT Press. References Spencer, Andrew. 1991. Morphological theory: an introduction Aitchison, Jean. 1987. Words in the mind: an introduction to the to word structure in generative grammar. Oxford, UK: mental lexicon. Oxford, UK: Blackwell. Blackwell. Bresnan, Joan (ed.) 1982. The mental representation of gram- Stowell, Tim, and Eric Wehrli (eds.), 1992. Syntax and seman- matical relations. Cambridge, MA: MIT Press. tics, Vol. 26: syntax and the lexicon. New York: Academic Cruse, D.A. 1986. Lexical semantics. Cambridge, UK: Press. Cambridge University Press. CHRISTINA L. GAGNÉ Lingua Franca Although the term ‘lingua franca’ is best known as a Genoa, whose dialects left a stamp on the language, for common noun, meaning a language used between peo- example, in the first and second singular pronouns mi ple who have no language in common, this entry is and ti, which may well be of Venetian origin. Lingua about the Lingua Franca in its proper noun meaning, Franca was diffused to the Western Mediterranean when i.e. the Italian-based contact language that was used the Genoese moved their commercial activities to the throughout the Mediterranean area from the fifteenth Iberian peninsula and North Africa, where it acquired to the nineteenth century. In this usage of the term, an important role as a medium of intercultural commu- ‘Franca’ refers to the Franks, the word that was used nication in places such as Tripoli, Tunis, and, especial- for Europeans in the Arabic-speaking world of the ly, Algiers. The presence of Turkish-speaking soldiers, Late Middle Ages. While ‘Lingua Franca’ literally Berber and Arabic-speaking Moors, and Christian means ‘European language’, its meaning was nar- slaves and renegades speaking a variety of European rowed down to ‘Romance-based pidgin’. Although languages turned Algiers into a multilingual city par Lingua Franca is sometimes referred to as Sabir, this excellence. Due to the Spanish presence in North may be confusing, since the latter is also used to refer Africa, a number of Spanish words were incorporated, to the French-based pidgin that came into existence which resulted in Spanish–Italian doublets such as fazer (presumably on the basis of Lingua Franca) in North (from Spanish) and counchar (from Italian), both mean- Africa in the late nineteenth century. ing ‘do’ or ‘make’. It is this presence of Spanish-derived Although Lingua Franca is traditionally categorized words that forms the major difference between the as a pidgin language, there is some reason to qualify western and eastern variety of Lingua Franca. this classification. As was already observed by Since it is no longer spoken, all information on the Schuchardt (1909), some of its linguistic features, such language is restricted to written sources. Some of as the generalized use of the infinitive, suggest that these are fictional works, such as poems and plays in Lingua Franca may perhaps be more accurately viewed which Oriental characters are depicted as speaking as a form of Foreigner Talk. Other linguists (e.g. some sort of restructured Italian. The problem with Minervini 1996) have claimed that it should rather be this type of source, of course, is to decide whether they seen as a second-language variety of Italian. And since reliably reflect the language as it was actually spoken. Italian and Spanish, the languages that formed the basis The best known among these literary sources is for Lingua Franca, were closely related dialects rather Molière’s (1670) Le bourgeois gentilhomme, which than separate languages five centuries ago, it might contains the famous lines Se ti sabir, ti respondir. Se perhaps more appropriately be categorized as a koiné, non sabir, tazir, tazir ‘If you know, answer. If you i.e. the product of dialect convergence. don’t know, keep quiet’. Another author who has made The Lingua Franca arose as a trade language in the extensive use of Lingua Franca for comical effect is context of the colonization of the Eastern the eighteenth-century Venetian playwright Carlo Mediterranean by the Italian city-states of Venice and Goldoni. Although Schuchardt (1909) felt Goldoni’s 625 LINGUA FRANCA Lingua Franca to be ‘colorless’, it is still worthwhile to feature that may be attributed to Italian and Spanish consult Kahane and Kahane (1976) for a survey of its influence. Direct as well as indirect pronominal linguistic features. objects are often marked by the preposition per, as in A more reliable source of information can be found mi star contento mirar per ti ‘I’m happy to see you’ in nonfictional texts, such as historical and geographi- (lit. I happy see PER you) (Dict.). Complex sentences cal works. The most extensive text of this type is a are quite common, with the subordinate clause being booklet entitled Dictionnaire de la langue franque ou attached to the matrix clause either by a conjunction, a petit mauresque (henceforth Dict.), published anony- relativizer, or a complementizer, or without any overt mously in Marseilles in 1830. It is a 112-page lan- linking element at all. The latter occurs especially with guage primer, containing a wordlist and eight short object sentences, as in mi pensar star meïo ‘I think it’s dialogues, meant to provide the French military with a better’ (Dict.). basic knowledge of the language when they embarked upon their invasion of Algiers. Another important doc- umentary source is Haedo’s (1612) Topographia e his- References toria general de Argel, a geography and history of Anonymous. 1830. Dictionnaire de la langue franque ou petit Algiers, partly based on an eyewitness account. Apart mauresque. Marseilles: n.p. from the Lingua Franca fragments, the book also con- Arends, Jacques. 1998. A bibliography of Lingua Franca. The tains some acute observations on the language situa- Carrier Pidgin 26. 4–5; 33–5. Arends, Jacques. 1999. Lingua Franca en de ‘Europese’ cre- tion in Algiers at the turn of the sixteenth century. ooltalen (Lingua Franca and the ‘European’ creoles). Haedo’s remarks leave no doubt that the Lingua Gramma-TTT 7. 173–90. Franca played an important role as a medium of Cifoletti, Guido. 1989. La Lingua Franca Mediterranea. interethnic communication among Moors, Turks, and Padova: Unipress. Europeans in that city. Haedo, Diego de. 1612. Topographia e historia general de Argel, Valladolid; republished (3 vols.), Madrid: Sociedad As to its main linguistic features, the following de Bibliófilos Españoles, 1927–1929. may be mentioned. While the lexicon is mainly Kahane, Henry, and Renée Kahane. 1976. ‘Lingua Franca’: the derived from Italian and Spanish, some words have story of a term. Romance Philology 30. 25–41. been taken from Arabic, such as taybo ‘good’ and Metzeltin, Michael. 1996. Lingua Franca. Kontaktlinguistik –– marfuz ‘bad’, and from Turkish, such as Yoldach Contact linguistics –– Linguistique de contact, Vol. 1, ed. by Hans Goebl et al., 554–8. Berlin and New York: De Gruyter. ‘Turk’. The morphology is more elaborate than is usu- Metzeltin, Michael 1998. Die linguae francae des Mittelmeers ally the case with pidgin languages. Agreement (The lingua francas of the Mediterranean) Lexikon der between the head and its modifier(s) is quite common, romanistischen Linguistik, Vol. 8, ed. by Günther Holtus, especially with regard to gender, as in barbero bono Michael Metzeltin and Christian Schmitt, 601–10. ‘(a) good doctor’ and bona bastonada ‘(a) severe Tübingen: Niemeyer. Minervini, Laura. 1996. La Lingua Franca Mediterranea: beating’ (both from Haedo). Although verbal inflec- Plurilinguismo, mistlilinguismo, pidginizzazione sulle coste tion is heavily reduced when compared to the lexifier del Mediterraneo tra tardo medioevo e prima età moderna languages, we find occasional use of the past partici- (The Mediterranean Lingua Franca: Multilingualism, lan- ple to express past tense, as in mi mirato iéri ‘I saw guage mixing, and pidginization along the Mediterranean him yesterday’ (lit. I seen (him) yesterday) (Dict.) The coast in the Late Middle Ages and the Early Modern Period). Medioevo Romanzo 20. 231–301. only other verbal form is the infinitive, sometimes Molière, Jean-Baptiste Poquelin de. 1670. Le bourgeois gentil- combined with an adverb such as bisognio ‘neces- homme. Molière, Oeuvres complètes, ed. by Robert sary’ to express future or irrealis, as in cosa bisognio Jouanny. Paris: Bordas, 1993. counchar? ‘What shall we do?’ or ‘What should be Schuchardt, Hugo 1909. Die Lingua Franca. Zeitschrift für done?’ (lit. What necessary do?) (Dict.). Although Romanische Philologie 33; translated as The Lingua Franca by Glenn Gilbert. Pidgin and creole languages: selected reduplication is generally rare in pidgins (as opposed essays by Hugo. Schuchardt, ed. by Glenn Gilbert, 441–61. to creoles), it occurs quite often in the Lingua Franca, Cambridge: Cambridge University Press, 1980. as in mucho mucho ‘very much’ (Haedo) and siéme Whinnom, Keith. 1977. Lingua Franca: historical problems. in siéme ‘together’ (Dict.). Pidgin and creole linguistics, ed. by Albert Valdman, The basic word order is S(ubject)-V(erb)-O(bject), 295–310. Bloomington: Indiana University Press. The entire Dictionnaire de la langue franque along with much although VS order is found as well, e.g. in sentences information on the Lingua Franca can be found at <http:// containing a focused constituent and in subordinate www.uwm.edu/~corre/franca/go.html> clauses. In contrast to many other pidgins, the copula JACQUES ARENDS is usually overtly expressed (by (e)star), as in qouesto star véro ‘That is true’ (Dict.). Pronominal subjects, See also Language: Contact—Overview; Pidgins especially third singular, often remain unexpressed, a and Creoles 626  LITERACY Literacy The term literacy is most commonly interpreted as both literacy and illiteracy should be based on degrees being the ability to read and write. However, focusing in a continuum rather than on categorical distinctions on decoding and encoding leads to meaningless tech- and universal quantitative rates, rendering the tradi- niques that obscure a broader understanding of this tional general distinctions between literacy and illiter- complex cultural phenomenon. Ever since the first acy, and between esthetically oriented literacy and western postwar warnings about the ‘literacy crisis’ functional or instrumental literacy useless. Any mean- and the widespread ‘illiteracy’ among schooled people ingful definition of literacy or illiteracy must refer to a of industrialized countries, questions have arisen that particular sociocultural context. challenge the validity of this focus. A further implication of this relativistic perspective Critical debates involving social history, sociolin- is that literacy as a subject of scientific inquiry concerns guistics, educational, and cultural studies have led to a both the myriad social practices of which the activities specific field of research that questions the psycholin- of reading and writing are an integral part and the mul- guistic view of literacy as a unified, abstract, and uni- titude of sociocultural concepts, values and functions of versal human cognitive capacity relating to the use of what is considered a literate performance in a given sit- written materials, and describes the variety of episte- uation. The focus, therefore, is on the contextual use of mological frameworks of the current research on language and not on abstract, context-free structures or approaches and attitudes toward literacy. on written products of communication. Moreover, if The challenge, therefore, has been to provide a reading and writing are essential elements of cultural broader understanding of the nature and consequences practices, i.e. cultural ways of doing things, and make of writing and reading as political, social, mental, sense within groups and communities, they must, per- physical, and linguistic actions, rather than merely force, be seen as deeply interconnected with and insep- individual standardized technomechanical skills. arable from each other. The greatest concern in literacy studies still focuses By focusing on everyday uses of reading and writ- on the fact that definitions, values, and functions ing in local communities, detailed ethnographic stud- ascribed to literacy are totally ambiguous unless they ies have played an important role in the investigation are viewed in specific sociohistorical contexts. The and description of the variety of events and communi- crucial question is what to consider as literacy in a cation practices that comprise the so-called ‘ecology’ given community or in a given historical epoch. From of literacy practices. Identifying different forms of lit- a pragmatic and ethnographic standpoint, for instance, eracy in different domains of life, such as computer lit- someone who simply signs legal documents may be eracy, academic literacy, workplace literacy, etc., and considered literate in rural areas, whereas someone developing the notion of hidden forms of literacy (in with the skills required to decode written text, but contrast to visible or dominant forms of literacy), whose ability to make sense of that text or to act upon ethnographic studies have also been the basis for it is not developed, is seen as illiterate in postindustri- insightful discussions about the strict divide tradition- al communities. From the same viewpoint, shop floor ally drawn between oral and literate cultures and sub- workers who can read, follow written instructions, and cultures, as well as about the relationship between write occasional short notes may be considered illiter- local forms of literacy and formal training. Some of ate in a technological and computer-based work con- the most telling findings of these studies concern the text, in which information must be fed into and taken complementary function of oral and written forms in out of computer terminals. literate communication practices; the processes of As these examples clearly demonstrate, recognizing informal learning in the acquisition of literacy prac- and manipulating patterns in print and linking these tices; the discrepancy between what counts as literacy patterns to oral language may not suffice to perform in school and in other mainstream institutions; and the successfully in a given sociohistorical or socioprofes- pervasive role of configurations of power and social sional context. Moreover, standardized test scores that knowledge structures in shaping institutional and non- quantify literate competence over time and across institutional literacy practices. space may conceal extremely different and continual- Another key issue in literacy studies has been the ly changing sociocultural realities. Hence, criteria for study of the consequences of literacy on language use 627 LITERACY and on intellectual and social life. In general terms, the naturalized or routinized by influential social groups is social view of literacy extended the notion of text and not equally acquired and used by an entire population, of written language usage conceived by theoretical lin- even with compulsory mass literacy training, which guistic traditions, demolishing an array of mythologies explicitly or implicitly embodies mainstream ideologi- inherited from philosophical and scientific discourses cal and normative agendas, entailing very different con- about the intrinsic value of writing due to the assumed sequences within particular cultures and subcultures. relationship between scriptural order and reason, and These studies also found that literacy practices had been between literate competencies and cognitive, techno- historically embodied within religious, ethical, and ide- scientific, and socioeconomic development. ological issues. The link between the logic of alphabetic writing Competing perspectives of the social and cultural and rationality or logical reasoning was made in the values and effects of literacy on the daily life of mod- 1960s and 1970s by English-speaking scholars, who ern societies were also drawn from the prescriptive integrated and extended the ideas about the cognitive agendas of postwar western national institutions and and social impact of literacy. These scholars attempted international organizations such as UNESCO. Since to demonstrate that a ‘literate mind’ is cognitively dis- the 1970s, the link between these agendas has been a tinct from a nonliterate one, since it is highly skilled in strong commitment for the establishment and expan- thinking and speaking abstractly, clearly, and reason- sion of a standardized literate competence, as a basic ably. Accordingly, it was also asserted that a literate human right to be achieved both in industrialized and society possesses a distinctive form of organization nonindustrialized areas around the world. Following a and is highly skilled in technoscientific development. line of reasoning and discussion that builds on com- Their historical, anthropological, and psychological monplace literacy mythologies intertwined with argu- studies about different, mainly nonindustrialized, cul- ments for competence in personal self-improvement, ture has led many scholars, concerned with how the citizenship, and social mobility in modern nations, acquisition of literacy affects individuals and societies, UNESCO’s studies and campaigns, and many national to corroborate and reinforce some traditional assump- educational policies inspired by them, have reinforced tions about literacy as a skill-specific competence, or a conventional assumptions about the ‘civilizing’ effects cognitive condition, and about the ‘great divide’ of literacy, while encouraging heated debates over the between speaking and writing and between oral and social purposes and political potential of literacy. literate cultures. Another implication of these assump- The studies and campaigns developed over the past tions is that literacy is identified as a neutral and pow- few decades have been based on assumptions about the erful technology, which positively transforms human usefulness and value of universal literacy training in mental capacities and social behavior. Accordingly, its any social context, and have been so devised that illit- appropriation by individuals and societies is primarily eracy and the aforementioned “literacy crisis” have a cognitive and technomethodological problem. As a become a strategic governmental issue, triggering a result, illiteracy is identified as a cognitive handicap number of local and national campaigns and massive on both individual and collective levels, or a tech- financial investments in many nations around the nomethodological insufficiency at an institutional world. These campaigns aimed to ‘eradicate’ illiteracy level (for instance, the use of inappropriate equipment, in the world before the year 2000. The analogy sites, and pedagogical materials). Transferred to com- between illiteracy and disease has been a recurrent munities, therefore, these assumptions are simplisti- theme, widely disseminated through official texts and cally constraining, reducing the idea of literacy by images. ignoring the fundamental issue of the sociopolitical It is generally recognized that the set of values and regulation of access to literacy, which is associated beliefs that underpinned institutional public discours- with social inequalities of power and control. es about literacy as a key element in the production of In contrast to this perspective, and drawing on com- progressive mentalities and productive individuals for parative socioanthropological studies, critical studies on capitalist economies and modern nation-state building literacy have found much evidence that different forms relates to the concepts of progress, development, and and levels of literacy are also shaped by social stratifi- emancipation inherited from post-Enlightenment lib- cation and control: even nominally literate contempo- eral social theories. Included here are beliefs in the lib- rary societies show institution-regulated and often erating and revolutionary consequences of literacy, pyramidal distributions of competence in basic formal such as democracy and political participation. and pragmatic conventions of writing. The resources of Concomitant to such official views and guide- literacy to which people have access in households and lines, however, significant alternative viewpoints communities are fundamentally unequal. Yet, as many have been brought into public and educational dis- cross-cultural and cross-historical comparisons within cussions. Drawing on critical literacy studies and on context-specific settings have demonstrated, what is a Marxist-oriented understanding of emancipative 628  LOCALIZATION OF LINGUISTIC INFORMATION education, many researchers have successfully light on the various forms of influence they may exert demonstrated that the acquisition of literacy may on the economic and political exclusion, or inclusion, entail both negative and positive aspects, and that it of any given contemporary social group. Concomitant is necessarily associated with ideological and politi- to this fundamental questioning is the attempt to pro- cal issues. Accordingly, they have tried to demystify mote a profound rethinking of the empowering prop- the fallacious equation of literacy with progress and erties underlying contemporary institutional and social mobility, as well as the official rhetoric and noninstitutional literacy practices. Needless to say, any current school practices that deliberately obscure the and all new commitments in this field must take into range of social, economic, and political forces that account educational issues and the fact that literate inform literacy practices and formal education. They competencies in postindustrial societies tend to be claim that by promoting homogenizing and manipu- transformed into a set of available consumer objects lative socialization, the acquisition of literacy has like, for instance, reading training within highly struc- become a disempowering process (in contrast to the tured and standardized interpretive schemata that fit in so-called ‘empowering literacy’), which implies pas- perfectly with standardized tests and performances. sivity, contributes nothing to personal improvement, and promotes social exclusion. Additionally, if liter- References acy is often present in unequal social structures, this inequality should be the most important motivator Barton, David, and Ivanic, Roz (eds.) 1991. Writing in the com- munity. Newbury Park, London and New Delhi: Sage for a reevaluation of the conceptions, functions, and Publications. uses to be ascribed to literacy within a given socio- Freire, Paulo. 1970. The adult literacy process as cultural action cultural context. for freedom. Harvard Educational Review 48. Having discarded a view of literacy as a context- Freire, Paulo. 1973. The pedagogy of the opressed. New York: neutral, skill-specific competence, works on literacy Herder and Herder. Goody, Jack. 1979. Domestication of the savage mind. acquisition and language education have also chal- Cambridge: Cambridge University Press. lenged linguistic assumptions about the neutrality and Goody, Jack. 1986. The logic of writing and the organization of transparency of writing as a modality of language society. Cambridge: Cambridge University Press. usage. Consequently, they have also questioned school Graff, Harvey J. 1979. The literacy myth: literacy and social conceptions of literacy that promote the acquisition of structure in the 19th century city. New York: Academic Press. Graff, Harvey J. 1987. The labyrinths of literacy: reflections on strict instrumental skills, passive attitudes, and pre- literacy past and present. London: Falmer Press. scribed behavior, as well as the traditional hierarchiza- Heath, Shirley B. 1983. Ways with words: language, life, and tion of literacy practices, textual genres, and linguistic work in communities and classrooms. Cambridge: modalities. In literacy training, emphasis has been Cambridge University Press. placed on the social and political dimensions of the Olson, David R. 1977. From utterance to text: the bias of lan- guage in speech and writing. Harvard Educational Review differences between different types of literacy, or dif- 47. 257–81. ferent forms of literacy and, more specifically, on the Olson, David R. 1986. Learning to mean what you say: toward whole issue of inequality between them. a psychology of literacy. Literacy, society and schooling, ed. In light of the different perspectives on literacy, and by Suzanne de Castell, Allan Luke and Kieran Egan. emphasizing the ideological dimension of any under- Cambridge: Cambridge University Press. Scribner, Sylvia, and Michael. Cole. 1981. The psychologie of standing of it, current research in this area focuses on literacy. Cambridge, MA: Harvard University Press. the variety of written materials and communication Street, Brian V. 1984. Literacy in theory and practice. Cambridge, practices of everyday life in different sociocultural New York and Melbourne: Cambridge University Press. networks and communities, in an attempt to throw INÊS SIGNORINI Localization of Linguistic Information The subject of localization is the relationship of func- the description of the anatomical structures necessary tion with neuronal structure. The main issues of local- to carry out such a function in the normal brain, and ization are the definition of the concept of a function, the description of the duty of a certain area of the 629 LOCALIZATION OF LINGUISTIC INFORMATION brain. Generally, one may differentiate between the normal function missing, but it represents a new state localizability of simple functions such as moving a leg of reorganization of the brain. In the 1960s the and complex functions such as language processing. American neurologist Norman Geschwind Simple functions have the greatest anatomical and (1926–1984) refined Wernicke’s model of language structural correlation whereas complex functions processing, and this so-called Wernicke–Geschwind apparently have relatively widespread cortical and model still forms the basis of current investigations on subcortical localization. normal and disturbed language function. Based on Before the nineteenth century, the common theory data from a large number of patients and a thorough of brain function, which was termed “holism”, understanding of cortical connectivity, Geschwind assumed that the brain simply was not a structure argued that regions of the parietal, temporal, and made up of discrete independent centers, each special- frontal brain lobes were critically involved in human ized for different functions, but must be looked upon linguistic capacities. This model holds that the com- as a single working unit. However, at the beginning of prehension and formulation of language are dependent the nineteenth century, scientists began to assign func- on Wernicke’s area, after which the information is tion to certain neuronal structures and started to favor transmitted over the arcuate fasciculus to Broca’s area the theory of localization of function within the brain. where it can be prepared for articulation. The first localization of linguistic function to a specif- The further expansion and improvement of this ic region and hemisphere of the brain that became model of language processing within the brain has been widely accepted is usually attributed to the French sur- made possible by the application of advanced neuro- geon Pierre Paul Broca (1824–1880) and the German physiological methods. Electrophysiological studies neurologist Carl Wernicke (1848–1905). They exam- using electric or magnetic stimulation and mapping ined the brains of individuals who had suffered a techniques provided a way to delineate the language stroke, become aphasic, and later died. Based on post- areas of the cortex prior to neurosurgical removal of mortem correlation, Broca suggested that the impaired brain tissue. These intraoperative mapping techniques articulation of speech was due to a lesion within the have contributed interesting localization information, left inferior frontal gyrus of the brain (Broca’s area). and provided evidence that a large region of the left Wernicke correlated lesions in the left superior tempo- hemisphere is clearly involved in language production ral gyrus (Wernicke’s area) with disorders in language and comprehension and that language localization comprehension (Figure 1). Wernicke, who was neither varies from patient to patient. Modern functional neu- a pure localizationist nor a pure holist, also developed roimaging techniques such as positron emission an elaborate model of language processing, where he tomography (PET), established in the early 1980s, and proposed that only basic perceptual and motor activi- functional magnetic resonance imaging (fMRI), estab- ties are localized to single cortical areas and intercon- lished in the 1990s, represent a major step toward bet- nections between these functional sites make more ter understanding the localization of linguistic complex intellectual functions possible. He stated that function. Both techniques are imaging procedures that different components of a single behavior are visualize local changes in cerebral blood flow and processed in different regions of the brain and thus metabolism that accompany language processing in advanced the first evidence for the idea of distributed normal subjects. Contemporary studies are offering processing and connectionism. In the first part of the new data on what brain areas might actually do and twentieth century, the idea of functional segregation how they might contribute to a network of neuronal again fell into disrepute, proposing that higher cogni- structures that collectively participate in language pro- tive abilities depended on the function of the brain as cessing. The results obtained by these neurophysiolog- a whole. However, at this time most neuroscientists ical methods support the notion that besides the were willing to accept some sort of cortical localiza- classical Broca’s and Wernicke’s areas, several addi- tion for sensory and motor function. In contrast, the tional distributed cortical and subcortical neuronal situation concerning memory and language was still structures of both hemispheres clearly make a signifi- an active battleground for the opponents and defend- cant contribution to language function (see Figure 1). ers of localization. Until the 1960s, most of the infor- Most of the areas consistently activated by linguis- mation about the localization of linguistic function tic processing lie within the left temporal, parietal, and was based primarily on patients with brain lesions. In frontal brain lobes (Figure 1) and partly in right hemi- this case, the language deficits resulting from brain spheric homolog areas. Extended areas in the left tem- injuries have been compared to the areas of the brain poral cortex are engaged in word retrieval, naming, which became lesioned. Even though the lesion can morphosyntactic processing, parsing, syntactic com- be accurately located, the function that is examined prehension and semantical analysis and also in the after injury does not reflect the simple equation of the articulation of speech. Moreover, several areas in the 630  LONG-DISTANCE DEPENDENCY Parietal many, years has now been seen to be oversimplified. lobe Areas all over the brain are recruited for language pro- 3 cessing, some are involved in lexical retrieval, some in Frontal grammatical processing, some in the production of lobe 2 8 Occipital speech, and some in attention and memory. These new 9 lobe findings are still too fresh for any overarching theories 1 5 that might explain how these areas interact. However, 4 6 there is a definite need for hybrid models of language 7 processing. Language is both localized and distrib- uted, which means that specific language processing Sylvian operations are carried out in particular brain locations Fissure organized in a distributed fashion. Temporal lobe References Figure 1. Left hemispheric cortical brain areas related to audi- tory linguistic processing. Broca’s area and parts of middle Broca, Paul. 1861. Remarques sur le Siége de la Faculté du frontal gyrus (1); dorsolateral prefrontal cortex and parts of Langage Articulé, Suivies d´une Observation d´aphémie superior frontal gyrus (2); supplementary motor area (3); pri- (perte de la parole). Bulletins de la Societé Anatomique 36. mary auditory cortex (4); Wernicke’s area (5); parts of middle Brown, Colin M., and Paul Hagoort (eds.) 1999. The neurocog- temporal gyrus (6); parts of inferotemporal gyrus (7); supra- nition of language. Oxford: Oxford University Press. marginal gyrus (8); angular gyrus (9). Finger, Stanley. 1994. Origins of neuroscience. New York: Oxford University Press. Gazzaniga, Michael S. (ed.) 2000. The new cognitive neuro- sciences. Cambridge, MA: MIT Press. left frontal cortex including Broca’s area are found to Geschwind, Norman. 1974. Selected papers on language and be concerned with word retrieval, verbal working the brain. Dordrecht: Reidel. memory, syntactic processing, semantic encoding and Kertesz, Andrew, (ed.) 1994. Localization and neuroimaging in retrieval, and also articulation, comprehension, and neuropsychology. San Diego: Academic Press. global attentional or executive functions. Right hemi- Kosslyn, Stephen M., and Richard A. Andersen, (eds.) 1992. Frontiers in cognitive neuroscience. Cambridge, MA: MIT spheric cortical areas that correspond to Broca’s and Press. Wernicke’s areas in the left hemisphere are associated Lashley, Karl, S. 1929. Brain mechanisms and intelligence. with prosodic elements that impart additional meaning Chicago: University of Chicago Press. to verbal communication and additional linguistic Ojemann, George A. 1991. Cortical organization of language. functions that have not been clearly defined up to now. Journal of Neuroscience 11. Stemmer, Brigitte, and Harry A. Whitaker (eds.) 1998. Moreover, the gyrus cinguli, the cerebellum, and sub- Handbook of neurolinguistics. San Diego: Academic Press. cortical regions like the basal ganglia and especially Uttal, William R. 2001, The new phrenology: the limits of local- the thalamus, which is presumably a lexical–semantic izing cognitive processes in the brain, Cambridge, MA: MIT interface, may not only play supportive roles during Press. linguistic processing. Wernicke, Carl, 1874. Der aphasische Symptomenkomplex: eine psychologische Studie auf anatomischer Basis, Breslau: The fact that there are so many new brain regions Cohn und Weigert; as Wernicke’s works on aphasia: a emerging in modern lesion and functional imaging sourcebook and review, translated by Gertrude H. Eggert. studies of language suggests that the classical The Hague: Mouton, 1977. Wernicke–Geschwind model, although useful for so SABINE WEISS Long-Distance Dependency The phenomenon of long-distance dependency occurs ‘free word order’ languages, sharply contrast with lan- in languages such as Latin, Russian, Japanese, Korean, guages such as English, which exhibit a relatively or Persian, in which the order of the major constituents fixed word order. In the following German (1) and of the clause is variable. These languages, known as Japanese (2) examples, for instance, the direct object 631 LONG-DISTANCE DEPENDENCY may either precede or follow the subject, a possibility In contrast to Japanese, Serbo-Croatian, in which word that is not available in English: order is much more flexible, allows both objects and subjects to be scrambled to the matrix clause: (1) ... weil niemand dieses Buch gekauft hat ... because nobody this book bought has (6) Verujem novac [ da Petar ------daje Milanu ] ‘... because nobody has bought this book.’ I-believe money that Petar gives to-Milan ... weil dieses Buch niemand gekauft hat ‘I believe that Peter gives money to Milan.’ (2) John-ga Mary-ni piza-o ageta (7) Verujem Petar [ da ------novac daje Milanu ] John-NOMINATIVE Mary-DATIVE pizza-ACCUSATIVE gave I-believe Petar that money gives to-Milan ‘John gave Mary pizza.’ Piza-o John-ga Mary-ni ageta Crosslingustic investigations have also shown that there are differences with respect to the syntactic posi- Ross (1967, 1986) coined the term scrambling to tions in which scrambled elements can be located refer to this variable word order and proposed a uni- within the matrix clause. In Persian, a constituent versal rule to account for it; the rule stated that two undergoing long-distance scrambling can be placed adjacent constituents can be permuted if they are con- both between the matrix subject and the verb (8) and stituents of the same minimal clause. It follows from in the clause-initial position (9). On the contrary, it Ross’s definition of the scrambling rule that this phe- seems that in Japanese this process is somehow nomenon is, as in the examples above, clause-bound. restricted to the clause-initial position, as shown above However, some languages additionally exhibit so- in (4), since long-distance scrambling to the interme- called long-distance scrambling, which involves the diate position within the matrix clause does not always movement of a constituent across a finite clause result in a fully grammatical sentence (10): boundary. In the following Russian sentence, the noun (8) Ali een ketab-ra fekr-mikone [ ke Mehry ------ be phrase ‘Petrov’ has scrambled to a higher clause out of Ali this book think that Mehry to the that-clause between brackets: Hassan dad ] (3) On skazal to Petrov stranno [ to ------ nam Hassan give he said that Petrovit-NOM is odd that us. ‘Ali thinks that Mary gave this book to Hassan.’ pomogal ] helped (9) Een ketab-ra Ali fekr-mikone [ ke Mehry ------ be Hassan dad ] It is claimed in the literature that scrambling is not (10) ??John-ga sono hon-o minna-ni [ Hanako-ga ---- a unitary phenomenon crosslinguistically. There are John-NOM that book-ACC everyone-DAT Hanako-NOM languages, English among them, that do not allow scrambling, languages that only have scrambling with- yonda to ] itta in the verb phrase (e.g. Hungarian), or clause-internal- read that said ly (e.g. German), and languages with a more flexible ?? ‘John, that book, said to everyone that Hanako word order in which a constituent may appear outside read.’ the clause it belongs to. These free word order lan- Another remarkable feature of long-distance scram- guages, on the other hand, also differ as regards the bling is that in some languages this grammatical long-distance scrambling operations they allow, since process can occur more than once within a single this phenomenon is subject to language-specific clause. The following Korean sentence, in which restrictions. In Japanese, for example, most direct scrambling has been applied to both direct and indirect objects can undergo long-distance scrambling, but objects, illustrates this possibility: when subjects are scrambled, the sentence turns out to be ungrammatical: (11) Kulim-uli Yenghi-eykeyj Chelswu-ka [ ai-ka ------ j Picture-ACC Yenghi-DAT Chelswu-NOM child-NOM (4) Sono hon-o John-ga Bill-ni [ Mary-ga------motteiru that book-ACC John-NOM Bill-DAT Mary-NOM have ------i cwu-ess-ta-ko ] sayngkakha-n-ta give think to ] itta that said ‘Chelswu thought that the child gave the picture to Yenghi.’ (5) * Sono hon-ga John-ga [------yoku ureteiru to ] that book-NOM John-NOM well sell that The phenomenon of long-distance scrambling has omotteiru recently received a great deal of attention. However, think there is no consensus at the moment about certain * ‘That book, John thinks (it) sells well.’ descriptive and theoretical aspects concerning this 632  LONG-RANGE COMPARISON grammatical process. German, for example, is tradi- involved: some authors argue that long-distance tionally analyzed as a language that does not have scrambling is a case of NP-movement––which also long-distance scrambling. In fact, a finite clause gives rise to passive sentences in English––while oth- boundary may never be crossed by a scrambled con- ers analyze it as an instance of wh-movement––which stituent in this language: forms English wh-questions. An alternative theory of (12) *...weil Hans den Wagen versprochen hat [ dass scrambling in generative literature rejects both any because Hans the car promised has that type of movement and the existence of a single basic structure. Under this assumption, the alternate con- er ------ reparieren würde ] stituent orders are all generated as basic structures he repair would without any derivational relationship between them. ‘... because Hans has promised that he would repair the car.’ References Nevertheless, German exhibits a structure, sometimes Corver, Norbert, and Henk van Riemsdijk (eds.) 1994. Studies referred to as the ‘Third Construction’, which closely on scrambling. Movement and non-movement approaches resembles long-distance scrambling: there are a spe- to free word-order phenomena. Berlin and New York: cial class of verbs—versprochen ‘try’, versuchen ‘try’, Mouton de Gruyter. Culicover, Peter W. 1996. On distinguishing movements. beginnen ‘begin’ or hoffen ‘hope’, among others—that Linguistic Inquiry 27. allow scrambling to the matrix clause out of their Fukui, Naoki. 1993. Parameters and optionality. Linguistic infinitival complements: Inqury 24. Grewendorf, Günther, and Joachim Sabel. 1994. Long scram- (13) ... weil Hans den Wagen versprochen hat [ ------ zu bling and incorporation. Linguistic Inquiry 25. reparieren ] Grewendorf, Günther, and Joachim Sabel. 1999. Scrambling in German and Japanese: adjunction versus multiple specifiers. ... because Hans the car promised has to repair Natural Language and Linguistic Theory 17. ‘... because Hans has promised to repair the car.’ Grewendorf, Günther, and Wolfgang Sternefeld (eds.) 1990. For some authors, this construction counts as an Scrambling and barriers. Amsterdam and Philadelphia: John Benjamins. instance of long-distance scrambling and therefore Miyagawa, Shigeru. 1997. Against optional scrambling. they classify German together with Japanese or Linguistic Inquiry 28. Russian as a long-distance scrambling language. Müller, Gereon, and Wolfgang Sternefeld. 1993. Improper Turning finally to the theoretical side of this phe- movement and unambiguous binding. Linguistic Inquiry 24. nomenon, many recent generative approaches to Nemoto, Naoko. 1995. Scrambling in Japanese, AGRoP, and economy of derivation. Lingua 97 scrambling assume that the alternate word order Ross, John R. 1967. Constraints on variables in syntax. Ph.D. arrangements in a language are the result of a syntac- Dissertation. Massachusetts Institute of Technology; as tic movement applied to a single basic structure. Yet Infinite Syntax! Norwood, NJ: Ablex, 1986. there is no agreement on which kind of movement is LUISA GONZA´ LEZ-ROMERO Long-Range Comparison Long-Range Comparison (LRC) is an area of linguis- In brief detail, LRC is aimed at establishing puta- tics that is open to extralinguistic concerns. This has tive relationships among or between large linguistic caused it to be popularized outside narrow circles of groupings, and pushing back the frontiers of linguistic specialists. The idea that all human languages and lan- reconstruction. Although he was not the first to note guage families might be ultimately relatable has prob- the relevant parallels, Modern LRC is often dated from ably been current since people started speculating Holger Pedersen’s coining of the term Nostratic (from about human language. There are numerous attesta- the Latin “ours”), to include Indo-European, Uralic, ˆ ˆ tions of such speculations, often linked to the tale of Altaic, and Semitic. In the 1960s Illic-Svityc compiled the Tower of Babel (Genesis 11:1–9), with various the first comparative dictionary of Nostratic. He added refinements. LRC is an essential link in this process. Kartvelian and Dravidian to Pedersen’s grouping. 633 LONG-RANGE COMPARISON Many LRC theories apply to whole continents, of the laryngeal theory in Indo-European studies, groups of continents, and even the whole of the Earth boosted by the discovery and interpretation of Hittite. (proto-World). One prominent figure in this field is The existence of similar segments in Semitic suggest- Joseph Greenberg, who has a long-standing involve- ed some sort of special relationship between Indo- ment in such scholarship. Greenberg’s earlier work on European and Semitic. The apparent parallels in ablaut African linguistic relationships is generally accepted between Indo-European and Kartvelian provide anoth- by linguists, although his later work on Eurasiatic er such example. In the last century, it was believed (partly, but not totally, covering the same areas as Illi - that Uralic and Altaic formed a superfamily (still Svity’s Nostratic) and Amerind (relating nearly all the referred to as Ural-Altaic in popular literature) based languages of North and South America) has come in primarily on typological features common to both: for much greater criticism, especially the latter; see progressive vowel harmony and agglutination. Other especially Matisoff (1990), Dixon (1997), Lass examples of LRC have often been based on similari- (1997:160–71), and the detailed treatments in ties in word order, which is often subject to rapid McMahon and McMahon (1995), Watkins (1997:3), change. Trask (1999). Another such proposal is Dene- Much of the controversy, and many of the prob- Caucasian, combining North Caucasian, Yeniseian, lems, may be illustrated by citing the development of Sino-Tibetan, and Na-Dene, which has also been the research into the history of Basque. As a language iso- subject of criticism. Greenberg’s disciple Merritt late, Basque is of particular interest here, as it is not Ruhlen has been more daring in his proposal of recon- included in many proposals for superfamilies which structing a single origin for all human languages do include almost everything from Kamchatka to the (1994). Some scholars (e.g. Dixon 1997:43) even Sahara. However, this has not stopped some scholars lament that there is a certain popular appeal to the con- from trying. Trask (1996:358-429) provides an cept of all human languages having a single ancestor overview, including extensive discussion, sometimes a (“Proto-World”). Dixon cites numerous references to little harsh, on failed attempts to fit Basque into vari- areal factors, typology, etc., to set up an alternative ous long-range proposals. ˆ ˆ view of linguistic development, which he dubs “punc- Russian scholars, such as the late V.M. Illic-Svityc, tuated equilibrium”. He continues its development, Dolgopolsky, Shevoroshkin, and Starostin, have con- and adds more criticism of LRC, in Dixon and tributed a great deal to recent work in LRC. Much of Aikhenwald (1999). the material is in Russian, but see Shevoroshkin and In some cases, LRC work is hampered by the fact Markey (1986). that many of the putative families on which the respec- Although it often appears that LRC concentrates on ˆ ˆ tive superfamilies are based are themselves on rather lexemes, it should be noted that Illic-Svityc listed shaky ground, or their exact relationships are still pages of useful suggestions on pronouns, affixes, mor- being worked out. Originally, Turkic, Mongolian, and phemes, etc. Nor does more recent work on Nostratic Tungus (some scholars would also include Korean and lack a discussion of morphology; cf. Bomhard Japanese) were grouped together within the Altaic (1996:90–101) and many of the articles in Salmons family, which often figures in various LRC theories, and Joseph (1998). To take one example, the use of a but recently the very status of Altaic itself has been dental stop to form some sort of preterite over wide seriously called into question. Meanwhile, Semitic has swathes of the Nostratic area, also extending to been found to be part of a much wider family conven- Japanese, certainly deserves some sort of discussion, tionally dubbed Afro-Asiatic. Afro-Asiatic in turn is even if much LRC work appears exaggerated. often compared to Indo-European; see especially the A further factor that is assuming increasing impor- work of Bomhard (1984, 1996). The status of Afro- tance and shedding light on purely linguistic issues is Asiatic within the field of LRC has also been open to the ongoing work on human genetics. This work is still some discussion: many Nostraticists would class it in its early stages, but is already yielding interesting under Nostratic, whereas other lines of research, e.g. results, e.g. possible support for Greenberg’s Amerind Greenberg’s Eurasiatic theory, would group it as a theory, see Cavalli-Sforza, et al. (1994). family equal in status to Nostratic with a more remote level of relationship. References The existence of typological similarities between the languages and language families compared often Baldi, Philip. 1990. Linguistic change and reconstruction makes LRC more difficult. Using typological parallels methodology Berlin and New York: Mouton de Gruyter. Bomhard, Allan R. 1984. Towards Proto-Nostratic: a new in genetic reconstruction is equivalent to e.g. suggest- approach. Amsterdam and Philadelphia: John Benjamins. ing that whales are fish based on similarities in form. ———. 1996. Indo-European and the Nostratic hypothesis. One excellent example is provided by the emergence Charleston, SC: Signum. 634  LOUISIANA CREOLE Cavalli-Sforza Luca L., Paolo Menozzi, and Alberto Piazza Renfrew, Colin, and Nettle, Daniel, (eds.) 1999. Nostratic: 1994 The history and geography of human genes. Princeton: examining a linguistic macrofamily. Cambridge: The Princeton University Press. McDonald Institute for Archaeological Research. Dixon, Robert M. W. 1997 The rise and fall of languages. Ruhlen, Merritt 1994. On the origin of languages Stanford: Cambridge: Cambridge University Press. Stanford University Press. Dixon, Robert M. W., and Alexandra Y., Aikhenwald, 1999. The Salmons, Joseph C., and Brian D. Joseph. 1998. Nostratic: sift- Amazonian languages. Cambridge: Cambridge University ing the evidence. Amsterdam and Philadelphia: John Press. Benjamins. Greenberg, Joseph 1987. Language in the Americas Stanford: Shevoroshkin, Vitalij V., and Thomas Markey 1986. Typology, Stanford University Press. relationship and time, Ann Arbor: Karoma. Lass, Roger 1997. Historical linguistics and linguistic change. Starostin, S.A. 1999. Subgrouping of Nostratic: comments on Cambridge: Cambridge University Press. Aharon Dolgopolsky’s The nostratic macrofamily and lin- Matisoff, James 1990. On megalocomparison Language 66(1). guistic palaeontology In Renfrew and Nettle (eds.), pp 106–20. 137–156. McMahon, April M.S., and Robert McMahon 1995. Trask, R.L. 1996. The History of Basque. London. Routledge Linguistics, genetics, and archaeology: internal and external ———. 1999. Why Should a language have any relatives? In evidence in the Amerind controversy”. Transactions of the Renfrew and Nettle (eds.), pp. 157–76. Philological Society 93(2).125–225. Watkins, Calvert. 1997. Thoughts on comparative linguistics in Pedersen, Holger 1967. The discovery of language: linguistic 1997. XVI International Congress of Linguists: Plenary science in the nineteenth century, translated by John Webster Sessions, Paris. Spargo.Bloomington: Indiana University Press. ROBERT ORR Louisiana Creole Louisiana Creole (LC) is now a dying language, but became known as Cajuns. The coexistence of LC as a remains spoken in the southern parts of the state. The minority language alongside Cajun French has had number of native speakers of LC was estimated at far-reaching consequences. Paradoxical though it may somewhere between 60,000 and 80,000 in the early seem, given the obsolescence of Cajun French itself, 1980s, but is no doubt lower today. most observers agree that LC decreolized and changed The first slave importations into what is today in the direction of Cajun during the twentieth centu- Louisiana took place in the second decade of the eigh- ry—many speakers have at the very least a passive teenth century. The geographic origin of African understanding of Cajun French. Since the Cajun set- slaves imported to French Louisiana is unusually well tlement has been limited along the Mississippi, this documented, and differs significantly from that of applies first and foremost to the prairie dialects rather Caribbean slaves. In all, 64% were from Senegambia, than to the riverine varieties. 31% from the Bight of Benin, and 5% from Bantu- A final milestone, of course, in the history of LC is speaking areas, making Mande, West Atlantic, and the transfer of the territory from French to American Gbe languages the main substrates of LC. sovereignty in 1803. Thus began the process of com- It is not clear, however—and this has been subject plete replacement with English of French and LC as to some debate—whether LC developed locally in the majority language. Louisiana, or whether it was brought in from else- Although once more widespread, LC is today spo- where, such as Africa or the Caribbean. On the one ken in a couple of linguistic enclaves together consti- hand, the language does display similarities to West tuting two main dialect areas—an eastern one along Indian French creoles, but no substantial immigration the Mississippi, and a western one around the Bayou therefrom is attested until the beginning of the nine- Têche. teenth century, when large numbers of refugees fleeing The best-studied riverine dialect is that of the the revolutionary turmoil in what is today Haiti upstream parish of Pointe Coupee, where there are per- arrived. At this time, though, LC is generally believed haps 1,000 speakers of LC. Although there are speak- to have already emerged. ers in New Orleans who have moved in from elsewhere Another migration of significance to the develop- in the state, the city has had an indigenous (rather ment of LC is the arrival from 1785 of groups of Gallicized) dialect, the last few elderly speakers of Acadians (ultimately) from Canada, who in Louisiana which are still alive. Other varieties of the eastern 635 LOUISIANA CREOLE dialect are spoken in Vacherie and Kraemer (Bayou- Some of the differences between the riverine Pointe Bœuf) in St. James. Another creolophone community Coupee variety and the more westerly Breaux Bridge relatively untouched by contact with Cajun or other (Pont-Breaux) dialect, which suggest a more far-reach- varieties of French was found in the 1990s in St. ing decreolization of the latter, include partial Tammany just north of New Orleans. An offshoot of (re)introduction of grammatical gender, a more LC completely unaffected by Cajun has also been doc- French-like system of adjectival comparison (using umented on the Mon Louis island in southern Alabama, /ke/ instead of verb serialization involving /pase/ but the last speaker of this variety died in the 1990s. ‘pass’), and changes in article usage and verbal mor- Related varieties were spoken into the twentieth centu- phology. Other comparatively recent changes in this ry in other localities in Mississippi, Alabama, and direction include negation placement partly following allegedly also in western Florida. French patterns, and a copula /(d)εt/ (< d’être). Some The nucleus of the FC-speaking area, however, is dialects also sport front rounded vowels and schwas, made up of the parishes of St. Martin, Lafayette, and normally absent in French-lexicon creoles. Iberia in the Têche area. Children speaking or under- Other features that set LC apart from most of its rel- standing LC may still be encountered in some commu- atives in the Caribbean include retention of a 2sg pro- nities in St. Martin, but they are becoming less and less noun /to/, and preposed pronominal possessors. numerous (they are no longer found in Pointe Coupee). Contact with Cajun and English has, of course, left its In addition, some few speakers are found in Lake mark on the LC lexicon as well, and Cajun is probably Charles in the westernmost part of the state. Before also responsible for the trilled realization of /r/ in LC. World War II, LC could also be heard in the parishes In the communities of Vacherie and Kraemer, the of Natchitoches and Jefferson Davis. Outside of white population has been heavily creolophone, Louisiana, LC is also said to be spoken by emigrant although a shift to English is under way. A similar sit- communities Sacramento (California), southeastern uation can be seen in villages in Pointe Coupee, St. Texas, and elsewhere. John Baptist, St. Martin, Acadia, Lafayette, and St- Since most of the above-mentioned Haitian refugees Landry. In all, about a fifth of the LC-speaking popu- and their slaves, just like the Cajuns, settled on the lation is believed to be made up of whites, but the prairie rather than along the Mississippi, similarities white ethnolects have been shown to be even closer to with Haitian Creole are more salient in the riverine French than their black counterparts. dialects than in the varieties centered around St. Martin. It is believed that certain structural characteristics are due to the Haitian immigration, such as the possessive References construction /zje a mwe ˜/ ‘my eyes’ (alternating with Klingler, Tom. 1992. A descriptive study of the Creole speech of presumably original /mo zje/) and the absolute posses- Pointe Coupee Parish, Louisiana. Ph.D. Thesis. University of Indiana. sive /mo ke ˜/ ‘mine’ (cf. Northern Haitian /ke ˜-a-m/). Neumann, Ingrid. 1985. Le créole de Breaux Bridge, Louisiane. As a result of the subsequent impact of Cajun, LC Hamburg: Helmut Buske Verlag. (even in its most basilectal forms) is—with the exception Neumann-Holzschuh. 1987. Ingrid. Textes anciens en créole of Réunionnais—the French-lexicon Creole the least louisianais. Hamburg: Helmut Buske Verlag. deviant from its lexifier. It is not obvious, though, that Valdman, Albert (ed.), 1997. French and Creole in Louisiana. New York: Plenum Press. decreolization is the only factor responsible for this––it Valdman, Albert, Tom Klingler, Margaret Marshall, and Kevin might be that a similar state of affairs was obtained even Rottet. 1998. Dictionary of Louisiana Creole. Bloomington: in the early days, and the structure of the Alabaman vari- Indiana University Press. ety suggests that this is at least in part the case. MIKAEL PARKVALL Lyons, John John Lyons is a specialist in semantics and linguistic publications include Introduction to theoretical lin- theory. Since the publication of his Introduction to the- guistics (1968), Semantics (2 volumes, 1977), oretical linguistics in 1968, Lyons has been recog- Language, meaning and context (1980), and Linguistic nized as an outstanding linguist. His major semantics (1995). Other publications are Chomsky 636  LYONS, JOHN (first published in 1970, with two other editions in its predecessor in that it covers several topics that were 1977 and 1991) and Language and linguistics (1981). not previously discussed and takes into account new Lyons’ work addresses a broad range of issues of fun- developments in the field. Its comprehensive style damental importance to linguistics. Most of his books brings unequivocal definitions of specific, complex are characterized by a clear and accessible style that notions and explains the relationship between linguistic appeals to a wide student readership. semantics and formal and philosophical semantics and Introduction to theoretical linguistics covers all contemporary pragmatics. major aspects of theoretical linguistics, from phonet- Linguistic semantics: an introduction begins with ics, semantics, and syntax to structural linguistics, uni- an explanation of what meaning is and introduces the versal grammar, and generative and transformational standard distinctions between language and speech, grammar. Although the book presupposes no previous langue and parole, competence and performance, knowledge, it rigorously addresses a wide range of form and meaning, and sentences and utterances. In topics, bringing the reader to an advanced level of addition, Lyons analyzes the different techniques that understanding theoretical linguistics. can be used to define the meaning of words. The book John Lyons’ volume on Chomsky (1970, 1977 2nd also deals with lexical semantics: it explains Peirce’s edition, 1991 3rd edition) is a guide to Noam distinction between tokens and types, as well as other Chomsky’s life and ideas, particularly addressing his fundamental notions such as homonymy, polysemy, earlier groundbreaking work. Lyons emphasizes that the synonymy, sense, reference, extension, intension, nat- short and relatively nontechnical Syntactic Structures ural kinds, and prototypes. Lyons also discusses sen- “revolutionized the scientific study of language”. tence meaning in terms of meaningful and John Lyons is particularly known for his seminal meaningless sentences, implication and negation in two-volume work, Semantics (1977). Volume 1 is a natural language utterances, declarative interrogative, general and comprehensive introduction to the field of imperative, and exclamative sentences, and the princi- semantics. It brings together the results of studies on ple of compositionality. It is the last part of the book, meaning and communication from many disciplines based on J.L. Austin’s theory of speech acts, and par- (including logic and philosophy), placing semantics in ticularly the chapter on context, that has been regard- the larger context of semiotics. It deals with language ed as the most remarkable. Lyons believes that a as a semiotic system, behaviorist semantics, logical satisfactory theory of context should include findings semantics, structural semantics, and semantic field from social sciences such as psychology, anthropolo- theory. A broad review of current semantic theories is gy, and sociology. also found in Language, meaning and context (1981). John Lyons’ role in the history of linguistics has Volume 2 of Semantics, which can be read inde- often been acknowledged. In 1995, a volume of essays pendently, deals with the main problematic issues in by distinguished scholars was published to honor the semantics of natural languages: homonymy, poly- Lyons’ outstanding contribution to linguistics. semy, conversational implicature, deixis and anaphora, Grammar and meaning, edited with an introduction by tense and aspect, mood and illocutionary force. The F.R. Palmer, is a collection of essays related to the wide range of topics analyzed in this volume and the major issues studied by Lyons over the years. originality of the approach make the book a standard reference work in linguistics. Designed as an elementary textbook, Language and Biography linguistics. An introduction (1981) is a general intro- Born in Manchester, UK, John Lyons studied at duction to linguistics and the study of language. It Cambridge, taught at London (1957–1961) and covers a broader range of topics than the classic Cambridge (1961–1964), then became Professor of Introduction to theoretical linguistics and presents the linguistics at Edinburgh (1964–1976) and Sussex main subfields of linguistics: the sounds of language, (1976–1984), and Master of Trinity Hall, Cambridge grammar, semantics, language change, psycholinguis- (from 1984 to September 2000). In 1987, John Lyons tics, language, and culture. was knighted for his exceptional contribution to the Linguistic Semantics: An introduction (1995) is a study of linguistics. critical discussion introducing linguistic semantics, defined by Lyons as the study of meaning systematical- ly encoded in the vocabulary/grammar of natural lan- References guage. The book was initially planned to be a second Lyons, John. 1968. Introduction to theoretical linguistics. edition of his Language, meaning and context (1981). Cambridge: Cambridge University Press. However, even though the general structure of the earli- ———. 1977. Semantics, 2 vols. Cambridge: Cambridge er book is preserved, this volume is very different from University Press. 637 LYONS, JOHN ———. 1981. Language and linguistics. An introduction. ———. (1995). Linguistic semantics: an introduction. Cambridge: Cambridge University Press. Cambridge: Cambridge University Press. ———. 1981. Language, meaning and context. Glasgow: Palmer, F.R. (ed.) 1995. Grammar and meaning. Essays in hon- Fontana/Collins. our of Sir John Lyons. Cambridge: Cambridge University ———. 1991. Chomsky, 3rd, revised and further enlarged edi- Press. tion. London: HarperCollins. First published, 1970. LAURA DANILIUC 638  M Machine Translation Exhilarating goals, serious disappointments, brain’, would be able to translate foreign languages significant successes: all have accompanied machine (Hutchins 1997:197–203). translation in its twentieth- and twenty-first century Early machine translations proceeded word for existence. The early goal––fully automated translation word from the original language into the target of a quality equal to that achieved by human language (in those days of the Cold War often translators––has proven elusive, and some researchers Russian to English), a process that ignores the doubt that it is feasible. Nonetheless, machine complexity of language structure and results in an translation and its sister, machine-aided human inaccurate and barely readable translation. With the translation, must be counted a success if measured development of more sophisticated tools for by their effectiveness in extensive use by multina- computerized syntactic analysis, many researchers tional companies, governments, and the European assumed that success was near at hand. Enthusiasm Commission. Computer tools that aid human for the goal of automatic, high-quality machine translators by automating some parts of the translation translation fueled generous government and private process, coupled with machine translation programs funding of research and development of translation whose output requires final human editing, have systems through the early 1960s. Dissenting voices together met a worldwide demand that is increasing such as that of Yehoshua Bar-Hillel (1960) that this faster than the supply of human translators. perfectionist goal was unrealistic were at first not The history of machine translation in modern heeded. times begins in 1947 with a memo from Warren In 1966, the National Academy of Sciences Weaver, then director of the Natural Sciences Automatic Language Processing Advisory Com- Division of the Rockefeller Foundation, to Norbert mittee (ALPAC) reported that the prospects for fully Wiener, a well-known computer scientist (then called automated high-quality, general-purpose machine cyberneticist): “I have wondered if it were un- translation were dim, causing immediate and thinkable to design a computer which would massive cutbacks in funding for research in the translate. Even if it would translate only scientific field. material (where the semantic difficulties are very By the early 1970s, interest in the field revived, notably less), and even if it did produce an inelegant but accurate and publishable translation untouched (but intelligible) result, it would seem to be worth by human hands was no longer the only goal; while” (Hutchins 1997:197). human–machine cooperation was envisioned and Only two years later, in 1949, the New York Times the creation of automated tools for translators reported that the computer of the United States Bureau became a special field within machine translation of Standards Laboratory, which it called the ‘electric research. 639 MACHINE TRANSLATION The appearance of interactive desktop computer and Japanese speakers. The problems of speech workstations greatly facilitated the development of recognition and speech generation, however, are even translator’s aids, and sophisticated ‘translator’s greater than those involved in written translation. Such workbenches’ were engineered. Current versions of projects always aim at clearly defined topic areas and these translator’s workbenches include large databases types of users, rather than at unrestricted use, but even of specialized vocabulary and high-quality desktop with this limitation, it is not foreseen that large publishing tools, as well as ‘translation memory’, systems for oral translation will be available for which enables the translator to store and reuse existing several decades. translations. Other areas of current research include the Currently used machine translation systems such development of desktop systems for the ‘interactive’ as Logos, Metal, Systran, and PARS can translate composition of specialized documents (for example, even complex texts in a limited and defined domain, business letters) in a target language by a human who or subject area. Nonetheless, they require the does not know that language. The system generates the services of humans to pre-or postedit the texts. In text by a process of asking the user more and more many important ways, the quality of machine detailed questions about what is meant, and choosing translation has not risen appreciably since the 1970s, from its inventory of phrases and structure and the need for human editing is projected to frameworks. continue. Machine translation continues to be a lively field At the other end of the spectrum, uses have been for research and development. Although completely discovered for the rough, often inaccurate, and automated high-quality translation will probably not largely unreadable ‘rough translations’ that result be achieved in the near future, more user-friendly, when context is not specified and texts are not pre- inexpensive systems will continue to be developed, or postedited. Such rough translations have proven particularly those aimed at special domains and for economical in helping users to quickly categorize special purposes. large numbers of foreign texts as either not useful or deserving of thorough translation. Inexpensive translation software for personal References computers, or freeware available on the Web, is widely used for obtaining the gist of e-mail, Web Arnold, Doug, et al. 1994. Machine translation. An introductory pages, and on-line discussion groups. These guide. Manchester/Oxford: Blackwell (url: http:// clwww.essex.ac.uk/MTbook/). systems, often lacking large lexicons or phrase Bar-Hillel, Yehoshua. 1960. The present status of automatic dictionaries, provide an even less exact gist than translation of languages. Advances in Computers 1. commercial systems. An example: French: Il fait 91–163. vraiment beau! English: He does indeed beauteous! Booth, Andrew D. (ed.) 1967. Machine translation. Amsterdam: Correct translation : The weather is really nice! North-Holland. Fortune City. 2001. Examples of nice translations made (Fortune City web site) But while many users by automatic translators available on the market. are disappointed at the poor quality of these ‘quick http:// www.fortunecity.com/business/reception/19/mtex. and dirty’ translations, others are satisfied to have htm even an approximate idea of the contents of the Henisz-Dosert, R., Ross Macdonald, and M. Zarechnak (eds.) text. 1979. Machine translation. The Hague: Mouton. Hutchins, John. 1986. Machine translation: past, present, As always, the market is driven by demand. future. Chichester: Ellis Horwood. Currently, machine translation systems into and ––––––. 1997. From first conception to first demonstration: the from English are available for a very large number nascent years of machine translation, 1947–1954. A of the world’s languages. Translation systems into chronology. Machine Translation 12. 195–252. and from the major European languages (French, ––––––. 2001. Home page. http://ourworld.compuserve.com/ homepages/WJHutchins. German, Italian, and Spanish) are available for a Kugler, M., K. Ahmad, and G. Thurmair (eds.) 1995. smaller number of world languages, while Translator’s workbench: tools and terminology for dramatically smaller numbers of systems are translation and text processing. Berlin: Springer. available to translate nonmajor languages into other Machine Translation (journal), Dordrecht: Kluwer Academic nonmajor languages. Publishing, 1986. Newton, J. 1992. Computers in translation: a practical Spoken language, rather than written, is the focus of appraisal, London: Routledge. much current research. The Verbmobil project of the Whitelock, Peter, and Kieran Kilby. 1995. Linguistic and German Ministry of Research and Technology, for computational techniques in machine translation system example, aims to prototype an on-the-spot oral design, 2nd edition. London: UCL Press. translation of business negotiations between German RUTH H. SANDERS 640  MALAGASY Malagasy Malagasy is one of two official languages of This century-long bilinguality led to profuse borrowing Madagascar (besides French). The population, from French. estimated at 17.5 million for mid-2004, speaks either the standard language, a regional dialect, or both. Dialects Dialects also occur in Mayotte (Comoros). Malagasy has a tripartite division into East History (geographical north, center, east, and southeast), West (west and southwest), and Intermediate (Betsileo, Malagasy belongs to the Southeast Barito subgroup Bara, and Antanosy, sharing features of both former) of the West Malayo-Polynesian subfamily of groups. The following table compares Menabe Austronesian. The language most closely related to Sakalava (West dialect, henceforth ‘Sakalava’) with Malagasy is Maanyan (in the southeast of Borneo). Merina (East) and the reconstructed ancestor language Estimated dates of migration from Borneo range from Proto-Malagasy (P-Mlg): 200 CE until the seventh century or later, based mainly on Sanskrit and Arabic loanwords exhibiting signs of P-Mlg > West/East P-Mlg Sakalava Merina English borrowing via Malay. Non-Malay introductory passages *lime lìme dìmy ‘five’ in some seventh century Old Malay inscriptions have been tentatively identified as being in a precursory *li- > li/di 〈 *kulit hòlitse hòditra ‘skin’ version of Malagasy, Maanyan, or close affiliate. *ele > ele/eli *kelek hèleke hèlika ‘armpit’ Bantu-speakers who subsequently arrived from the 〈 *tina’i tinày tsinày ‘intestines’ African mainland, adopting Malagasy, exercised substantial influence on its phonology and vocabulary. *ti- > ti/tsi *witik vìtike vìtsika ‘ant’ In the ninth to fifteenth centuries, contacts with Islamic *te > te/ti *katen hàte hàtina ‘itch’ traders led to considerable lexical borrowings from *-m,*-n,*ŋ > -Ø/-na *lalan làla làlana ‘path’ Swahili, Arabic, and Malay. In the fifteenth century, the Antaimoro in the southeast began writing, implementing *-r, *-t > -tse/-tra *epat èfatse èfatra ‘four’ the Arabic script soon also adopted by other dialects. In the early seventeenth century, the Menabe Sakalavas Phonology (western lowlands) gained hegemony over the west and northwest that lasted until the eighteenth century, Merina Malagasy has 19 primary consonantal promoting their dialect to corresponding prominence. segments, spelled p, t, k, b, d, g, f, s, h, v, z, l, r, m, n, Around 1797, the reunited Merina in the highlands ts, j /dz/, tr /t.ɹ/, and dr /d. ɹ/ (including four proceeded to expand, annexing the remaining high- affricates), and nine prenasalized consonants lands, the north and the east. Before the French conquest functioning as single segments: mp, nt, nk, mb, nd, ng in 1895, the Merina kingdom encompassed the entire /ŋg/, nts, nj, ntr, ndr. Other dialects additionally have island except the south and part of the west. The Merina velar ñ /ŋ/. In southern central-highlands dialects, ts is dialect (formerly Hova) was becoming standard in much /ð/. In Betsimasaraka, s is /∫/. Some base-initial of Madagascar, enhanced by schooling activities of consonants undergo nasal sandhi with the final nasal British missionaries encouraged by Radama I (reigned of prefixes. 1810–1828). This also led to numerous English bor- Bisegmental consonant clusters are mainly limited rowings. The same missionaries introduced Latin-script to loanwords and include r, l, or s as one component spelling, made official by Radama I in 1820. (artìsta ‘artist’, jeografìa ‘geography’, anglìsy Under French rule (1896–1960), Merina remained ‘English’), also word-initially (frantsày ‘French’, the standard dialect of Malagasy, but the language krismàsy ‘christmas’). Frequently, clusters were of government, education, and the media was French. separated by an anaptyctic i (cf. adirèsy ‘address’, After independence, particularly after the “Malagasi- alikàola ‘alcohol’, Madagasikàra ‘Madagascar’) or, fication” campaign launched in 1975, (Merina) by vowel harmony, an o (boròsy ‘brush’, dokotèra Malagasy became the language of school education, and ‘doctor’) or a (kàratra ‘card’). But in casual speech, an the predominant language of government and the media. unstressed interconsonantal vowel is sometimes But French remains the language of higher education. dropped (fòtsiny [fùtsini] > [fùzni] ‘only’). 641 MALAGASY There are five simple vowels spelled a, e, i or y, o of root-passive verbs); the latter attaches to a host /u/, and ô /o/, and six diphthongs, ai or ay /ay/, ao /aw/, word similar to a suffix: ìa /i/, ià /ya/, òa /u/, and oà /wa/ (y stands for i word- finally). The ô, restricted to loanwords (hôtèly ‘hotel’, Person and Number Nominative Oblique Genitive tômôbìlina ‘automobile’), is often printed without diacritic, leading to ambiguity. In casual speech, the 1st singular (iz)àho àhy =ko/=o 2nd singular ianào anào =nào/=ào dipthongs can contract to [e] (< ai, ià), [i] (< ìa), [o] (< 3rd singular ìzy àzy =ny/=ny ao, oà), or [u] (< òa). 1st plural inclusive isìka antsìka =ntsìka/ Malagasy does not normally have final consonants. =tsìka Historical word-final coda were either dropped (*lepas 1st plural exclusive izahày anày =này/=ày > lèfa ‘liberation, escape’; *lalow > làlo ‘pass’; *apuy > 2nd plural ianarèo anarèo =narèo/ =arèo àfo ‘fire’) or retained as ‘weak ending’ with a dummy 3rd plural ìzy irèo àzy irèo =n’ìzy vowel spelled a (in some dialects e), pronounced as irèo/=ny short schwa [F] or [əF ]: *-k, *-p > -ka; *-r, *-t > -tra; *- m, *-n, *-ŋ> -na (some dialects > -ña, and some deleted it). At suffixation, the historical situation is variedly Authentic adjectives (kèly ‘small’, màina ‘dry’, tsàra reflected in the morphophonemics (lèfa → fandefàsana ‘good’) are distinguished from adjectival verbs: the ‘broadcast’; làlo → lalòvina ‘passed’; tsìdika → latter feature tense (madìo ‘is clean’, nadìo ‘was clean’, fitsidìhana ‘short visit’; lèlaka → lelàfina ‘licked’). hadìo ‘will be clean’) and voice (diòvina ‘be cleaned’). Stress (mainly indicated in dictionaries and language Both form degrees of comparison via combination with primers) is distinctive, cf. tànana ‘hand’ (< *taŋan) vs. independent markers: tsàra kokòa (good more) ‘better’, tanàna ‘town’ (< *tana’-an). Words with a ‘weak tsàra indrìndra (good extremely) ‘best’. In the ending’ have either stress on the antepenultimate comparative, the marker is optional: Natsìro [kokòa] syllable that shifts upon suffixation (àkatra → akàrina nòho ny poàra ny pàoma (was-tasty [more] than the ‘be raised’) or penultimate stress that does not (*ja’it > pear the apple) ‘The apple was tastier than the pear.’ zàitra → zàirina ‘be sewn’; *ta’an > tàna > fitànana The numerals are irày, ròa, tèlo, èfatra, dìmy, ènina, ‘endurance’). Words without a ‘weak ending’ have fìto, vàlo, sìvy ‘1...9’, fòlo ‘10’, ròapòlo ‘20’, tèlopòlo either penultimate stress that shifts upon suffixation ‘30’, etc., zàto ‘100’, ròan-jàto ‘200’, etc. Teens are (fèno → fifenòana ‘filling’; dìa → fandiàvana formed by adding ambìn′ny fòlo after the digit (but ‘threshold’) or ultimate stress that does not (*umei > iràika ambìn′ny fòlo ‘11’). Numerals follow numerated omè → fanomèzana ‘gift’), unless an additional vowel nouns (òlona tèlo ‘three persons’), but precede units of appears (*behay > bè → fanabeàzana ‘upbringing’). measure (pàoma tèlo kilào ‘three kilos of apples’). When the stress shifts onto an i/y, this becomes e either The verbs have an elaborate paradigm of forms that when the preceding vowel is e (kèly → fikelèzana does not include a plural. ‘reduction’), or sometimes reflecting historical *e Tense is expressed by alternating an optional (*alem > àlina → fialèmana ‘dinner time’). prefixal m- (present) with n(o)- (past) or h(o)- (future): manadìo ‘cleanse’/nanadìo ‘cleansed’/hanadìo ‘will cleanse’; mitèny ‘speak’/nitèny/hitèny; and indràmina Morphology ‘be borrowed’/nindràmina/hindràmina. The optional o The morphology of Malagasy involves flexion (e.g. in appears when the prefixless present-tense form has an tense alternation), affixation (voice markation, initial consonant: soràtana ‘be written’/ nosoràtana/ nominal derivation), alternating clitics (personal hosoràtana. pronouns), and root alternation (deictics). Voice is intertwined with some other categories. The The noun is practically invariant. Number is active voice is formed with the prefix maN- (action indicated with a complex system of demonstratives verbs), mi- (action and status verbs), or ma- (status and that also expresses proximity: ìo pàoma ìo ‘this apple’; adjectival verbs) to the stem (manòratra àho ‘I write’; irèo pàoma irèo ‘these apples’; irèry pàoma irèry Mitèhina ny làkana àho ‘I shove the boat [with a ‘those apples’; Marikìvy ìo ‘This one is sour.’ There is pole]’; Matòry àho ‘I sleep’; Mànana tràno aho ‘I also a series of locatives for ‘visible’ with i- (ìo, ìty, ìto, have a house’). Causative verbs prefix maha- (fàly ìtsy, iròa, etc.), somewhat visible with e- (èo, èty, ...), ‘happy’ → mifàly ‘rejoice, be merry’ → mahafàly ‘be or vague with a- (ào, àty, ...), etc. The definite articles pleasing’; tònga ‘reach’ → mahatònga ‘cause’). The are ny for nouns, and i for proper names. active imperative has the suffix -a (manoràta ‘write!’; Personal pronouns distinguish nominative mitehèna ‘shove! [with pole]’; matorìa ‘sleep!’). (subject), oblique (object, possessive pronoun), and Passive forms are frequently used and the most genitive forms (possessive adjective, actor-argument variegated. There are some 30 so-called root-passive 642  MALAGASY status verbs that topicalize the undergoer in an Names with the article Ra- undergo nasal sandhi (n+r > uninflected form (àzo ‘received, understood’; hèno ndr), and a hyphen replaces the apostrophe (ny bòkin- ‘listened to’, hìta ‘seen’). A pronominal actor is in the dRasòa ‘Rasoa’s book’). Preceding adjectives merge genitive: Hènoko ìzy ‘He is listened to by me.’ with the genitive marker: ny bòky vaovàon′ny Passive forms of prefixed-active verbs have the suffix mpampiànatra ‘the teacher′s new book’. -ina or -ana: Tehènina ny làkana ‘the boat is shoved Malagasy sentences are mainly predicate-initial: [with a pole]’; Totòfana tàny ny làvaka ‘the hole is filled Vaovào ny bòky ‘The book is new’; Àhy ny bòky up with earth’. Pronominal actors are in the genitive: vaovào ‘The new book is mine’; Mpampiànatra àho Soràtano ny taratàsy ‘The letter is written by me.’Active ‘I am a teacher’; Tèlo ihàny ny bòky ‘There are only derivations of root-passive verbs have suffixed passives three books’; and Mitèny àho ‘I speak.’ Transitive (tsìnjo ‘be perceived/ noticed’ → mitsìnjo ‘observe/ sentences are typically Verb–Object–Subject (VOS): watch’ → tsinjòvina ‘be observed/ watched’). Mivìdy ny bòky àho ‘I buy the book’ and Manòratra For verbs with two nonactor arguments, a taratàsy i Rìta ‘Rita writes a letter.’ Indirect objects ‘stationary’ (medium, goal) and a ‘mobile’ one precede the subject: Mivìdy bòky ho an′ny zàza i Rìta (instrument, theme), the former can be topicalized ‘Rita buys a book for the child.’ either by the above passives, or with the so-called Subject fronting with emphasizing is achieved with circumstantial passive, formed by circumfixation of the particle no (Ny mpampiànatra no nivìdy ny bòky aN-...-ana or i-...-ana (mandràkotra ‘cover’ → ‘The teacher [is the one who] bought the book’) that andrakòfana ‘be covered [with]’; vìdy ‘buy’ → emphasizes any fronted term: Omàly no novidìna ny ividìana ‘have [something] bought [for one]’). The bòky ‘The book was bought [precisely] yesterday.’ ‘mobile’ argument is then topicalized by a passive with Subject fronting with the particle dìa emphasizes the prefix a- (mandròso ‘serve [food] [to someone]’ → verb: Ny mpampiànatra dìa nivìdy ny bòky ‘The rosòana ‘be served [guest]’/aròso ‘is served [food]’). teacher bought [didn′t steal] the book.’ It also The above passives all partake in the (m)/n(o)/h(o) emphasizes a sentence-final subject: Hivìdy ny bòky tense alternance and form imperatives with -o (or -y dìa ny mpampiànatra ‘The teacher [noone other] will when the base has -o-): henòy ‘listen!’; andràsana ‘be buy the book.’ waited for’ → andràso ‘wait!’; rosòy ‘serve [to The relative marker is izày: ny bòky izày novidìn′i someone]!’/arosòy ‘serve [a dish]!’. Rìta ‘the book that was bought by Rita.’ It is often Finally, there is the resultative passive with voa- optional: Hanòratra taratàsy ny mpampiànatra [izày] (mandàza ‘say, declare’ –– voalàza ‘be/get said’; mifìdy nivìdy ny bòky ‘The teacher who bought the book will ‘choose’ –– voafìdy ‘get chosen’) and the incidental write a letter.’ resultative with tafa- (mivèrina ‘return’ –– tafavèrina The above article still leaves out many features, ‘[unexpectedly] be/get back’). Both have a future (with particularly numerous irregularities and exceptions, but ho-) but no past tense, and have no imperative. also various intricate regularities, that make Malagasy Nominalization of verbs is productive. Agent nouns an extremely interesting object of linguistic study. derive from active verbs by inserting -p- after the prefixal m- (mandèha ‘go’ → mpandèha ‘passenger, References voyager’; mitèny ‘speak’ → mpitèny ‘speaker’). Nouns Adelaar, Alexander K. 1995. Asian roots of the Malagasy: a denoting some aspect of that which is expressed by the linguistic perspective. Bijdragen tot de Taal-, Land- en verb can be formed by replacing prefixal m- by f- Volkenkunde 151. 325–56. (fandèha ‘way of walking’; fitèny ‘manner of speaking’). Andriamanantsilavo, Seth, and William Ratrema. 1981. Ny fi- Abstract nouns are formed by additional suffixation of - tsipi-pitenenantsika (The grammar of our language), Vol. I. Antananarivo: Librairie mixte. (an)ana (fambolèna ‘cultivation’; fidìrana ‘admission’; Dahl, Otto C. 1991. Migration from Kalimantan to Madagascar. fandehànana ‘departure’, fitenènana ‘conversation’). Oslo: Norwegian University Press. Mahdi, Waruno. 1988. Morphophonologische Besonderheiten und historische Phonologie des Malagasy (Morphophono- Syntax logical particularities and historical phonology of Malagasy). Berlin and Hamburg: Reimer Verlag. Typically, an attribute follows the modified Paul, Ileana (ed.) 1998. The Structure of Malagasy, vol. II, Los element––ny tràno lehibè ‘the big house’––while a Angeles: UCLA Department of Linguistics. predicate precedes the subject––Lehibè ny tràno ‘The Pearson, Matthew, and Ileana Paul (eds.) 1998. The structure of house is big’––also demonstrating that existential Malagasy, Vol. I. Los Angeles: UCLA Department of clauses are formed without a ‘to be’ equivalent. Linguistics. Rajaonarimanana, Narivelo. 1994. Grammaire moderne de la In possessive atribution with nominals (pronouns see langue malgache. Paris: L’Asiathèque. above), the genitive marker n′ is preposed to the definite Randriamasimanana, Charles. 1999. Clausal architecture and article (n′ny bòky ‘of the book’, n′i Rìta ‘Rita′s’). movement verbs in Malagasy. Selected papers from the Eighth 643 MALAGASY International Conference on Austronesian Linguistics, ed. by creation of a national language). Zeitschrift für Phonetik, E. Zeitoun and P.J.-K. Li. Taipei: Academia Sinica. Sprachwissenschaft und Kommunikations for schung 37. Rasoloson, Janie, and Carl Rubino. 2004. Malagasy. The 618–29. Austronesian languages of Asia and Madagascar, ed. by Simon, Pierre. 1988. Ny fiteny fahizany, reconstitution et A.K. Adelaar and N. Himmelmann. London and New York: périodisation du malgache ancien jusqu′au XIVe siècle. Routledge. Paris: Institut des Langues et Civilisations Orientales. Schmidt, Bernd. 1984. Die Malgassisierung und das ‘malagasy Vérin, Pierre, Conrad P. Kottak, and Peter Gorlin. 1969. The iombonana’ –– Bilanz der Verwendung und Schaffung einer glottochronology of Malagasy speech communities. Nationalsprache (Malagasification and the ‘communalization Oceanic Linguitics 8. 26–83. of Malagasy’ campaign –– net results of implementation and WARUNO MAHDI Malay-Indonesian and Malayic languages Malayic is a subgroup of the Western Malayo- classification. It is now clear that Malayic is more Polynesian branch of the Austronesian language closely related to the Chamic languages, spoken in family. Malayic languages are spoken throughout the Cambodia and Vietnam, than to any of these Malay-Indonesian archipelago, from Sumatra in the languages. One factor hindering classification is that west to New Guinea in the east, and also on mainland many languages have borrowed heavily from Malay- Southeast Asia, mostly in the Malay Peninsula and in Indonesian, which has served as a regional lingua parts of Thailand. franca for many centuries. This may make such Malay-Indonesian is a member of the Malayic languages appear to be much more closely related to subgroup. The indigenous name for the language is Malayic than they actually are. Similarly, no clear Bahasa Melayu (literally, ‘the Malay language’), but criteria have been established for distinguishing some varieties used in Indonesia are also called between Malayic languages and dialects of Malay- Bahasa Indonesia (‘the Indonesian language’, see Indonesian. Scholars in the field have therefore below). Similar forms of Standard Malay-Indonesian preferred using the neutral term ‘isolect’ to refer to any serve as the national languages of Indonesia, Malayic speech form which has a name of its own and Malaysia, Brunei, and Singapore; the latter three are is regarded by speakers as distinct from other varieties. particularly similar to each other. A large number of There is great variation among Malayic isolects, Malay dialects and Malay-based pidgins and creoles many of which are not mutually intelligible. They fall are also spoken throughout the region. into several categories. Some, like Riau Malay (spoken The number of speakers of Malay-Indonesian is in the Riau archipelago in Indonesia) or Kedah Malay estimated at over 250 million, making it by far the (spoken in the Malaysian state of Kedah), are thought to most widely spoken language in Southeast Asia, and a be direct descendants of Proto-Malayic, a hypothetical major world language. Most speakers do not use language reconstructed on the basis of modern isolects. Malay-Indonesian as a native home language; the Other isolects, however, have had a more complex number of native speakers is difficult to estimate, but history, and owe their emergence to language contact is probably over 50 million. This figure is rapidly and language shift. For example, Betawi, the language of increasing, as more and more people in Indonesia, the indigenous ethnic group of Jakarta, is based on Malaysia, and Brunei shift from their ancestral home Malay, but has incorporated lexical and grammatical languages to Malay-Indonesian. elements from Balinese, Javanese, Sudanese, and Chinese, languages spoken by the ancestors of today’s speakers. Some isolects have developed from pidginized Classification forms of Malay, collectively known as Bazaar Malay, While there is wide agreement about the existence of which originally served only for interethnic commu- the Malayic subgroup and the Malay-Indonesian nication, not as a first language. Baba Malay, spoken by language, linguists have not been able to agree on their acculturated Chinese communities in Malacca, Penang, classification. Malayic used to be classified together and Singapore, is thought to have developed from with Javanese, Sundanese, Madurese, Achenese, and Bazaar Malay, which gradually became the speakers’ Lampung as in a putative ‘Malayo-Javanic’ branch, first language. Most isolects spoken in eastern Indonesia but strong doubts have been cast on the validity of this are also believed to have developed from forms of 644  MALAY-INDONESIAN AND MALAYIC LANGUAGES Bazaar Malay. This complex situation has also con- sometimes uncoordinated) effort of native and Western tributed to the difficulty in classifying Malayic isolects. scholars. The great Malay scholar Raja Ali Haji (c. 1809–1870) composed a grammar and a dictionary of Malay. Later, the Dutch scholar C.A. van Ophuysen Standard Malay-Indonesian (1854–1917) standardized the language for use at History schools throughout the Dutch Indies. In 1928, a congress The cradle of Malay civilization and of the Malay- of nationalist students declared the Malay language, Indonesian language was in south-central Sumatra. under the name Bahasa Indonesia, as the national Many scholars believe, however, that its hypothetical language of the Indonesian nation. During the Japanese ancestor Proto-Malayic was spoken in western Borneo. occupation (1942–1945), the modernization of Malay- The original place name Malayu (=Malay) has been Indonesian received an impetus, as it was widely used in identified with the former Malay kingdom of Jambi in the administrative and educational systems and in the central Sumatra. The Chinese monk I Ching, who mass media. Indonesia declared its independence in visited the area in the seventh century AD, reported on 1945, whereby Malay-Indonesian (under the name a place called ‘Moloyu’; later Javanese inscriptions and Bahasa Indonesia) became its national language. When manuscripts also refer to the area of Jambi as ‘Malayu’. Malaysia, Singapore, and Brunei followed suit, Malay- The Sejarah Melayu (‘Malay Annals’), the canonical Indonesian (under its traditional name Bahasa Melayu) work of classical Malay literature, traces Malay origins became their national language as well. In 1972, the to Palembang, a city south of Jambi, which historians spelling of Malay-Indonesian was reformed and and archeologists identify with the center of the ancient harmonized, and a joint council (known by its acronym maritime empire of Sri Vijaya. MABBIM) has been coordinating language planning The earliest direct evidence of Malay comes from a activities in these countries ever since. handful of seventh-century AD inscriptions, found in southern Sumatra and on the nearby island of Bangka, Typical Features and associated with Sri Vijaya. While not all scholars The sound system The sound system of Standard agree that the language of these inscriptions is the direct Malay-Indonesian is relatively simple. There are 18 ancestor of modern standard Malay-Indonesian, it is primary consonants (given in their conventional ortho- commonly referred to as Old Malay. The inscriptions are graphy): b, d, j, g, p, t, k, c, m, n, ny, ng, l, r, h, s, w, y. written in a formal language that borrows heavily from Sometimes the glottal stop is also analyzed as a Sanskrit; there is no direct evidence of the language phoneme. (In this orthography, c and j represent palatal ordinary people spoke in their everyday lives. Old Malay stops, while the digraphs ny and ng represent a single inscriptions dating from the ninth century were also consonant each, the palatal nasal and velar nasal, respec- found in areas where Malay was not indigenous, like tively). The vowels are a, e, i, o, u, plus a mid-central Java and the Philippines, showing the early spread of vowel (confusingly written e, like the mid-front vowel). Malay as a literary language throughout the region. There are also a few loan phonemes, like f and z, which The use of Old Malay as a literary language in Java first entered the language via loanwords. In Standard and in the Philippines did not survive long. However, in Malay-Indonesian, there is no distinctive word accent. Sumatra it has continued uninterrupted. Even among peoples who speak rather different languages, like Word formation Beyond borrowing and coining new Rejang and Lampung, Malay (written in an Indian- roots, Malay-Indonesian words are formed by three derived script called rencong) continued to serve as the principal processes: affixation, reduplication, and com- major literary language. The oldest extant Malay pounding. Affixes include prefixes (makan ‘eat’ vs. di- manuscript is a recently rediscovered fourteenth- makan ‘eaten’), suffixes (kerja ‘work’ vs. kerja-kan ‘do century work originating from southern Sumatra. Some [something]’), and circumfixes, which attach themselves letters and longer works from the sixteenth century are simultaneously to the beginning and end of roots (baik preserved in collections in the West, and from the ‘good’, ke-baik-an ‘goodness’). Redupli-cation fills seventeeth century onward, Malay manuscripts become many functions. For example, anak means ‘child’, while numerous. The contents of these works are varied, and anak-anak means ‘a group of children’; jalan means range from legends, chronicles, and religious treatises, ‘walk’ or ‘road’, jalan-jalan is ‘go for a walk, go out’. to legal documents and letters. The language of these Compounding combines two existing roots into one manuscripts, while showing some variation across time, word; for example, matahari ‘sun’ is made up of the space, and style, is nevertheless remarkably uniform, words mata ‘eye’ and hari ‘day’, and literally means ‘the and has been termed Classical Malay. eye of the day’. In addition to these three processes, Modern Standard Malay-Indonesian came into clipped words (made up of parts of existing words) and existence in the nineteenth century, as a joint (although acronyms are very common, especially in Indonesia. 645 MALAY-INDONESIAN AND MALAYIC LANGUAGES Word classes Most scholars agree that Malay- During the first half of the second millennium AD, Indonesian has both function words and content words, the Malays, who were originally animists and then and that content words include at least two categories: Hindus and Buddhists, gradually converted to Islam. nouns and verbs. Some analyses also posit adjectives, With the new religion, they also acquired the Arabic adverbs, and other classes. Part of the discrepancy stems alphabet, which formed the basis for the Malay from different semantic, morphological, and syntactic alphabet known as Jawi. The earliest example of Jawi criteria used by different scholars to classify words. writing is a fourteenth-century inscription from Another factor has been Eurocentricity: scholars Terengganu in the Malay peninsula, and the earliest (especially in the past) tried to fit Malay-Indonesian into Jawi manuscripts are two letters written by the sultan categories created by European scholars for European of Ternate to the king of Portugal in 1520 and 1521. languages, which were often incompatible with the Since the seventeenth century, Romanized alphabets actual categories of Malay-Indonesian. In fact, it is have gradually replaced Jawi, but Jawi is still used possible that Malay-Indonesian roots are precategorial, sometimes, especially in Malaysia and Brunei. with the category or word class determined by affixes or by the place the word fills in the sentence. Other Malayic isolects Word order The basic word order in Malay- The number of distinct Malayic isolects probably runs Indonesian is, like English, Subject–Verb–Object: Tuti into the hundreds. A few of the major ones are listed makan pisang (lit. Tuti eat banana) ‘Tuti is eating below. bananas.’ However, according to the circumstances Iban: The Iban people of northern Borneo were and the particular emphasis desired by the speaker, formerly known as Sea Dayaks. They speak a number these constituents can be ordered in any way. Also, of closely related isolects, mostly in the Malaysian state unlike in English, the subject and the object can be left of Sarawak, but also in the neighboring state of Sabah unsaid, if they are already understood (or deemed so and across the border in Kalimantan (Indonesian by the speaker). Focusing on the object (as in English Borneo). The total number of speakers is estimated at passive sentences) and not on the subject is often the upward of 500,000. Iban is more divergent from Malay- normal way of expressing things in Malay-Indonesian. Indonesian than other major Malayic isolects. For this Thus, the usual way to say that ‘someone ate the reason, it is often classified as a separate language. bananas’ would be pisang itu dimakan (orang), which Minangkabau: This important Malayic isolect is used literally means something like ‘the bananas were eaten by about 7 million people. Native to western Sumatra, it (by people)’. Modifiers normally follow the head, as in is now spoken via immigration outside its historical rumah besar (lit. house big) ‘a big house’ and rumah homeland, principally in the Indonesian province of itu (lit. house that) ‘that house’. Possession can be Riau and in the Malaysian state of Negri Sembilan. indicated by a simple juxtaposition of nouns: rumah Often classified as a separate language, Minangkabau Tuti (lit. house Tuti) ‘Tuti’s house’, rumah saya (lit. has a long literary tradition, and has influenced standard house I) ‘my house’. Indonesian through the numerous authors of An important difference between Malay- Minangkabau origin who played an important role in the Indonesian and Indo-European languages like English creation of modern Indonesian literature. is that words are not inflected, and many categories Banjarese: Banjarese probably came into existence expressed by inflection in Indo-European languages as a lingua franca used among the Dayaks and Malays are simply left unspecified in Malay-Indonesian. Thus, of southern Borneo and their Javanese overlords. It is rumah can mean either ‘house’ or ‘houses’, and makan now spoken as a first language by several million can mean ‘eat’, ‘eats’, ‘am eating’, ‘have eaten’, ‘ate’, people in the city of Banjarmasin and in other etc., depending on the context. locations throughout the southern parts of Borneo. Riau-Johor Malay: This dialect group is spoken in Writing Systems the southern part of the Malay Peninsula and in the Riau All early scripts that were used for writing Malay were archipelago in Indonesia. Of all modern vernaculars, it based on Indian writing systems. The oldest Malay bears the closest resemblance to the national standard inscriptions were written in a slightly modified version language of Malaysia, Singapore, and Brunei. of the Pallava script of southern India. The earliest Kelantan-Patani Malay: This group of closely Malay manuscript was written in the Kawi script, used related dialects is spoken by several million speakers for writing Old Javanese, while later Malay on both sides of the Thai–Malaysian border. Kelantan- manuscripts from Sumatra were written in the local Patani Malay is characterized by radical phonological rencong script, an angular alphabet also ultimately changes, which make it difficult or impossible to derived from Indian writing systems. understand by speakers of other Malay dialects. 646  MALKIEL, YAKOV Ambonese Malay: This creole is spoken natively on indigenous inhabitants of Jakarta, but has been the island of Ambon in Maluku, where it is in the influenced by Standard Indonesian. With at least 5 process of replacing the indigenous languages. It is million native speakers and many more second- also widely used as a lingua franca (and increasingly as language speakers, it is one of the largest Malayic a first language) in other parts of Maluku. Ambonese isolects. Jakarta Indonesian is becoming the language Malay probably developed from a form of Bazaar of choice for educated urban speakers throughout Malay, which eventually became the first language of Indonesia. some speech communities. Closely related isolects are spoken in and around Kupang (on the island of Timor) References and Larantuka (on the island of Flores). Adelaar, K. Alexander. 1992. Proto-Malayic: the reconstruction Manado Malay: This creole originated on the island of its phonology and parts of its lexicon and morphology. Ternate, as evidenced by its Ternate lexical Pacific Linguistics Series C-119. Canberra: The Australian component. It is spoken natively in the city of Manado National University. in North Sulawesi, and is gradually replacing the local Asmah Haji Omar. 1983. The Malay peoples of Malaysia and Minahasa languages throughout this province. Closely their languages. Kuala Lumpur: Dewan Bahasa dan Pustaka. Collins, James T. 1998. Malay, world language: a short history. related varieties are spoken in northern Maluku and in Kuala Lumpur: Dewan Bahasa dan Pustaka. many parts of northern and central Sulawesi. Sneddon, James Neil. 1996. Indonesian reference grammar. St. Jakarta Indonesian: The urban koine of Jakarta is Leonards, NSW: Allen & Unwin. based on Betawi, the Malayic isolect used by the URI TADMOR Malkiel, Yakov Born in Ukraine, Yakov Malkiel was formed on the appropriate methodology of this discipline, on intellectually in Germany, where he received his some of the schools and tendencies of modern doctorate in Romance Linguistics. With the beginning Linguistics, and on some of their main figures of World War II, he moved to the United States, the (Leonard Bloomfield, Leo Spitzer, etc.). country in which he combined his work as a Professor It is also related to the general synchronic Linguistics, of Romance Philology in the University of Berkeley the study of (a) language(s) at a given point in time. In (California) with the development of most of his this sense, it is necessary to highlight his interesting scientific career. studies on what he denominated irreversible binomials, His wide cultural and linguistic knowledge, his that is, those idiomatic expressions formed by two enormous capacity of work, and his astonishing elements whose order cannot be permuted (odds and domain of several languages allowed him to produce an ends, but *ends and odds); on the special uses that the enormous bibliography, written in German, Italian, languages make of the letters (in the construction of French, Portuguese, Spanish, and English. Although acronyms, in the construction of fixed expressions, etc.); this bibliography only includes a dozen books, it or on the relationship that exists between grammatical contains several hundred articles and essays. In fact, gender on the one hand and sex and size on the other. Malkiel himself published his own Autobibliography in But the scientific effort of Yakov Malkiel was 1988 (Yakov Malkiel: a tentative autobibliography), in primarily centered on diachronic (historical) linguistics, which more than 800 titles are registered. Such a vast the field in which, in turn, we may distinguish two areas: scientific production comprises, as expected, very In the first place are his works on the methodology diverse topics, which makes it difficult to find a and epistemology of diachronic Linguistics and its superficial description for this linguist’s work lines and different fields of study (etymology, forms and causes contributions. Nevertheless, as a guideline, it can be of the linguistic change, studies on phonology and affirmed that the scientific work of Yakov Malkiel is historical morphology, etc.). In this area, Malkiel tried related to three large thematic fields that, presented to give a new impulse to a type of linguistic inves- from less to more importance, are: tigation that was in clear decadence after the almost In the first place, it is related to the historiography absolute triumph, at the beginning of the twentieth of Linguistics, the field in which he published essays century, of the synchronic perspective defended by 647 MALKIEL, YAKOV Ferdinand de Saussure and his followers (the so-called the Romance languages are of special importance, like structuralists) and, later, by the Generativism spawned the case of those that he denominated hispanic interfixes: by Noam Chomsky. morphemes without meaning that would appear between To get that revitalization of the historical studies, he the root and the suffix of certain Spanish words (like established new conceptual distinctions, separating the pan-ad-ero ‘baker’, where pan is the root that means true diachronic Linguistic, that according to him should ‘bread’, -ero is a suffix that expresses ‘agent’, and -ad- focus on the study of specific linguistic processes, from is an intermediate element lacking any meaning). the Glottodynamics, that should study the universal In definitive, Yakov Malkiel can be characterized as constants that rule linguistic change and evolution. one of the main figures of contemporary diachronic Moreover, he substituted the proper term “diachronic Linguistics, to which he dedicated most of his efforts Linguistics” or “historical” for that of genetic and his vast knowledge, something noticeable in his Linguistics, convinced that evolutionary Linguistics numerous publications which are so filled with should not be limited to describe mere historical knowledge and data that they are quite inaccessible to developments, but rather that it has to be able to discover the uninitiated in the particular subject matter. This the decisive causes of these developments. Along those dedication to a linguistic area that has been neglected lines, his theory on the multiple causation of linguistic over the last decades may explain that most of the change is especially remarkable, according to which main studies in the development of the modern most of these changes have their origin in several Linguistics omit in their pages the reference to this interrelated factors, factors that can depend on different figure, of unquestionable scientific merits and of linguistic planes (phonology, morphology, grammar, or obligated reference when someone plans to work in semantics) or, even, in the extralinguistic reality. the field of diachronic Linguistics. At the same time, as he himself affirms in the foreword of one of his most important works (From Biography particular to general Linguistics), he tried to harmonize genetic Linguistics with the new theoretical perspec- Yakov Malkiel was born in Kiev (Ukraine) in 1914, to a tives coming from structuralism, thus changing the Jewish family. Because of the civil war, he emigrated to atomistic type of study developed by traditional Germany, where he studied at the University Friedrich- historical Linguistics for another that is more in Wilhems in Berlin. In 1938, he obtained his Doctorate in accordance with the general lines of contemporary Linguistic Romance degree with suma cum laude. linguistic thought. In fact, it could be affirmed that Because he was jewish, he was forced to emigrate again Malkiel was, because of his way of analyzing the data in 1940, this time to the United States. Here, he worked and of understanding the reality of language, among the for a short while at the University of Wyoming, and later structuralist followers of Bloomfield, although he never he moved to the University of Berkeley (California), lost his scientific individuality and maintained, with where he successively held the positions of Lecturer, respect to the latter, important theoretical differences. Assistant Professor of Spanish and Portuguese, Professor In the second place, the dedication of Malkiel to the of Romance Philology and Professor Emeritus of field of historical Linguistics is reflected in the Linguistics and Romance Philology. In 1946, he founded publication of countless specific studies on diachronic the Romance Philology magazine, of which he was phonology and morphology. In those works, almost Editor in Chief for many years. In 1952, he participated always centered on the Roman languages––and, among in the creation of the Department of Linguistics of the those, especially Spanish, Portuguese, and Italian––he University of Berkeley, where he carried out his tried to combine the explanation of particular facts with educational activities until his death on April 24, 1998. the formulation of general theories, something that The merits of his career were recognized with connects these works with those described previously numerous distinctions, such as three Guggenheim and that also allows his results to be extrapolated to the Awards and an honoris causa doctorate from the analysis of any other language or group of languages. universities of Chicago (1966), Illinois (1969), Paris From among those essays, one can cite the ones (1983), Berlin (1983), Georgetown (1987), Oxford dedicated to the explanation of the evolution of certain (1989), and Salamanca (1994). sounds or groups of sounds in the Romance languages: those related with symbolic values of the sounds References (phonosymbolism) and, above all, those focused on the study of diverse elements derivative from the Romance Malkiel, Yakov. 1958. Los interfijos hispánicos: Problema de lingüística histórica y estructural. Miscelánea homenaje a languages. Among the latter, those in which he argues André Martinet. La Laguna: Universidad de La Laguna. about the possible existence of diverse types of empty ––––––. 1966. Genetic analysis of word formation. Current morphs (morphologic elements lacking of meaning) in trends in linguistics. Vol. III. La Haya: Mouton. 648  MANNER OF ARTICULATION ––––––. 1968. Essays on linguistic themes. Berkeley: ––––––. 1993. Etymology. Cambridge: Cambride University University of California Press. Press. ––––––. 1982. Perspectives on historical Linguistics. ––––––. 1983. Introduction. In Malkiel. From particular to Amsterdam/Philadelphia: John Benjamins. general Linguistics. ––––––. 1983. From particular to general Linguistics. Essays Rico, Francisco. 1969. Yakov Malkiel. Anuario de Estudios 1965–1978. Amsterdam/Philadelphia: John Benjamins. Medievales 6. ––––––. 1988. Yakov Malkiel: a tentative autobibliography. Hall, Robert A. 1987. Linguistics and pseudo-linguistics. Berkeley: University of California Press. Amsterdam/Philadelphia: John Benjamins. ––––––. 1989. Theory and practice of romance etymology: Hymes, Dell, and John Fought. 1981. American structuralism. studies in language, culture and history. Variorum: Longdon. The Hague: Mouton. ––––––. 1990. Diachronic problems in phonosymbolism: edita JOSÉ CARLOS MARTÍN CAMACHO and inedita, 1979–1988. Amsterdam/Philadelphia: John Benjamins. ––––––. 1992. Diachronic studies in lexicology, affixation, See also Etymology; Historical Linguistics; Indo- phonology: edita and inedita, 1979–1988. Amsterdam/ European 4: Romance; Language Change; Struc- Philadelphia: John Benjamins. turalism Manner of Articulation Consonants are formed by creating a constriction in through the mouth, but with the constriction close the vocal tract; the manner of articulation describes enough to cause turbulence in the airstream, the type of constriction, whereas the place of producing a hissing type of noise called frication. The articulation indicates where the constriction takes English fricatives are [f v], [θ] as in thick, [
] as in place. Sounds can also be made with different the, [s z], [ʃ] as in shoe, and [] as in garage. The airstream mechanisms. fricatives [s z ʃ ] are called sibilants. Fricatives are A primary aspect of the manner of articulation is also known as spirants. the degree of the constriction, that is, whether the air Affricates consist of a stop immediately followed has a free passage through the mouth or whether there by a fricative made at the same place of articulation. is an obstruction. By varying the degree of obstruction, English has the affricates [tʃ] as in itch and [
] as in different types of sounds can be produced. Different badge. Approximants have an articulatory constriction sounds can also be made by allowing air to pass that is narrower than the vowel [i], but not close through the nasal passage; such sounds are known as enough to cause frication. Any approximant can be nasal. Also, some sounds are lateral, with the air turned into a fricative by narrowing the constriction passing out through the sides of the vocal tract, but not until frication begins. English has the approximants through the center. [l], [] as in Mary, [w] as in wine, and [j] as in yes. Vowels are produced with a very open vocal tract. Some dialects of English pronounce, e.g. which with a For consonants, we distinguish three degrees of voiceless approximant as well, symbolized as []. stricture: stops, fricatives, and approximants. Fricatives and approximants differ only in that Obstruents include the oral stops, fricatives, and approximants are slightly more open than fricatives. affricates. Nasals and approximants are nonobstruent For this reason, where no approximant symbol is given sounds, called sonorants. on the International Phonetic Alphabet chart, an Stops have a complete closure in the oral cavity. approximant can be shown by using the “lowered” Oral stops involve closure of both the oral and nasal diacritic [ ] with a fricative symbol; for example, a cavities, with no air passing through either one. voiced bilabial fricative would be symbolized as [], Nasal stops have air passing out through the nasal indicating that it is like a fricative, but with a more cavity, but not through the oral cavity. In English, [p, open vocal tract. t, k, b, d, and g] are oral stops, and [m, n, and ŋ (as Two types of trill are commonly found. In one, in long)] are nasal stops. Nasal sounds are discussed the tip of the tongue is near the front teeth; in these, further below. the tongue tip vibrates very quickly against the roof Fricatives are made with an incomplete con- of the mouth several times. The other trill is uvular striction in the oral cavity, allowing air to pass [R]; with it, the uvula strikes against the dorsum of 649 MANNER OF ARTICULATION the tongue. Snoring often involves an uvular trill the air pressure to rise in the pharynx and oral cavity; made while breathing in. on release of the oral closure, air flows out with a Trills are not made by consciously controlling the rather hollow, popping sound. This airstream motion of the tongue; rather, the tongue is placed in mechanism is called glottalic; glottalic egressive stops the appropriate position and tension, air is blown are called ejectives [p’ t’ k’]. Many native North through the gap, and aerodynamic forces cause the American languages have ejectives. tongue to vibrate rapidly against the roof of the Although it is possible to make glottalic ingressive mouth. stops, they are not at all common in languages. More Taps are often described as a trill of one vibration. often, we find implosives [  ]. These are made by The tongue is flicked quickly against the roof of the raising the larynx, simultaneously making a stop, and mouth. A stop, by contrast, is a more deliberate action. then, as the oral stop is released, lowering the larynx In North American dialects of English, the tap [] is a so that the vocal folds vibrate as they move typical variant of /t/, occurring between vowels, as in downward. city. Clicks are made by making a velar closure as in [k] Laterals are made with a central constriction, but and simultaneously making an oral closure further with the sides of the constriction open, allowing air to forward (e.g. [p] or [t]). Then, the tongue body is escape. English has only the one lateral [l]. pulled back, lowering the air pressure in the oral Nasals are made with a velic opening allowing air cavity. The release of the oral closure is a click; the to pass out through the nasal passage. If the velum is airstream is velaric ingressive. The click symbols are raised against the pharyngeal wall (velic closure), the bilabial [ ], dental [|], alveolar [!], palatoalveolar [
], nasal cavity is shut off from the rest of the vocal tract, and alveolar lateral []. As ordinary consonants, clicks and no air can go through the nasal cavity. If the velum are found only in languages of southern Africa; is lowered (velic opening), air can pass through the however, English speakers often use the dental click [|] nasal cavity. to express mild displeasure (often written tsk! tsk!), The term nasal, used alone, means nasal stop. and the lateral click [] is used to urge on horses. Otherwise, the type of sound must be specified, as in Glottalic egressive sounds are not ordinarily used in nasal fricative, nasal lateral, etc. Nasal fricatives and language. approximants occur, but they are usually the result of being next to a nasal stop or vowel. Nasal stops are ordinarily voiced. Further Reading Air flowing out of the body is called egressive. Air Catford, J.C. 1977. Fundamental problems in phonetics. entering the body is called ingressive. Most speech is Bloomington, IN: Indiana University Press. egressive, but some sounds in some languages are Catford, J.C. 1988. A practical introduction to phonetics. made with an ingressive airstream. Oxford: Oxford University Press. In addition to the direction of airflow, there are Gimson, A.C. 1980. An introduction to the pronunciation of English. London: Edward Arnold. three different airstream mechanisms that are used to Hardcastle, William J., and John Laver, (eds.) 1997. The set the air into motion (Chart 1). Most sounds are handbook of phonetic sciences. Oxford: Blackwell. made by pushing air out of the body by contracting the International Phonetic Association. 1999. Handbook of the lungs. These sounds are called pulmonic. Pulmonic International Phonetic Association. Cambridge: Cambridge egressive stops are called plosives. Pulmonic University Press. Ladefoged, Peter. 1993. A course in phonetics. Ft. Worth, TX: ingressive sounds are not normally used in language Harcourt Brace Jovanovich, (see below): Ladefoged, Peter. 2000.Vowels and consonants: an introduction Sounds can be made by making an oral closure to the sounds of languages. Oxford: Blackwell. (such as a [p]), lowering the larynx, forming a glottal Ladefoged, Peter, and Ian Maddieson.1996. Sounds of the stop, and then raising the larynx. These actions cause World’s Languages. Oxford: Blackwell, Laver, John. 1994. Principles of phonetics. Cambridge: Cambridge University Press. Minifie, Fred, Thomas J. Hixon, and Frederick Hixon. 1973. CHART 1 Normal aspects of speech, hearing, and language. Englewood Cliffs, NJ: Prentice-Hall. Stops Pulmonic Glottalic Velaric Pullum, G.K., and W.A. Ladusaw. 1996. Phonetic symbol Egressive Plosives Ejectives –––– guide. Chicago: University of Chicago Press, ptkbdγ p’ t’ k’ Rogers, Henry. 2000. The sounds of language: an introduction Ingressive –––– Implosives Clicks to phonetics. Harlow, England: Pearson.  |!
 HENRY ROGERS 650  ¯ ORI AND THE POLYNESIAN LANGUAGES MA Ma¯ori and the Polynesian Languages Maori is the indigenous language of New Zealand, and languages of this family. The languages named are since 1987, it has been an official language of the related to one other as shown in the family tree, which country. It was brought to New Zealand 800 to 1,000 shows the subgroups (subfamilies) within the family. years ago by voyagers from Central Polynesia, Through the comparative method, a great deal is specifically from the Southern Cook Islands, and the known about the structure of PPN. The following table area of modern French Polynesia. It is thus very sets out examples of the data which allow the closely related to the languages of these areas and less reconstruction of the sound system and vocabulary of closely related to the other languages of the whole PPN. Polynesian family of languages. This language family, which forms the most easterly group within the Tonga Samoan Hawaiian Ma¯ori PPN English Austronesian language family, contains some 30 po¯ po¯ po¯ po¯ *po¯ ‘night’ languages, which are spoken by approximately valu valu walu waru *valu ‘eight’ 750,000 people in communities varying in size from kai ‘ai ‘ai kai *kai ‘eat’ more than 300,000 (Samoan) to just 100 or 200. The tolu tolu kolu toru *tolu ‘three’ fale fale hale whare *fale ‘house’ family encompasses all the indigenous languages fa¯ fala hala whara *fara ‘pandanus’ spoken within the Polynesian Triangle, as well as ‘a¯ ala ala ara *ʔara ‘awaken’ many spoken within Micronesia and Melanesia. The hala sala hala hara *sala ‘wrong’ Polynesian Triangle is a vast geographic area formed hala ala ala ara *hala ‘way’ by lines joining New Zealand, Rapanui (Easter la‘a¯ la¯ la¯ ra¯ *laʔa¯ ‘sun’ ua lua lua rua *rua ‘two’ Island), and Hawaii, and the languages to be found fafo fafo waho waho *fafo ‘outside’ here include Maori, Rarotongan, Tahitian, Rapanui, fefine fafine wahine wahine *fafine ‘woman’ Hawaiian, Tongan, and Samoan. All the languages of mo‘unga mauga mauna maunga *maʔuŋa ‘mountain’ this family are descended from Proto-Polynesian (PPN), a language spoken in the region of Tonga and The asterisk indicates that the forms so marked are Samoa some 2,500 years ago. As the Polynesian reconstructed by the methods of historical linguistics peoples dispersed and settled in the other regions of and not actually attested. The ‘g’ of Samoan and the the Pacific, the language they carried with them ‘ŋ’ in PPN both designate the ‘ng’ sound of English diversified and, over time, became the modern ‘sing’. The ‘ symbol in Tongan, Samoan, and PPN Tongan Nuclear Polynesian Samoan Eastern Polynesian Central Eastern Polynesian Rapanui Mäori, Rarotongan, Tahitian, Hawaiian 651  ¯ ORI AND THE POLYNESIAN LANGUAGES MA Hawaiian, as well as the ʔ in PPN, designate the glottal categories such as tense and the role that words play stop, often called a ‘catch’ or ‘break,’ articulated in the within sentences are indicated by means of separate larynx. The macron over some vowels (a¯) indicates particles: that these are long. I haere ‘went’, kua haere ‘has gone’, ka haere ‘will An area of the grammar of PPN and of subsequent go, goes’, me haere ‘must go’. developments in its descendants that is of particular Similarly, whether nouns are subjects, objects, interest to scholars studying language typology is the possessors, and so on is marked by particles: in the matter of ergativity. Ergative languages, such as Maori example above, ‘The man ate the fish’, the fact Tongan, mark the subjects of clauses differently that ‘the fish’ is the object of ‘eat’ is shown by the depending on whether the verb is transitive or not, particle i. whereas nonergative languages, such as Maori, treat One area of Maori grammar in which a suffix plays subjects in the same way, irrespective of the nature of a role is the formation of passive verbs: the verb. horoi ‘wash’, horoia ‘be washed’; inu ‘drink’, inumia ‘be drunk (of liquids)’; kai ‘eat’, kainga ‘be eaten’. Tongan: There are 17 such suffixes, although many of these Na‘e ‘alu e tangata are very rare. However, the question of which suffix Past go the man goes with which verb and how this phenomenon is ‘The man went (intransitive verb ‘go’)’ best described has been a subject of some discussion within theoretical linguistics (see especially Bauer Na‘e kai ‘e he tangata e ika 1997:477–9). Past eat subject marker the man the fish In Maori, as in most of the languages of the ‘The man ate the fish (transitive verb ‘eat’)’ Polynesian Triangle, the predicate phrase (usually a Ma¯ori: verb) occurs first in sentences in which no special emphasis is placed on any one part. It is followed by I haere te tangata the subject and any other phrases, such as objects and Past go the man adverbial phrases. The examples given above illustrate ‘The man went (intransitive verb ‘go’)’ this feature for both Maori and Tongan. These languages, however, also have devices for placing I kai te tangata i te ika particular emphasis on some part, usually by placing Past eat the man object marker the fish the phrase in the initial position: ‘The man ate the fish (transitive verb ‘eat’)’ Na¯ te tangata te ika i kai Given that both patterns occur in the languages of the Of the man the fish past eat family, the question arises as to which of these patterns ‘It was the man who ate the fish (not someone else)’ predominated in PPN and by what mechanisms the descendant languages developed the patterns presently With the exception of some eight nouns, all attested (see especially Chung 1978). designating people, which lengthen the third last The study of the PPN vocabulary and of the structure vowel (wahine ‘woman’, wa¯hine ‘women’), nouns in of the family produces hypotheses about the settlement Maori show no change to form plural. Rather, number of the Pacific that accord well with those derived from is shown in the accompanying articles: archeology. Polynesians entered the Triangle from te whare nga¯ whare eastern Fiji and settled the Tonga-Samoa area before a the house the plural house group moved further eastward to modern French ‘the house’ ‘the houses’ Polynesia. From there, peripheral areas such as Rapanui, Hawaii, and New Zealand were discovered and settled. te¯ra¯ puke e¯ra¯ puke Maori itself is, in terms of its sound system, that hill those hill grammar, and vocabulary, typical of the family as a ‘that hill’ ‘those hills’ whole. Maori has ten consonants, p, t, k, m, n, ŋ (spelt The Maori pronoun system shows two features ‘ng’), r, w, h, and f (spelt ‘wh’), and five vowels, a, e, typical of Polynesian languages: i, o, and u, which may be short or long. Furthermore, syllables are restricted in shape to at most one ● Separate forms not only for singular (one consonant followed by up to three vowels. Thus, in person) and plural (several people) but also for Maori, consonants occur only singly and never at the dual (two people): koe ‘you (one)’, ko¯rua ‘you end of a word. Maori has very little in the way of (two)’, and koutou ‘you (several)’. inflection, that is, the use of devices such as suffixes ● Separate forms for ‘we’ depending on whether for grammatical purposes. Rather, grammatical the person spoken to is included (inclusive forms) 652  ¯ ORI AND THE POLYNESIAN LANGUAGES MA or not (exclusive forms): ta¯ua ‘you (one) and extended into primary and secondary schooling, me’, ma¯ua ‘him or her and me’, ta¯tou ‘you (two catering to approximately 20,000 pupils from or more) and me’, and ma¯tou ‘them and me’. preschool to high school levels. Maori Studies is a popular subject at all New Zealand Universities and The possessive constructions in most Polynesian can be pursued up to the doctoral level. languages, including Maori, show two parallel sets of In 1987, the passage of the Maori Language Act markers. Which is appropriate depends on the made Maori the only de jure official language of New relationship between the possessor and the thing Zealand and founded the Maori Language possessed, in particular, whether the possessor Commission. This body, with its office and staff in exercises independent control over the thing Wellington, develops policy and promotes activities possessed: designed to give substance to the official status of the ta¯ku pukapuka but to¯ ku ingoa language. Part of the Commission’s work has centred my book my name on the development of vocabulary. Maori is now used ‘my book’ ‘my name’ in official documents, in broadcasting, and in education as never before, and deliberate expansion of Na¯ He¯mi te¯nei kai but No¯ Pita te¯nei whare its vocabulary to equip it for these domains has been Of James this food Of Peter this house essential. In developing new vocabulary, the ‘This food is James’s’ ‘This house is Peter’s’ Commission and other authorities have been guided Spread as it is over a country of the size of New by the principle that terms should be created with the Zealand, Maori shows some dialect diversity. This is, resources already found in the language. Thus, however, mostly a matter of vocabulary and is by no borrowing from English is ruled out, and the creation means so great that mutual intelligibility is impaired. of new compounds or the specialisation of existing Once the language of the entire Maori population, it words are preferred strategies. has been subjected to severe pressure from English, Maori is the vehicle of a considerable traditional especially during the second half of the twentieth literature. Oral traditions of a variety of genres, such as century. Of the approximately 520,000 people (15% of settlement histories, aetiological myths, fairy stories, the New Zealand population) who identified songs, and poetry, were committed to writing by themselves as ethnically Maori in the 1996 census, Maori elders themselves and are accessible partly in only some 26% claimed that they could conduct a publications and partly in manuscripts in libraries and conversation on everyday matters in Maori. Other museums. Oratory and song are two oral genres that studies have shown that there has probably been some are still very much alive today and play central roles in overreporting by many respondents and that there may Maori culture and community life. At the same time, well be only as few as about 10,000 really fluent creative writing, especially for school use, is using speakers of Maori. Further, the native speakers of Maori within new literary genres, such as the short Maori are concentrated in the older generations, and story and even the novel. the natural transmission of the language from parents The maintenance and development of Maori is a to children within families is essentially broken. In very important strand in New Zealand’s present parallel to the decrease in the population speaking movement toward a more equitable, multicultural, and Maori, there has also been a reduction in the domains inclusive social structure, as past wrongs and where it is regularly used. In most parts of the country, assimilationist policies are redressed. Maori is regularly used only during ceremonial occasions, such as the welcoming of guests and at funerals and church services. References The awareness that the continued survival of Maori Bauer, Winifred. 1997. The Reed reference grammar of Ma¯ori. had become precarious and the strongly held feeling Auckland: Reed. that the language is a central feature of Maori ethnic Biggs, Bruce. 1963–1976. The languages of Polynesia. Current and cultural identity have led, since the mid-1970s, to trends in linguistics, Vol. 8, ed. by T. Sebeok, 466–505. The Hague: Mouton. a wide range of initiatives in language maintenance Biggs, Bruce. 1969. Let’s learn Maori. Wellington: Reed. and development. In response to the recognition that Chung, Sandra. 1978. Case marking and grammatical relations the natural transmission of the language could no in Polynesian. Austin: University of Texas Press. longer be relied on to ensure its acquisition by Krupa, Victor. 1982. The Polynesian languages. London: succeeding generations, preschools were founded in Routledge & Kegan Paul. Lynch, John. 1998. Pacific languages an introduction. which Maori was the sole language spoken and in Honolulu: University of Hawai’i Press. which instruction took place within a Maori cultural Williams, H. W. 1971. A dictionary of the Maori language, 7th environment. Since the early 1980s, this initiative has edition. Wellington: Government Printer. 653  ¯ ORI AND THE POLYNESIAN LANGUAGES MA The Ma¯ori Language Commission’s website at http://www. See also Austronesian; Comparative Method; tetaurawhiri.govt.nz/ describes the Commission’s activities Endangered Languages; Genetic Relationship; and contains links to other relevant sites. Language Planning; Typology RAY HARLOW Marshallese and Micronesian Languages Malayo-Polynesian is one branch of the Austronesian Micronesia. Many scholars question the position of language family, and has two sub-branches: Western Yapese in this group since no one has yet Malayo-Polynesian (which includes a large number of demonstrated that it must belong to this group. Several languages, including some of the major languages of scholars, including Blust, have recently proposed that the family) and Central-Eastern. The Central-Eastern this language belongs to the Oceanic subgroup. The sub-branch in turn consists of two major branches: approximately 1,300 speakers of Kapingama-rangi Central Malayo-Polynesian and Eastern Malayo- include 700 on Kapingama-rangi and several hundred Polynesian. The Eastern Malayo-Polynesian in turn is in Porakiet village on Ponape. It is somewhat subdivided into two sub-branches: South Halmahera- intelligible with Nukuoro, with 55% lexical similarity. West New Guinea and Oceanic. The Oceanic sub- There are roughly 550 speakers of Nukuoro, with 125 branch contains over 450 languages, or about half of on Ponape. The chief languages of the Remote the total for the Austronesian family as a whole. Most Oceanic sub-branch include: Ikiribati (Gilbertese) Oceanic languages have lost word-final consonants with over 60,000 speakers, mainly in Kiribati; Trukese and have simplified consonant clusters. The most with about 45,000 speakers, including second- important sub-branch, Remote Oceanic, has a number language speakers; and Marshallese (over 30,000 of branches and sub-branches. The more important speakers). groups of languages belonging to Remote Oceanic are Geographically, the Micronesian Islands lie within Micronesian and Central New Hebrides. The Remote four major archipelagos: the Marianas, the Carolines, Oceanic sub-branch covers most of Micronesia and all the Marshalls, and the Gilberts. There are 19 languages of Polynesia (including Fijian, Rotuman, Samoan, in the Micronesian subgroup including Carolinian Maori, Tongan, Tahitian, Hawaiian, and Rapanui or (Saipan Carolinian), English, Kapingamarangi, Kusaie Pascuense). However, a number of scholars question (Kosrae, Kosraean, Kusaiean), Mokil (Mokilese), the validity of the Remote Oceanic subgroup. Mortlock (Mortlockese), Namonuito, Ngatik, Nukuoro The languages of most of the Micronesian islands (Nukoro, Nuguor), Paafang, Pingelap (Pingilapese), (except the languages of Palau and the Marianas) are Ponape (Pohnpeian, Ponapean), Puluwat, Satawal more closely related to those of eastern Melanesia (the (Satawalese), Sonsorol, Truk (Ruk, Trukese), Ulithi, Solomon Islands and Vanuatu). Linguistic analysis Woleain (Wolean), and Yapese. Kiribati is the language points toward eastern Melanesian origins for the of the Kiribati (Gilbert) island chain, of neighboring earliest eastern Micronesian peoples. The Micronesian Banaba (Ocean) island, Niu, and it is also spoken in the subgroup of the Remote Oceanic sub-branch includes Line Islands. Kusaie is the language of 5,000 all the languages of Micronesia except for Palauan, inhabitants of Kosrae State (Federated States of Yapese, and Chamorro (which all belong to the Micronesia), and many inhabitants also speak Western Malayo-Polynesian branch and have affinities Ponapean. Nauruan is the traditional language of with the western Austronesian languages of Indonesia Nauru, with approximately 4,000 speakers. English has and the Philippines) and Nukuoro and Kapingama- been the strongest influence on Nauruan (Australia rangi (which are Polynesian). There are approximately administered the island from 1920 to 1968), and there 73,000 total speakers of Chamorro, of whom 60,000 are signs that Nauruan is being abandoned in favor of are on guam and 13,500 on the Northern Mariana English. Ponape, with about 22,000 speakers, is the Islands, Micronesia. The 15,000 speakers of Palauan major language of Pohnpei State (Federated State of or Palau primarily reside in Belau, guam, and Western Micronesia). Ngatikese, Pingelapese, and Mokilese are Carolines, Micronesia. And there are roughly 5,000 counted as dialects of Pohnpeian or as closely related speakers of Yapese on Yap, Caroline Islands, languages. Ponape has a ‘high language’ with a 654  MARSHALLESE AND MICRONESIAN LANGUAGES partly separate vocabulary, used in speaking about terms of these ritual languages are metaphors or people of high rank. Truk (22,000 speakers) is the metonymies. major member of a dialect chain with a total of 40,000 Students and youngsters have also developed a speakers. These languages are spoken in the Caroline ‘student language’ or slang in Majuro and on Ebeye Islands (Federated States of Micronesia), and have (Zewen 1977). It is characterized by nonstandard numerous English and Japanese loanwords from their vocabulary, arbitrary coinages, and figures of speech colonial history. that often quickly fall out of fashion or use. Marshallese has the largest number of speakers Linguists have differed over the vowels of (about 30,000) of any Micronesian language. Marshallese. Certainly, Marshallese has a very high Marshallese is primarily spoken in the Republic of the number of vowel phonemes, particularly in Marshall Islands, but also on Nauru. English and comparison with Ponapean (with which it has roughly Japanese have had the strongest effect on Marshallese. 33% lexical similarity). This situation seems to be the Japanese administration followed after World War I, result of vowel fusions after certain consonants have and the United States took over from 1945 to 1986. disappeared according to Marshallese sound laws. The Marshall Islands comprise 29 coral atolls and Byron Bender (1968) argues that there were three five single coral islands scattered over about 180,000 vowel phonemes (a e I ), other vowel sounds being square miles of the Pacific Ocean. Kwajalein Atoll is conditioned by surrounding consonants. The usual the world’s largest atoll formation, comprising 93 spelling recognizes nine. Other scholars distinguish 12 islets. Considering this vast area and scattered land vowel phonemes, and add a distinctive feature of area, the Marshallese language is amazingly length, giving a total of 24. There are three groups of homogenous, but there are two main dialects that consonant sounds, according to their effect on sometimes differ in vocabulary, the choice or the neighboring vowels: the plain or light consonants, the pronunciation of a word, and in the vowels of words heavy consonants, and the rounded consonants. whose first two consonants are identical. These two Marshallese has lost the complex series of numeral dialects coincide with the two major chains of islands: classifiers that are found in the other Micronesian the eastern, sunrise, or Ratak chain and the western, languages. ‘Cilcino’ (six) originates as ‘cilu’ (three) sunset, or Ralik chain. However, at focal centers like plus ‘cilu’ (three); ‘7’ as ‘3 + 3 + 1’. Majuro, or Ebeye on Kwajalein, one finds speakers of With respect to Marshallese phonology, there are both dialects. Zewen (1977) found evidence of the two eight basic patterns of syllables. Words, when dialects merging into a kind of Marshallene ‘koine’ in monosyllabic, consist of one of these patterns; when the District Centre, as well as on Ebeye. polysyllabic, they consist of a sequence of these The Ralik dialect has obtained a higher status patterns. Most Marshallese words are dissyllabic. because the Bible was first translated into Ralik, and the With respect to morphology, Marshallese is not an early missions were located in that chain. The eastern inflected language in which word classes can be defined (Ratak) dialect is more conservative. The differences by the type of inflections they take. Other criteria have between the two dialects appear in the vocabulary and to be considered such as the meaning of the word, and the phonetic realization, rather than in the grammar. For the valence or systematic combinability with other instance, words that begin with a vowel have a different words. For instance, nouns may be determined by phonetic onset in the two dialects. The Ratak dialect has demonstrative pronouns. They may be main or quite a number of words that consist of a reduplicated secondary elements of phases or may be preceded by a basic morpheme. There are also minor dialects within preposition. In clauses they may be subject, predicate, or the major dialects: Mejit, for example, is divergent in object. With the exception of personal and nonpersonal the east, as in Ujelang in the west. proper nouns, they may take ‘possessive suffixes’ or Terms associated with precolonial customs and may be constructed with possessive classifiers. activities such as native clothing, handicraft, and navigation are today referred to as ‘Kajin Etto’ (old language), although understanding of these terms References varies in the present generation (Zewen 1977:4–6). In Abo, T., B.W. Bender, A. Capelle, and T. DeBrum. 1976. precolonial times, ritual languages were mandatory for Marshallese–English dictionary. Honolulu: University Press certain activities: for instance, special words had to be of Hawaii. used on the northern voyages, especially to Rongerik. Bender, B.W. (ed.) 1984. Studies in Micronesian linguistics. Bender, B.W. 1968. Marshallese phonology. Oceanic Word taboos were associated with tattooing or Linguistics 7. 16–35. particular places: for instance, on the small islands of Bender, B.W. 1984. Object marking in Marshallese. In Bender, Kwajalein Atoll, ‘kako’ (rooster) was taboo. Most pp. 443–65. 655 MARSHALLESE AND MICRONESIAN LANGUAGES Blust, R. 1990. Summary report: linguistic change and Kroeber, A.L. 1911. Phonetics of the Micronesian language of reconstruction methodology in the Austronesian language the Marshall Islands. American Anthropologist 13. 380–93. family. Linguistic change and reconstruction methodology, Lingenfelter, S.G. 1986. Social anthropology and translation. ed. by P. Baldi, 133–53. New York: Mouton de Gruyter. Notes on Translation 116. 30–46. Clark, R. 1987. Austronesian languages. The world’s major Matsuoka, S. 1929. Study of the language of the Marshall languages, ed. by B. Comrie, 899–912. New York: Oxford Islands. Tokyo: Kyodo Kenkyusha. University Press. Rehg, K. 1984. The origins of ‘compensatory lengthening’ rules Combs, M.R. 1984. Glides: the problem of definition and in Micronesian languages. In Bender, pp. 53–59. analysis, including a discussion of the treatment of Zewen, F.-X. 1977. The Marshallese language: a study of its semivowel glides in Micronesia and an acoustic phonetic phonology, morphology, and syntax. Unpublished Ph.D. analysis of glides in selected Trukic dialects. Ph.D. Thesis. Dissertation. Hamburg: Universitat Hamburg. Honolulu: University of Hawaii. CAMILLA COCKERTON Martinet, André The main representative of functionalist structuralism, Accordingly, he sees language structure only as a logical André Martinet is one of the most enlightening figures of complement and manifestation of language functioning. twentieth-century linguistics. He spent his childhood in Within this functional perspective, Martinet clearly the peripheral Provencal region of Savoy, which insists on the reliance on concrete linguistic facts and endowed him with an early sensitivity for the problems the need to study specific languages exclusively in their of bilingualism and language social aspects, while own functional framework, as each language during his brilliant classical education he acquired corresponds to a specific organization of reality and practical knowledge of a number of languages. His human experience: ‘What we wish to examine is the degree was in Germanic linguistics but he also acquired whole of man’s linguistic activity, and as this activity broad knowledge of the general European linguistic and only takes place within the framework of specific philological tradition. During his education in Paris, languages, our primary task is to study these languages Berlin, and Copenhagen, as well as later while he was as so many different objects.’ This rational and prudent teaching at Columbia University, he established direct attitude, which is in sharp contrast to perhaps more contacts with all relevant structuralist trends in both attractive, bold theoretical assumptions of some other Europe and America (developed by the Prague School, formalistic and structural linguists, has often been the Copenhagen School, and American structuralism), labeled as Martinet’s ‘realism’. Martinet maintains that showing appreciation of all elements of other schools the task of linguistic analysis is not to seek that could fit his own view of language. Influenced manifestations of universal categories in languages, but primarily by the structuralism of Ferdinand de Saussure to discover the unique categories articulated by each and that of the Linguistic Circle of Prague (particularly language. Accordingly, in his view, theory should be Nikolay Trubezkoy), Martinet further developed the always guided by analysis, not the other way around, no structuralist theoretical tenets enriching them by matter how unfamiliar the resulting theory might achievements of American linguistics and Danish appear, as ‘the scientific treatment of an object requires, glossematic theory. However, adopting a clearly first and foremost, that the object’s integrity should not functional view, he specifically emphasizes the functions be sacrificed to methodological exigencies.’ of language in general, or the functions of linguistic A cornerstone of the functionalist approach and elements as the main object of linguistic enquiry in Martinet’s main contribution to general linguistics is his opposition to structuralism and formalism of the important notion of ‘double articulation’ referring to majority of American structural linguists in which the two fundamental levels of language structure, that of system of a language is explicitly studied in abstraction meaningful units (primary articulation) and that of from its functions. For him, language is essentially a expression units (sounds) which are meaningless in human institution and an instrument of communication, themselves but serve to distinguish the meaningful while the method of linguistic analysis is based on the elements (secondary articulation). The concept of principle of communicative relevancy of linguistic units ‘double articulation’ (or ‘duality of patterning’ as it is and their role in the transmission of information. termed in American linguistics to avoid confusion with 656  MARTINET, ANDRÉ the articulation of sounds) allows a small number of variations’, and syntax is ‘loosely understood as the meaningless phonological elements to be combined to survey of all data pertaining to the combinations of distinguish an infinite set of meaningful utterances. The significant elements’. Syntax is not completely minimal linguistic forms that are meaningful are predictable but dependent on compatibility or conven- monemes, corresponding roughly to what American tions, which one just has to know and cannot calculate linguists call morphemes, which Martinet further in advance. Martinet, however, has been criticized for analyzes as units of grammatical meaning (morphemes) avoiding to a certain extent the problem of semantic and units of lexical meaning (lexemes). Thus, the word analysis by arguing that all linguistically relevant books is composed of two monemes: a lexeme book and aspects of meaning must already be manifested at the a morpheme -s denoting plurality. Moneme has both an level of expression, while its other aspects easily expression and content and consists in the association escape objective analysis and require nonlinguistic between a signifier (vocal expression) and a signified (external) explanation and therefore should not be the (semantic content). In line with the structuralist subject matter of linguistics. tradition of the phonological Prague School the basic An important part of Martinet’s work is devoted to unit of the sound system is called a phoneme: it is a historical linguistics and diachronic problems of lan- minimal, contrastive sound unit that distinguishes one guage change. In his seminal work Économie des meaningful utterance from another (e.g. pin/bin). changements phonétiques (1955; Economy of sound Phonemes may be further analyzed in terms of a set of changes) he formulates an original, well-grounded underlying distinctive features specifying the ways the theory of sound change as arising, in part, from internal sound is physically produced. However, Martinet pressures within a phonological system. He explains rejects Roman Jakobson’s phonemic theory that there is that linguistic evolution is ‘governed by the permanent a universal inventory of these distinctive features and conflict between man’s communicative needs’ (using as that they are based on binary oppositions. many units as possible, as different from each other as Martinet made major contributions to the field of possible) and ‘his tendency to reduce to a minimum his general and descriptive phonology further developing mental and physical activity’ (using as few units as the phonemic theory of the Prague School through possible). By economy, Martinet means the principle of elaboration of such notions as phonemic frequency, ‘least effort’ and the tendency to reconcile these two functional load, neutralization, and segmentation of contrasting elements to attain an improved functional phonemes. His second doctoral dissertation La load and help maintain an acceptable level of phonologie du mot en danois (1937; Phonology of communication. However, optimum economy and Danish) was one of the first complete phonological complete symmetry can never be reached because of descriptions of a language. He regards phonology as physiological factors such as inertia and asymmetry of ‘functional phonetics’, which has to interpret real the speech organs (e.g. distinction between front and phonetic facts from the point of view of linguistic back articulation). Martinet distinguishes syntagmatic functions of sound differences and their functional economy (contextual assimilating pressure exerted by effects and not as different realizations of an ‘abstract adjacent linguistic units) from paradigmatic economy system’. (dissimilating pressure exerted by the units in the Using the function of moneme in the utterance as system which might have appeared in that particular the starting point, Martinet also developed a coherent place), and determines the effect of different variable system of general morphological and syntactic synchronic factors (such as frequency, correlative analysis focusing particularly on the sentence structure relationships, the notion of empty place in a pattern, and predicate function. While rejecting logical and etc.) on language development. In this way, he tried to psychological criteria, he analyzes sentences exclu- reconcile the apparently conflicting viewpoints of sively by applying the criterion of linguistic function synchronic and diachronic linguistics by drawing upon (revealing linguistic ‘choices’ of speakers by means of his findings in one field to illuminate the other. commutation). He establishes a hierarchical relation- However, he has been often criticized on account of ship between monemes, the minimal units of the first deliberate neglect of external historical and social articulation according to their function in the sentence aspects of language change. Although it is true that by classifying them into free monemes and bound or Martinet primarily considered the inner causes of functional monemes. The word can consist of one language change, he also underlined the role of external nonanalyzable item or moneme, e.g. head and of explanation in the rate of change and its specific combined items or synthemes such as headache. In historical context, and in this way has convincingly terms of communicative relevancy, morphology is succeeded in elucidating many mechanisms of sound considered to be the ‘presentation of the variations of change making his work indispensable in any serious the form of signifiers and of the conditioning of these study on language evolution. 657 MARTINET, ANDRÉ In addition to his important theoretical achievements, Société de linguistique de Paris in 1931; President, Martinet also wrote about specific problems of many Société européenne de linguistique, 1966–1999; different languages frequently discussing honorary member, Linguistic Society of America methodological and practical questions of language (LSA) in 1973, Union Mundial pro Interlingua (UMI) learning and acquisition, while at the same time working in 1998; editor of Word 1946–1960; founder and editor on the development of international auxiliary languages. of La Linguistique since1965. Martinet died in With his central position between European and Châtenay-Malabry on July 16 1999. American structuralist schools and explicit appreciation of positive outcomes of traditional comparative References philology, Martinet was one of the rare linguists who tended to bridge the gap between different fields of Fried, Vilém and René Dirven (eds.) 1987. Functionalism in linguistic study. His clearly formulated functional theory linguistics. Amsterdam: Benjamins. Fuchs, Catherine, and Pierre Le Goffic. 1992. Les linguistiques based on the empirical-deductive approach exerted an contemporaines: repères théoriques (Current linguistics: important influence on further development of European theoretical standpoints). Paris: Hachette. linguistics, while his method, which always remains Lepschy, Giulio C. 1970. A survey of structural linguistics. close to the actual reality of specific languages, has been London: Faber and Faber. used for phonological and syntactical descriptions of Leroy, Maurice. 1963. Les grands courants de la linguistique moderne, Brussels: University Press of Brussels; as The more than hundred languages. main trends in modern linguistics, transl. by Glanville. Price. Oxford: Basil Blackwell, 1967. Mahmoudian, Mortéza. (ed.) 1979. Linguistique fonctionnelle: Biography débats et perspectives. (Functional linguistics: discussions André Martinet was born in Saint-Albans-des- and perspectives). Paris: PUF. Martinet, André. 1955. Économie des changements Villards, Savoy on April 12, 1908. He obtained degree phonétiques: Traité de phonologie diachronique (Economy in Germanic linguistics from Université de Sorbonne of sound changes: Diachronic phonology), Bern: Francke and Ecolé Pratique des Hautes Etudés, Paris (1932), Verlag. and a Ph. D. for two dissertations in Germanic ———. 1960. Eléments de linguistique générale, Paris: linguistics on consonant gemmination in Germanic Armand Colin; as Elements of general linguistics, transl. by Elizabeth Palmer. London: Faber and Faber, 1964; Chicago: languages and on phonology of Danish in 1937. He University of Chicago Press, 1982. was Chair of Phonology, Ecolé Pratique des Hautes ———. 1962. A functional view of language. Oxford: Etudés, Paris in 1938–1946. During World War II, Clarendon. while imprisoned in German camp, he carried out an ———. 1975. Studies in functional syntax (Études de syntaxe impressive phonetic and phonological survey on fonctionnelle). Munich: W.Finck. ———. 1989. Fonction et dynamique des langues (Function contemporary French pronunciation published in and dynamics of languages). Paris: Armand Colin. 1945. He moved to New York, where he was director ———. 1993. Mémoires d’un linguiste. Vivre les langues of interlinguistic research of International Auxiliary (Memoirs of a linguist. Living the languages). Paris: Quai Language Association (IALA) 1946–1948. He was Voltaire. Professor at Columbia University and was an active Sampson, Geoffrey. 1980. Schools of linguistics, Stanford, CA: Stanford University Press. member of the Linguistic Circle of New York Tobin, Yishai. 1988. Phonetics versus phonology: the Prague 1948–1955. He was Chair of Phonology, Ecolé school and beyond. The Prague School and its legacy, ed. by Pratique des Hautes Etudés and Professor of general Yishai Tobin. Amsterdam and Philadelphia, PA: Benjamins. linguistics at Sorbonne, Paris in 1955–1978; member, ANITA SUJOLDZIC Mass Media and Language Broadcasting (TV, radio), print media (newspapers, media. Telecommunications and news agencies, the magazines, books, pamphlets, posters), film, theater, two other areas of mass communications, differ from and electronic products such as CDs, records, tapes, them in a number of ways, but the boundaries are being and videos are the most important instances of mass blurred by modern, especially digital, information 658  MASS MEDIA AND LANGUAGE technologies. Important characteristics that they all clusters of co-occurring features. Recent ones define it share are the unidirectionality of the information flow in terms of semantic potentials, the typical meanings from a source to a target and the constraints they expressed in such contents, and the typical linguistic impose on feedback. Broadcast and print media have expressions used to do that end. Expressions range developed response channels, which differ in kind, from text or discourse properties to features of immediacy, and effect as a result of the technology they pronunciation. Such a wide definition is to better able are based on. They are often referred to as news media to capture the diverse means of expression that are since they normally carry news, current affairs, and used in restricted reported domains to shape content sports. But news need not be a defining criterion as and to express underlying ideological stances. music, film, or weather channels show. Although they Focusing more narrowly on linguistics, one should will be ignored, one should add that the linguistic begin by saying that, in view of the diversity of linguistic effects of entertainment media may surpass those of the models, it is functional ones, pragmatics, text and informational and educational media and may further discourse linguistics, socio-and psycholinguistics that endanger aid endangered languages. have an affinity to the concerns of media studies. Four Before turning to linguistic aspects, one must points should be emphasized. The first is that linguistics emphasize the fact that they are (sociopolitical) can be an ancillary discipline or an independent player. institutions with, as a special characteristic that emanates In the latter capacity, it pursues its own agenda, using from their purpose, developed mechanisms for the flow, media output as primary data for the descriptive study of categorization and packaging of information, and a language. In the former, it helps highlight aspects of the history of professionalism. In terms of communicational communicational process that are of interest to media function, they can thus be defined as domains whose studies. Second, media output does not occur in isolated goal it is to make content public, accessible to debate. units but as instances of higher order units, such as But they are better referred to as hyperdomains since all program formats, or as products of a medium’s domains, e.g. politics, law, business, lifestyle, are institutional structures, such as news and current affairs. providers of content. Their sociopolitical role, especially Output is thus subject to ever more abstract norms and their relationship with sources, and their impact on expectations. Third, output has a presentative and audiences remain controversial. What is clear is that they representative side. The former highlights style or, to use are not mere transmitters of information; they may Hallidayan terms, the ‘interpersonal’ function, the latter indeed have a symbiotic relationship with sources (e.g. the way reality is construed, the ‘ideational’ function. In in sport). Minimally, they should be called co- other words, presentation is about the way media output orchestrators of what is ‘in the air’ in the public arena. targets its audience, representation about the way it Turning to language, co-orchestration means that (re-)constructs what it is about. Both presentation and the media collect, select, manipulate, shape content, representation are supported by ideologies and attitudes and create a packaged, marketable product that is that they reaffirm. Finally, one must distinguish output as recognizable to large audiences and helps maintain a a product, the result of complex institutional processes media outlet’s audience share. Language is the central, and output as a process, the way it is experienced by an but not the only, component in these processes, the audience. Linguistics can shed light on both, including others being visuals, film, sound effects, presentation, the points mentioned above, viz. the embeddedness of and layout. Language is crucial in the packaging of the public idiom inside a network of source and target raw input into output and in mediating it to an domains. audience. The inference that media language must be Studies on sports coverage have revealed that the homogeneous, however, is unwarranted in light of the technology, i.e. radio, TV, or print, and the mode of diversity of source domains, professional routines, communication, i.e. speech or writing, have effects on technologies (print, radio, or TV), audience the idiom, which are independent of such factors as stratification and habits, and the need for outlets to program formats. Radio, for instance, turned out to be create recognizable and consumable products. It is more action-oriented, focusing more on detailed spatial better to speak of media languages or to borrow from expressions than TV and the press, even though the time media sociology the term public idiom, which is better constraints on radio commenting are much heavier. able to refer to the existence of abstract norms and Combining discourse and text linguistics with Content professional routines that bear upon the use of Analysis, studies on representation have left no doubt language. The linguistic concept of register has been about the existence of ideological bias in media used to describe such differences as they were seen to coverage. They have revealed mechanisms whereby emanate from a particular configuration of situational ‘hidden meanings’ are transmitted that reinforce features or context-of-situation. Early definitions of attitudes such as racism, in the absence of an audience’s this term highlighted lexicogrammatical features or direct experience with ethnic groups. Studies closer to 659 MASS MEDIA AND LANGUAGE mainstream linguistics have looked at how language is Generally speaking, research has moved away from used to recast reality into a socially coherent consensus narrow linguistic concerns (grammatical structures, or, less often, conflictual position. ‘Critical Linguistics’ lexical choices) to broad concepts that permit the or Critical Discourse Analysis, outgrowths of Hallidayan understanding of ideological bias. It is here that the functional grammar, have thus revealed how lexical audience dimension has been reintegrated, whereas fields, grammatical systems like transitivity or modality older research was largely confined to comprehension can be used to create ideological bias. Gender analyses and recall studies. Like all texts, media texts can be seen in this context. Text and discourse analysis are assumed to contain slots for the reader, which have looked at macro-and microproperties of media activate in recipients shared concepts, schemata, texts. Van Dijk argues forcefully that texts express, or prototypes, and the like. Mass media, thus, do not can be reduced to, macrosemantic propositions and are target passive recipients; they activate recipients’ condensed in conventionalized (text-) syntactic super- worldview. Although this is an attractive position, the structures. He exemplified his approach in a study on the implication is that mass media address individuals, international press coverage of the assassination of while media sociology rightly operates with layers of Lebanon’s President General Gemayel, which revealed recipients that reinforce or cushion off media impact. the presence of a bias in favor of ‘Western’ formulations Related research has looked at linguistic details to of the event and the dominance of local (as against show how texts incorporate features of the language of global) journalistic routines. They showed up in the the audience (‘audience design’) to enhance their shared macrostructure of media texts. From a representational impact and to maintain or increase methodological perspective, he showed how (under- audience size. lying, social) ideologies and (concrete acts of) discourse Studies of interviews, talkback shows, etc., have are connected through (cognitive) models that revealed patterns similar to the ones in monologs. discourses such as editorials aim to activate to bring over Research tends to highlight the product––messages the writer’s intent. Studies on the relationship of a text’s that recipients are most directly exposed to––although structure with news values and narrative theory have they interpret findings by relating them to institutional shown that news values take precedence over the or general social parameters. Studies of the institu- chronology of events. Reports, for instance, take the tional context of the public idiom, i.e. the formulation form of an inverted pyramid of information structure, of communicative and language norms, reveal the where the most important, often the most recent, institutions’ internal structures, such as the contribution information comes first, with background being delayed. of advisory bodies, language guardians, or of editorial This anticlimactic structure can be traced to the policies. To turn to linguistics as an independent player, American Civil War, where the emphasis was placed on one should note the traditional interest of descriptive ‘recency’ as against ‘explanatory background’. It follows and applied linguistics in media language to study the that reports do not need to be read to the end. Editorials, ‘real’, as against the ‘potential’, in language use. The in contrast, which argue for or against some position, design of reference and teaching materials have been have a climactic structure and require full-text reading. important outcomes. Corpus linguistics has added a A traditional concern of presentation studies has new methodological dimension to comparative been with the choice of language (variety). Choice was research. seen to respond to such factors as program policy, Although linguistics has important contributions to intelligibility, and to favor standard or prestige make to mass media studies, one should not forget that languages. This turned out to be true only in specific factors to do with ownership, legal status, business situations, as when public service media enjoy a strategies in light of the globalization of communi- monopolistic position, when broadcasting focuses on cation, political control, public accountability, etc., high cultural products or on nation-building. Britain’s dominate research. BBC between the 1920s and 1950s is a case in point. Competitive settings as in the United States of America or federal ones as in Germany promote a References wider range of varieties. It has also been shown that Bell, Allan, and Peter Garrett (eds.) 1998. Media discourse. standard varieties reflect the tone of officialdom and of Oxford: Blackwell. elites, and support representational bias. A strict Burger, Harald. 1990. Sprache der Massenmedien. Berlin: separation between presentation and representation is Walter de Gruyter. Dijk, Teun A. van. 1988. News as discourse. Hillsdale, NJ: thus unrealistic and generalizations about either must Lawrence Erlbaum. be tempered in light of differences between intended ———. 1998. Opinions and ideologies in the press. Media norms and practices, and between intended discourse, ed. by Allan Bell, and Peter Garrett, 21–63 communicative effects and its actual realization. Oxford: Blackwell. 660  MATHEMATICAL LINGUISTICS Halliday, Michael A.K., and Ruqaiya Hasan. 1976. Cohesion in Leitner, Gerhard. 2000. Lexical frequencies in a 300 million word English. London: Longman. corpus of Australian newspapers: analysis and interpretation. Fowler, Roger. 1991. Language in the news. Discourse and International Journal of Corpus Linguistics 5(2). 1–32. ideology in the press. London: Routledge. Leitner, Gerhard, and Markus Hesselmann. 1996. ‘What do you Leitner, Gerhard. 1997. The sociolinguistics of communication do with a ball in soccer?’ Medium, mode and pluricentricity media. Handbook of sociolinguistics, ed. by Florian Coulmas, in soccer reporting. World Englishes 15(1). 83–102. 187–204. Oxford: Blackwell. GERHARD LEITNER Mathematical Linguistics Mathematical linguistics is a branch of linguistics frequency of the occurrence of labial, front, palatal, studying the features and properties of language from sonorant, occlusive, fricative, and voiced consonants in the mathematical point of view. It is usually divided into 109 world languages of different language families, e.g. several branches: (1) quantitative or statistical Indo-European, Finno-Ugric, Samoyedic, Tungus- linguistics, dealing with unique language entities on the Manchurian, Turkic, Paleo-Asiatic, etc. He built phonetical, lexical, syntactical, and semantical levels; phonostatistical distances between the languages inside (2) algebraic linguistics; and (3) machine translation. each language group (e.g. Germanic, Romanic, Algebraic linguistics use qualitative mathematical Slavonic, Ugric, etc.) and each language family. methods like: algebraic methods, theory of algorithms, Lexical statistics investigates the sizes of corpus, graph theory, topology, and mathematical logic. the zones of words with high, medium, and low Mathematical linguistics develops formal apparatus for frequencies, the laws of their distribution, especially the description of natural and artificial languages. The Zipf law. Lexical statistics gave rise to glot- discipline was created in the 1950s to define language tochronology, created by an American linguist Mooris phenomena logically, precisely, and exactly. Its Swadesh in the 1950s. He tried to make statistical fundamental based on the ideas that language is a comparisons between the rates of change within the functional mechanism, which can be studied through fixed list of words of two languages to find out the the frequency of occurrences of speech units. More time these two languages separated. The often than not, under mathematical linguistics, its glottochronological method is based on two pivotal statistical branch is understood. Mathematical assumptions: (1) in every language, there exists a list linguistics unites the study of quality and quantity in of 100 core words; (2) the rate of disappearing words language covering such fields as phonological, lexical, is constant. This method was much criticized. The grammatical, semantic, and stylistic statistics. The main methodological fault lies in the fact that it is hard frequency of occurrence of linguistic units and features, to establish the main 100 word core of a language and their rank, order, laws of distribution, correlation, the false relation of the decay of carbon-14 with the stability, and variance coefficients are studied. disuse of words in a language. Nevertheless, if one Considerable attention is given to the definition of the takes the whole word stock of two languages, then one units of population, random sampling, necessary, and can speak about their difference, without resorting to sufficient sample size. The forms of frequency the precise time of language separation. To establish distributions of linguistic units are studied. the proximity of two languages, one should compute Phonological statistics studies the frequency of the correlation coefficients of their phonological, occurrence of letters, phonemes, and phonemic groups morphological, syntactical, and other categories. It is in languages as it is and in different languages world the task of statistical language typology and over. One can establish the phonostatistical distances dialectology. Grammatical statistics researches the between the sound pictures of different writers or quantitative relationships of morphological categories different world languages. In 1937, a Japanese linguist and parts of speech. Masao Onishi was one of the first to compare the Syntactic statistics determines the frequency, frequency of occurrence of groups vowels and distribution, and relationships of sen-tence structures consonants of ten languages: English, German, French, and types of sentences. Semantic statistics describes Spanish, Italian, Russian, Chinese (Cantonese), Bengali, the quantitative properties of semantic fields of Sechuana, and Japanese. Yuri Tambovtsev compared the different words. 661 MATHEMATICAL LINGUISTICS Stylistic statistics studies quantitative features of last 30 years, natural language processing has been style, correlation, and disputed authorship. The developing rapidly. The work that has been done in this application of statistical methods to problems of area has resulted in a large variety of systems constructed disputed authorship has a rather long tradition in for specific applications. Very often, this particular field English with respect to the works of Shakespeare, Mark is called computational linguistics, and is considered to Twain, Hamilton, etc. It is often called stylometry. The be a separate discipline in its own right. This strand of study of the statistics of sound or word usage helps to mathematical linguistics is close to machine translation catch the nuances of the text. The statistical spectra of and consists in using the computer as a tool in testing sounds or words then and now may give a history of the theoretical linguistic models like transformational development and changes of a language as it is and grammars, Montague grammars, generalized grammars, language habits of a particular writer. The use of etc. Computational linguistics include speech synthesis, stylometry in authorship attribution is based on the speech recognition, constructing concordances, etc. It is hypothesis that there are quantifiable features of style, advisable to include in mathematical linguistics as a characteristic to a particular writer. Authorship discipline all areas of linguistics where various statistical attribution is usually performed in the following way: is counts and analyses are required. the literary opus under investigation more similar to the candidate author “A” or more similar to the candidate References author “B” based on the accepted computed features? Brainerd, Barron. 1974. Weighing evidence in language and Parametric and nonparametric statistical tests are used literature: a statistical approach. Toronto: University of to measure similarity, e.g. t-test (or Student’s-test), chi- Toronto Press. square test (or Pearson test), Mann–Whitney, Wilcoxon, Burrows, J. F. 1987. Computation into criticism. Oxford: and some other tests. Clarendon Press. Butler, Christopher. 1985. Statistics in linguistics. Oxford: Basil Typological statistics quantifies the so-called Blackwell. language universals. In the 1960s, Joseph Greenberg Herdan, Gustav. 1966. The advanced theory of language as deduced 11 typological indexes to put languages in choice and chance. Berlin: Springer-Verlag. some order, e.g. index of synthesis, i.e. the ratio of the Kenny, Anthony. 1986. A stylometric study of the New number of morphemes to the number of words, or Testament. Oxford: Clarendon Press. Ledger, Gerard R. 1989. Re-counting Plato. A computer index of complexity of morphological structure; index analysis of Plato’s style. Oxford: Clarendon Press. of agglutination, i.e. the ratio of the number of Onishi, Masao. 1937. Articulation bases of ten major agglutinative inner morphological links and the languages. A study from the view-point of occurrence of number of words with inner morphological links; sounds. Study of Sounds 6 [in Japanese]. index of derivation, i.e. the ratio of the number of roots Oostdjik, Nelleke. 1991. Corpus linguistics and the automatic analysis of English. Amsterdam: Rodopi. and the number of words; index of inflation, i.e. the Piotrovsky, R. G., K. B. Bektaev, and A. A. Piotrovskaya. 1977. ratio of the number of nonreflexive morphemes and Mathematical linguistics [in Russian]. Moskva: Vys. Shkola. the number of words, etc. These measures were further Tambovtsev, Yuri. 1994. Dynamics of functioning of phonemes modified and complemented by different linguists to in sound chains of the languages of different structures [in eliminate the shortcomings. Russian]. Novosibirsk: Novosibirsk State University. ––––––. 1994. Typology of orderliness of sound chains in Some linguists include corpus linguistics, a branch of language. Novosibirsk: Novosibirsk State University. linguistics concerned with the study of language use by Teshitelova, Marie. Quantitative linguistics. Amsterdam: John means of large text corpora into mathematical Benjamins. linguistics, because they use statistical methods. For the YURI TAMBOVTSEV Maya and Mayan Languages Sharer’s The Ancient Maya (1994) begins with a the Maya, in reference to the Maya people, and as clarification of the more standard uses of the terms adjective, as in Maya writing. Mayan, on the other Maya and Mayan, citing the first as both noun, as in hand, is used with reference to the language family, 662  MAYA AND MAYAN LANGUAGES and the individual languages that make up that family, include the grammatical category aspect, the such as K’iche’ Mayan, Yucatec Mayan, etc. These pronominal system, the concept of ergativity, the conventions will be adhered to in the present essay. numerical system and numerical classifiers, and The Mayan language family consists of 28 extant finally the phonological system. languages and have approximately 6 million Mayan speakers covering a fairly contiguous geographical area encompassing southern Mexico, most of Aspect Guatemala, and parts of Belize, El Salvador, and The term aspect is frequently overlooked in language Honduras. The number of speakers of individual teaching and learning. Comrie (1976) comments that in Mayan languages ranges considerably, with speakers spite of the usage of more traditional terminology, for of K’iche’, for example, spoken in southern example, tense, with regard to the temporal relations Guatemala, numbering more than 1 million, while expressed in sentences, such as John painted the house others, such as Itzá, spoken in the Peten of Guatemala last summer and John was painting the house last are on the verge of extinction. The one group of summer, the distinction between these two is not one of Mayan speakers not geographically bound to the rest tense, but rather of aspect. Both sentences occur in the of the family, the Huastecs, migrated north some 4,000 past; thus; an additional explanation to account for the years ago along the Gulf Coast to present-day difference in meaning must be given. Aspect is the Veracruz. It is believed that at approximately the same grammatical term used to describe how a situation or time, the Yucatecan branch of the Mayan family action is viewed. When the meaning expressed by the separated from the main body. This subgroup consists verb is viewed in its entirety as completed, the term of four languages spoken throughout the Yucatan perfective is used to indicate the aspectual category. Peninsula, northern Guatemala, and parts of Belize. When it is not viewed as completed, and makes no Yucatec Mayan, by far the largest of this subgroup in reference to the beginning or end of the action or state, terms of speakers, numbering approximately 1 the term imperfective is used. million, is the language of the three of the four Mayan languages do not have tenses, but denote surviving hieroglyphic books, or codices. temporal distinctions by means of a number of aspect The main Mayan language body subsequently split markers and particles. Bricker et al. (1998) list the into eastern and western subgroups, each with several following aspect categories for Yucatec Mayan: offshoots, including, ultimately, the Cholan languages, the principal languages of Maya hieroglyphics. Much Assurative future je’el…e’ Je’el in wilikech. of recent inquiry and investigation into Mayan and I will see you. Completive t-…(-aj) Tin wilajech. among the Maya people has been generated by interest I saw you. in the decipherment of the ancient Maya writing Compulsive yaan Yaan in wilikech. system, as explored thoroughly in Coe’s Breaking the I must see you. Maya code (1992). Definite future k-…-j Kinj ilikech. In order to gain an appreciation of the richness of It is I who will see you. Mayan, it is important to define a few concepts and Desiderative taak Taak in wilikech. linguistic elements commonly addressed in the I need to see you. literature on this indigenous language family. Each Durative táan Táan in wilikech. will be defined and examples will be cited within the I am seeing you. section in which they appear. The language examples Habitual/Incompletive k-…(-ik) Kin wilikech. I see you. cited herein, while limited to Yucatec Mayan, will Inceptive jo’op’ Jo’op’ in wilikech. serve to illustrate a number of features found cross- I began seeing you. linguistically throughout the entire Mayan family. For Obligative k’abéet K’abéet in wilikech. more detailed explanations of a number of I need to see you. grammatical features in several Mayan languages, see Proximate perfective táant…e’ Táant in wilikeche’. I just saw you. Robertson’s The history of tense/aspect/mood/ voice in the Mayan verbal complex (1992). The examples cited Additional aspect markers include constructs such as a will utilize the standardized alphabet agreed to by La continuous past (laayli’…e’), a remote future (Bíin…), Academia de la Lengua Maya de Yucatán, A.C., etc. By means of these elements, Mayan speakers Mayáon, A.C., and others, in 1984, an important point indicate their perceptions of the internal temporal as many recent publications and dictionaries, let alone components of a situation, without reference to older and colonial texts and materials, frequently do external criteria needed for tense distinctions such as not. The topics presented in the following segments present, past, and future. 663 MAYA AND MAYAN LANGUAGES Pronominal System and Ergativity accusative case marking pattern found in English, for example, the subject of transitive and intransitive The pronominal system of Mayan consists of three verbs is treated equally, that is, they are the same; it is sets of pronouns, commonly referred to as Set A, Set the object pronoun that is marked differently. In B, and the Independent Set. Set A pronouns in Yucatec Mayan, an ergative–absolutive family of languages, serve as subjects of transitive verbs, as subjects of the subject of intransitives and the object of transitives intransitive verbs in the incompletive aspect, and as are marked in the same manner, and different from the possessives. Set B pronouns serve as direct objects of subject of transitives. transitive verbs and subjects of intransitive verbs in the completive aspect. The Independent pronouns are used A ergative for emphasis or clarification at the beginning of a sentence, with topical marker -e’, also at the beginning nominative { S of a sentence or clause, and as indirect objects. Whereas the pronouns of Sets A and B are bound to an aspect particle or word, or to a verb, the Independent accusative O } absolutive pronouns are free of such attachments, and hence the Many of the Mayan languages have developed a split designation. ergative system when certain constructions follow the Set A pronouns ergative–absolutive pattern of the ancestral Common Singular Plural Mayan language, and others follow a nominative– accusative pattern. This split in the usage of pronouns 1st person in- k- is aspectual in nature, with an incompletive aspect 2nd person a- a-...-e’ex 3rd person u- u-...o’ob following the nominative–accusative configuration, and completive, the ergative–absolutive. Set B pronouns Numerical System and Numerical Classifiers Singular Plural The Maya developed a vigesimal numbering system, 1st person -en -o’on 2nd person -ech -e’ex that is, based on 20 rather than a decimal, base 10 3rd person -Ø -o’ob system. Unlike English, the numbers in Maya cannot be used in isolation. For example, in response to the question How many brothers do you have?, you Independent pronouns cannot say simply, One. The numbers are always used Singular Plural with a numerical classifier, denoting animacy, 1st person teen to’on measure, or shape. A similar phenomenon in English 2nd person teech te’ex can be seen with measurements and group 3rd person leti’ leti’o’ob classifications such as a pinch of salt, a flock of geese, and the like. A list of some of the more commonly The uses of Set A and Set B pronouns summarized used classifiers includes the following. above illustrate a linguistic phenomenon common to perhaps one quarter of the world’s languages, that of -cháach a handful of something ergativity. Ergativity is the “term used to describe a -chi’náab a measure from the thumb to the end of the forefinger grammatical pattern in which the subject of an -kúuch a load of something, typically carried on intransitive clause is treated in the same way as the your back object of a transitive clause, and differently from -kúul for plants transitive subject (Dixon, Ergativity, 1994:1)”. In -lúub a measure of distance of approximately English, this could be rendered as She see him and Him four kilometers -náab measure from thumb to end of little sees, and He sees it, where him is the direct object of the finger; a handspan transitive verb in the first example, and the subject of -p’éel for inanimate objects the intransitive in the second, complementary to the use -p’u’uk a mouthful of He as subject of the transitive in the third example. -sáap a measure of depth, equivalent to a fathom The following figure shows a representation of -téen/-púul number of times something is done -ts’áan a pair of things Dixon’s universal syntactic–semantic primitives, in -ts’íit long, thin shape which the letter A symbolizes the subject of transitive -túul for people, animals, etc. verbs, with S representing the subject of intransitives, -wáal for flat, thin objects, such as tortillas and O, the object of transitives. In the nominative– -xéet’ a piece of something 664  MAYA AND MAYAN LANGUAGES Phonology An additional element of Mayan phonology of note is the predominance of stops among both consonants Tozzer (1921), addressing issues of Mayan pronun- and vowels, a phenomenon whereby the flow of air ciation from one of the earliest grammars written on through the oral cavity is briefly interrupted. An Yucatec Mayan, that of Beltran de Santa Rosa dating example of a glottal stop in English can sometimes be back to 1746, says (translated): heard replacing the t sound in the word button, with Because in this language you don’t speak entirely like the flow of air momentarily restricted at the vocal you write, nor do you pronounce many words according cords. All five of the vowels in Mayan may to the rules (and hence, some who speak it sound like incorporate a glottal stop, as in ma’, no, and si’, foreigners or are judged as not speaking as they should; firewood. All of the vowels may be reduplicated, in being thus, that they speak in accordance with the rules which the vowel is interrupted by the stop and then of the grammar) be advised that the use of synalepha rearticulated, as in ba’alo’ob, things, and tu’ux, and syncope is so necessary, that without hyperbole one where. may state, that all of the being and attractiveness of this The occurrence of stops on the consonants ch, p, language is the use of them and the main part of its grammar is their explanation. and t, indicated in written form by the addition of an apostrophe, ch’, p’, t’, is said to be phonemic in Syncope is no other thing than swallowing some word Mayan, as the articulation of these stops can serve to or some syllable, some vowel or consonant. And this distinguish a word with them from another word figure so graces the Mayan tongue that without it it without, as in maak, to cover, and maak’, to eat seems that her words become strange, unpleasant and in quickly, chak, red, and ch’ak, to cut with an ax. her cadence, ugly. In such a manner, that I can without A final distinctive phonological element common recklessness say that almost one half of her words are to many of the Mayan languages is that of pitch. syncoped or are syncopable. Yucatec Mayan is a tonal language, with three Indeed, there are a number of phonetic changes in distinctive tones. One is long and high, and the other Mayan that clearly distinguish the written form from the is long and low, as in xuul, end, and xúul, planting spoken. Three such phenomena are syncope, the loss or stick, and miis, cat, and míis, broom. The third tone suppression of one (or more) sounds in the interior of a is realized on short vowels, and is considered word, synalepha, the fusion into one word from the final neutral. syllable of one word with the initial of the next, and It is sincerely hoped that the preceding essay has finally, apocope, the cutting off or omitting of the last provided you with a very brief but enticing sound or syllable of a word. Occurrences of each can be introduction to the Maya and their languages, and that seen in the following examples. as a result of the planting of this infinitesimal grain of linguistic sand, your interest and enthusiasm for linguistics grows and develops into your own unique Syncope ● The transitive verb with a pronominal object may lose the -i- of -ik in the pearl. incompletive and the -a- of -aj in the completive: Kin wilikech.
Kin References wilkech. ● All polysyllabic transitive verbs lose the Bricker, Victoria, Eleuterio Po’ot Yah, and Ofelia Dzul de Po’ot. vowel of the temporal endings before the 1998. A Dictionary of the Maya language as spoken in -o’on of the 1st person plural, -e’ex of the Hocabá, Yucatán. Salt Lake City, UT: University of Utah 2nd person plural and -o’ob of the 3rd Press. person plural: Tu ya’alajo’on.
Tu Coe, Michael D. 1996. The Maya. NewYork: Thames and ya’aljo’on. Hudson, Inc., 5th edition, 1993. ● In words of two syllables containing two Coe, Michael D. 1992 Breaking the Maya code, NY: Thames similar vowels, the second is lost when the and Hudson, Inc. plural is used, a verbal pronoun is used, or Comrie, Bernard. 1976. Aspect. Cambridge: Cambridge a demonstrative pronoun is used: xanab
University Press. xanbo’ob Dixon, R.M.W.1994. Ergativity, Cambridge: Cambridge Synalepha ● Contraction of durative and A pronoun: University Press. táan in
tin, táan a
tan Hofling, Charles Andrew. 1991. Itzá Maya texts with a ● Contraction of terminative and A pronoun: grammatical overview. Salt Lake City: University of Utah ts’o’ok in
ts’in, ts’o’ok a
ts’a Press. ● The negative ma’ and the nominal Landa, Diego de. 1975. Relación de las cosas de Yucatán; pronoun: ma’ in
min as Landa’s Relación de las cosas de Yucatán, a Apocope ● Word final -l is lost in most cases: Tun translation, translated by Alfred Tozzer. Millwood, NY, weenel
Tun weene. Kraus Reprint Co. ● j
Ø optionally in many cases Mayáon, A.C. 1984. Ts’iibjuumt’aan u ti’al u ts’íibta’al maaya t’aan (Alphabet for Writing Yucatec Maya). Felipe 665 MAYA AND MAYAN LANGUAGES Carrillo Puerto, Quintana Roo: Comité Ejecutivo de Sharer, R.J. 1947. The ancient Maya. Stanford: Stanford Quintana Roo. University Press, 5th edition, 1994. Anonymous. 1996. Popul vuh: the Mayan book of the dawn of Suárez Molina, Victor. 1945. El español que se habla en Yucatán: life translated by Dennis Tedlock. New York: Simon & Apuntamientos Filológicos. Mérida, Yucatán: Ediciones de la Schuster. Universidad Autónoma de Yucatán, 3rd edition, 1996. Robertson, John S. 1992. The history of tense/aspect/mood/ Tozzer, Alfred M. 1977. A Maya grammar, NY: Dover voice in the Mayan verbal complex. Austin: University of Publications, Inc. Texas Press. MICHAEL VROOMAN McCawley, James David An enthusiastic student, James David McCawley was Blackwell. He was a dedicated and accessible teacher, admitted to the University of Chicago in 1954 at age sharing his library, notes, and insights. The courses he 16. He took his MS in mathematics four years later. taught ranged from introductory linguistics courses, He was awarded a Fulbright Fellowship in 1959–1960 syntax and semantics, to ‘tense and aspect’, ‘discourse to attend Westfälische Wilhelms University in structure’, and ‘speech acts’ to ‘tone in Bantu Münster, Germany. Although the fellowship was to languages’ and ‘Japanese phonology’. His inventory study mathematics, McCawley’s interest began to be was so extensive that rather than attempting a term-to- drawn more toward languages. He studied a variety of term rotation of classes offered, Dr. McCawley would languages, including Dutch. After the Fulbright year, circulate a list of topics and ask students to choose. he returned to the University of Chicago, where he His stature as an internationally recognized scholar continued his exploration of languages, taking and his outstanding ability as a mentor won him a Japanese and falling in love with it. During this named chair at the University of Chicago. He became period, he read Noam Chomsky’s seminal work the Andrew MacLeish Distinguished Service Syntactic structures. Linguistics as a field of study Professor of Linguistics and of East Asian Languages beckoned. He applied to MIT and was accepted in and Civilizations. their first Ph.D. class in Linguistics. In 1965, he He was widely sought after as a speaker and as a completed his dissertation under the direction of guest faculty member. He accepted visiting or Noam Chomsky: The accentual system of modern temporary appointments in 17 Universities, including standard Japanese. the University of Illinois, University of Gööteborg, McCawley returned to the University of Chicago, Australian National University, University of Hawaii, where he received tenure in 1969 and became a full University of New Mexico, Central Institute of professor in 1970. In the summer of 1971, he met then English and Foreign Languages (Hyderabad, India), Ph.D. candidate Noriko Akatsuka. After a whirlwind Guangzhou Institute of Foreign Languages, and courtship, during which Dr. McCawley went to Japan National Tsing Hua University (Hsinchu, Taiwan). to formally ask for her hand, they were married on He served the Linguistic Society of America in a Beethoven’s birthday, December 16, 1971. Dr. number of positions: as a member of the Nominating Akatsuka joined Dr. McCawley at the University of Committee (January 1974 to December 1976; Chicago. Chairman 1976); as a member of the Executive Dr. McCawley’s interests were broad-ranging. He Committee (January 1978–December 1980); as Vice- published on elements of phonetics and phonology as President (1995); and as President (1996). He became well as on syntax and semantics. His early training in a member of the American Academy of Arts and mathematics put him at ease with the abstract formal Sciences in 1983, and served a term as its President. Dr. systems of syntactic form and semantic integration. McCawley was associate editor of five professional One of his most influential works attempted to make journals; consulting editor for four; and on the editorial these mysteries accessible to his colleagues and board of another five. students: Everything that linguists have always Dr. James D. McCawley brought joy and enthusiasm wanted to know about logic (but were ashamed to ask). to his study of linguistics. He regularly hosted parties to Chicago: University of Chicago Press, and Oxford: celebrate linguistics with students and colleagues. He 666  MCCAWLEY, JAMES DAVID championed Hanggul Day, the day the Hanggul Institute of Technology in 1965 for his thesis on ‘The alphabet was officially adopted by Korean, as the accentual system of modern standard Japanese’. He linguistic holiday. He composed linguistic doggerel for worked as a translator of Russain math books for the University of Chicago departmental holiday University of Chicago 5–9/59, 8/60–1/61; for the commemorating the movement of the offices from mechanical translation group at MIT 1–7/62, 2–6/63; as Goodspeed Hall, annually celebrated with songs and a teaching assistant for German at MIT 9/62–1/63; and counterscholastic papers. James D. McCawley had a as a research linguist for IBM Watson Research Center, famous alter ego, Quang Phúc Ðông, who produced fine Yorktown Heights, NY, 6–9/64. He took up a position as studies in scatological linguistics. His students honored Assistant Professor of linguistics, University of him with a festschrift in this little developed genre: Chicago, 9/64 and held this position until 9/69 when he Studies out in left field: defamatory essays presented to advanced to associate; then, in 9/70, he became a full James D. McCawley on the occasion of his 33rd or 34th professor of Linguistics. He was appointed the Currently birthday. Andrew MacLeish Distinguished Service; during his In addition to linguistics and music, Dr. McCawley tenure at Chicago, he held visiting appointments at had a passion for good food. He was a gourmet and an Tokyo Center for Advanced Studies in Language, excellent cook. A condition for guest lectures was 6–9/67; University of Illinois (Linguistic Institute), often an eater’s tour of the host city. He shared his 6–8/68; Scandinavian Summer School of Linguistics, knowledge of and pleasure in Chinese cuisine through 8/69;University of Göteborg, 9–12/69; University of his book: The eater’s guide to Chinese characters. Michigan, 1–4/70; Fifth International Seminar on Dr. McCawley was also a staunch Libertarian. He Theoretical Linguistics, Tokyo, 8/70; University of ran as the Libertarian candidate for the Trustee of California at Santa Cruz, 6–8/71; Australian National University of Illinois in 1976, 1978, and 1980. University (Canberra), 6–8/73; University of Hawaii Throughout his career, he refused to accept federal (Linguistic Institute), 7–8/77; University of Illinois at funding for his research or to review other scholar’s Chicago, 1–3/78; University of New Mexico (Linguistic proposals for such studies. Institute), 6–8/80; Central Institute of English and James D. McCawley is known for his pioneering Foreign Languages (Hyderabad, India), 7–8/80; work on the interface of syntax and semantics, for his University of Maryland (Linguistic Institute), 7–8/82; skepticism of theory, and for his demand for rigorous Georgetown University (Linguistic Institute), 6–8/85; treatment of data. His publication record is Linguistics, Institute on Transformational-Generative outstanding. A full list can be found on his web-page: Grammar, University of Delhi, 8–9/85; Guangzhou http://humanities.uchicago.edu/depts/linguistics/facult Institute of Foreign Languages, 9/87; University of y/mccawleycv.html He set the framework for studies Arizona (Linguistic Institute), 6–8/89; and National of tonal and near-tonal systems in works such as Tsing Hua University (Hsinchu, Taiwan), 2–6/94. He ‘Some tonal systems that come close to being pitch served as editor for 15 journals, and was president, vice- accent systems but don’t quite make it’ (1970) and president, and held other posts within the Linguistic ‘What is a tone language?’ (1978). He provided a clear Society of America; May 1983, he was appointed fellow critique of syntactic theory in works such as (1982) of American Academy of Arts and Sciences. He was Thirty million theories of grammar and (1988) The also three time Libertarian candidate for Trustee of the syntactic phenomena of English. At the time of his University of Illinois. In 1991, he received an honorary death, Dr. McCawley had several works in progress, Ph.D. from University Göteborg, He authored 225 including two books. His students have prepared him scholarly articles and seven books; he was also another festschrift. His lasting contributions, both in dedicatee of Zwicky, Salus, Binnick, and Vanek (eds.), print and through the impact of those he has trained in Studies out in left field: defamatory essays presented to the field, form a modern linguistics respectful of data James D. McCawley on the occasion of his 33rd or 34th and meticulously formal. birthday (Edmonton: Linguistic Research, 1971; reprinted 1992, Amsterdam and Philadelphia: Benjamins), and of Brentari, Larson, and MacLeod Biography (eds.), The joy of grammar (Amsterdam and James David McCawley was born James Quillan Philadelphia: Benjamins, 1992); and one forthcoming. McCawley, Jr in 1938, in Glasgow, Scotland; post- McCawley died on April 10, 1999. World War II, he emigrated as a minor to the United States, acquired citizenship, and changed his name to James David McCawley. He earned an M.S. in References mathematics from University of Chicago in 1958 and Chomsky, Noam. 1957. Syntactic structures. The Hague: earned a Ph.D. in linguistics from Massachusetts Mouton and Company. 667 MCCAWLEY, JAMES DAVID McCawley, James D. 1970. Some tonal systems that come McCawley, James D. 1982. Thirty million theories of grammar. close to being pitch accent systems but don’t quite make it. Chicago: University of Chicago Press. Papers from the 6th regional meeting, Chicago Linguistic McCawley, James D. 1984. The eater’s guide to Chinese Society, 526–32. characters. Chicago: University of Chicago Press. McCawley, James D. 1978. What is a tone language. Tone: a McCawley, James D. 1988. The syntactic phenomena of linguistic survey, ed. by V. Fromkin, 113–31. New York: English. Chicago: University of Chicago Press. Academic Press. Zwicky, Arnold, et al. 1971. Studies out in left field: defamatory McCawley, James D. 1981. Everything that linguists have essays presented to James D. McCawley on the occasion of always wanted to know about logic (but were ashamed to his 33rd or 34th birthday. Edmonton, Alberta: Linguistic ask). Chicago: University of Chicago Press and Oxford: Research. Blackwell. JUDITH M. MAXWELL Meaning The linguistic study of meaning has traditionally meaning always involves an idealization and concentrated on investigating conceptual or logical generalization of the meanings of the language code, meaning, known as ‘sense’. However, logical or abstracted from particular contexts. Such an conceptual meaning is not the only kind of meaning idealization was thought necessary in order to make that sentences or utterances convey. In terms of child any progress in studying linguistic meanings development, the interpersonal meanings of language scientifically and logically. However, it is apparent are primary. And utterances might tell you who the that semantics on its own cannot give anything like a speaker/writer is (idiosyncraey), their age or when full meaning to a sentence; in other words, semantics they were speaking (age), what activity they are underdetermines meaning. One most obvious problem engaged in (occupation), where they come from area is deictic terms (from deixis ‘pointing’), whose (provenance), their relationship with the hearer/reader meaning depends on knowing who uttered them (to (status), and whether they are speaking or writing or whom), when they were uttered, and where they were combinations of these, such as reading aloud (mode) uttered. The pronouns I and you, and the adverbs now, (Crystal and Davy 1969). It is clear that all these are yesterday, and here are clear examples. But deixis is social and interpersonal meanings, except the last, much more widespread than its occurrence in a few which is textual. isolated lexical items. Most tense markers or aspectual particles are time deictics. Past, present, future, and more complex tenses refer to periods of time relative Semantic and Pragmatic Meaning to the time of utterance. ‘I saw him here yesterday’ is The distinction between interpersonal and conceptual incomplete in meaning unless we know the time of meaning is in partial correspondence with the utterance so that ‘saw’ and ‘yesterday’ refer to a period distinction between pragmatic and semantic meaning. of time previous to that time. In addition, we have to Pragmatics answers the question ‘What did the know where the speaker is to understand ‘here’, and speaker A mean by uttering sentence X to hearer B in who the speaker is to understand the referent of ‘I’. context C?’, whereas semantics answers the much A related aspect of the pragmatic/semantic simpler question ‘What does sentence X mean?’ In distinction is the difference between reference and pragmatics, meaning is a matter of what effect the sense (Frege 1892). Reference is the relationship speaker intends to produce in the hearer by making an between a phrase and the thing in the world that it utterance, which makes it interpersonal. Semantics has identifies. The sense, however, is the concept to do with the meanings of sentences and how these represented by the meaning of the phrase. The meanings can be computed by the meanings of their versatility of language lies partly in the fact that any component parts (clauses, phrases, words, and one sense can be used to refer to many different morphemes) (Leech 1983). entities. Woman, whose sense is, let us say ‘adult Of course, we naturally encounter and learn female’, might be used to refer to any one of the two meanings in pragmatic and social contexts, especially billion women in the world. Just as the relationship in our first language. Investigation of semantic between sense and referent is one-to-many, so is the 668  MEANING reverse relationship between referent and sense––in institutional contexts. Each context has certain other words, there are many different phrases with meaning potentials––likelihoods that certain meanings different senses that could be used to refer to the same will occur in them. individual: The former president of the United States, Ideology can be seen operating directly in the the man from Arkansas, this consummate orator, the meanings of words and clauses. According to sexual partner of Monica Lewinsky. Benjamin Lee Whorf (1956) and his theory of While reference in normally conventional, in that a linguistic relativity, the world does not come to our particular phrase guides us quite clearly to identify the consciousness in ready-made categories, but different referent, this is not the case with metaphor––where the languages dissect the continuity of our experience in reference is unconventional. And names have no or different ways. Whorf’s hypothesis, unfashionable for little sense—we simply have to learn the pairing many years due to the influence of universalism in between each individual person or thing and the linguistics, has lately been rehabilitated to some extent language form that refers to them. (Lee 1996; Gumperz and Levinson 1996). This tradition, taken over later by Critical Discourse Analysis, emphasizes that language and its meanings Coded and Inferred Meanings are not a transparent medium through which reality is Words or lexical items can be seen as signs making up observed, but distort and construct that reality, giving the linguistic code. These signs consist of a signifier us the concepts through which we think. The (phonological or graphological form) and signified languages we speak hand down to us ready-made (meaning), like two sides of a piece of paper categories, which we regard as commonsense. They (Ferdinand de Saussure 1960). In general, these thereby unconsciously carry with them an ontology or pairings are thought to be arbitrary; otherwise, ideology of which we may not be aware. languages would resemble each other much more The classic case is of kinship categories. In Thai, (there is nothing about the form dog that makes it more for example, the primary criterion for categorizing suitable than the French chien to act as a signifier of siblings is by seniority rather than by sex as it is in ‘dog’). Hence, in terms of semiotics (the study of English; elder sibling /phî:/ and younger sibling signs), most linguistic signs are symbols (involving no /nó:ŋ/. And this difference in ontology has ideological resemblance between signifier and signified) rather implications. Senior siblings in Thai culture have than icons (based on resemblance) (Peirce 1940:104). greater rights and responsibilities than in most Western This arbitrariness is not absolute, however. Some cultures, traditionally giving orders and making signifiers are obviously iconic, based on ono- demands of younger siblings, but also taking care of matopoeia, e.g. splash, and at a more subtle level non- them and their welfare. Furthermore, Thais use these English speakers can guess quite reliably which word words as terms of address and second person pronouns tin or thunder is most suitable as a signifier for the with friends and acquaintances as well as with blood signified ‘thunder’. siblings, so that in a friendly conversation with a new However, linguistic communication in any acquaintance, when it is not obvious from appearances language is also heavily reliant on non-coded which of the conversational partners is the older, at meanings, created by pragmatic inference, and some some point the question has to be asked about relative languages are perhaps more reliant on inferences from age, to facilitate polite discourse and correct context than others. Chinese, for example, without a interpersonal behavior. highly developed system of tense for time reference, More generally, we may just not take much notice leaves it to be inferred from the context. of things if we have no words for them in our language. In Chinese, there is a common word for the groove that runs between the middle of the upper lip Meaning, Social Context, and Ideology and the nose. There is no such word in English. While some inferences and interpretations might Presumably, this contrast reflects the unimportance of depend upon supplying general knowledge, many this aspect of reality to English speakers, and its meanings are dependent more specifically on importance to folk physiognomy in China. awareness of the context of utterance. Apart from Sometimes, of course, ideology is more obviously conversation and narrative, discourse types generally encoded in words and then they may become the site occur in specific institutional contexts––lectures in an of ideological struggle. Was the movement of white academic institution, sales reports in a business people into Australia or Israelis into Palestine context, golf score-cards at a golf course, and so on. ‘settlement’ or ‘invasion’? These words and their The concepts of register and genre are an attempt to meanings then become contested terms. There can systematically link discourse types with their often be attempts to redefine the meanings of terms as 669 MEANING in ‘property is theft’. Alternatively, one could try to head. He illustrates this with the sentence ‘That Judas replace a fashionable term with another that better perished by hanging himself there is no certainty’, suits one’s ideology. For example, economic growth which begins with the presupposition ‘Judas perished might be replaced with cancer, since from an by hanging himself’, leading us to anticipate that the ecological standpoint, growth in mature economies sentence will end with the word doubt, but reversing actually harms the environment and threatens our life- our expectation of certainty by ending with the word support systems. In discourse, there is often an ‘certainty’! However, Fish showed that it is not the ideological struggle to get one’s meanings and subjective meanings of the individual reader that metaphors accepted as the conventional ones. constitute either what is literature or how it should be read, but rather the inter-subjectivity of the interpretive community, into which readers are inducted by Structuralism and Deconstruction learning the conventions of how to read and make A parallel emphasis on the determination of meaning meaning. by code has been found in studies of literary meaning. Bakhtin too, emphasized the intersubjectivity of Structuralism emphasized that a speaker or writer does meaning by insisting that it only exists in dialog. not impart meaning to utterances, but that the code, the Meaning is like a spark between two electric terminals linguistic system underlying them, produces it. This (the speakers), and hence ‘word is a two-sided act. It is conviction led to literary analysis aimed at revealing determined by whose word it is and for whom it is the underlying structures, which, rather like the meant’. Intertextually, words bear the traces of their grammar or morphology of the linguistic code, past uses by previous speakers, and language is determine the shape of the texts that realize them (see, therefore never in one’s own words but other people’s for example, Vladimir Propp’s (1958) Morphology of words: ‘the trouble with words is that you don’t know the folk tale). whose mouth they’ve been in’. These uses and traces Structuralism gave way to deconstruction under the tend to be evaluative and ideological, part of social influence of Roland Barthes and Jacques Derrida. struggle. It follows from this view that ‘the analysis of Barthes (1970) emphasized the radical ambiguity of meaning should be based not only, and even not as the sign as part of the code and the play of signifiers. much, on linguistics as on metalinguistics which For him, literary meaning could and should not be studies the word not within the system of language and constrained within the concept of unity as not in a text which is removed from dialogical structuralism tried to pretend. Barthes valued literature intercourse, but precisely within the sphere of that was less than easy to read, less than lisible, dialogical discourse itself’. Bakhtin’s theories raise because it breaks with the coded conventions. In this radical questions about the distinction between type of writable or scriptible literary text, the ‘code’ analytic (definitional) and synthetic (descriptive) items are echoic, intertextual, fragments of voices meaning and the tension between meaning stability from other texts and from other codes. Derrida (1967) and meaning change. with his notion of differance (difference/deferring), suggested that signifiers have shifting signifieds Meaning, Metaphor, and the Experiential whose meaning is never complete, because the text at Hypothesis hand always has reverberating connotations and each use of a text further develops its meaning. More Another alternative to subjectivity can be found in the extremely, for him, language became a play of experiential hypothesis that grew out of studies of signifiers without any signifieds––with nothing metaphor. In its infancy, during the heyday of existing outside the text, and no meaning or truth Chomskyan linguistics, the science of semantics independent of language (Jefferson and Robey 1982). attempted to marginalize metaphor as an anomaly, because nonliteral meanings posed challenges to semantic rules like selectional restrictions––the rules Reader Response, Dialogism, and that stipulate what semantic pairings were allowable Intersubjective Meaning between phrases in a syntactic relationship. Noam Less extreme than Derrida’s emphasis on radical Chomsky held his famous sentence ‘colourless green instability and subjectivity of meaning are the theories ideas sleep furiously’ to be semantically ill-formed of Stanley Fish (1980) and Mikhail Bakhtin and because sleep should be restricted to human or animal Valentin Volosinov (1973). For Fish, meaning is ‘an subjects, green is only allowed to premodify concrete, action made upon a reader rather than a container from not abstract, things, and so on. However, it is possible which a reader gets a message’. Meaning is the for an inventive reader to interpret this sentence experience of the process of reading in the reader’s metaphorically: ‘Although ideas about protecting the 670  MEANING environment are boring, ignoring and suppressing Existence and fact are independent of belief, them will only delay an angry social reaction.’ knowledge, perception, modes of understanding, and Besides such challenges, George Lakoff and Mark every other aspect of human cognitive capacities. No Johnson (1980) made clear that metaphor could not be true fact can depend on people believing it, on their ignored as it was all-pervasive, even in the language of knowledge of it, on their conceptualization of it, or on any other aspect of cognition. (Lakoff 1987:164) those who try to reject it. For example, John Locke in this passage uses the metaphors ‘move’, ‘mislead’, Lakoff cannot accept this assumption, because, and ‘cheat’. although there is a reality out there, we do not have unmediated access to it, as our thinking is inescapably But yet, if we would speak of things as they are, we must mediated by the metaphors we use. Even scientific allow that ... all the artificial and figurative application theories and models are basically metaphorical of words eloquence hath invented, are for nothing else but to insinuate wrong ideas, move the passions, and hypotheses that can only approximate descriptions of thereby mislead the judgment, and so indeed are perfect reality, and a new scientific model will initiate a cheat. [Essay concerning human understanding Book 3, programe of research trying to establish what features Chapter 10, p.105] of the model do or do not apply. On the other hand, Lakoff rejects subjectivism and relativism, because The claims of Lakoff’s conceptual metaphor theory infant bodily experiences are universal; hence, our (Lakoff 1987) are that abstract thought is fully conceptual metaphors are held in common. possible only through the use of metaphor, and that Experientialism, then, steers a middle course between these conceptual metaphors originate in our bodily objectivist philosophical traditions, which assume that infant experiences. We have certain preconceptual truth is something we can access independent of any experiences as infants, such as body movements, our description of it, and subjectivism, which believes that ability to move objects, to perceive them as wholes truth is a matter of individual belief, relative to and retain images of them; and certain image schemas circumstances. However, unlike Critical Discourse that recur in our everyday bodily experience, e.g. analysts, Lakoff sees the body, not culture and containers, paths, balance, up and down, part and ideology, as the determinant of thought and ontology. whole, and front and back. The hypothesis claims that Objectivist views of meaning disregard the idea that most abstract concepts arise from these preconceptual ideology and meaning exist intersubjectively in a physical experiences and schemas by metaphorical dialectical relationship with each other, or that our projection. For example, abstract concepts like amount physical experience of embodiment determines thought. are conceptualized by projection from the bodily According to Lakoff, the objectivist hypothesis claims experience of up and down, producing a number of that reality and fact are independent of observation or lexicalized metaphors: perception, or, for that matter, language. Within the objectivist paradigm of semantics, there arose a truth- The number of books printed each year keeps on going conditional approach to meaning: knowing the meaning up. My income rose/fell last year. of a sentence is equivalent to knowing the conditions If you’re hot, turn the heat down. under which that sentence would be true. Clearly, if He is underage. language constructs reality rather than merely reflecting it, the truth-conditional theory of meaning becomes Many of the basic links in conceptual metaphors circular to some extent. This is related to another can be traced back to metonymic relations such as problem, which is that truth–conditional semantics has cause and effect. Hence, for example, if we have more tended to restrict itself to analytical statements rather things in a pile it will be higher than when we have than synthetic statements. A synthetic statement is one less things, a possible origin of this set of lexical that is contingently true or untrue, such as ‘Beckham items. And there is a well-established set of transferred to Real Madrid’, whereas an analytical metaphorical vocabulary in English, and other statement is one that is necessarily true, such as ‘boys are languages, which conceptualizes anger as heat. The male’. It is doubtful whether theories of meaning, which origin of this is quite obviously one of cause and exclude synthetic statements, can be useful models for effect––when we become angry, our bodies do the meaning of natural languages in real–world contexts. actually rise in temperature. In the light of the lexical evidence for this References hypothesis, Lakoff went on to develop an Bakhtin, Mikhail, and Valentin Volosinov. 1973. Marxism and Experientialist philosophy, rejecting the Objectivist the philosophy of language. Cambridge, MA: Harvard and Subjectivist paradigms. The problem with University Press. Objectivism is the independence assumption, namely: Barthes, R. 1970. S/Z. Paris: Seuil. 671 MEANING Crystal, D., and D. Davy. 1969. Investigating English style Lakoff, George, and Mark Turner. 1989. More than cool reason: Harlow: Longman. a field guide to poetic metaphor. Chicago: University of Derrida, Jacques. 1967. La voix et la phenomene. Paris: Seuil. Chicago Press. Fish, Stanley. 1980. Is there a text in this class?. Cambridge: Lee, Penny. 1996. The Whorf theory complex. Amsterdam: Cambridge University Press. Benjamins. Frege, Gottlob. 1892. Uber Sinn und Bedeutung. Zeitschrift fur Leech, G.N. 1983. Principles of pragmatics. Harlow: Longman. Philosophie und philosoph. Kritik 100 25–50. Peirce, C.S. 1940. The philosophy of Peirce: selected writings, Goatly, Andrew. 1997. The language of metaphors. London: ed. by J. Buchler. London. Routledge. Propp, Vladimir. 1958. The morphology of the folk-tale. Gumperz, John, and Steven C. Levinson. (eds.) 1996. Bloomington: Indiana Research Centre in Anthropology. Rethinking linguistic relativity. Cambridge: Cambridge Saussure, F. de. 1960. Course in general linguistics. London: University Press. Peter Owen. Halliday, M.A.K., and Ruqaiya Hasan. 1985. Language, Thomson, John B. 1984. Studies in the theory of ideology. context and text. Geelong: Deakin University Press. London: Polity Press. Jefferson, Ann, and David Robey. (eds.) 1982. Modern literary Whorf, B.L. 1956. Language thought and reality: selected theory. London: Batsford. writings of Benjamin Lee Whorf, ed. by John B. Carroll. Johnson, Mark. 1987. The body in the mind. London: Cambridge, MA: MIT Press. University of Chicago Press. ANDREW GOATLY Lakoff, George. 1987. Women, fire and dangerous things. Chicago: University of Chicago Press. Lakoff, George, and Mark Johnson. 1980. Metaphors we live See also Context; Ideology and Language; by. Chicago: University of Chicago Press. Metaphor Medicine and Language The description of language in medical encounters is investigation (treatment and advice); and (VI) one of the oldest and most prominent topics in discourse termination (small talk and closing). Each part of the analysis (the study of the sequences and organization of encounter is associated with conventional discourse language in context). The topic is interesting for sequences: question and answer in the history and theoretical and applied reasons: Theoretically, the physical, for example, and imperatives in the delivery description of language in medical encounters of treatment and advice. Byrne and Long found, contributes to our understanding of institutional however, that the discourse of the medical encounter is interaction, symmetrical and asymmetrical roles and highly asymmetrical: it is the physician who relationships as created and reflected by discourse, and interactionally controls most of the discourse. The specialized sequences within the interaction of medical physician asks the questions, controls the topics and encounters (Drew and Heritage). Practically, the their development, deflects or ignores patient topics or description of language in medical encounters allows contributions that he or she deems irrelevant, provides linguists to help medical professionals improve the amount of medical information that he or she communication and to help patients and families work deems appropriate, and determines the amount of effectively with medical professionals. social talk in openings and closings. The institutional In Western industrialized societies, the medical power and authority of the physician, as well as the encounter has a recognizable event structure. One of relatively powerless institutional position of the the first articles on the medical encounter by British patient, then, are created, reflected, and maintained by researchers Byrne and Long examined more than the asymmetrical discourse practices of the encounter. 2,000 medical encounters and identified the typical The asymmetry of a medical encounter––the control parts of an encounter: (I) relating to the patient of the physician over the discourse––was a topic of (greetings and small talk); (II) discovering the reason much investigation in early research on language in for the encounter (the patient complaint); (III) medicine. Some of these asymmetries were troubling conducting a verbal or physical examination, or both because they were seemingly related to issues of gender, (the history and physical); (IV) consideration of the class, and ethnicity. For example, a number of studies patient’s condition (delivery of diagnostic opinion and described the ways physicians ignored topics mentioned information); (V) detailing treatment or further by female patients, especially when those topics moved 672  MEDICINE AND LANGUAGE away from the medical world and toward the patient’s presentational sequences used in the faculty clinic personal world with its complicated psychological and (note that the contribution from the patient social conditions (Mishler; Waitzkin). Paget argued that significantly affects the course of the discourse): sequences like the following show how female patients’ topics can be ignored by physicians who are in control (4) PT: Have a hysterectomy and that, I’m that, if there’s an alternative. I’m terrified of of the discourse: operations. (1) PT: (sigh) I dunno. DR: Uh, okay, well, there certainly is an alternative, I’m thinkin maybe it’s a hormone deficiency or yeah, we can treat this by just freezing it here something. in the office and that usually will take care of DR: Let me just look at the scar. it about 90% of the time. No-no that’s awright. Specific discourse practices such as questioning and During the interaction of physicians and female patients, the presentational or persuasional organization of these kinds of asymmetries are found in topic control information can thus have a direct effect on decision- and development (Davis). Other studies noted the ways making in the medical encounter. The troubling that medical care seems to be practiced differently when implication of this particular study was that the patients are from different social classes and ethnicities. discourse of medical encounters seemed to differ on Fisher studied encounters in two different gynecological the basis of social class. clinics: a community clinic staffed by resident doctors- Much of this early research on the asymmetrical in-training and attended by mostly poor minority women discourse of the medical encounter focused on and a faculty clinic staffed by university physicians and question–answer sequences. Several classic studies attended by mostly middle-class white women. She showed how physicians control the discourse of the found that women in the community clinic received encounter by using questions that constrain patients to more treatment recommendations for hysterectomies brief answers. One study looked at the discourse of ten than women in the faculty clinic and, in fact, were not medical encounters and found that patient-initiated told that options other than hysterectomy were available. questions (that is, utterances in question form that In the decision-making sequences in the faculty clinic, introduced new information or a new topic) comprised physicians often used questioning strategies: for less than 1% of the discourse in a medical encounter example, a question such as Now did Karen explain to (Frankel). Another study also found that physicians you the abnormal, what the abnormal Pap smear established question–answer discourse almost business is? can function to give patients a chance to immediately after the beginning of the encounter: in this display their knowledge and competence as background study, physicians typically interrupted the patient’s for collaborative decision-making. Patients, too, in the account of the reason for the visit to begin asking faculty clinic used questioning strategies: for example, a questions after only 6 to 16 seconds (Beckman and question such as Is that necessary? opens up the Frankel). A third study defined questions more broadly as discourse to the discussion of alternative treatment requests for information in any form but still found that options. Physicians and patients in the community clinic, patient questions comprised only 9% of the discourse in however, often did not use questioning strategies, the 21 medical encounters that were analyzed (West). thereby reducing the chances that the discourse of the These early studies of the language of the medical encounter would develop into collaborative decision encounter were focused primarily on the ways making. The discourse in the community clinic was physicians used discourse to maintain asymmetrical more oriented to persuasional strategies justifying the control of the discourse, with particular attention to the physician’s choice of treatment, as in the following way that question–answer sequences turned the medical sequences (note that there are no contributions from the encounter into an interview (rather than a more patients in these sequences): symmetrical consultation or discussion). However, (2) DR: It (cancer) could come back…and for recent research in conversation analysis has been somebody your age, that’s had your family, focused on the ways that both physicians and patients you’re sure that you don’t want children, I’d actively and collaboratively construct the interaction of recommend a hysterectomy. the medical encounter as a type of institutional (3) DR: What you should do if you don’t want any discourse (ten Have; Maynard). Sometimes, this more children is have a hysterectomy. No construction achieves the asymmetrical discourse of the more uterus, no more cancer, no more babies, encounter as described above, with both physicians and no more birth control, and no more periods. patients working interactively to produce asymmetrical These persuasional sequences, with no mention of discourse. In a study of general practice encounters, for other treatment options, contrast with the example, Heath found that patients rarely responded to 673 MEDICINE AND LANGUAGE the delivery of a diagnostic opinion by a physician, not contextual dimensions such as power and expertise. contributing an utterance at all after the delivery of One study looked at the ways that both physicians and diagnostic news, or offering only a minimal response: patients made claims to power over the discourse of medical encounters (Ainsworth-Vaughn). In encounters (5) DR: Er, yes, (0.3) this one’s blocked, the other one’s in private practice between physicians and patients with not. long-standing relationships, the patients asked almost PT: (1.2) (6) DR: You’ve got, erm (0.8), bronchitis. 40% of the questions, thereby controlling the discourse PT: Er to a significant degree. An early study by Tannen and DR: (4.5) (doctor begins to write prescription) Wallat looked at the ways in which physicians construct I’ll give you antibiotics to take for a week. frames for the delivery of medical information based on professional expertise. Two recent studies looked at the In both sequences, the physicians offer the patients an ways in which physicians acknowledge and utilize lay opportunity to respond to the diagnostic news (note the expertise in the discourse of medical encounters as well: pauses, with the number of seconds timed in in a children’s disability clinic, families who display lay parentheses), but the patients do not take the expertise about the condition of their children were able interactional opportunity to develop the topic. This to move the discourse toward more collaborative kind of minimal response preserves the asymmetrical decision-making, while families who do not display interactional and institutional power and role of the what physicians deem an appropriate lay expertise physician as the expert provider of diagnostic concerning their children’s condition had medical information. Patients, however, have their own encounters with a traditionally asymmetric discourse interactional means of questioning or disputing a (Barton). These contextual dimensions are thus physician’s diagnostic news. Although they typically interactionally dynamic in their effect on the discourse remain silent or contribute only minimally to the of the medical encounter. discourse right after the delivery of diagnostic news, Analysis of the discourse of the medical encounter patients recycle the topic of diagnosis when they provides important information for the description disagree with it. For example, a patient may repeat of institutional interaction in general and symptoms, emphasize their severity, or question the physician–patient interaction in particular. This diagnosis during a different part of the encounter, often research has been used in a variety of applications. during the treatment recommendations or closing. In The curriculum of many medical schools now includes the following sequence, for example, the patient information about medical communication that draws initially had not responded to the delivery of diagnostic on linguistic research. Community outreach programs, news, but then produced an account of the severity of such as support groups, also draw on linguistic his symptoms after the physician had turned to the research to introduce families to ways of communi- treatment activity of writing a prescription: cating effectively in the institutional context of (7) PT: That I could understood (.) because it (.) medicine (e.g. asking questions about diagnostic it’s the headaches was the thing that’s labels, alternative treatments, and so on). Even got me, (0.4) More than anything else. consumer publications about medical communication (1.2) More than the devil in hell because have begun to draw on the research on medical they were getting more or less discourse to advise patients and families in their permanent yer know. (1.2) They were coming efforts to work toward establishing collaborative even when I was never—pain in the back of discussion and decision-making in medical care. me neck. (28.00) DR: Right well I’ll tell what we’ll do Mister Tarrett. References Through their own interactional means, then, patients Ainsworth-Vaughn, Nancy. 1998. Claiming power in are able to question or challenge the physician’s doctor–patient talk. New York and Oxford: Oxford. Barton, Ellen. 1999. The social work of diagnosis: evidence for diagnostic opinion, even when they initially appear to judgments of competence and incompetence. Constructing have accepted it asymmetrically. A similar pattern was (in)competence: disabling evaluations in clinical and social found in another study of the delivery of diagnostic interaction, ed. by Dana Kavarsky, Judith Duchan and news in a child assessment clinic: patients and families Madeline Maxwell. Mahwah, NJ: Lawrence Erlbaum. did not initially dispute the diagnostic opinion, but Barton, Ellen. 2000. The interactional practices of referrals and accounts in medical discourse: expertise and compliance. they found interactional ways to return to the topic if Discourse Studies 2. they disagreed with the diagnosis (Maynard). Beckman, Howard, and Richard Frankel. 1984. The effect of Recent ethnographic research has looked at the ways physician behavior on the collection of data. Annals of that the discourse of physicians and patients reflects Internal Medicine 101. 674  MEILLET, ANTOINE Byrne, Patrick, and Barrie Long. 1976. Doctors talking to Heath, Christian. 1992. The delivery and reception of diagnosis patients: a study of the verbal behaviours of doctors in the general-practice consultation. In Drew and Heritage. consulting in their surgeries. London: HMSO. Maynard, Douglas. 1991. Interaction and asymmetry in clinical Davis, Kathy. 1988. Power under the microscope. Dordrecht, discourse. American Journal of Sociology 97. Holland: Foris. Maynard, Douglas. 2002. Bad news, good news: conversational Drew, Paul, and John Heritage (ed.) 1992. Talk at work: order in everyday talk and clinical settings. Chicago: interaction in institutional settings. New York and University of Chicago Press. Cambridge: Cambridge University Press. Mishler, Elliot. 1984. The discourse of medicine. Norwood, NJ: Fisher, Sue. 1993. Doctor talk/patient talk: how treatment Ablex. decisions are negotiated in doctor-patient communication. Paget, Marianne. 1993. On the work of talk: studies in In Fisher and Todd. misunderstandings. In Fisher and Todd. Fisher, Sue, and Alexandra Todd (eds.) 1993. The social Silverman, David. 1987. Communication and medical practice. organization of doctor–patient communication, 2nd edition. London and Thousand Oaks, CA: Sage. Norwood, NJ: Ablex. Tannen, Deborah, and Cynthia Wallat. 1993. Doctor/mother/ Frankel, Richard. 1979. Talking in interviews: a dispreference for child communication: linguistic analysis of a pediatric patient-initiated questions in physician–patient encounters. interaction. In Fisher and Todd. Everyday language: studies in ethnomethodology, ed. by Waitzkin, Howard. 1991. The politics of medical encounters: George Psathas. New York: Irvington. how patients and doctors deal with social problems. New ten Have, Paul. 1989. Talk and institution: a reconsideration of Haven, CT: Yale University Press. the ‘asymmetry’ of doctor–patient interaction. Talk and West, Candace. 1984. Routine complications: troubles with talk social structure: studies in ethnomethodology and between doctors and patients. Bloomington: Indiana conversation analysis, ed. by Deidre Boden and Don University Press. Zimmerman. Berkeley: University of California Press. ELLEN BARTON Meillet, Antoine Antoine Meillet’s intellectual and scientific life was the linguistic relationship between Greek and Latin but organized around two fundamental axes: comparative also detail the development of both the Greek and the linguistics and general linguistics. Up until 1906, his Latin linguistic system. publications showed his interest in philology in the Ferdinand de Saussure’s work Mémoire sur le lineage of Ferdinand de Saussure, as well as his wish système primitif des voyelles dans les langues indo- to describe and analyze culture in the semantic européennes (Report on the primitive system of perspective of Michel Bréal. After 1906, Antoine vowels in Indo-European languages) (1878) made Meillet became more familiar with the linguistic such an impression on Meillet that he used this work theories of the French sociologist Emile Durkheim as a basic source and often quoted from it. He (1858–1917). Then, he widened the fields of borrowed the notion of “system” from Saussure and it investigation and took an increasing interest in the became the foundation of his work. The Introduction à definition of the general laws of linguistics. l’étude comparée des langues indo-européennes Antoine Meillet gave a new vigor to comparative (Introduction to the comparative study of the Indo- linguistics. For him, the analysis of the grammar of a European languages), which explains and develops language went hand in hand with the study of the relationships between various Indo-European lan- history of this language. In his historical studies, he guages, is one of the most brillant illustrations of the was concerned with achieving the highest degree of Saussurean approach. accuracy possible. He always maintained that linguistic He also developed a theory of linguistic differ- forms cannot and must not be analyzed individually, entiation in order to explain how linguistic changes that but the whole encompassing systems have to be taken occur at the point of origin of a given language family into account. Therefore, the exhaustive description of have less impact on languages in areas that are given languages was absolutely indispensable. These geographically remote from the point of origin. These ideas are clearly formulated in his Traité de grammaire more remote languages are thus more likely to retain comparée des langues classiques (Essay on archaic characteristics from their ancestor language. comparative grammar of classical languages). In the The impressive number of his contributions to the French edition, more than 700 pages refer not only to study of various languages (Armenian, Celtic, German, 675 MEILLET, ANTOINE Greek, Iranian, Latin, Polish, Slavic, etc.) shows the even Leonard Bloomfield in the United States fruitfulness of his method and his way of thinking. His admitted that his research on sentence structure were analyses were of such high quality that they still remain inspired by Meillet’s ideas. Since Antoine Meillet’s a model for research. work was both innovative in its theoretical approach Antoine Meillet’s theories on meaning have also and rigorous in its implementation, and since it remained influential to this day. During the first period exhibited an interest in all languages, it can still be of semantic studies, roughly from 1883 to 1931, used as a stepping stone for further linguistic research. Darwinian evolutionism gave rise to two opposing arguments. In France, Arsène Darmesteter was a defender of the organicist thesis, which maintained that Biography languages and words must be considered as living Antoine Meillet was born in Moulins, France on organisms, similar to plants or animals. Thus, he November 11, 1866. He graduated at the Sorbonne in argued that while the laws governing meaning may be Paris, in 1885. In 1890, he conducted one year of described as fixed and timeless entities, words are born, fieldwork in the Caucasus, studying modern Armenian. live, change, and die according to natural selection. In Then, he became director of comparative Indo- Essai de Sémantique (Essay on semantics), published European studies at the School of Advanced Studies in in 1897, Michel Bréal took up the opposite point of Paris. He received his Ph.D. in 1897 for his works on view, saying that language has no reality outside of Old-Slavonic in Recherches sur l’emploi du génétif- human activity. For Bréal, the semantic laws are accusatif en vieux-slave (Researches on the Use of psychological or intellectual laws and govern genitive-objective Case in Old-Slavonic), his first thesis, language-specific patterns of words and sentences. and on the Indo-European root *men in De indo- This approach resonated with Antoine Meillet, for europaea radice *men “mente agitare” (About the whom semantics always had a historical dimension. In Indo-European Root *men), his supplementary thesis in “Comment les mots changent de sens” (“How words Latin. He was Professor of Armenian at the School of change meaning”), part of Linguistique historique et Advanced Studies in Paris in 1902. At the end of World linguistique générale (Historic linguistics and general War I, he contributed as an expert to the definition of linguistics), he argued that modifications in the linguistics groups, languages, and political boundaries. meaning of words depend on three interdependent sets Then, he became secretary of the Société Linguistique of factors: linguistic ones (language structures), de Paris (Paris Linguistics Society), before being historical ones (cultural contacts), and social ones. elected to the Académie des Inscriptions et Belles- Throughout his career, he increasingly focused on Lettres (French Academy of Inscriptions and Letters) in the idea that language and the social environment are 1924. He retired in 1932. Meillet died in closely intertwined. In his studies on the period of Châteaumeillant, France on September 21, 1936. bilingualism in France at the time of the Germanic invasion, he pointed out that antagonistic forces exist among communities, which must be taken into References consideration in the description of the linguistic Auroux, Sylvain (ed.) 1988. Antoine Meillet et la linguistique landscape. Due to these efforts, he was considered by de son temps. Histoire-Epistémologie-Langage X, (2). some as the founder of a French sociological school. 5–348 While this may not be completely accurate, his ideas Meillet, Antoine. [1903] 1936. Esquisse d’une grammaire certainly revolved around language as a social fact and comparée de l’arménien classique (Outline of a comparative he worked on topics that even today remain partly grammar of classical Armenian), Vienne: Impr. des PP. Mêkharistes. uncharted territory. –––––– [1903] 1978. Introduction à l’étude comparative des While Antoine Meillet is especially known as a langues indo-européennes (Introduction to the comparative comparative and general linguist, his book Les study of the Indo-European languages), Alabama: Langues dans l’Europe nouvelle (The tongues in New University of Alabama Press. Europe), 1918, and Les Langues du monde (The –––––– [1913] 1975. Aperçu d’une histoire de la langue grecque (Outline of a history of the Greek language), Paris: tongues of the world), coedited with Marcel Cohen in Klincksieck. 1924, also showed his great interest in dialect study –––––– [1915] 1931. Grammaire du vieux perse (Old-Iranian and typology. grammar), Paris: Champion. Famous for his publications and his personality, –––––– [1921] 1982. Linguistique historique et linguistique Antoine Meillet became a leader of a linguistic school. générale (Historic linguistics and general linguistics), Paris: Champion and Genève: Slatkine Reprints. Many people were influenced by his theories and his –––––– [1925] 1979. Traité de grammaire comparée des works, like Émile Benveniste, Marcel Cohen, Gustave langues classiques (Essay on comparative grammar of Guillaume, Louis Hjelmslev or André Martinet, and classical languages), Paris: Champion. 676  METAPHOR –––––– [1928] 1984. La méthode comparative en linguistique Swiggers, Pierre. 1996. Lexicon grammaticorum, 622–4 historique (The comparative method in historic linguistics), Tübingen : Max Niemeyer. Paris: Champion. THIERRY PONCHON –––––– [1936] 1951. Linguistique historique et linguistique générale II (Historic linguistics and general linguistics II), See also Bloomfield, Leonard; Martinet, Andrè; Paris: Klincksieck. Saussure, Ferdinand de Metaphor Metaphor and metonymy are two closely related conversational maxims. Some important papers concepts. They have been treated for centuries as types representing these and other modern theories can be of figures of speech. Therefore, their study was thought found in Ortony (1993). to be the business of literary scholars, rather than of grammarians or linguists. Metaphor is defined from The Cognitive Theory of Metaphor this ‘classical’ perspective as a figure of speech in which one word is used to indicate something different This theory has revolutionized the scientific study of from the literal meaning, so that one thing or idea is metaphor and metonymy, has dominated it for the past likened to a different thing or idea. In a sentence like 20 years, and has attracted large numbers of linguists, He had a heart of stone, the word stone is not used in psychologists, and other kinds of scientists to this field its normal physical sense; rather, it conveys the idea of of research. (For the distinction and connections insensitivity, by likening this moral attitude to the between metaphor and metonymy, see Metonymy.) hardness of stones. The classical view confines The theory was first proposed in Lakoff and Johnson metaphor to ornate, flowery language, and assumes (1980). On empirical grounds, its authors criticized the that the interpretation of any metaphorical use of basic assumptions of the traditional theory of language can be wholly reduced to literal meaning, metaphor and of most of the modern ones, especially which is its ‘real’ meaning (i.e. the ‘real’ meaning of the assumption that the meaning of metaphorical the phrase of stone in this context is ‘insensitive’). expressions can be reduced to ‘literal’ meaning, their Conventional, automatic metaphorical expressions like ‘real’ meaning, and that metaphors had no serious The foot of the mountain are thus regarded as conceptual import. The Cognitive Theory of Metaphor established literal meanings of words and cease to be (CTM) evolved as an essential part of the approach metaphorical––they are called ‘dead’ metaphors. known as Cognitive Linguistics (see People, George Lakoff; Ronald Langacker); a brief introduction is Ungerer and Schmid (1996). Let us now review some Some Precognitive Modern Theories of of the main tenets of the CTM: Metaphor (1) Metaphor makes it possible for people to The contemporary theory of metaphor that has had the understand new, complex, or abstract areas of strongest impact on linguistics is the cognitive theory experience (‘domains’) on the basis of more accessible of metaphor (see below). Among precognitive modern domains, which are understood in their own terms; for theories of metaphor, we can single out the instance, the abstract domain of life is often ‘interaction theory’ and the ‘pragmatist’ position, understood as a journey along a path (e.g. He’s at a which have also been followed by numerous linguists. crossroads [in life]). The most basic metaphors have The interaction theory, due to Max Black, claims that as input or ‘source’ domains universal physical notions metaphor is the result of a semantic interaction or like ‘verticality’, ‘container’, ‘path’, etc., known as tension between a metaphorical term and its context. ‘image schemas’ (Johnson 1987). The tension between stone and heart of leads to (2) Therefore, metaphor is not just a matter of construing the former as ‘insensitivity’, its ‘real’ language use but also, and fundamentally, a matter of meaning. The pragmatist position, due to Paul Grice thought, of conceptualization. Creative, conscious, and John Searle, holds that metaphor has to be reduced unconventional metaphors are claimed to be usually to the ‘literal’ meaning by applying a series of extensions of automatic, unconscious, conventional pragmatic principles of conversation, like metaphors. 677 METAPHOR (3) It is therefore important to distinguish between This would be done by a different metaphor, ANIMALS conceptual metaphors and metaphorical expressions. AS PEOPLE, as in It’s a noble dog. A given conceptual metaphor may be expressed by all A very recent development within cognitive kinds of linguistic and non-linguistic means (words, linguistics is the theory of ‘blending’ (due to Gilles sentences, gestures, etc.); some of these linguistic Fauconnier and Mark Turner), which regards expressions may cease to be used metaphorically but metaphor and metonymy as just two manifestations of the conceptual metaphor may still motivate other a general mapping ability. expressions. For instance, in Old English sad The CTM has been applied to the study of all aspects originally meant ‘full of food’ and metaphorically ‘full of language (lexicon, especially polysemy, grammar, of sensations’, the indirect source of its present sense discourse, and conversation) and to many other ‘sorrowful’. Sad stopped being a metaphorical disciplines, from artificial intelligence to the study of expression when its original physical sense was lost. literature, philosophy, anthropology, law, religion, etc. However, the conceptual metaphor PEOPLE ARE CONTAINERS FOR SENSATIONS, which indirectly References motivated its metaphorical senses, is alive in countless other expressions (full of joy, I am fed up with his Barcelona, Antonio (ed.) 2000. Metaphor and metonymy at the behavior, sorrowful, etc.). crossroads. A cognitive perspective. Berlin and New York: Mouton de Gruyter. (4) Metaphors consist of fixed multiple simultaneous Cameron, Lynne, and Graham Low (ed.) 1999. Researching and projections (or ‘mappings’) from the so-called ‘source’ applying metaphor, Cambridge and New York: Cambridge domain onto the so-called ‘target’ domain. A sentence University Press. like I see what you mean is an expression of a Gibbs Jr., Raymond W.1994. The poetics of mind. Figurative conceptual metaphor that maps the conceptual structure thought, language and understanding. Cambridge and New York: Cambridge University Press. of SEEING onto that of UNDERSTANDING, and which –––––– and Gerard Steen (eds.) 1999. Metaphor in cognitive makes immediately comprehensible, without linguistics, Amsterdam and Philadelphia: John Benjamins. necessarily involving any pragmatic rules of Johnson, Mark. 1987. The body in the mind: the bodily basis of interpretation, scores of other conventional expressions meaning, imagination, and reason. Chicago and London: of the same metaphor like Your arguments are The University of Chicago Press. Kövecses, Zoltán. 1990. Emotion concepts. New York: transparent, or Your theory has thrown light on this Springer-Verlag. problem. (This is evidence against pragmatist theories.) Lakoff, George. 1987. Women, fire and dangerous things. What Metaphorical mappings are fairly systematic, but they categories reveal about the mind. Chicago and London: The are only partial (about this, see Lakoff 1993). University of Chicago Press. (5) The mapping is always unidirectional: only the Lakoff, George. 1993. The contemporary theory of metaphor. In Ortony, pp. 202–51. source is projected onto the target domain. Therefore, Lakoff, George, and Mark Johnson. 1980. Metaphors we live simultaneous bidirectional metaphorical projections by. Chicago and London: The University of Chicago Press. do not exist, as the interaction theory claims. In the Lakoff, George, and Mark Turner. 1989. More than cool reason: conceptual metaphor PEOPLE AS ANIMALS (as in Don’t a field guide to poetic metaphor. Chicago and London: The snap at me / Their love nest has been discovered), we University of Chicago Press. Ortony, Andrew, (ed.) 1993. Metaphor and thought. Cambridge project an aspect (aggressive behavior, living place) of and New York: Cambridge University Press. some animals (dogs or birds) onto some aspects of Ungerer, Friedrich, and Hans-Jörg Schmid. 1996. An introduction people (anger, meeting point), but no aspect of people to cognitive linguistics. London and New York: Longman. is mapped onto animals by virtue of this metaphor. ANTONIO BARCELONA Metathesis ‘Metathesis’ refers to several related processes in transposition of any two linguistic units, such as which the linear order of two elements is reversed, as syllables or words (She will go —> Will she go?), it when ask is pronounced aks in dialectal English. most commonly denotes the transposition of two Although in the broadest sense metathesis is the sounds: consonant with consonant, consonant with 678  METATHESIS vowel, or vowel with vowel. In some cases, not two grammatical category, e.g. singular vs. plural or whole sounds, but just a feature of the sounds, for definite vs. indefinite. example vowel length, is transposed: basile:os ‘of a In the absence of grammatical conditioning, king’ in Ancient Attic Greek corresponds to basileo:s synchronic metathesis is sporadic and abrupt, in Ancient Ionic Greek (the colon indicates a motivated by speech errors (‘slips of the tongue’) or by lengthened vowel). A thorough description of the influence of words related in form or meaning. For metathesis must begin with the following four example, the English cases of irrelevant ~ irrevelant parameters: and integral ~ intregal/intrical may be due to the influence of revelation and intricate, respectively. As (1) Synchronic ~ diachronic. Synchronic these examples show, sporadic cases of synchronic metathesis occurs within one chronological metathesis often occur when the inverted sounds are period. Diachronic metathesis takes place from not adjacent to one another, and they may involve the one time period to another, e.g. from Middle transposition of two sounds or the relocation of one English to Modern English. sound into another syllable. (2) Adjacent ~ nonadjacent. Adjacent metathesis If sporadic synchronic metathesis occurs in certain occurs when two contiguous sounds are words with sufficient frequency among speakers in a transposed. With nonadjacent or ‘long- given speech community, it may become a permanent distance’ metathesis, the transposed sounds diachronic sound change over time. Sporadic may be separated by one or more intervening diachronic metathesis includes: (1) the long-distance sounds. (reciprocal) metathesis of two sounds, often l or r: e.g. (3) Regular ~ sporadic. Regular metathesis applies Latin parabolam ‘word’ developed into Spanish consistently, to many different words. Sporadic palabra; (2) the transposition of one sound into a metathesis is restricted to only a few words and neighboring syllable: Latin capra ‘goat’ became occurs haphazardly. dialectal Italian crapa; and (3) the inversion of two (4) Abrupt ~ gradual. Abrupt metathesis adjacent sounds, as with Old English bridd and completely transposes sounds in a single step. Modern English bird. Gradual metathesis can best be understood as a A sporadic diachronic metathesis can become type of chain reaction, a series of sound regular (permanently invert the same two sounds in changes over time that eventually inverts two many different words) if it fulfills a specific structural sounds. Gradual metathesis therefore uses one purpose. For example, a preference for syllables or more intermediate stages between the ending in a vowel may have caused the change from original order and the final transposed order. Proto-Slavic *melko to South and West Slavic mleko A closer look at metathesis reveals four major types of ‘milk’ (the asterisk indicates that the relevant word inversion, pairing regular or sporadic with synchronic form is not actually attested, but reconstructed on the or diachronic, each one a unique combination of the basis of modern words). Similarly, less favored above parameters with other linguistic causes and consonant sequences such as /dl/ and /nr/ are often tendencies. metathesized in order to locate the more sonorant One type of synchronic metathesis is an abrupt (vowel-like) sound in the syllable-final position and regular process that applies to adjacent sounds in order the less sonorant sound at the beginning of the next to mark certain grammatical relationships between syllable, e.g. Latin titulum ‘title’ most likely words. For example, in Rotuman, a Malayo- developed first into *tidle, which then metathesized Polynesian language of the central Pacific, metathesis into Old Spanish tilde; Latin generum ‘brother-in- marks the distinction between definite and indefinite: law’ became *yen.ro and then Spanish yer.no. In /hosa/ ‘the flower’ vs. /hoas/ ‘some (unspecified) contrast to these examples, a sporadic metathesis flower’. Other languages where metathesis has a without structural motivation, as in the irrelevant ~ grammatical function include Clallam, Leti, Cherokee, irrevelant example, cannot become regular, i.e. this Zoque, and Kwara’ae. same process will not spread to other words with l and Regular synchronic metathesis is generally the v in similar positions. product of a previous diachronic metathesis. Speaker Regular diachronic metathesis is often the result of variation between metathesized and unmetathesized a series of interrelated sound changes over time, a forms, initially nothing more than variation in the gradual chain reaction of regular changes whose last pronunciation of a word during an earlier stage creates a transposed version of the original chronological stage of the language, is eventually sequence. For example, French Breton became *brtõ, sorted out over time by assigning each of the two which then developed into Le Havre French bertõ. variants to one of two options within a specific This is an example of ‘pseudometathesis,’ since the 679 METATHESIS two sounds e and r are not transposed directly. Rather, *agéka~a~ > [agékhaa~] ‘my eye’. As before, the h the e was first lost and then an e was inserted into a actually moved through the vowel a in several steps, different position. If only the original word and the but the result resembles a one-step metathesis. final result were compared, the modern word would seem to be the product of metathesis, when in fact it results from several nonmetathetical processes. References A gradual diachronic transposition is often the case Alexander, James D. 1985. R-Metathesis in English: a with consonant + vowel sequences, where i or u diachronic account, Journal of English Linguistics 18. 33–40 reduces to j or w, respectively. These sounds can then Besnier, Niko. 1987. An autosegmental approach to metathesis merge with adjacent consonants—they become in Rotuman. Lingua 73. 201–23 Blevins, Juliette, and Andrew Garrett. 1998. The origins of ‘secondary articulations’ of these consonants—and consonant–vowel metathesis. Language 74(3). 508–55 later give rise to vowel developments on the other side Grammont, Maurice. 1933. Traité de phonétique. Paris: of the consonant. Examples of such chain processes Delagrave; 8th edition, 1965. are as follows: Ancient Greek phéresi > phéresj > Hock, Hans Henrich. 1985. Regular metathesis. Linguistics 23. phérejs ‘you carry’; Latin sapui > *sabwi > *sabwi > 529–46 Hogg, Richard M. 1977. Old English r-metathesis and *sawbe > *sowbe > Portuguese soube ‘I knew’. The generative phonology. Journal of Linguistics 13. 165–75 vowels thus gradually wandered ‘through’ a Hume, Elizabeth. 1988. Metathesis in phonological theory: the consonant. The result, again, resembles the product of case of Leti. Lingua 104. 147–86 metathesis. Consonants may also reduce to secondary Ultan, Russell. 1978. A typological view of metathesis. articulations before reappearing on the opposite side Universals of human language, Vol. 1, ed. by Joseph H. Greenberg. Stanford: Stanford University Press. of a vowel. For example, an h may temporarily Vennemann, Theo. 1988. Preference laws for syllable structure become a breathy-voice articulation on the and the explanation of sound change. Berlin: Mouton de neighboring vowel and then resurface as a full Gruyter. consonant on the other side: Cayuga /akekaha~/ > KENNETH J. WIREBACK Metonymy Metaphor and metonymy are two closely related The Cognitive Theory of Metonymy concepts. Both have been treated for centuries as This theory emerged alongside the cognitive theory of figures of speech, and are as such regarded to be the metaphor, but the nature of metonymy and its business of rhetoric or literary studies rather than relevance for thought and language have only recently linguistics. The traditional theory of metonymy begun to be investigated in depth. In the standard form defines it as the ‘use of the name of one thing for that of the Cognitive Theory of Metonymy (CTMy), as in of another associated with or suggested by it’. Lakoff and Turner (1989), metonymy is defined as a (Webster’s new twentieth century dictionary, 2nd mapping with a primarily referential purpose, in which edition.) In The White House issued a statement, the the source and the target are conceptual entities in the phrase the White House is used instead of the phrase same domain. ‘Mapping’ means ‘conceptual the President. This traditional theory treats metonymy projection’. In the previous example, the concept of as a matter of language use; its conceptual nature is the White House (the ‘source’) is mapped onto the overlooked. The meaning of metonymic expressions is concept of the President (the ‘target’): the President is also assumed to be wholly translatable into literal primarily conceptualized as someone who lives in that meaning. Metonymic expressions are considered to be building. This allows the phrase the White House to necessarily referential (in the technical sense of refer to, i.e. to designate, the person in office as the designating an extra-linguistic entity). Most twentieth- President of the United States at a given time. Finally, century treatments of metonymy have essentially a place and the people or things located in it are so maintained the same views; a good illustration is closely linked in experience that they can be said to be Ullmann (1962). in the same ‘domain’ or area of experience. 680  METONYMY Metonymy can link a part to a whole, as in We need journeys are in two different domains. In the a couple of extra hands in our factory (HANDS [BODY metonymy CONTAINER FOR CONTENT, as in He PART] FOR PEOPLE), a whole to a part, as in The Times drank a glass, the source (the glass) and the is here (NEWSPAPER COMPANY FOR REPORTER), or a part target (the contents of the glass) are in the same to a part, as in He has a good pen (PEN [INSTRUMENT] domain, namely, the domain of ‘containment’. FOR ACTION [WRITING]). This tripartite classification This criterion, however, does not always work continues the traditional theory, except that the latter clearly. For instance, the source and the target reserves the term synecdoche for PART-FOR-WHOLE of PEOPLE ARE ANIMALS (Richard is a lion) are in metonymies. the same domain, the domain of LIVING BEINGS. In the CTMy, metonymy is a primarily conceptual This and other cases have led some linguists to phenomenon, rather than just a matter of language use, claim that an absolute distinction between so that conceptual metonymies have to be distin- metaphor and metonymy cannot be maintained; guished from metonymic expressions; metonymy is they are better regarded as the two ends of a systematic, as it responds to general patterns like continuum. WHOLE FOR PART or to more specific patterns like Metaphor and metonymy often interact in INSTRUMENT FOR ACTION. And the meaning of complex ways. One of the most intriguing of metonymic expressions cannot be simply reduced to these is the fact that a great many metaphors literal meaning. have a metonymic basis: the metaphor MORE IS Some of the main areas of current debate in the UP (A high number/Prices soared) is motivated CTMy are the following: by the metonymic activation of quantity by (1) Referential function. Some cognitive linguists verticality, through their frequent experiential still hold the traditional view that every association (e.g. in pouring liquids into metonymic noun phrase must be referential. containers and watching their level rising). But most cognitive linguists note that, even The CTMy has been shown to be highly relevant for though referring is a very frequent function of the study of cognition and reasoning, lexical metonymy, metonymic noun phrases are often semantics, grammar, and discourse. A large number of also used nonreferentially. For instance, in John categories have been shown to have a metonymic is a good hand at fencing, the predicate noun prototype (a kind of model of the category); see Lakoff phrase does not designate an individual; it just (1987). The CTMy provides a useful framework for the mentions a property. study of lexical polysemy, in which metonymy is a (2) The nature of the relationship between source fundamental force. But, most importantly, the CTMy and target. Some linguists avoid treating has demonstrated the pervasive metonymic motivation metonymy as a mapping process, because, of a great many grammatical structures, like raising, unlike metaphor, metonymy apparently does conversion, generic sentences, etc., of certain types of not consist of systematic multiple mappings. pragmatic inferences like indirect speech acts and An approach that has recently won widespread conversational implicatures, and of discourse support is that of regarding the metonymic strategies. source as a ‘cognitive reference point’, which, in an appropriate context, can cause the mental Further Reading activation of the target (see Kövecses and Radden 1998; Langacker 1999): In The White Barcelona, Antonio (ed.) 2000. Metaphor and metonymy at the crossroads. A cognitive perspective. Berlin and New York: House issued a statement, a LOCATION causes Mouton de Gruyter. the mental activation of the PERSON LOCATED. It Barcelona, Antonio. 2003. Metonymy in cognitive linguistics. also imposes a certain perspective on that An analysis and a few modest proposals. Motivation in person. As we saw earlier, this imposition is a language: studies in honor of Günter Radden, ed. by Hubert type of mapping . Cuyckens, Thomas Berg, René Dirven, and Klaus-Uwe Panther, 223–55. Amsterdam and Philadelphia: John (3) The distinction from, and interaction with, Benjamins. metaphor. The key property distinguishing Gibbs Jr., Raymond W. 1994. The poetics of mind. Figurative them, according to the standard view in the thought, language and understanding. Cambridge and New CTMy, is that metaphor is a mapping across York: Cambridge University Press. different domains, whereas metonymy is a Goossens, Louis, Paul Pauwels, Brygida Rudzka-Ostyn, Anne- Marie Simon-Vanderbergen, and Johan Vanparys. 1995. By mapping within the same domain. Consider the word of mouth. Metaphor, metonymy and linguistic action metaphor LIFE IS A JOURNEY, as in I have gone a in a cognitive perspective. Amsterdam and Philadelphia: long way (said about one’s life): life and John Benjamins. 681 METONYMY Kövecses, Zoltán, and Günter Radden. 1998. Metonymy: Langacker, Ronald. 1999. Grammar and conceptualization. developing a cognitive linguistic view. Cognitive Linguistics Berlin and New York: Mouton de Gruyter. 9-1. 37–77. Panther, Klaus-Uwe, and Günter Radden (eds.) 1999. Lakoff, George. 1987. Women, fire and dangerous things. What Metonymy in language and thought. Amsterdam and categories reveal about the mind. Chicago and London: The Philadelphia: John Benjamins. University of Chicago Press. Panther, Klaus-Uwe, and Linda Thornburg. 1998. A cognitive Lakoff, George, and Mark Johnson. 1980. Metaphors we live approach to inferencing in conversation. Journal of by. Chicago and London: The University of Chicago Press. Pragmatics 30. 755–69. Lakoff, George, and Mark Turner. 1989. More than cool reason: Ullmann, Stephen. 1962. Semantics. An introduction to the a field guide to poetic metaphor. Chicago and London: The science of meaning. Oxford: Blackwell. University of Chicago Press. ANTONIO BARCELONA Mexico The mosaic of languages in Mexico is among the richest that time, it lost much of its territory to what is now the and most complex in the Americas. It is the largest southwestern region of the United States, the area Spanish-speaking country in the world. However, from the states California to Texas. Today, this region Spanish has only dominated in the region that is today has the largest number in the Unite States of first- and Mexico for the past five centuries. For millennia before, second-language speakers of Spanish. the mountain valleys and tropical forests of Middle America supported a density of population and spectrum Development and Characteristics of Mexican of cultural developments that nurtured one of the richest Spanish arrays of native languages in the Americas. Mexico, therefore, concentrates the largest number of speakers of Spanish came rapidly to dominate in Mexico from one of the most widely spoken languages in the world the earliest colonial times. This dominance resulted while preserving the largest number of speakers of the not just from the power of the Spanish Conquest most ancient native American languages. from 1519 to 1521. Of most consequence were the devastating diseases brought to the native population by the Spanish. At the beginning of the sixteenth Demographic and Geographic Profile of century, just before the Conquest, the native Mexican Languages population of Mexico amounted to approximately 25 There are a third of a billion speakers of Spanish in the million. Only three decades later, well over 90% of world so that every one in 18 people in the world that population had died from European diseases speaks this language. In Mexico, there are 90 million against which natives had no immunity. The first-language speakers of Spanish, almost 90 percent population of Mexico did not again reach its pre- of the country’s population of 104 million (estimate, Conquest level until the middle of the twentieth 2002). More than a fourth of all Spanish speakers in century. the world, therefore, lives in Mexico. Mexico City, the The Conquest had a devastating and permanent capital, concentrates a population of over 20 million, effect on the linguistic balance of Mexico. The making it the fourth largest metropolis in the world. It Spanish-speaking population enjoyed robust growth, is thereby the largest Spanish-speaking city in the its language coming to the forefront in all the colony’s world. Speakers of native languages in Mexico amount commercial, cultural, military, educational, and to slightly less than 10% of the population. They speak religious activities. Native languages survived only almost 300 different languages. Well over half the precariously among geographically and socially population of Mexico is a racial mixture of Spaniard isolated remnants of the original native population. and Amerindian, known as mestizo. Just under a third The Spaniards who came to Mexico in the early of the population is mainly of Amerindian descent. colonial period amounted to less than 100,000 people. The speakers of the numerous languages of Mexico However, they were mostly male and interbred widely occupy an area of more than 750 million square miles. with the surviving native women, producing the Mexico is four times the size of Spain, where Spanish mestizo population. It was this growing population that originated. The area of present-day Mexico, however, became the main vehicle for the advancing dominance is only half of what it was in the nineteenth century. At of Spanish. 682  MEXICO Distinctions of Spanish in Mexico extensively in Spanish. Such is the origin of the often- used Mexican expression, ojalá (May it be!). Many The use of Spanish in Mexico has developed differences words in Mexico beginning with “al” such as alcalde from other Hispanic countries in terms of vocabulary, (mayor), algebra (algebra), and algodón (cotton) pronunciation, and grammar. These distinctions have originate from Arabic. occurred based on the historical origins of the language The most distinctive source for the vocabulary of in central and southwestern Spain and its evolution Mexican Spanish, however, derives from native amid the dynamics of Mexican development. languages, especially Nahuatl. This was the language The foundations of Spanish throughout Latin of the Aztecs, who ruled the part of Mexico that the America originated with the language as it was spoken in Spanish initially conquered and occupied. From the central and southwestern Spain during the sixteenth and Aztec language, Mexican Spanish has received its seventeenth centuries. The kingdom of Castile, in the words for chocolate (chocolate; from the Nahuatl central mestizo of the Iberian Peninsula, led the conquest word, chocolatl), tomate (tomato; from tomatl), and of colonies in the Americas. In addition, soldiers, coyote (coyote; from coyote). These words have merchants, administrators, settlers, and missionaries set entered the vocabulary of the entire world. The out from the southwestern areas of Andalusia, primarily strongest influence on modern vocabulary the port of Seville, and Estramadura. development in Mexican Spanish has been English, The dominance of Castile in the propagation of the contributing such words as béisbol (baseball), cóctel language was so extensive that the Spanish language is (cocktail), and sándwich (sandwich). often also referred to as castellano or Castilian. The As throughout the Spanish-speaking world, the base of Spanish in Mexico and throughout Latin standard, accepted version of the language in Mexico America reflects patterns dominant in Castile and continues to adhere to rules formulated by the Spanish southwest Spain at the time of colonization. Variations Royal Academy, in Madrid. (Somewhat similar to the have emerged since then from that base. way in which French-speaking countries follow norms While in Spain today the “c” may have a “th” sound, established by the French Academy, in Paris.) There in Mexico it is pronounced now only as an “s.” The are, nonetheless, regional Mexican language sound of Castilian “x” at the time of colonization was distinctions. These occur along the border with the “sh” but has now evolved in Mexico to a breathed “h” United States, around the industrial region of sound. (Galician and Portuguese still maintain the Monterey, the Caribbean coast, southern Amerindian earlier tendency.) The original pronunciation, therefore, areas, rural regions, and Mexico City. Throughout the of the first syllable of Mexico sounded like “mesh” but latter part of the twentieth century, the United States has now become “meh”. In addition to the way in which has increasingly become a center for the speaking of speakers of Mexican Spanish tend to abbreviate or Mexican Spanish due to the economic diaspora of eliminate certain sounds and syllables, they also use a millions of Mexicans to that country. distinct musicality and intonation, raising the pitch of a phrase or sentence toward the end. Mexican Spanish is spoken more slowly, musically extending sounds. There Beyond Spanish is a tendency to emphasize consonant sounds and reduce or “swallow” unaccented vowel sounds. Less than 10 million Mexicans use the several hundred As the largest Spanish-speaking country, Mexican native languages still surviving in the country. These publishers and media outlets considerably influence languages are grouped into 11 linguistic families: Spanish-language communications. This is especially Algonquin, Chiapanecan-Mangue, Chinantecan, true for Mexican television and radio, which massively Huave, Mayan, Mixe-Zoque, Oaxacan, Tarascan, penetrates into the Latino population of the United Tlapanecan, Totonacan, and Uto-Aztecan. The major States. Mexican television networks, such as Univision languages still have several hundred thousand to and Telemundo, produce novelas (series dramas) that are several million speakers. broadcast around the world. Nonetheless, dominance of Colonial missionaries established written forms for Spanish-language media and communications remains the native languages, using the Roman alphabet. Thus, with Spain, and, in the Americas, is divided between writing today of native languages reflects native Mexico and Argentina. The prominence of Mexican pronunciation as transcribed into equivalent colonial literature was recognized in 1990 when Octavio Paz won Spanish alphabetic sounds. Pre-Conquest written forms the Nobel Prize in literature. were limited to pictograms and/or ideograms. Many of The vocabulary of Spanish derives essentially from these have only recently been deciphered, and often Latin. However, since the Moors occupied much of the only partially. Hostility to native languages appeared Iberian Peninsula from the Middle Ages until the end throughout early Mexican history, their speaking of the fifteenth century, words of Arabic origin occur prohibited and many documents in them burned. 683 MEXICO The Mayan languages, which in Mexico appear along native languages. This development is especially true in a southern belt of the country from west of the state of the state of Chiapas with its armed rebellion against the Chiapas then along the border with Guatemala and central government of Mexico. The growth of the Belize and then into the peninsula of Yucatan, have nearly indigenista movement throughout the twentieth century, two million first- and second-language speakers. The beginning with the victory of the Mexican Revolution of Mayan languages spoken in Mexico are principally Chol- 1910, supported extensive anthropologic, archeologic, Chontal, Tzeltal, Tzotzil, Mocho, and Yucatecan. In the and linguistic studies of the ancient native languages. ancient language, glyphs were used for syllable sounds. Many Mexican and US universities now offer courses in The Mayan languages generally follow a syntactical the study of these languages. Publishing and media firms sequence of verb–object–subject. The great literary epic, of varying sizes now exist in these languages. Popul-Vuh, recounting the Mayan version of the creation of the universe did not, however, originate in a Mayan References language of Mexico but of Guatemala. Some publishing Atlas lingüístico de México, Serie estudios de dialectología and radio broadcasting occurs in Mayan languages where mexicana dedicada a Pedro Henríquez Ureña, 4. Mexico there is a more dense concentration of such speakers. City: Colégio de México, 1990. In the southwestern part of Mexico, around the area Borah, Woodrow Wilson. 1951. New Spain’s century of of Oaxaca, there are half a million speakers of Zapoteco depression, Ibero-Americana, Vol. 35. Berkeley, CA: (the “cloud” people) languages. These are highly tonal University of California Press. Boyd-Bowman, Peter. 1973. Patterns of Spanish migration to (i.e. they use different levels of intonational pitch to the New World, 1493–1580. Buffalo: Council on distinguish words) and require an alphabet of 39 letters International Studies, State University of New York. to capture their range of sounds. There are under Campbell, Lyle, et al. 1978. Bibliography of Mayan languages 100,000 speakers of Mixe languages, and a quarter of and linguistics, Albany: State University of New York. million speakers of Mexteco languages. This region also Carlos Castillo. 1987. University of Chicago Spanish-English English-Spanish Dictionary, 4th edition, 1–43, Chicago: has several thousand Chinanteco speakers. Oaxaca and University of Chicago Press. the area around it has the greatest concentration of Grace, Lee Ann. 1976. The effect of bilingualism on sixteenth- speakers of native languages in Mexico. century Mexican Spanish. Dissertation, State University of In the central part of Mexico, from Nayarit into the New York at Buffalo. state of Mexico and through Puebla down to Vera Hill, Jane H., and Kenneth H. Hill. 1986. Speaking Mexicano: dynamics of syncretic language in Central Mexico. Tucson: Cruz, there are well over a million first- and second- University of Arizona Press. language speakers of Nahuatl (Aztecan) languages. Lochhart, James. 1991. Nahuas and Spaniards: postconquest The eastern portion of the region has several hundred Central Mexican history and philology, Nahuatl Studies thousand speakers of the Totonacan languages. Unlike Series, 3. Stanford, CA: Stanford University Press. Mayan languages, Nahuatl does not impose a fixed Marcus, Joyce. 1992. Mesoamerican writing systems: propaganda, myth, and history in four ancient civilizations. word order on subject, verb, and object. Moreover, Princeton, NJ: Princeton University Press. somewhat like German, it is an agglutinative language: Moreno de Alba, José G. 1994. La pronunciación del español en Nahuatl strings morphemes together into words that México. Mexico City: Colegio de México. can accumulate into more than ten syllables. Perisinotto, Giorgio. 1975. Fonologia del español hablado en la Due to extensive migration from the impoverished ciudad de México. Mexico City: Colegio de México. areas of southern Mexico into the United States, the EDWARD A. RIEDINGER native languages of Mexico now have hundreds of See also Maya and Mayan Languages; Nahuatl thousands of speakers in the latter country. Political and and Uto-Aztecan Languages; Spanish and Iberoro- cultural consciousness has given renewed vigor to the mance Languages Middle (Classical) Japanese Middle Japanese is a historical stage in the history of Middle Japanese can be estimated as ninth to twelfth the Japanese language between Old Japanese and CE, roughly corresponding to the Heian period, Early Modern Japanese. The chronological span of although ninth century data in many respects are 684  MIDDLE (CLASSICAL) JAPANESE transitional from Old Japanese to Middle Japanese, in forms of many of (but not all) earlier man’yôgana pretty much the same way as thirteenth century, is signs and cursive forms of Chinese characters, which transitional between Middle Japanese and Early roughly stand for stems of words. The syllabic system Modern Japanese. Middle Japanese is a language is polyphonic like man’yogana; one syllable can be predominantly based on the dialect of the upper and written by more than one sign. This syllabic system of middle nobility in the capital Heiankyô (present day writing is a direct predecessor of modern hiragana, Kyôto). In contrast to Old Japanese, there is no where each syllable is rendered by only one sign. The evidence for any kind of dialectal variation, and the predecessor of modern katakana syllabary also existing sources preserve only bits and pieces of originated in this period, although its usage was language samples as spoken by commoners and even mainly limited to Buddhist texts and to indicating the lesser nobility. reading of Old Japanese texts written in man’yôgana. It is necessary to keep in mind that Old Japanese and Only in the twelfth century katakana gains somewhat Middle Japanese do not represent a direct continuity in wider usage and penetrates even some literary texts. time, as both of them are based on geographically close Finally, there was another writing system with limited but in some respects different dialects. Nevertheless, usage called kunten (‘Japanese readings [rendered by] both can be defined as belonging to the Western dialect dots’), chiefly used for ‘Japanicizing’ the reading of group of Central Japanese. texts originally written in Chinese. It was based on placing dots at certain corners or sides of a Chinese character that indicated particles or suffixes of Middle Sources Japanese. As a matter of fact, there were several The written sources of the Heian period composed different kunten systems, invented by various Buddhist completely or partially in Middle Japanese can be sects, which usually kept them secret from members of divided into three major groups: (1) commentaries on other sects. Buddhist and Confucian classics; (2) dictionaries; and (3) literary works. The first group of texts, the Phonology commentaries, is considerably influenced by Chinese, and mostly uses a cumbersome system of The sound system of Middle Japanese underwent two writing that only approximately represents the major radical changes as compared to the Old language. In addition, these texts contain a number of Japanese system. The first major change resulted in archaisms from Old Japanese, no longer essential for more than double of the number of consonants when Middle Japanese. The second group, dictionaries, certain consonants (palatalized and labiovelar) were although containing valuable lexical and phonetic introduced to the system due to the large influx of materials with clues on vocabulary and pronun- Chinese loanwords. The second major change was due ciation, respectively, tells us next to nothing about to the weakening and subsequent loss of certain the morphology (word structure and word formation) consonants between vowels that ultimately destroyed and syntax (sentence structure) of the language. the Old Japanese requirement that no two vowels Therefore, the present paper is based predominantly occur adjacent to each other. on the data found in literary texts, although Middle Japanese had the following consonants: -p, dictionary data are also included for the description f, f’, b [mb], b’ [mb’], t, t’, d [nd], d’ [nd’], k, k’, kw, g of Middle Japanese phonology, and the commentary [ŋg], g’ [ŋg’], gw [ŋgw], m, m’, n, n’, -N, s, s’, z [nz], z texts are used for the description of the writing [nz’], w, y, r, r’. Similar to Old Japanese, no consonant system. is possible in the syllable-final position; thus, only Extant literary texts represent both poetry and syllables of CV[V] and V structure are found in the prose. The Japanese written language up to the Meiji native Middle Japanese vocabulary. In Chinese loan- period (1868–1911 CE) was predominantly based on words, however, nasal sonorants -m, -n, -ŋ and Middle Japanese, although starting from the probably voiceless -p, -t, -k as well are possible. Muromachi period there have been texts produced in Consonants b, d, g, z, and r occur only in word-medial the vernacular. For this reason, Middle Japanese is position in the native vocabulary. Between vowels, -f- often called ‘Classical Japanese’. shifted to -w- and consequently disappeared before the high vowels /i/ and /u/ or went to -y- before /e/. The same fate befell the original intervocalic -w-. In certain Writing System sets of verbs and adjectives, intervocalic -k- and -s- Several writing systems were in use in the Heian also disappeared before high vowels /u/ and/or /i/. period. Literary texts are written in a mixed system The set of Middle Japanese vowels, on the other including both syllabic signs representing cursive hand, shrank compared to Old Japanese, as only five 685 MIDDLE (CLASSICAL) JAPANESE vowels a, i, u, e, o are present. The vowels /e/ and /o/ developments in Middle Japanese that are not found in are found in Middle Japanese only in post-consonantal Old Japanese are indicated below in bold typeface). position, that is, they never appear as a word or syllable initial. Nominal Morphology From the Chinese–Japanese dictionary Ruiju Myôgishô (1081 CE), we have fairly good infor- Most nominals (with the exception of some pronouns) mation on the Middle Japanese accentual system. It in Middle Japanese have no formal markers, was a system that was based on pitch and possessed distinguishing them from other parts of speech, e.g. two major characteristics: register (high (H) or low yama ‘mountain’, fito ‘person’, yuki ‘snow’, midu (L)) and locus (location of the drop in pitch in a given ‘water’, fi ‘fire’, futa ‘2’, towo ‘10’. word with a High–Low contour), which are also Some Middle Japanese pronouns have two stems: typical for the modern Kyoto dialect. Like in Modern unextended, and extended with suffix -re, which may Kyoto, nominals were characterized by both register be treated as a formal marker of these pronouns, and locus, while verbs and adjectives exhibited only distinguishing them from other parts of speech, and register. also from other pronouns that do not have the special stem in -re. It is necessary to note that Middle Japanese has a unified system of personal-reflexive Lexicon pronouns (...self), which does not exist in Old There are many loanwords from Chinese in Middle Japanese. Japanese that penetrated different lexical domains, Personal-reflexive pronouns: wa-/ware 1/3ps and but they are most numerous among the sociopolitical 1/3pp (rare), maro 1ps (attested only once in OJ), wa- and religious strata of the Middle Japanese ga mi 1/3ps, mi 1ps, midukara 1/3ps, wono-/wonore vocabulary: dairi ‘forbidden city’, kugyau ‘nobility’, 1/2/3ps and 1/2/3pp, wono-ga mi 1ps, wonodukara sou ‘monk’, kyau ‘sutra’. Many Chinese loanwords 1/2/3ps, nandi 2ps (pejorative), [mi]masi 2ps, omafe have doublets in the native vocabulary, e.g. (native 2ps (honorific), gozen 2ps (honorific), watakusi 3ps. word is given in parentheses): sekai (yo) ‘world’, kyau Demonstrative pronouns: proximal ko/kore ‘this’, (miyako) ‘capital’, guu (miya) ‘palace’, sinwau (miko) mesial so-/sore ‘that’, distal ka/kare ‘that over there’, ‘prince’, daizin (otodo) ‘minister’. There are also a-/are. Demonstratives kore and sore can be used as 3p words coined by the Japanese from Chinese elements, pronoun ‘he’ or ‘she’. Demonstrative pronouns e.g. chûnagon ‘middle councilor’, dainagon ‘senior indicating place or direction: proximal: koko, konata councilor’. There are also loanwords from Sanskrit ‘here’, mesial: soko, sonata ‘there’, distal kanata, borrowed via Chinese or Korean: butu [but] ‘Buddha’ kasiko, anata ‘over there’. Interrogative pronouns: ta- < Middle Chinese but < Sanskrit Buddha, fati ‘bowl /tare ‘who’, nani ‘what’, ika, ikani, ikaga, ikade for alms’ < Middle Chinese pat or Old Korean *pati ‘how’, iduku/iduko ‘where’, iduti ‘where to’, idure (cf. Middle Korean pali) ‘id.’ < Sanskrit pa:tra or ‘which’, itu ‘when’, iku/ikuda/ikura ‘how many’, pa:tri ‘bowl’. nado/nazo/nadote ‘why’, nadeu/nandeu ‘what kind’, ‘why’. Collective pronouns: mina, subete ‘all’. In contrast to Old Japanese, only the following Morphology native numerals are attested in Middle Japanese: fito- Middle Japanese morphology underwent some ‘one’, futa- ‘two’, mi- ‘three’, yo- ‘four’, itu- ‘five’, cardinal changes as compared to Old Japanese mu- ‘six’, nana- ‘seven’, ya- ‘eight’, kokono- ‘nine’, morphology. First, a number of Old Japanese towo ‘ten’, fatati ‘twenty’, misoti ‘thirty’, yosoti morphological markers are no longer used in Middle ‘forty’, musoti ‘sixty’, momo ‘hundred’, ti ‘thousand’, Japanese texts. Second, a number of new and yorodu ‘ten thousand’. Native numerals above ten morphological markers appear in Middle Japanese that have disappeared from the language being replaced by did not exist in Old Japanese, partly due to the fact that Chinese loanwords, e.g. zifu iti [nzip it] ‘eleven’, Old Japanese and Middle Japanese are based on literally ‘ten’ + ‘one’. The numerals of Chinese origin geographically close but not identical dialects. Third, below ‘eleven’ occur only with classifiers of Chinese some constructions of Old Japanese consisting of origin and are never used independently. The system clearly identifiable word elements underwent the of classifiers is in its infancy, although it is already process of reduction in Middle Japanese and conflated richer than the Old Japanese system. The following into single unanalyzable suffixes , e.g. the Old classifiers are well attested: -tu (objects used with Japanese phrase V-am-aku posi ‘wants to do V’ digits), -ti (objects used with tens and hundreds), -ri became a verbal suffix -amafosi (morphological (people), -fe (layers and folds), -ka (days), -nin elements and their functions that represent new (persons), -do (times), -ba (roofs), -sudi (long thin 686  MIDDLE (CLASSICAL) JAPANESE objects), -tokoro (high-ranking persons), -zyau of ability to perform an action, e.g. ye-yom-az-u PREV- (wards). recite-NEG-FIN ‘[he] cannot recite’. There is also one circumfix in Middle Japanese: na….so, encircling an infinitive form and making a negative imperative: na- Verbal Morphology yak-i-so NEG-burn-INF-IMP ‘do not burn’. Verbal The verbal morphology of Middle Japanese underwent affixes differ in form after consonantal stems and after three great changes as compared to Old Japanese. vowel stems (with the possible differentiation between First, in sharp contrast to Old Japanese, all Middle forms found after weak and strong vowel verbs). Thus Japanese adjectives behave like verbs and not like e.g. the attributive suffix has the following forms: -u nouns. In other words, Middle Japanese adjectives after consonantal verbs, -uru after weak vowel verbs, have a comparatively well-developed verb-like and -ru after strong vowel verbs: yuk-u fito ‘person paradigm that differentiates between attributive and who goes’, kof-uru fito ‘person who loves’, and mi-ru final forms, e.g. aka-ki isi (red-ATTR stone) ‘red fito It is impossible to provide a list of all verbal affixes stone’ and isi aka-si (stone red-FIN) ‘stone is red’. here, so only the most important are listed: infinitive Second, auxiliaries of Old Japanese that followed the -i/-Ø, negative infinitive -azu/-zu, finite -u, attributive infinitive form of verbs became suffixes in Middle -u/-uru/-ru, imperative -e/-yo, negative -an-/-az-/-n-/-z- Japanese. Third, tense marking has disappeared from , tentative -am-/-m-, iterative -af-, passive- -are/-rare-, the language, while ‘retrospective’ constructions filled causative/honorific -ase-/-sase-, debitive -ubey-, the void. The last two changes can be demonstrated by negative debitive -umazi-, retrospectives -ikyi and a single example: old Japanese verb-i-kyi verb-INF- -iker-, perfectives -ite-/-it-/-te-/-t- and -in-/-n-, PAST became Middle Japanese verb-iki verb-RETR. perfective-progressive itar-/-tar-, conjectural -umer-, Middle Japanese verbs are divided into several optative -amafosi/-mafosi, desiderative -abaya.‘person classes: consonantal verbs, with roots ending in a who sees’. consonant (yuk- ‘go’, kir- ‘cut’, omof- ‘love, think’, It is impossible to provide a list of all verbal affixes nokos- ‘leave’, etc.), vowel verbs, with roots ending in here, so only the most important are listed: infinitive a vowel (including weak vowel verbs that lose the final -i/-Ø, negative infinitive -azu/-zu, finite -u, attributive vowel under certain conditions like kofi- ‘love’, -u/-uru/-ru, imperative -e/-yo, negative -an-/-az-/-n-/-z-, tasuke- ‘save’, and strong vowel verbs that never lose tentative -am-/-m-, iterative -af-, passive- -are/-rare-, their root vowel, like mi- ‘see’, ke- ‘kick’, etc.), and causative/honorific -ase-/-sase-, debitive -ubey-, irregular verbs (ko- ‘come’, se- ‘do’, ar- ‘exist’, sin- negative debitive -umazi-, retrospectives -ikyi and ‘die’, in- ‘go [away]’). In addition, there are defective -iker-, perfectives -ite-/-it-/-te-/-t- and -in-/-n-, verbs: n-, to ‘be’, to ‘say’ that have only a limited perfective-progressive itar-/-tar-, conjectural -umer-, number of forms and are irregular as well. optative -amafosi/-mafosi, desiderative -abaya. Middle Japanese adjectives are traditionally divided In addition to this quite complex verbal into two classes: one with root ending in -si and morphology, there is also the no less complex system another ending in any other syllable. The only of auxiliary verbs that mostly follow the infinitive distinction between these two classes can be seen in form of verbs. Auxiliary verbs can be subdivided into the final form, where -si adjectives do not take the several classes (within each class, only the most final marker -si to avoid the doubled syllable, e.g. aka- frequent are listed): honorific (tamaf-, ofas[e]-, ‘be red’ becomes aka-si red-FIN ‘is red’, but utukusi- ofos[e]-, imas-), humble (tatematur-, tuka[u]matur-, ‘be beautiful’ becomes utukusi ‘is beautiful’. Since tamafe-, kikoye[sase]-, mawos-), polite (faber-, very few suffixes can directly follow adjectival stems, saburaf-), assertive (nar-), cooperative-reciprocal (af-), adjectives form a special secondary conjugation type, directive (ide-/ idas-, ire-/ir-, ko-, yuk-/ik-, age-/agar-, consisting of adjectival infinitive form -ku plus the kudas-, yar-, watar-, yose-/yor-), and resultative (fate- verb ar- ‘exist’ that contracts to -k-ar-, e.g.: aka-ku ar- , ok-, tuke-/tuk-, tome-). be red-INF exist > aka-k-ar-, to which negative -az- can be further added: aka-k-ar-az-u be red-INF-exist- References NEG-FIN ‘[it] is not red’. Ikeda, Tadashi. 1980. Classical Japanese grammar illustrated Middle Japanese verbs can take both prefixes or with texts. Tokyo: Tôhô Gakkai. preverbs and suffixes, but while a word is limited to Takeuchi, Lone. 1987. A study of classical Japanese tense and only one prefix, multiple suffixes may be added. There aspect. Copenhagen: Akademisk Forlag. are three verbal prefixes, uti-, kaki-/kai-, and mote-, Vovin, Alexander. 2003. A reference grammar of classical Japanese prose. London and New York: Routledge/Curzon. with mostly unclear or poorly investigated functions, ALEXANDER VOVIN and one preverb ye-, which, in combination with a following negative form of the verb, conveys the lack See also: Japanese; Old Japanese 687 MIDDLE ENGLISH Middle English Middle English is the name given to the varieties of after Chaucer’s life (born c. 1340–1346, died 1400), English spoken between the Norman Conquest of up to the introduction of printing. (Caxton brought out 1066 and the introduction of the printing press to the first printed edition of The Canterbury tales in England by William Caxton in 1476. These historical 1478.) It was during the Late Middle English period events are simply convenient demarcation points; in that the pronunciation changed in a complex process reality, many of the linguistic changes characterizing commonly referred to as the Great Vowel Shift. Middle English started before and extended beyond Furthermore, when vowels occurred at the end of a these dates. word and were unstressed, they were completely lost in all dialects, and since word endings often had grammatical significance, the grammar changed as Background well. Late Middle English also marks the beginning of In 1066, William, Duke of Normandy, also known as standardization, the complete replacement of French William the Conqueror, invaded England, defeated the by English as the official, legal, and parliamentary English armies at Hastings, and declared himself King language, and the rise in literacy. of England. Throughout the Middle English period, England was ruled by royals who were of French Dialects descent: the Norman line from 1066 to 1154 and the Plantagenet line from 1154 to 1485. The demographic One consequence of the political and cultural effect of the Norman Conquest was not overwhelming, dominance of the Anglo-Normans was that during the and people who spoke only French probably never first three centuries of Middle English, no regional exceeded 2% of the total population. The social and variety of the language achieved prestige and cultural consequences of this event, however, were prominence comparable to that of West Saxon during substantial. The positions of power and prestige and the Old English period. Each geographical area had its the centers of learning and writing passed from Anglo- own scribal and literary traditions. Dialects differed Saxon to Norman hands. Virtually every aspect of the both in sound structure and vocabulary, as well as in history of the language during Middle English has to word and sentence structure. The main north–south be described and studied both on its own terms and in divisions of Middle English are known as Northern, the context of continuing bilingualism and imitation of Midland, and Southern Middle English. Within the Continental sources. south, a separate dialect, Kentish, developed in the Early Middle English usually covers the time from counties of Kent and Sussex. The Midland dialect area the mid eleventh to the mid thirteenth centuries. splits further into two linguistically distinct varieties: Relatively few records written in English survive from East Midland and West Midland. that period; most of the historical documentation and literary production during the Early Middle English Orthography period was in Latin or in Anglo-Norman, the variety of French used by William the Conqueror, his Another consequence of the Norman Conquest was successors, and their courts. For English, it was a time the replacement of the ‘Insular’ script used by the of rapid reduction of grammatical markers and Anglo-Saxons with a Continental form of the Roman borrowing of vocabulary from Anglo-Norman. alphabet known as ‘Carolingian’ script. The letters Central Middle English stretches roughly from the <æ> and <ð> were preserved only in the earliest texts. mid thirteenth to the mid fourteenth centuries, up to The Old English yogh <> was used to represent both the beginning of the Hundred Years’ War (1337–1453) [j], as in ow ‘you,’ and [ç, x], as in hei (e) ‘high,’ and the epidemic of bubonic plague known as the rou(h) ‘rough’; the continental <g> was adopted for Black Death (1348–1350). The Central Middle [g] and [ ]. The letter thorn <þ> was very similar in English period is characterized by further attrition of shape to the letter <y>; hence, ye, yt stand for the, that. inflections, vowel length adjustments, and the (ye, yt were retained in printers’ fonts during the continuing introduction of new loanwords from fifteenth and sixteenth centuries, which gave rise to French, although the use of French was already on the pseudo-antiquarian spellings like Ye Olde Antique decline. Late Middle English is the time during and Shoppe.) New digraphs (two-letter combinations) 688  MIDDLE ENGLISH were introduced for the spelling of many long vowels Old English [talə] Middle [ta:l(ə)] and diphthongs: <ou, ow> for [u:], <ee> for [e:], <oo> talu ‘tale’ English tale for [o:], etc. To avoid the visual confusion of adjacent Old English [nɔzə] ‘ Middle [nɔ:z(ə)] vertical strokes, the letter <u> was replaced by <o> nosu nose’ English nose when it was next to <m, n, u, v, w>. Thus, sume Old English [brεkən] Middle [brε:k(ə)] ‘some’, pronounced [sυm(ə)], huni ‘honey’, brecan ‘break’ English brek(e) pronounced [hυni], and luue ‘love’, pronounced [lυv(ə)], were respelled to some, hon(e)y, loue. New The lengthening often did not take place when the consonant letter combinations introduced by the second, unstressed, syllable of the word was not lost. Anglo-Norman scribes were: <ch, cch, tch> for [], This produces patterns of historically related words like <gg(e)> for [], <gh> for [ç, γ, x], <wh, qu, quh> for game ~ gammon, late ~ latter, shade ~ shadow, throat earlier <hw>, and <sh, ssh, ss> for [ʃ]. ~ throttle, nose ~ nostril. An important spelling consequence of this change is that, in the modern language, words that contain the orthographic sequence Sound System -VCe (any vowel + any consonant + final -e) are usually interpreted as containing a long (or tense) vowel. Middle English was a time of rapid change of the The opposite tendency, toward shortening, was also quality and the quantity of many vowels. The southern a continuation of Old English changes. Long vowels dialects of Late Middle English, which served as the were shortened before two or more consonants, thus: basis of the emerging fifteenth-century standard, had Old English go¯dspell ‘gospel’> Middle English five short vowels and seven long vowels: [gɔsp(ə)l], Old English cre¯pte [kre:pte] ‘crept’ > Short vowels Long vowels Middle English crepte [krεpt(ə)]. These shortenings are responsible for Modern English patterns like cheap I υ i: u: ~ Chapman, dear ~ darling, five ~ fifth, wise ~ e: o: wisdom. This change is also connected with the ε ɔ ε: ɔ modern spelling convention, whereby consonants are a ɑ: doubled when adding suffixes only after short vowels, The colon marks vowel length. In dictionaries and as in beg ~ begged, drum ~ drummer, fat ~ fatter, sit ~ edited manuscripts, length may be indicated by a sitting, red ~ redden (compare fate ~ fated, site ~ macron (¯) over the vowel letter. Whether [ɑ:] or [ɔ:] siting). is used in a particular word is an important dialect The most far-reaching single event in Middle criterion: in the south, Old English [ɑ:] was raised and English sound change, which profoundly affected the rounded to [ɔ:], while the northern dialects retained general shape of the language, was the loss of final the low vowel: compare Northern Middle English unstressed vowels. In early Middle English, final stan(e), mair with Southern Middle English ston(e), unstressed vowels lost their qualitative distinctions more, which developed from Old English sta¯n ‘stone’, and merged into the mid-central vowel schwa [-ə], ma¯ra ‘more’. usually spelled <-e>. By c. 1400, all final schwas, Middle English inherited or developed a range of including a number of grammatical endings, had diphthongs that were to become an important disappeared from the pronunciation; the <-e> component of the changes included in the Great Vowel remained in the language as a spelling convention Shift. The diphthongs of Late Middle English were: indicating the quantity of the vowel in the preceding [ij] as in stil(e) ‘stile’, [ej] as in we ‘way’, [ew] as in syllable, e.g. bite ~ bit, cut ~ cute, mate ~ mat, mete ~ newe ‘new’, [iw] as in stiward ‘steward’, [uw] as in ful met, note ~ not. ‘fowl’, [ow] as in bowe ‘bow’, and [aw] as in lawe During Middle English, the consonantal system ‘law’. Toward the end of the period, the long vowels reached its present-day state except for the later /i:, e:, u:, o:/ had already started changing to /əj, ij, əw, (seventeeth century) introduction of // as in measure. uw/ in the south. Long consonants, as in bedd ‘bed’ and sterre ‘star,’ Middle English continued the Old English tendency were simplified. A combination of factors, most to make vowel length dependent on its environment. A notably the influx of French loanwords, resulted in an change started in the north toward the end of the enriched system of consonant distinctions. The voiced twelfth century, and gradually reached the south by the fricatives, [v, ð, and z], which in Old English could end of the fourteenth, whereby stressed vowels in open appear only in the middle of a word, when flanked by syllables (syllables with no final consonant) were vowels, became independent phonemes, allowing lengthened in words of two syllables. Examples of this minimal pairs such as fast - vast, thigh - thy, and seal development are: ‘acquatic mammal’ - zeal. The consonant /h/ remained 689 MIDDLE ENGLISH stable only initially before vowels. The chart below person singular present tense, the singular subjunctive shows the consonant inventory of late Middle English: for all persons, present and past tense, the singular imperative, etc. The phonetic loss of unstressed vowels Labial Dental Alveolar Palatal Velar resulted in essentially the modern system except for the survival of -(e)st and -(e)th for the second- and Voiceless p t k third-person singular, present tense. The modern third- stops Voiced stops b d g person singular present tense suffix -(e)s, an inflection Voiceless f θ s ʃ h, x of obscure origin, possibly influenced by fricatives Scandinavian, was originally restricted to northern Voiced v ð z texts. It spread to the Midlands and the south during fricatives the fourteenth century and reached London by c. 1400. Affricates ,  Nasals m n Many formerly ‘irregular’ verbs began to develop Liquids l, r regular past tense forms ending in -ed: cleave ~ clove Approximants j w ~ cleaved, help ~ holp ~ helped, leap ~ lope ~ leaped, walk ~ welk ~ walked. During the Middle English period, many complex verb forms and auxiliary verbs continued to develop and Word and Sentence Structure enrich the verbal system. Except for the passive The weakening and loss of final unstressed vowels in progressive forms (is/was being written), all other Middle English resulted in a very significant reduction progressive forms, the perfect, and the pluperfect forms of distinctive grammatical markers. Until the twelfth (has arrived, had arrived) were in use by Late Middle century, English nouns and their determiners agreed in English. The basic modal auxiliary verbs, can, may, grammatical gender: se dæ ‘the day’ (masc.), seo must, were also in place, and the earlier modal spræc ‘the speech’ (fem.), and þæt word ‘the word’ auxiliaries shall and will were often used as pure future (neuter). The Old English gender system was tense markers. The verb do was used as a tense marker, grammatical, because the fact that a noun was a causative verb, and a substitute verb, but not as an feminine or masculine did not necessarily reflect auxiliary verb in its modern functions. Questions were feminine or masculine traits of the object referred to. formed by simple inversion of the verb and the subject: Most of the gender marking was lost during the Seist thou hit me? ‘Are you saying it to me?’ Negation Middle English period, and for the few elements that in the early texts could involve attaching the particle ne still show gender, such as the pronouns he, she, it, the ‘not’ before the verb, but gradually the negative function gender reflects actual sex differences. While Old was taken over by the originally emphatic naht ‘naught, English inflected nouns differently depending on not’ from nawiht ‘nothing’, often positioned whether they occurred as subjects or objects, this immediately after the verb. Multiple negation was distinction was also gradually leveled out and had common: Of fule spaches hie ne mai nauht þolien. ‘(Of) disappeared completely by the end of the Middle foul speaking they not can/may not tolerate.’ English period. The plural and the possessive –es In the course of Middle English, the order of the endings, used only for one subgroup of masculine and sentence elements became essentially identical to that neuter nouns in Old English, spread to all nouns. The of Modern English, Subject–Verb–Object, in both difference between singular and plural adjectives main and subordinate clauses. Fixing the position of (marked by -e for the plurals) was abandoned after the the subject and the object with respect to the verb is end of the fourteenth century. related to the loss of grammatical markers. Unlike The personal pronouns in Middle English had the Modern English, where only certain adverbs such as same distinctions as in Modern English: singular/ ever, never, and sometimes can intervene between an plural, subject/object, first/second/third person, and auxiliary and its main verb, Middle English allowed three genders for the third-person singular. The most the object to be enclosed between the two parts of the important innovation in the pronoun system of Middle verb: I ne haue nowt but mi swerd ibrout. ‘I not-have English was the adoption of the Scandinavian nothing except my sword brought.’ pronouns they, them, their, which replaced Old English hie, hem, hire. Throughout Middle English, there were Vocabulary and Word Formation two separate second-person pronouns: thou, thee, thine/thy (singular) and ye, you, your (plural). The most serious linguistic consequences of the Verb forms in Early Middle English were marked Norman Conquest were in the area of vocabulary and for number, person, tense, and mood. Many of the word formation. It is estimated that approximately inflections were identical, however: -e was the first- 10,000 words were borrowed from Anglo-Norman and 690  MIGRATION AND LANGUAGE French into Middle English. Seventy-five percent of References these early loanwords are still in use. Among them are The richest ME database is the Middle English Dictionary, (a) everyday words: air, beast, close, dangerous, feast, based on the analysis of a collection of over three million flower, jealous, journey, liquor, mountain, noble, river, citation slips. Web access is available through the electronic soil, tender, very, etc.; (b) legal, administrative, and Middle English Compendium, which includes also a hyper- political terms: army, assembly, council, defense, bibliography of Middle English prose and verse and an associated network of electronic resources. judge, liberty, navy, parliament, record, servant, Blake, Norman (ed.) 1992. The Cambridge history of the soldier, tax; and (c) words from the spheres of English language: Vol. II, 1066–1476. Cambridge: literature, art, science, and medicine: beauty, color, Cambridge University Press. grammar music, poet, physician, romance, surgeon. Brunner, Karl. 1887/1963. An outline of Middle English New words continued to be produced by grammar, translated by Grahame Johnston. Oxford: Blackwell. compounding: bedtime (c. 1250), hangman (c. 1393), Fisiak, Jacek (ed.) Studies in Middle English linguistics. Berlin coal-black (c. 1250), to overcast (c. 1225), and and New York: Mouton de Gruyter. worldly-wise (c. 1415) are some examples, with the Jones, Charles. 1972. An introduction to Middle English. New earliest recorded dates in parentheses. Mixed-source York: Holt, Rinehart & Winston. compounds began to appear too: breast-plate (1386), Jordan, Richard. 1925/1970. Handbook of Middle English grammar: phonology, translated by Eugene Joseph Crook, freemason (1376), and knight errant (1350), The Hague: Mouton. combining English with French roots, while in Kurath, Hans, Sherman M. Kuhn and Robert E. Lewis (eds.) commonweal (1330), cornerstone (1300), and 1952–2001. Middle English dictionary. Ann Arbor: gentleman (1275), the French root is followed by an University of Michigan Press and London: Oxford University English root. The inventory of Old English prefixes Press. Electronic access: http://ets.umdl.umich.edu/m/med/. McIntosh, Angus, M.L. Samuels, Michael Benskin, with the and suffixes was increased by a large number of assistance of Margaret Laing, and Keith Williamson. 1986. Romance ones: en-, mis-, re-, sub-, able-, -ance, -ess, A linguistic atlas of late mediaeval English, 4 vol. Aberdeen -ment, -ous, which could produce mixed-origin words and New York: Aberdeen University Press. such as talkative, unknowable, wizard (English roots + McSparran, Frances (ed.), 2002. Middle English Compendium. Romance suffixes), and colorless, cheerful, Ann Arbor: University of Michigan press. Electronic access: http://ets.umdl.umich.edu/m/med/. spousehood (Romance roots + English suffixes). Minkova, Donka. 1991. The history of final vowels in English, Related to the fact that Middle English lost final Berlin and New York: Mouton de Gruyter. unstressed vowels was the new tendency of freely Mossé, Fernand. 1968. A handbook of Middle English, converting words from one part of speech to another, translated by James A. Walker. Baltimore: Johns Hopkins i.e. nouns could simply be used as verbs, verbs as Press. Mustanoja, Tauno F. 1960. A Middle English syntax. Helsinki: nouns, adjectives as verbs, etc., without any particular Société Néophilologique. marking. Examples of this process are the verbs child Wright, Joseph, and Elizabeth Mary Wright. 1928. An ‘give birth’ (c. 1200), calm (c. 1399), cripple (c. 1300), elementary Middle English grammar, 2nd edition. London: tame (c. 1315), and word ‘speak’ (c. 1205), all formed Oxford University Press, Reprint 1967. from earlier nouns and adjectives. DONKA MINKOVA Migration and Language In some sense, probably all the world’s languages are indigenous languages of the Americas and Australia migrants. In a few cases, languages are spoken in a have been largely replaced by Western European particular region because the ancestors of their languages during the last few centuries. Such speakers were the first people ever to live there: thus, replacements may occur gradually if the migrants have Austronesian languages were spread across previously techniques of food production that enable them to uninhabited Oceanic islands between about 1600 BCE outreproduce the original inhabitants (demic diffusion and 1300 CE. Often, however, languages are carried or the wave of advance model): thus, Bantu-speaking by new settlers into a region already inhabited by farmers spread over much of southern Africa between speakers of different languages: hence, for instance, about 3000 BCE and 500 CE, into territory previously 691 MIGRATION AND LANGUAGE occupied by hunter–gatherers and herders speaking often correspond to linguistic divisions: for instance, non-Bantu languages. Alternatively, a relatively small Rh-negative individuals are considerably more but highly organized group may impose a new common among the Basques than other Europeans, language on an existing population through military and Basque is a linguistic isolate, unrelated to its Indo- conquest (élite dominance), as seems to have occurred European neighbors. He argues that the Basques may in the Roman empire, giving rise to the Romance represent a pre-Indo-European people, confined to a languages. small mountainous location by later population Migration may have a range of linguistic incursions. consequences of great theoretical interest. If it results Besides the mere existence of language families, in two languages coming into contact, one may die out other linguistic information can contribute to an because of deliberate suppression or because the understanding of early migrations. According to the speakers choose to shift to the other language on age–area hypothesis (formalized by Isidore Dyen as socioeconomic grounds. If both remain in existence, migration theory), the location of greatest diversity there may be lexical borrowing, particularly into the within a language or family is likely to be the original language perceived as having less prestige (thus, the homeland from which the family spread, because the Roman occupation of Britain resulted in extensive deepest divisions reflect the earliest splits; moreover, borrowing of vocabulary from Latin into Brittonic this theory minimizes the number of movements Celtic). In cases of long-term stable multilingualism in required to explain the current distribution. Thus, a community, the languages may converge in English dialectal diversity is greater within Britain structure, as has been observed in Kupwar in India than in the more recently settled English-speaking between Indo-European and Dravidian languages. parts of the world. Linguistic paleontology (the study Finally, a pidgin may arise when speakers of different of reconstructed cultural vocabulary) may also provide languages have some strong need to communicate, but clues to the area of origin of a family. A third line of cannot for some reason acquire each other’s languages evidence is provided by borrowed vocabulary, fully, as, for instance, in trading situations or on slave revealing contact with other groups: the movements of plantations in the colonial era. the Gypsies from India to Iran, the Caucasus, Anatolia, Another particularly significant result of migration and then to the Balkans have been traced by Kaufman is the creation of language families. When a speech in various sets of loanwords in Romani. community splits because of movement of some A recent strand of research relating to migration and speakers to new areas, the language spoken by the language has considered the distribution of linguistic different groups will diverge as various changes diversity around the world. Johanna Nichols, drawing accumulate in each population (resulting from internal on techniques of the population sciences, has sought to drift or external influence) and are unable to propagate quantify structural and genetic diversity of languages in to the other populations. Over time, this diversification different areas and has distinguished between residual produces a “family tree” of languages (the zones (characterized by high genetic density and Austronesian languages seem to provide a particularly structural diversity, often in mountainous regions clear-cut example). However, if contact is maintained serving as linguistic refuges: the Caucasus is an between the speech communities, changes will be able example) and spread zones (whose low genetic and to diffuse between neighboring groups, producing a structural variation results from the spread of a single dialect continuum of related speech forms, rather than language or family across the area: an instance is the a clearly defined tree (as is still seen to some extent in Eurasian steppe). She has also argued that the high Western Europe within parts of the Romance and diversity in the Americas implies earlier and more Germanic groups). numerous settlements than those proposed by Joseph Such language families provide the basis for Greenberg, whose ‘Amerind’ people are claimed to comparative historical linguistics. However, their have entered the Americas approximately 14,000 years existence has also been used by several nonlinguists as ago. Daniel Nettle, however, adopting Robert Dixon’s evidence for prehistoric population movements. Thus, evolutionary–biological model of punctuated equilibria from an archeological perspective, Colin Renfrew has for language change, has opposed her theory that suggested connecting the Indo-European family with languages ramify at a constant rate over time. He the spread of farming from Anatolia across Europe and proposes that the initial movement of people into a large India after approximately 6500 BCE and linking the uninhabited area (such as the Americas) would promote controversial Nostratic language grouping with still rapid diversification as communities occupied available earlier movements from the Fertile Crescent. Luigi habitat niches, followed by gradual areal convergence Cavalli-Sforza has noted that genetic discontinuities or local language replacement, resulting in a decline in 692  MISKITO AND MISUMALPAN LANGUAGES linguistic and genetic diversity. A Neolithic transition in Nettle, Daniel. 1999. Linguistic diversity. Oxford: Oxford an area would provoke a faster reduction of diversity, as University Press. Nichols, Johanna. 1990. Linguistic diversity and the first farming groups advanced through previous hunter- settlement of the New World. Language 66. gatherer territory (as in Eurasia and Africa, where initial Nichols, Johanna. 1992. Linguistic diversity in space and time. human settlement and development of food production Chicago and London: University of Chicago Press. were comparatively early). Renfrew, Colin. 1987. Archaeology and language: the puzzle of Indo-European origins. London: Jonathan Cape. Renfrew, Colin, 1999. Nostratic as a linguistic macrofamily. References Nostratic: examining a linguistic macrofamily, ed.by Colin Renfrew and Daniel Nettle. Cambridge: McDonald Institute Campbell, Lyle. 1998. Historical linguistics: an introduction. for Archaeological Research. Edinburgh: Edinburgh University Press. Sapir, Edward. 1949. Time perspective in Aboriginal American Cavalli-Sforza, Luigi. 2000. Genes, people and languages. culture: a study in method. Selected writings of Edward London: Allen Lane. Sapir in language, culture and personality, ed. by David Dixon, Robert. 1997. The rise and fall of languages. Mandelbaum. Berkeley: University of California Press. Cambridge: Cambridge University Press. MARISA LOHR Dyen, Isidore. 1956. Language distribution and migration theory. Language 32. Greenberg, Joseph. 1987. Language in the Americas. Stanford: See also Archeology and Language; Genetic Rela- Stanford University Press. tionship; Pidgins and Creoles Miskito and Misumalpan Languages Miskito is the most widely spoken language of a Nicoya Peninsula of modern Costa Rica. These Central American language family known as migrants were known by the Spanish as the Chorotega Misumalpan. John Mason coined the term and they left behind two now extinct languages–– Misumalpan using the first two letters of its three Mangue and Subtiaba. Uto-Aztecan migration branches: Miskito, Sumo, and Matagalpan. Miskito occurred slightly after the Otomanguean migration. and Sumo are spoken in the Mosquito Coast region of The first Nahau-speaking people arrived in the ninth Eastern Nicaragua and Honduras by approximately and tenth centuries and settled in what is now El 150,000 and 8,000 people, respectively. The now Salvador and Northwestern Nicaragua. The second extinct Matagalpan languages, Matagalpa and wave of migrants, whose languages came to be known Cacaopera, were spoken in the Central Highlands of as Pipil or Nicarao, arrived in the twelfth century and Nicaragua and Southeastern El Salvador. settled on the Western shore of Lake Nicaragua. Misumalpan is a relatively small language family By the time of the Spanish conquest, Misumalpan that in the pre-Columbian period was surrounded languages had virtually been eliminated from the geographically by larger and more widespread Pacific Lowlands. However, the western branch of language families from Mesoamerica (Otomanguean, Misumalpan languages continued to be widely spoken Uto-Aztecan, and Mayan) and lower Central America in the Central Highlands. Matagalpa, also known as (Chibchan). Before the ninth century, Misumalpan Pantasmas, survived throughout the colonial period languages are presumed to have been spoken in all of but by the end of the nineteenth century it had been the three major ecological zones of Nicaragua: (1) the replaced by Spanish. Nevertheless, an Indian ethnic Pacific Lowlands, (2) the Central Highlands, and (3) identity and certain communal institutions and land the Caribbean Coastal Plain. However, starting at this tenure systems continue into the present in the time, two major waves of migration from Mesoamerica Nicaraguan province of Matagalpa. In the Morazán resulted in Mesoamerican languages displacing province of Southeastern El Salvador, Cacaopera Misumalpan languages to the East. survived into the twentieth century but at present, it Speakers of the Otomanguean language family who has been completely replaced by Spanish. The case of originated in Soconusco, Chiapas comprised the first Cacaopera has puzzled linguists because it occurred wave that settled in the Pacific Lowlands and the well out of the Nicaraguan homeland of Misumalpan 693 MISKITO AND MISUMALPAN LANGUAGES languages. Together, Matagalpa and Cacaopera This Miskito historical ascendance explains in part comprise Matagalpan, the name used to refer to the the relative homogeneity of Miskito as well as its western branch of Misumalpan languages. extensive borrowing of vocabulary from English and Although several distinct dialects of Miskito exist, West African languages, whose influence was brought Miskito is more homogenous than Sumo and to the region as a result of English slavery and Matagalpan and it is therefore considered an isolate subsequent West Indian migration to coastal Central within the language family. Sumo and Matagalpan are America precipitated by banana, mining, and lumber more internally differentiated and are also more businesses. While Miskito also borrows from Spanish, closely related to one another than either is to linguists have found that Sumo and Matagalpan Miskito. Therefore, they are grouped by linguists in a languages demonstrated a clear preference for subfamily called Sumalpan. In turn, Sumo has two borrowing from Spanish, while Miskito has distinct branches––Northern Sumo and Southern historically borrowed primarily from English. This Sumo. Three extant dialects of Northern Sumo, undoubtedly results from the coastal and external trade Tuahka, Panamahka, and Tawahka, continue to be orientation of the Miskito as opposed to the upriver spoken today while Southern Sumo has a single and subsistence agricultural pattern of the Sumo, who extant representative, Ulwa, and other extinct dialects live on the eastward-moving frontier between Spanish- such as Kukra. Tuahka, also spelled Twahka, is speaking Western Nicaragua and the Mosquito Coast. spoken by a few hundred people in and around the In a volume about Mosquito Coast Creole English, village of Wasakin along the Bambana River in the Johm Holm extensively studied the influence of Mosquito Coast. Panamahka is the most common Mosquito Coast English and West African languages Sumo dialect and it is spoken by about 6,000 speakers on Miskito. He concluded that Miskito grammar and living in the area of the Waspuk River and its phonology was not significantly influenced by African tributaries in Nicaragua. Tawahka is spoken in languages but lexical borrowing from among others, Honduras by about 1,000 people around the Patuka Fon, Twi, Bambara, Igala, Yoruba, and Kongo, is River. Finally, the Southern Sumo language Ulwa is unmistakable. For example, Miskito people tell spoken in the village of Karawala on the Rio Grande ‘anansi’ stories (from the Twi word for spider) and de Matagalpa by about 350 people. Ulwa is in the their sorcerers are known as ‘sukias’ (from the Fon final stages of extinction as residents of Karawala, word sukuna). Whereas the Sumo word for river most of whom define themselves as ethnically Ulwa, mouth is ‘tikipas’, the Miskito word is ‘bila’––derived are becoming Miskito, English, and Spanish speakers, from the Kongo ‘bwiila’. Mosquito Coast English and as opposed to the long-standing pattern of Miskito- Miskito are deeply interpenetrated. Unlike Sumo and Ulwa bilingualism. Matagalpan languages, Miskito was able to project Whereas Sumo and Matagalpan languages have words into Caribbean Creole Englishes spoken in geographically receded over the last 1,000 years, Belize, Jamaica, and elsewhere as well as British Miskito expanded during the colonial period. Believed English itself. For example, the English word dory to have originated in the Cape Gracias, a Dios region (canoe), which has an Oxford English Dictionary at the terminus of the Coco River that forms the entry, entered British English through Miskito. eastern border between Nicaragua and Honduras, The ascendancy of Miskito people also explains in Miskito is now spoken by over 100,000 people living part the structural similarities between Misumalpan along the coast and major rivers between the languages. In other words, linguists have noted the Escondido River in Nicaragua and the Black River in anomalous fact that based on the large degree of Honduras. As a result of the dislocations caused by the lexical differentation, Misumalpan languages have Contra-Sandinista War, Miskito speakers now live in clearly been separated from a common ancestor for a large numbers in Managua, Miami, and Port Arthur, long time. In fact, based on the method of Texas. In alliance with the British who became glottochronology, Morris Swadesh estimated that 43 commercially and militarily active in the region minimum centuries have passed since they diverged. starting in the 1640s, Miskito-speaking people became Yet, despite their distant point of divergence, they a strong force in the Western Caribbean allowing them possess remarkably similar grammatical structures, to expand their area of influence and settlement well which at first glance would suggest a more recent beyond their traditional boundaries. As a result, common origin. With regard to the similarities in Miskito place names can be found as far as the grammatical structure between Miskito and Sumo, the Caribbean coastal areas of Costa Rica and Panama. widespread and historically deep pattern of Hence, the Sixaola River that divides Costa Rica multilingualism in the Mosquito Coast explains in part comes from the Miskito ‘Siksa’ (Black) and ‘Awala’ this incongruity. The majority of Sumo speakers, who (River). have historically been in a subordinate position to 694  MISKITO AND MISUMALPAN LANGUAGES Miskito speakers, also speak Miskito as a second the Votic branch of Chibchan that includes Costa language. Rica’s Guatuso. Collete Craig has intensively studied The specific nature of multilingualism in the and sought to preserve this language, which has now Mosquito Coast is directly related to the noted ethnic been almost completely replaced by Creole English. hierarchy and division of labor in the region that Adolfo Constenla Umaña, a Costa Rican scholar, has developed in the late nineteenth and early twentieth made a well-documented case for the plausibility of a centuries when North American banana, lumber, and shared common ancestor between Chibchan and mining industries operated heavily in the area. During Misumalpan. In an ambitious synthetic effort to this time, English-speaking Creoles, Afro- categorize languages of the Americas, Joseph Nicaraguans who emphasize their cultural and social Greenberg presents 226 etymologies in his case for a connections to the West Indies, were the dominant high-level ‘Chibchan-Paezan’ family that includes locals who occupied the highest status occupations in Misumalpan as well as distant languages that have not the enclave economy of the region. Miskito and been previously thought to be related such as Spanish-speaking laborers from the Pacific occupied Yanomami, Tarascan, and Chimu. the next rung followed by Sumo Indians at the bottom With the exception of Miskito, Misumalpan of the hierarchy. Susan Norwood has noted a languages are in danger of becoming extinct in the sociolinguistic landscape in which the lower status near future. Despite the tireless efforts of Nicaraguan groups tend to learn higher status languages in and North American linguists such as Collete Craig for addition to their own langauges. Hence, Miskito Rama and Ken Hale and Tom Green for Ulwa, both speakers typically learn Creole English and/or these languages are not likely to last another Spanish, while Sumo speakers tend to learn Miskito generation. Sumo, whose Northern variant, is now and Creole and/or Spanish when possible. Miskito sometimes called Mayangna, stands in a better speakers rarely learn Sumo, and English speakers position, particularly in light of recent legal victories often have some proficiency in Miskito but none in for Sumo communities in Nicaragua. Sumo. Spanish speakers have by far the highest degree of monolingualism. Norwood found that some Northern Sumo speakers in the Ulwa village of References Karawala spoke Panamahka, Ulwa, Miskito, English, Campbell, Lyle. 1997. American Indian languages: the and Spanish. Such situations of intense multilin- historical linguistics of native America. New York: Oxford gualism and social stratification have had clear University Press. linguistic consequences, including the aforementioned Constenla Umaña, Adolfo. 1990. Una hipótesis sbore la localización del protochibcha y la dispersión de sus structural similarities. descendientes. Filología y Linguística (San Jose, Costa In 1920, German linguist Walter Lehmann Rica) 13. 129–61. convincingly demonstrated the unity of Misumalpan Craig, Colette, and Kenneth Hale. 1992. A possible Macro- languages. Since then, linguists have speculated about Chibchan etymon. Anthropological Linguistics 34. its relationship with other language families. Until 173–201. Green, Thomas. 1999. A lexicographic study of Ulwa. Ph.D. recently, linguists have commonly accepted the so- Dissertation. MIT. called ‘Macro-Chibchan hypothesis’, that postulates Greenberg, Joseph. 1987. Language in the Americas. Stanford: that Misumalpan and Chibchan languages share a Stanford University Press. common origin. This contention is based on an Hale, Ken, and Danilo Salamanca. 2002. Theoretical and evaluation of the distribution of Chibchan languages in universal implications of certain verbal entries in dictionaries of the Misumalpan langauges. Making lower Central America as well as comparisons of dictionaries: preserving indigenous languages of the extant Chibchan languages to Misumalpan languages. Americas, ed. by William Frawley, Kennetch Hill and Recently, linguists such as Ken Hale, Danilo Pamela Munro. Berkeley: University of California Press. Salamanca, and Lyle Campbell have withdrawn their Holm, John. 1978. The Creole English of Nicaragua’s Miskito support from the inclusion of Mismulpan languages in Coast: its sociolinguistic history and a comparative study of its lexicon and syntax. Ph.D. Dissertation. University of Macro-Chibchan based on the inconclusiveness of the London. evidence. Lehmann, Walter. 1920. Zentral-Amerika, Teil I, Die Sprachen Arguments for a Chibchan connection point to the Zentral-Amerikas in ihren Beziehungen zueinander sowie presence of two clearly Chibchan languages both to zu Sud-Amerika und Mexiko. Berlin: Dietrich Reimer. the North and South of the Misumalpan heartland. Mason, John. 1940. The native languages of Middle America. The Maya and their neighbors, ed. by Clarence Hay. New Paya, also known as Pech, is a Chibchan language York: Appleton Century Company. spoken by about 300 people in southern Honduras. Norwood, Susan. 1993. El Sumo, Lengua Oprimida. Wani 14. Rama, spoken by about 25 people on Rama Cay close 53–64. to Bluefields, Nicaragua, has been identified as part of BARON PINEDA 695 MODERN LINGUISTICS Modern Linguistics The discovery of Sanskrit in 1786 by Sir William Jones The beliefs pervading linguistic thought up to then is a key event in the History of Linguistics because it are altered late in the century due to the neogrammarian initiates the age that extends up to the present time. The doctrine (1870–1900) defended by scholars such as era of Modern Linguistics, as it is named, thus covers Hermann Osthoff, Karl Brugmann, and Hermann Paul. two centuries that, as will be seen, differ greatly in their The neogrammarians, thus, initiate a new phase in the linguistic methodology and orientation. development of Comparative and Historical Linguistics, Despite being the age of Comparative and Historical which has to be understood, in general, as a reaction Linguistics, the nineteenth century must be considered a against the research carried out previously. Notice, for heterogeneous era where three stages can, at least, be example, that they reject, among others, Humboldt and identified: the early, the mid-, and the late nineteenth Schleicher’s idea about the organic nature of language, century. The initial phase of the century is, for example, and that the grammatical laws they postulate do not particularly important for being the period when the allow, in contrast to the earlier ones, any kind of birth of this new linguistic tradition takes place. The irregularity; hence, their need to develop the concept of resemblance of Sanskrit to classical languages such as analogy. Latin and Greek and also to several Germanic languages The age of Comparative and Historical Linguistics brings to light at the end of the eighteenth century two comes, nevertheless, to its end at the beginning of assumptions that, as illustrated in the works by Rasmus the twentieth century with the posthumous publication Rask (1787–1832), Jacob Grimm (1785–1863), and of Ferdinand de Saussure’s Cours de linguistique Franz Bopp (1791–1867), become the hallmark of the générale (1916), which, as deduced from its linguistic research initiated in the early Romantic age: most basic contents, means a return to Descriptive (a) the existence of a common ancestor language, named Linguistics: (a) as a system of interdependent Indo-European, for these apparently unconnected phonological, syntactic, and lexical elements, a lan- languages; (b) and consequently, their possible guage has form and substance; (b) the linguistic comparison, as well as, by extension, that of all lan- competence of a speaker is different from his real guages over the world. Notice, in fact, that the first sound linguistic production; hence, the distinction between law intended to reconstruct the phonological system of langue and parole; and (c) the study of language can Indo-European, known as Grimm’s law, and the first be approached either from a diachronic or from a linguistic typology established by Wilhelm von synchronic perspective. Humboldt (1767–1835) according to the morphological The influence of this work is, in fact, quickly seen pattern the Indo-European languages exhibit, go back to in Europe, where conferences, publications, and this specific period. schools with a clear structural orientation constantly The mid-nineteenth century is, in turn, remarkable appeared during the first half of the century: among because, transferring to the linguistic situation the others, the Prague School with Nikolaj Trubetzkoy and growing interest of the time in history and the natural Roman Jakobson, the Copenhagen School with Louis sciences, it entails the full establishment of the Hjelmslev and the London School with John Rupert comparative and historical method of linguistic Firth. Although also pervading the American thought analysis. As a consequence, the strictly contrastive up to the 1950s, the structural linguistic trend takes a interest of the previous stage is supplemented in this slightly different course in the United States owing to intermediate phase of the Romantic age, as reflected in the following facts: (a) the existence of unknown its most significant contributions, with an enormous American-Indian languages, which arouses an desire to investigate the historical and genetic evolution anthropological interest in linguists such as Franz of languages: August Schleicher’s Genealogical Tree Boas and Edward Sapir; and (b) the influence of Model (1821–1868) is, for instance, designed in behaviorist psychology, which developed around the consonance with the Darwinian theory to establish the hypothesis that all kinds of analysis must be based on genetic relationships between Indo-European and its describable facts, excludes the semantic component of descendants, and Johannes Schmidt’s Wave Model language from the structural theories of outstanding (1872) is alternatively devised to explain the fact that linguists such as Leonard Bloomfield. genetically unconnected languages can and, in fact, do The publication of Noam Chomsky’s Syntactic share several linguistic phenomena. structures in 1957 finishes, nevertheless, with the 696  MODIFICATION structural theoretical framework and initiates, languages, and the twentieth century chiefly consequently, a new period in the History of Descriptive concerned, in turn, with the description of the Linguistics. The era of Transformational-Generative linguistic system. Grammar, as it is called, signifies a sharp break with the linguistic tradition of the first half of the century both in References Europe and America because, having as its principal objective the formulation of a finite set of basic and Bierwisch, Manfred. 1971. Modern linguistics. Its transformational rules that explain how the native development, methods and problems. The Hague and Paris: Mouton. speaker of a language can generate and comprehend all Botha, R.P. 1992. 20th century conceptions of language. its possible grammatical sentences, it focuses mostly on Oxford: Blackwell. syntax, and not on phonology or morphology, as Koerner, E.F.K. 1976. Towards a historiography of linguistics. structuralism does. Its situation changes, however, from 19th and 20th century paradigms. History of linguistic the 1980s onward because, with the increasing interest thought and contemporary linguistics, ed. by Herman Parret. Berlin and New York: Walter de Gruyter. in meaning awakened in this particular period, the Marcos Marín, Francisco. 1990 Introducción a la Lingüística: Transformational-Generative model ceases to be the Historia y Modelos. Madrid: Síntesis. only theoretical framework of the time and instead Mohrmann, C., A. Sommerfelt, and J. Whatmough. (eds.) 1961. coexists with several new approaches to the study of Trends in European and American linguistics 1930–1960. language, labeled as functional for regarding, in sharp Utrecht: Spectrum. Mohrmann, C., F. Norman, and A. Sommerfelt. (eds.) 1963. contrast to the latest version of the Chomskian theory, Trends in modern linguistics. Utrecht: Spectrum. semantics and pragmatics as the most basic levels of Mounin, Georges. 1967. Histoire de la linguistique dès origines linguistic analysis. In any case, the last modification of au XXe siècle. Paris: Presses Universitaires de France. Transformational-Generative Grammar nowadays Mounin, Georges. 1972. La linguistique du XXe siècle. Paris: continues to be the leading linguistic theory from Presses Universitaires de France. Newmeyer, Frederick J. 1998 Language form and language among all the formal ones that, for their part, focus function. Cambridge, MA: MIT Press. mainly on the structural components of language and Pedersen, Holger. 1931. The discovery of language: linguistic show, consequently, an almost complete disregard for science in the nineteenth century. Bloomington: Indiana meaning. University Press; 2nd edition, 1962. In sum, the age of Modern Linguistics lasts two Robins, R.H. 1967. A short history of linguistics, London: Longman. centuries, which, although heterogeneous in Waterman, J.T. 1963. Perspectives in linguistics: an account of themselves, differ greatly in their linguistic interests: the background of modern linguistics. Chicago: Chicago the nineteenth century mainly devoted to the University Press. comparison and historico-genetic evolution of BEATRIZ RODRÍGUEZ-ARRIZABALAGA Modification Two properties semantically motivate the syntactic asterisk preceding the hypothetical expression); noun categories clause and noun phrase, respectively: phrases whose head is a name instead of a noun predication and reference. Clauses involve a undergo determination and modification under certain predication, that is, they make a judgment about an circumstances, as in the big Jack. entity, as in The weather was changeable. Noun Determination is a function of the noun phrase that phrases exhibit the property of reference, whereby is performed, in English, by grammatical categories linguistic entities make reference to extralinguistic (as opposed to lexical categories, which include the entitites, as in the green pencil. Fully referential noun noun, the adjective, the verb, and the adverb): the phrases may be both determined and modified. closed classes of the article (a and the), the Pronominal noun phrases do not admit either demonstrative (this-these/that-those), the quantifier determination or modification (*that she or *outgoing (some, much, etc.; also numerals like one-first, etc.) he, for instance, are ungrammatical in Contemporary and the possessive adjective. Unlike determination, English, which is indicated by the presence of the modification is not exclusive of the noun phrase. 697 MODIFICATION Modification is a function of the noun phrase, the recursive (i.e. repetitive) character of modification: adjectival phrase, the adverbial phrase, and the sometimes, modifiers contain heads that are modified prepositional phrase, which is realized by the syntactic themselves, as in the woman with the glittering categories of phrase and clause. Instances of jewellery, where the prepositional phrase that modifies modification in the noun phrase, the adjectival phrase, the nominal head woman governs a noun phrase in the adverbial phrase, and the prepositional phrase are, which the nominal head jewellery is modified by the respectively: the sharp edge, difficult to handle, adjectival phrase glittering. incredibly fast, and well below zero. These examples Examples like the sharp edge, difficult to handle, illustrate the contrast between phrasal and clausal and incredibly fast also show the difference between modification: the adjectival phrase sharp modifies the premodification and postmodification in English. The nominal head edge, the adverbial phrase incredibly linear order is modifier–modified in the sharp edge and modifies the adverbial head fast, and the adverbial incredibly fast whereas in the adjectival phrase difficult phrase well modifies the prepositional below, whereas to handle the order qualifies as a modified–modifier. the embedded clause to handle modifies the adjectival The former linear order is called premodification and head difficult. Phrasal and clausal modification give the latter is called postmodification. Also of structural rise to different degrees of phrase elaboration: phrasal import is the question of the syntactic (dis-) continuity modification turns out simple phrases, such as the of modifiers. When modifier and modified are noun phrase a paltry ten pounds; whereas clausal adjacent, the linear arrangement is described as syntac- modification produces complex phrases, as is the case tically continuous, as in a man who wore a panama with the adjectival phrase impossible to forget. hat called in, where no constituent has been inserted In English, the inventory of phrases as modifiers between the relative clause and its antecedent. If includes the noun phrase, the adjectival phrase, the modifier and modified are not adjacent, the linear adverbial phrase, and the prepositional phrase (in other arrangement is termed syntactically discontinuous, as words, all phrasal syntactic categories with the in a man called in who wore a panama hat, where the exception of the verb phrase). Noun phrases as modi- verb phase separates the modified (the antecedent) and fiers of other noun phrases belong to two classes: they the modifier (the relative clause). Whereas relative are either nominal compounds (lexicalized or not), as order in the phrase and adjacency represent syntactic in the picture pen, or possesssive modifiers in the phenomena, the phenomenon of double (or multiple) genitive, as in Sally’s promotion. Adjectival phrases modification is essentially semantic since the structural modifying nominal heads fall into the descriptive and arrangement of the modifiers with respect to the head classifying subtypes: descriptive adjectival phrases reflects the degree of semantic integration of the entity focus on a certain property of the noun that functions and the properties that are associated with it (typically, as a nominal head, as in the high wall; classifying in the context of several adjectives modifying a single adjectival phrases assign a label of type or class to the noun, as in a small round green box). noun in function of nominal head, as in the regimental Considering clauses as modifiers, they can be either headquarters. Even though both descriptive and finite or nonfinite, depending on whether the verbal classifying adjectival phrases perform the function of predicate agrees with the subject or not. Finite clauses restricting the semantic scope of the modified, in the function of modifier include the relative clause, descriptive adjectival modifiers in English admit the nominal modifier clause, and the adjectival intensification with very (a very high tower), whereas modifier clause. Relevant examples of relative clauses classifying adjectival phrases do not (*very regimental modifying nominal heads are the man who married headquarters). Typically, adverbial phrases modify my sister (unreduced), and the woman with the adjectival heads in adjectival phrases, as in walking-stick, the manager promoted by the new surprisingly difficult. Adverbial phrases also function directors, and the girl riding that bike (reduced). as modifiers of other adverbial phrases, as is the case Nominal modifier clauses can be ilustrated by means with very deeply. Instances of adverbial phrases as of examples like the conclusion that the plan is modifiers of noun phrases include adverbial genitives feasible. Adjectival modifier clauses contain, typically, like today’s in today’s top ten. Prepositional phrases to-infinitive––thus, nonfinite––clauses in adjectival modifying nominal heads typically involve possessive constructions like easy to misinterpret. postmodification as in the branch of the tree. Morphologically, modifying categories, particularly Postmodification by means of prepositional govern- adjectives, may be either marked or unmarked, that is, ment is also present in noun phrases like the shop with variable or invariable. In Contemporary English, the dusty window. An instance of a prepositional modifiers are morphologically unmarked or invariable phrase modifying an adverbial head is far away from (with the exception of the genitive in position of here. All the examples given above focus on the premodification -John´s mistake- and postmodification 698  MOHAWK AND THE IROQUOIAN LANGUAGES -a mistake of John´s). Consequently, the tendency is Dik, Simon. 1989. The theory of functional grammar, part 1: for the modified element to carry the morphological the structure of the clause. Dordrecht: Foris. Dik, Simon. 1997. The theory of functional grammar, part 2: mark (the distinctive feature). In other historical stages complex and derived constructions, ed. by Kees Hengeveld. of the English language, such as Old English, the Berlin and New York: Mouton de Gruyter. modifying adjective showed agreement of gender, Downing, Angela, and Philip Locke. 1992. A university course number, and case with the nominal head, thus bearing in English grammar. New York and London: Prentice-Hall. the morphological mark. Langacker, Ronald. 1991. Foundations of cognitive grammar II: descriptive application. Stanford: Stanford University Press. Van Valin, Robert D., and Randy LaPolla. 1997. Syntax: References structure, meaning and function. Cambridge: Cambridge University Press. Aarts, Flor, and Jan Aarts. 1988. English syntactic structures. Functions and categories in sentence analysis. New York and JAVIER MARTÍN ARISTA London: Prentice-Hall. Crystal, D. 1980. A dictionary of linguistics and phonetics. See also Determiner; Predication; Syntactic Cate- Oxford: Blackwell; 3rd. edition, 1991. gory Mohawk and the Iroquoian Languages Mohawk is a language of the Iroquoian family. The well, and the survivors, a mixture of Huron, Petun, family consists of two major branches: Southern Erie, and Neutral, moved west to Sandwich, Ontario, Iroquoian and Northern Iroquoian. Southern Iroquoian and became known as the Wyandot. Many is represented by just one language, Cherokee, now subsequently moved southward, ultimately ending up spoken primarily in North Carolina and Oklahoma. in Oklahoma. The Wyandot dialect of Huron was last Northern Iroquoian has several sub-branches. The spoken in the mid-twentieth century. French first offshoot of Northern Iroquoian developed into missionaries among the Huron, particularly during the Nottoway, Meherrin, and Tuscarora. The Nottoway seventeenth and eighteenth centuries, left rich records people were first encountered by Europeans near the of the language. Wyandot texts and lexical and Virginia coast in 1650. The language, which grammatical material were recorded in the early disappeared during the mid-nineteenth century, is twentieth century. Petun, Neutral, Wenro, and Erie are known through just two wordlists from the early part known only through some names. of that century. The Meherrin people were first Five of the remaining Northern Iroquoian groups, encountered in 1650 near the North Carolina coast, but the Seneca, Cayuga, Onondaga, Oneida, and Mohawk, by 1730 they had merged with the Tuscarora. All that formed a strong political alliance known as the League remains of their language are two town names. The of the Five Nations Iroquois. Their territories stretched Tuscarora were first encountered in eastern North essentially from the western edge of modern New York Carolina. Early in the eighteenth century, most began State, where the Senecas were known as the Keepers of to move northward, where their descendants reside the Western Door, to the eastern edge, where the today in two locations: near Niagara Falls in eastern Mohawks were the Keepers of the Eastern Door. New York State and at Six Nations in southern Another group to the north, now known as the Ontario. Few speakers remain. Laurentian, was not part of the League, nor was a group The second offshoot of Northern Iroquoian was to the south, the Susquehannock. The Laurentians met Huron. The earliest mention of the Huron people is in Jacques Cartier in 1534 at the mouth of the St. Champlain’s account of his 1615 visit to what is now Lawrence River, but they had disappeared from the southern Ontario. The Huron Confederacy, consisting area by the time of Champlain’s arrival in 1603. of four tribes, was decimated in 1649 by attacks from Vocabulary lists of their languages remain from the the Five Nations Iroquois. Some survivors fled toward time of Cartier’s voyage, and it is from them that we Québec City, where their descendants live today at have the name Canada¸ a term that persists in the Lorette. The language is no longer spoken there. modern Five Nations languages today meaning Others settled with other Iroquoian groups in the area. ‘settlement, town’. The Susquehannock to the south, Some of these groups were subsequently defeated as also known as the Andaste, suffered during the 699 MOHAWK AND THE IROQUOIAN LANGUAGES colonial period from European diseases and attacks leaving just vowel length. Open, stressed syllables are from colonists and the other Iroquois until the last lengthened, as in kí:ken ‘this’. Length is marked survivors were murdered in 1763. Their language is orthographically with a colon:. known through a wordlist recorded by a Swedish Nouns are based on a noun stem, which may be a missionary in his journal published in 1696. noun root or a nominalized verb stem. Morphological The Five Nations languages are all still spoken nouns begin with a noun prefix. In unpossessed nouns, today primarily by elders, although all communities the prefix encodes the gender of the referent, like the have language classes. Seneca is spoken in three Neuter o- in o-tsíhkw-a’ ‘fist, knot, knob, puck’. If the communities in western New York: Cattaraugus, noun is possessed, the prefix encodes the person, Allegany, and Tonawanda. Cayuga is spoken at Six number, and gender of the possessor, as in akw- Nations in southern Ontario, where many Cayugas atsíhkwa’ ‘my puck’. Different sets of possessive fled after the American Revolution. Other Cayugas prefixes are used for Alienable and Inalienable moved into Oklahoma, where the language was possession. The prefix akw- ‘my’ in ‘my puck’ spoken until the late twentieth century. Onondaga is indicates Alienable possession (meaning that the spoken in central New York south of Syracuse and at possessed is not an integral part of the possessor). The Six Nations. Some Oneidas remain in New York prefix k- ‘my’ in k-atsihkwà:ke ‘my fist’ indicates State, but most moved to southern Ontario near Inalienable possession (meaning that the possessed is London, and to Wisconsin near Green Bay. There are an integral part of the possessor). Inalienable six main Mohawk communities: Six Nations in possessions include most body parts, but not hair or Ontario; Tayendinaga near Deseronto in Ontario; internal organs such as the heart or stomach. Separated Ahkwesahsne with territory in Ontario, Québec, and body parts such as eyelashes or fingernails are New York State; Kanehsatake at Oka to the northwest Alienable. Terms for most kinsmen do not describe of Montreal; Kahnawake on the southern shore of the possession, but rather the relationship, such as St. Lawrence across from Montreal; and Wahta at rake’níha ‘my father’, literally ‘he is father to me’, or Gibson in Ontario. There are several thousand the reciprocal atiara’sè:’a ‘my cousin’, literally ‘we speakers of Mohawk, more than any of the other two are cousins to each other’. Northern languages. The ages of the youngest Morphological nouns end in a noun suffix. The speakers vary from community to community, but most common is -a’ as in otsíhkw-a’ ‘knob’. Various there are immersion schools in four of them, at Six grammatical endings can be added to words serving as Nations (Ohsweken), Ahkwesahsne, Kanehsatake, nominals, as in Kahnawa’kehronon’kénha’ ‘former and Kahnawake. A number of speakers of Iroquoian Kahnawake residents’: ka-hnaw-a’=ke=hronon’= languages reside outside of these communities as kenha’ NEUTER - rapids - NOUN.SUFFIX = LOCATIVE = well. RESIDENTIAL = DECESSIVE. Although the languages are not mutually Verbs are potentially the most complex words and intelligible, their basic structures are similar. They are by far the most frequent in speech. All verbs contain illustrated here with examples from Mohawk. a pronominal prefix, a verb stem, and, apart from The consonant inventory is small: t, k, kw, ts, s, n, r, commands, an aspect suffix. This structure can be y, w, h, ʔ. There is a notable absence of labials (such as seen, for example, in katerohrókha’ ‘I watch’: k- p, b, or m). The vowel inventory is similarly small: i, aterohrok-ha’ I-watch-HABITUAL. The pronominal ˜ . The Mohawk communities have agreed on e, a, o, ˜u,  prefix represents the core arguments of the clause, a practical orthography, which represents the that is, the one or two major persons or objects consonants as t, k, kw, ts, s, n, r, i, w, h, and ’ (with i involved. These specify person (first, inclusive, for the glide [j] and apostrophe for glottal stop). The exclusive, second, or third), number (singular, dual, or vowels are spelled i, e, a, o, on, and en, with plural), and in third person, gender (masculine, nasalization indicated by a following n. Stress neuter-zoic, or feminine-indefinite). The grammatical originally occurred on the penultimate syllable of a roles of the arguments are specified as well, but not in word, although the pattern can be obscured by the terms of subjects and objects, but rather as addition of vowels within the word. The stressed grammatical Agents (typically those performing syllable carries distinctive tone (pitch). An acute actions and controlling situations) and grammatical accent marks high or rising tone (ó), and a grave Patients (typically those affected by the situation but accent marks falling tone (ò). The falling tone appears not in control). The categories are semantically based, where a stressed vowel immediately preceded a but they are crystallized in the lexicon and the laryngeal (such as h), as in *iahratóhrok’, which grammar, so that speakers have no choices about became iahatò:roke’ ‘he climbed under there’. If a degrees of agency as they speak. Verb forms are syllable ended in a laryngeal, the laryngeal was lost, simply learned with the appropriate prefixes. The 700  MOHAWK AND THE IROQUOIAN LANGUAGES intransitive verb eniakwaterohrókha’ ‘we’ll go watch sentences in themselves: wa’onkwanahskwaién:ta’ne’ it’ contains the Agent pronominal prefix iakw- ‘we ‘we got a pet’ (wa’-onkwa-nahskw-a-ient-a’-n’ all’. The intransitive verb ionkwén:ten ‘we are poor’ FAC T UA L -1. P L U R A L . PAT I E N T -domestic.animal- has the Patient pronominal prefix ionkw- ‘we’. The STEM.JOINER-have-INCHOATIVE-PUNCTUAL). The same Agent and Patient components of transitive verb can be part of a larger sentence: pronominal prefixes are often fused. The transitive wa’onkwanahskwaién:ta’ne’ è:rhar ‘we got a dog’. verb shakwá:iatskwe’ ‘we called him’ has the Verbs can serve other syntactic functions as well. They transitive pronominal prefix shakwa- ‘we/him’. The can function syntactically as nominals, much like transitive verb enionkhiia’tahserón:ni’ ‘she will dress nouns. Some are lexicalized as nominals, so that us up’ contains the pronominal prefix ionkhi- ‘she/us’. speakers understand them first as names for entities, The verb stem may itself be complex. This stem such as tewa’á:raton ‘lacrosse stick’ (te-w-a’ar-a-t-on aterohrok ‘watch’ contains a Middle prefix -ate-. Verb CISLOCATIVE-NEUTER-net-STEM.JOINER-be.in-STATIVE ‘it stems may also contain an incorporated noun stem, has a net in it’). Some are used alternately as like -nahskw- ‘domestic animal’ in ranahskwiióhne’ predicates or nominals, like tahontsihkwà:’eks ‘they ‘he was a beautiful animal’: ra-nahskw-iio-hne’ hit the puck, they play lacrosse’ or ‘lacrosse players’ MASCULINE . AGENT -animal-be.beautiful. STATIVE - PAST (ta-hon-tsihkw-a-’ek-s CISLOCATIVE-MASCULINE. (‘he was beautiful in the way domestic animals are’). PLURAL.AGENT-puck-STEM.JOINER-hit-HABITUAL). Full Incorporated nouns qualify the meaning of the verb. verbs are also used for many functions served by Nouns are incorporated both to create single words adjectives and adverbials in other languages. for recurring concepts, as above, and to manipulate Because the verb provides a full grammatical the flow of information. When speakers wish to direct skeleton of the sentence, word order is used for purely special attention to an object, they generally des- pragmatic purposes. Nouns are strikingly rare in ignate it with a separate, independent noun. If the connected speech; sentences most often consist of just object is an established part of the scene, or incidental a verb and various particles. When independent to the point at hand, it may be backgrounded by nominals are present in a clause, all possible incorporation. constituent orders can occur, although not all would be Verbs may contain various additional prefixes and pragmatically felicitous. After various orienting suffixes. Among the prefixes are a Contrastive for particles, the most important elements tend to occur unexpected situations, a Coincident for similarity or early in the clause, with successively more predictable simultaneity (‘the same, when’), a Partitive ‘so’ that and peripheral information expressed later. appears in a variety of syntactic constructions, a Some of these structures can be seen in the excerpt Negative ‘not’, a Translocative ‘thither’, a Factual below from an anecdote told by Watshenní:ne’ Sawyer typically used for past events, a Duplicative that of Kahnawake. (Abbreviations include M for indicates various kinds of ‘two-ness’, including MASCULINE, N for NEUTER, PL for PLURAL, PRT for repetition of an event or a shift in position or state, a PARTITIVE, SJ for STEM JOINER, and ST for STATIVE.) Future tense ‘will’, an Optative ‘might, should, would, could’, a Cislocative ‘hither, there’, and a Repetitive Nè:ne ó:nen ‘again, back’. it is now Among the suffixes are an Inchoative ‘become’, ‘Now then several Causatives (‘cause’), Instrumental Applicatives (‘do with …’), Benefactive Applicatives wa’-onkwa-nahskw-a-ién:t-a’-ne’ è:rhar. FACTUAL-1.PL.PATIENT-ANIMAL-SJ-LIE- (‘do for’), Reversives (‘un-’), Distributives (‘here and INCHOATIVE-PUNCTUAL dog there’), and Purposives (‘go in order to …’). All verbs except commands contain one of the three basic we acquired a dog. aspectual suffixes: Habitual, Punctual, or Stative. The Ra-nahskw-iió-hne’ ken’=k nì-:r-a è:rhar. Habitual is used for recurring events and, with some M.AGENT-animal- little=just PRT-M.SG. dog verbs, for ongoing activities. The Punctual is a be.nice.ST-PAST AGENT- Perfective, used for events viewed as wholes. The be.a.size Stative is used for states. With some verbs, it is also He was a beautiful little dog. used for activities in progress, and with some verbs it is also used as a Perfect. The Habitual and Stative may Butch ronwá:-iat-s-kwe’ be followed by a postaspectual suffix: Past, 3.PL/M.SG-call-HABITUAL-PAST Continuative, or Progressive. His name was Butch. Because all verbs contain pronominal reference to their core arguments, they can serve as complete Toka’ kí:ken Ka-hnaw-a’=ke=hró:non’ 701 MOHAWK AND THE IROQUOIAN LANGUAGES and this NEUTER-rapids-NOUN.SUFFIX= sok i-en-ionkhi-ia’t-énhaw-e’ place=RESIDENTIAL then TRANSLOCATIVE-FUTURE-INDEFINITE/1.PL-body- And folks in Kahnawake, take-PUNCTUAL thi ionkw-én:ten shen’s ki’ wáhi’ and then she’d take us over there that 1.PL.PATIENT-be.poor formerly just TAG we were poor then, you know, en-iakw-ate-rohrók-ha-’ kí:ken. FUTURE-1.PL.EXCLUSIVE.AGENT- this énska=k ki’ nà:’a MIDDLE-watch-PURPOSIVE-PUNCTUAL one=only just I guess to watch the game.’ I guess we only had one wa’-t-hon-tsihkw-a-’ék-st-ha’ FAC T UA L - D U P L I C AT I V E - M . P L . AG E N T -puck- S J -hit- References INSTRUMENTAL.APPLICATIVE-HABITUAL Chafe, Wallace. 1976. The Caddoan, Iroquoian, and Siouan ball (‘they hit the puck with it’, i.e. ‘they played languages. The Hague: Mouton. Froman, Frances, Alfred Keye, Lottie Keye, and Carrie Dyck. lacrosse with it’). 2002. English-Cayuga Cayuga-English dictionary. Toronto: Ahkwesáhs=ne n-en-t-hón:n-e-’ University of Toronto. Ahkwesahs=place PARTITIVE-FUTURE-CISLOCATIVE- Michelson, Karin, and Mercy Doxtator. 2002. Oneida-English English-Oneida dictionary. Toronto: University of Toronto. M.PL.AGENT-go-PUNCTUAL Mithun, Marianne. 1979. Iroquoian. The languages of native They would come from Ahkwesahsne, America: historical and comparative assessment, ed. by Lyle Campbell and Marianne Mithun. Austin and London: ta-hon-tsihkw-à:-’ek-s University of Texas. CISLOCATIVE-M.PL.AGENT-puck-hit-HABITUAL Mithun, Marianne. 1999 and 2001. The languages of native the lacrosse players North America. Cambridge, UK: Cambridge University. Rudes, Blair A. 1999. Tuscarora-English English-Tuscarora aw-ent-a-tokenhti-’=ke io-t-ohetst-on n=entie’. dictionary. Toronto: University of Toronto. N-day-SJ-be.holy-NOMINALIZER N.PATIENT- the=noon Trigger, Bruce G. (ed.) 1978. Northeast: handbook of North =place MIDDLE-pass-ST American Indians, Vol. 15. Washington: Smithsonian Sunday afternoon. Institution. Woodbury, Hanni. 2003. Onondaga-English English-Onondaga Sok en-ionkhi-ia’t-a-hserón:ni-’ istèn:’a dictionary. Toronto: University of Toronto. then FUTURE-INDEFINITE/ mother MARIANNE MITHUN 1.PL-body-SJ-prepare-P So then my mother would dress us up See also Noun Incorporation Mongolian The language known as Mongolian, one of several Socialist Republic and in parts of China, Mogul in languages belonging to the Mongolian subgroup of the Afghanistan, Buryat in the Buryatia Autonomous Altaic languages, is based on the Khalkha language. It Republic and in parts of China, Dagur in Heilongjiang is spoken in northeastern Asia by approximately 2.5 and the Inner Mongolian Autonomous Region of China, million people in the Mongolian People’s Republic, Santa in the Gansu region of China, and Monguor and commonly known as Outer Mongolia, where it is the Bao’an, both in the Qinghai region of China. official language. In addition, about 2 million people in As an Altaic language, Mongolian shares several the Inner Mongolian Autonomous Region of the characteristics with other languages in this family. People’s Republic of China and up to half a million Some of these include agglutination, the use of people in other areas of northern China speak Khalkha postpositions instead of prepositions, vowel harmony, Mongolian. Other languages that belong to the the placement of modifiers before what they modify, Mongolian subgroup of the Altaic family include Oirat the absence of a relative pronoun, the absence of a in Outer Mongolia and China, Kalmyk along the lower verb meaning ‘to have’, the absence of grammatical Volga River in the Kalmykia Autonomous Soviet gender, and the absence of articles. Despite these 702  MONGOLIAN similarities, some linguists do not accept the thesis of harmony occurs within a word, including the suffixes a common ‘Altaic’ origin. They claim that they may be that have been added. When vowels harmonize, all of due to sociolinguistic contact and mutual interaction the vowels within a word agree—they must all either over millennia, which would make the similarities be front (/e/, /y/, /œ/) or back (/a/, /o/, /u/), with /i/ typological (due to mutual contact) instead of genetic being a neutral vowel. Because of this, many suffixes (due to a common origin). This is complicated by the have two versions: one with a front vowel and one obscurity of the origin of the ‘Altaic’ -speaking with a back vowel. Vowel harmony is demonstrated in peoples, the extreme nomadism of these peoples, and, the ablative, instrumental, and comitative cases in the simply, linguistic change. above chart; because of vowel harmony, the form As an agglutinating language, Mongolian forms *ger-aas is ungrammatical because the suffix does not words by adding morphemes (meaningful elements have a front vowel as the stem word. Vowel harmony that cannot be split into smaller meaningful elements) rules, however, do not apply in four instances: to a base, or root, word with each morpheme compound words, where each member retains its own corresponding to a single lexical meaning or vowels (tserendulmaa ‘Tserendulmaa’ (a woman’s grammatical function. An illustration of Mongolian name)); certain suffixes (oj-gwe ‘treeless, bare of agglutination is the word bari-ld-aa-tʃi-d-t ‘to the forests’); foreign names (germaani ‘Germany’); and wrestlers’, which is composed of the following parts: recent borrowings (pjoneer ‘pioneer’ (Boy Scouts)). bari- ‘to seize’ + -ld- ‘each other’ + -aa- NOUN- The basic word order of Mongolian is subject– FORMING SUFFIX + -tʃi- ‘an occupation’ + -d- object–verb. Modifiers (adjectives, adverbs, PLURAL + -t LOCATIVE. Mongolian also has a determiners, etc.) generally precede the word that they number of cases (these manifest themselves as suffixes modify and do not show agreement, e.g. no matter that attach to nouns); these cases indicate the function what case (nominative, genitive, dative, etc.) or of the word in a sentence. The following is a chart of number (singular or plural) a noun is, the adjective the declension of the words mal ‘cattle’ and ger ‘yurt’. will stay the same. Although Mongolian has various plural markers, e.g. nom ‘book’, nom-uud ‘books’, Case ‘cattle’ ‘yurt’ zaluu ‘youth’, zaluu-tsuud ‘youths’, a noun does not Nominative (indicates the subject) mal ger take a plural suffix when modified by a number, e.g. Genitive (indicates possession ‘of’) mal-iin ger-iin negen nom ‘one book’, xoyor nom ‘two books’, gurav Dative-Locative (indicates the mal-d ger-t nom ‘three books’. Because the number already indirect object or location ‘in’/’at‘) indicates whether there is one or more, a plural marker Accusative (indicates the mal-iig ger-iig direct object) would be redundant. The following sentence illustrates Ablative (indicates direction mal-aas ger-ees many of the features just described. away from ‘from’) Instrumental (indicates means mal-aar ger-eer bi xojor sajn mori-toj bajn or agency ‘with’/’by‘) I two good horse-COMITATIVE exist Comitative (indicates mal-taj ger-tej ‘I have two good horses.’ accompaniment ‘with’) The adjective sajn ‘good’ shows up before the noun As illustrated by this chart, many suffixes in mori-toj ‘horse’, which is in the comitative case, but it Mongolian correspond to prepositions in English. For does not show agreement. If it were to show example, the Mongolian word mal-iin would be agreement, one would expect to find the comitative translated into English as ‘of the cattle’. It is possible suffix added to sajn. Also, the noun mori-toj ‘horse’ is to combine more than one of the above suffixes, modified by the number xojor ‘two’, but it does not resulting in constructions such as nœxœr- ‘friend’ + - have a plural marker. The equivalent sentence with the iin ‘of’ GENITIVE + -d ‘at’ LOCATIVE = nœxr-iin-d plural marker would be ‘at my friend’s house’. Mongolian also has some bi sajn mori-d-toj bajn postpositions (these are like prepositions except that I good horse-PLURAL-COMITATIVE exist they occur after nouns), which follow nouns in the ‘I have good horses.’ nominative, genitive, ablative, or comitative cases: This sentence also shows how Mongolian expresses with nominative: dotor ‘in’, e.g. tasalgaan dotor ‘in the possession without a verb ‘to have’: the verb meaning ‘to room’; be/to exist’ is used with the possessor in the nominative with genitive: tuxay ‘about’, e.g. nom-iin tuxay ‘about case and the possessed in the comitative case. The literal the book’; etc. translation would be ‘I am with good horses.’ One of the striking features of Mongolian (and For comparative structures (‘NOUN is Altaic languages in general) is vowel harmony. Vowel ADJECTIVE-er than NOUN’), Mongolian puts the 703 MONGOLIAN first noun in the nominative case, the second noun and in modern Tungus languages, whose ancestors (‘than NOUN’) in the ablative case, and then the borrowed words from ancient Mongolian. Some adjective. The sentence ‘Today is warmer than characteristics of ancient Mongolian include the yesterday’ would be translated œnœœdœr œtʃigdr-œœs preservation of the consonants [p], [f], and [h] in initial dulaan (today yesterday-ABLATIVE warm). position; the division of all vowels and consonants into Mongolian differs from other Altaic languages in front and back; the preservation of [γ], [g], [b], and [w] verb endings: Mongolian verbs are not marked for in intervocalic (between vowels) position; and the person. Rather, verbs only inflect for tense (past vs. presence of grammatical categories. The middle period nonpast) and for aspect (perfective vs. imperfective); lasted from around the twelfth century to the sixteenth the bare verb stem serves as an imperative (a century CE. The language of this period is known as command). There are only four basic verb endings: Classical Mongolian, and the Mongolian script comes Past Nonpast into use at this time. Distinguishing traits of middle, or Classical, Mongolian include the disappearance of [γ], Perfective -v -laa [g], [b] and [w] between vowels (this leads to the Imperfective -dee -na development of long vowels in Mongolian); the loss of In addition to these suffixes, one can add –aarai to [h] in initial position in some words; and the partial loss a bare stem to form a polite request: yav ‘go!’ becomes of grammatical categories. The contemporary period, sajn yawaarai ‘please go well!’ which begins around the sixteenth century CE, follows A verb with one of the basic verbal endings cannot be middle Mongolian and continues to the present. This negated. The negation of a verb is achieved by attaching period is characterized by the loss of [p], [f], and [h] in the negative suffix, -gwe, to a noun form of a verb. There initial positions; the distinction between short and long are three noun-forming endings for verbs: -san for past vowels; and the absence of grammatical categories. tense, -x for nonpast, and -dag for customary or repeated The Mongols have used several different systems to actions. Hence, for instance, the sentences ten ire-v ‘he write their language. The Secret History of the Mongols, came’ and bi mede-n ‘I know’ would have the following a major literary work that dates from 1240 CE and negative counterparts: ten ir-sen-gwe ‘he did not come’ marks the beginning of the Middle Mongolian period, and bi mede-x-gwe ‘I do not know,’ not *ten ire-v-gwe or was written in Chinese characters. The classical *bi mede-n-gwe. In English, the closest literal translation Mongolian script was in use by the thirteenth century. of these phrases would be ‘without his having come’ and Most scholars believe that the Mongols borrowed this ‘without my knowing’, respectively, both gerund alphabet from the Uigurs who had adopted it from a structures. However, in order to negate a command, one version of the Syriac alphabet. The letters of this need only add the negative word bitgii, as in bitgii yav Mongolian alphabet have initial, medial, and final forms ‘Don’t go.’ The verb bajn ‘to be/to exist’ is also not (the shape of the letter changes depending on whether it negated like other verbs; it has its own negative is the first letter, a middle letter, or the final letter of a counterpart, biʃ ‘not to be/not to exist’— ter tend biʃ ‘he word), and it was originally written horizontally from is not there’. right to left. Later, this changed to a vertical, left-to- Mongolian does not have relative pronouns. In right format, perhaps under the influence of Chinese. order to form a relative clause, verbs transform into The Mongolian script was used in the Mongolian verbal nouns and are then used as attributive adjectives People’s Republic until 1941 when a new Cyrillic-based to modify nouns. For instance, the relative alphabet was adopted. There were two reasons stated in construction ‘the work which we did’ would be official documents for this adoption: first, a big gap rendered bidnii xii-sen adil (‘we do-PAST work’). between the written language (using the traditional The verb xiix ‘to do’ is turned into a noun with the Mongolian script) and the spoken language had addition of the past-tense morpheme –san (in this developed; and second, the Mongolian script was not case, –sen due to vowel harmony). Then it, along with suitable for assimilating foreign words. Despite this the other part(s) of the clause (in this instance, there is adoption, the Mongolian script continues to be used by only a subject pronoun, bidnii ‘we’), is placed before Mongolian speakers in China and in private the noun that this modifies, adil ‘work’. correspondences by older people in Outer Mongolia. The history of the Mongolian language is usually divided into three periods: the ancient period, the References middle period, and the contemporary period. The ancient period lasted until about the twelfth century Campbell, George L. 1991. Compendium of the world’s languages London and New York: Routledge; 2nd edition, CE. Although texts written in ancient Mongolian have 2000. not been found, evidence exists in Chinese texts Deny, J. 1952. Langues Turques, Langues Mongoles et Langues wherein a few ancient Mongolian words are attested Toungouzes In Meillet and Cohen. 704  MONTAGUE, RICHARD Meillet, A., and Marcel Cohen (eds.) 1952. Les Langues du Sinor, Denis. 1952. Langues Mongoles, In Meillet and Cohen. Monde. Paris: Center national de la recherche scientifique. Troxel, D.A. 1953. Mongolian vocabulary. Washington: United Poppe, Nicholas. 1965. Introduction to Altaic linguistics. States Government Printing Office. Wiesbaden: Otto Harrassowitz. ROBERT A. CLOUTIER Poppe, Nicholas. 1970. Mongolian language handbook, Washington: Center for Applied Linguistics. See also Altaic; Vowel Harmony Montague, Richard Richard Montague is one of the most influential achieving an accurate, adequate, and philosophically philosophers of the past century and can be deservedly satisfactory scientific account of natural language. The considered the founder of contemporary formal initial words of “Universal grammar” (1970a) mark a semantics. As an undergraduate at the University of clear declaration of intentions: California at Berkeley (1948–1950), he was attracted to a number of disciplines, but particularly to mathe- “There is in my opinion no important theoretical difference between natural languages and the artificial matics, philosophy, and Semitic languages, all of languages of logicians; indeed I consider it possible to which he pursued very rapidly to an advanced level. comprehend the syntax and semantics of both kinds of The most important single influence on the direction languages within a single natural and mathematically of his work was the logician Alfred Tarski. While still precise theory.” a graduate student at Berkeley, Montague had already acquired considerable national and even international In this paper, he states that a grammar consists of a reputation. Between 1954 and his formal dissertation syntactic algebra and a semantic algebra that are defense in 1957, he authored six, and coauthored subject to a homomorphism condition: there is a another four, significant papers in mathematical logic, homomorphism mapping elements of the syntactic including researches on Boolean algebras, proof algebra onto elements of the semantic algebra. This theory, model theory, and axiomatic set theory. In requirement elegantly captures Frege’s Composi- 1955, he joined the faculty of the University of tionality Principle in a formal setting: the meaning of California at Los Angeles and expanded his interests a complex expression will be a function of the into other areas in the field of logic such as abstract meaning of its parts and the operation or operations recursion theory, predicate logic, and the model theory combining them. of higher-order logics. In the early 1960s, he Montague also introduced the method of fragments. increasingly focused on the application of Unlike in generative grammar, where structural sophisticated logical methods to traditional problem descriptions and rules are proposed for a single areas in philosophy: determinism (in “Deterministic construction or constructions but the rest of the theories,” published in 1962); the concept of scientific grammar is left undefined (or assumed to be generated explanation; the so-called hangman or surprise by standard mechanisms), Montague presented a examination paradox (“A paradox regained,” 1960, formal and precise treatment of several “fragments” or with David Kaplan); conditional or derived obligation; subsets of expressions of English. The monostratal events (“On the nature of certain philosophical syntactic formalism he proposed also departed from the entities,” 1960); and indirect discourse (“That,” 1959, standard multistratal proposals preferred within with Donald Kalish). generative-transformational approaches. In the latter, From the mid-1960s until his untimely death, syntactic derivations are based on a set of base rules Montague embarked on an ambitious program to and additional transformations that change the linear apply some of the precise tools of mathematical logic arrangements of terminals, respectively deriving deep and model theory to the analysis of natural language. and surface structure representations. Montague’s This was a concern that started to emerge in his syntax is modeled on the recursive bottom-up rules introductory logic text, Logic: techniques of formal preferred by logicians and categorial grammarians. reasoning (1954, with Donald Kalish). The main These rules derive well-formed expressions of different advantage of this mathematical approach was categories, which in turn are assigned an interpretation. 705 MONTAGUE, RICHARD There are two possible methods for providing a 1974). The groundbreaking nature of Montague’s semantic interpretation for a fragment of English. The approach can only be properly understood in the first one, illustrated in “English as a formal language” context of a comparison between the linguistic (1970b), is the method of direct interpretation. landscape of that time and the field of formal Expressions of a language are directly assigned semantics nowadays. This discipline has undergone denotations in a model. The second method (the indirect radical changes that are almost entirely attributed to method) is developed in “The proper treatment of Montague’s legacy, namely to the combination of the quantification in ordinary English” (1973). It consists of ideas and analysis contained in his most influential translating the expressions of English into a formalized papers and the developments by his disciples and logical language (the language of typed intensional followers, within the enterprise commonly known as higher-order logic) on a first step and providing an Montague Grammar. interpretation for this language. Although from a technical point of view, this intermediate step should be dispensable, in practice using a formalized Biography representation language has several advantages. First, it Richard Montague was born in Stockton, California, makes possible a systematic analysis of semantic on September 20, 1930. He received his B.A. from ambiguity. One or more disambiguated formulae are University of California, Berkeley in 1950 and Ph.D. assigned to each ambiguous English sentence. Second, for a dissertation on mathematical logic, tutored by translating English into a rigidified uniform language Alfred Tarski, (University of California, Berkeley) in increases perspicuity and the ability to represent the 1957. During his graduate years, he also worked with logical properties of natural language in different realms Professors W.J. Fischel in Arabic, and Paul Marhenke (tense, modality, quantification) using well-known and Benson Mates in philosophy; between 1950 and precise formal devices. Finally, Montague always 1953, he held the Howison Fellowship in Philosophy, maintained a nonpsychological objectivistic stance, and and for two succeeding years he was a teaching considered the syntax, semantics, and pragmatics of assistant in mathematics. In the spring semester of natural languages as branches of mathematics on par 1955, he joined the faculty of the University of with geometry or number theory. Nevertheless, in the California, Los Angeles, as Acting Instructor in following decades several connections have been drawn Philosophy, and in the following years he advanced with some of the psychological concerns of linguistic very quickly through the academic ranks. He served as theories, mostly in that the intermediate semantic a member of the United States National Committee for representation language can be viewed as a the International Union of the History and Philosophy characterization of the speakers’ semantic competence of Science and, for the years 1966 and 1967, as (or at least of some of its properties). Chairman of the national Subcommittee for Logic, Numerous innovative semantic proposals by Methodology, and Philosophy of Science. He also Montague have become standards in the field. First, he served as Secretary of the Association for Symbolic treated noun phrases as uniformly denoting Logic from 1966 until his death and as a consulting (generalized) quantifiers, against the received view in editor for the Journal of Symbolic Logic since 1958. logic that considers proper names as constants and He died on March 7, 1971, in his home in Los existential, and universals as quantifiers. Second, Angeles, at the hands of persons still unknown at the Montague used lambda-abstraction and the lambda- time of writing operator to represent the compositional combination of expressions as functions and arguments. Finally, he generalized the Frege/Carnap distinction between References sense/intension and reference/extension to a full van Benthem, Johan, and Alice ter Meulen (eds.) 1997. The intensional semantics. The intension of an expression handbook of logic and language. Cambridge, MA: MIT is a function from possible worlds to its extension. Press. Intensionality is pervasive in natural language, not Dowty, David, Robert Wall, and Stanley Peters. 1981. Introduction to Montague semantics. Dordrecht: D. Reidel only in opaque contexts such as the complements of Publishing Company. verbs like believe but also in the transitive Gamut, L.T.F. 1991. Logic, language and meaning. Chicago: complements of verbs such as seek, or in modifiers University of Chicago Press. such as presumed, alleged, or former. Gutiérrez-Rexach, Javier (ed.) 2003. Semantics: critical Montague’s most significant articles with concepts in linguistics. London/New York: Routledge Publishing Company. applications in linguistics and philosophy were Lewis, David. 1972. General semantics. Semantics of natural posthumously compiled by Richmond Thomason in language, ed. by Donald Davidson and Gilbert Harman, the volume Formal philosophy (Yale University Press, 169–218. Dordrecht: D. Reidel Publishing Company. 706  MOOD Montague, Richard. 1970. English as a formal language. ––––––. 1974. Formal philosophy. Selected papers of Richard Linguaggi nella Società e nella Tecnica, ed. by B. Visentini Montague, edited and with an introduction by Richmond et al., 189—224. Milan: Edizioni di Comunità. Thomason. New Haven: Yale University Press. ––––––. 1970. Universal grammar. Theoria 36. 373–98. Partee, Barbara, 1973. Some transformational extensions of ––––––. 1973. The proper treatment of quantification in Montague grammar. Journal of Philosophical Logic 2. 509–34. ordinary English. Approaches to natural languages, ed. by J. Partee, Barbara (ed.) 1976. Montague grammar. New York: Hintikka, J. Moravcsik, and P. Suppes, 221–47. Dordrecht: Academic Press. D. Reidel Publishing Company. JAVIER GUTIÉRREZ-REXACH Mood The grammatical meaning of mood (often also referred need and dare (including might, could, would, should). to as modality or by the adjective modal) is a complex These have counterparts in German (e.g. mögen one, but one that we can break down into two main ‘may’, wollen ‘want’, können ‘can’) and in French characteristics. In its narrow meaning, it denotes a (such as pouvoir ‘can’, devoir ‘must’, vouloir ‘want’). particular aspect of verbal morphology, usually In this sense, there is an obvious interplay of mood indicated by an inflectional marking on a verb form, with tense, the grammatical distinctions of time made, tied in closely with tense and aspect relations. But in and aspect, relating to the internal temporal structure the larger context of sentences (“form”) and meaning of a situation (see also the paragraph on epistemic vs. (“function”), it refers to the grammatical category deontic modality). expressing the type of sentence. To make matters more It must be said, however, that from a typological complicated, both types are sensitive to the specific perspective, this characterization of mood is less than modal function of individual elements, in addition to clear: not all mood forms are used equally in all both the narrow and wider understanding of mood. languages––neither with respect to type of expression In traditional grammar, mood expressed in verbal (e.g. verbal morphology vs. separate expressions) nor morphology commonly distinguishes the indicative variety of expression; one extreme in the latter (John plays soccer.), the subjunctive (John would play category is Tuyuca, a language spoken in Brazil and soccer.), the imperative (Play soccer, John!), and (in Colombia, where every sentence is modal: Tuyuca has Classical Greek or Hungarian) the optative (roughly, I five evidential modals (see below) that characterize wish John played soccer.). These distinctions can be even every declarative sentence in a specific modal found in many European languages. Other languages form. The English sentence He played soccer, for express the same moods by the use of particles (e.g. example, can be rendered by díiga apé-wi ‘He played Ngiyambaa, a language from Australia or Jacaltec soccer (I saw him play)’, díiga apé-ti ‘I heard the from Guatemala), and also clitics (such as Luiseño game and him, but I didn’t see it or him’, or díiga apé- from California). This state of affairs warrants a wider yi ‘I have seen evidence that he played: his distinctive conception of mood already than simply referring to shoe print on the playing fields. But I didn’t see him verbal paradigms. Moreover, even those languages play’, for example. that express mood with verbs can be further And comparing German and English, for a differentiated. The classical languages (Classical moment, we can see another factor of the interplay of Greek, Latin) and the modern Romance languages different types of modality. The German equivalent of beyond French (e.g. Spanish, Italian) have particular the English sentence He should have come home is Er verbal paradigms, i.e. inflectional endings on the main hätte nach Hause kommen sollen, where the first verb. The North European languages (such as French, boldfaced word is the subjunctive form of the verb ‘to English, or German) use modal verbs, but German and have’, while the second one is the modal form for French, for example, can also express moods through ‘should’, both expressed by one modal form in English verbal inflection, namely the subjunctive. (see also the paragraph on epistemic vs. deontic From a European point of view, the modal verbs are modality below). On the other hand, languages like the clearest example of the narrow conception of French use the subjunctive often for purely mood. The traditional modal verbs of English are may, morphosyntactic reasons, rather than semantic ones, can, must, ought (to), will, shall, and to some extent namely when embedded under particular verbs. 707 MOOD Apart from verbal mood classes, mood also example (viz. Could you pass the salt, please?, which expresses the speaker’s attitude. Among the requests an action, akin to an imperative or order). classifications of mood distinctions that have been In sum, mood in its narrow sense denotes a proposed in the literature, the one between epistemic particular distinction in the (often verbal) paradigm of and deontic use is the most common one. Epistemic languages, either by encoding inflectional modality is the area of mood concerned with morphology, modal verbs, particles or clitics express- knowledge and belief (incl. ‘possibility’, ‘probability’, ing a particular mood. In the broader context, modality ‘certainty’), and it may also express the speaker’s relates to the speaker’s attitude (e.g. epistemic vs. degree of commitment to what he says (see also deontic modality) and the function of a sentence evidential mood below). The deontic use of mood or uttered (viz. illocutionary force). There is a rich modality expresses ‘permission’, ‘obligation’, and literature that can be found on each of these topics, ‘prohibition’. For example, John may play soccer both from the point of view of a single language as tomorrow is ambiguous between the possibility of well as typologically, and ongoing research keeps John playing soccer the next day (the epistemic providing new insights. meaning viz. ‘possibility’) and the ‘permission’ given to John to play soccer (by the speaker or someone else; Further Reading the deontic use). Another proposal considers a three-way distinction of Austin, John L. 1962. How to do things with words. London: mood between illocutionary force (the ‘communicative Oxford University Press. Bybee, Joan. 1985. Morphology: A study of the relation intention of an utterance’; see below), status (‘degree or between meaning and form. Amsterdam: John Benjamins. kind of reality’), and modality (in the sense of Foley, William A., and Robert D. van Valin. 1984. Functional declarative, interrogative, etc.). Evidential and evaluative syntax and universal grammar. Cambridge: Cambridge distinctions are also sometimes regarded as mood University Press. distinctions. We have already seen an instance of the Jespersen, Otto. 1924. The philosophy of grammar. London: Allen & Unwin. former above, indicating the kind of evidence a speaker Kiefer, Ferenc. 1987. On defining modality. Folia Linguistica has for what he says (e.g. direct vs. indirect evidence). XXI(1). 67–94. Evaluatives, on the other hand, can be defined as attitudes Lyons, John. 1977. Semantics, Vol. 2. Cambridge: Cambridge toward known facts, such as Hixkaryana (a Carib University Press. language spoken in Northern Brazil), where nomokyaha Palmer, Frank R. 1979. Modality and the English modals. Cambridge: Cambridge University Press. hampini ‘He’s coming––be warned!’ vs. nomokyatxow Palmer, Frank R. 1986. Mood and modality, 1st edition. hampe ‘They are coming! I don’t believe it!’. Cambridge: Cambridge University Press; 2nd edition,1990. The larger picture that mood or modality expresses Portner, Paul. 1999. The semantics of mood. GLOT is the illocutionary force of an utterance, which may International 4(1). 3–9. be that of a request for action or information, or an Sadock, Jerrold M. and Arnold M. Zwicky. 1985. Speech act distinctions in syntax. Language typology and syntactic order, a warning, a promise, etc. Here, a strict description, Vol. 1: clause structure, ed. by Timothy Shopen. characterization is even more difficult than with the ed. Cambridge: Cambridge University Press. narrower meanings of mood, because the illocutionary Searle, John R. 1983. Intentionality. Cambridge: Cambridge force of an utterance is in general independent of its University Press. grammatical form, or sentence type. In this von Wright, E.H. 1951. An essay in modal logic. Amsterdam: North-Holland. connection, we often speak of sentence type referring KLEANTHES K. GROHMANN to the form of a sentence (the above-mentioned moods ‘declarative’, ‘interrogative’, etc.), where it is clear See also Aspect; Speech Acts; Tense and Aspect that not every interrogative asks for information, for Marking; Tense: Syntax Moore and the Gur Languages The Gur languages (sometimes called Voltaic/ phylum, are spoken in interior West Africa, in Burkina Voltaique languages, especially by French-speaking Faso, Mali, Niger, and the northern parts of Côte scholars), a family in the immense Niger-Congo d’Ivoire, Ghana, Togo, Benin, and Nigeria. The Gur 708  MOORE AND THE GUR LANGUAGES language family comprises subgroups including Oti- Mampruli, and Nanuni, is at the south of the Gur area. Volta and Grusi. Most of the Gur languages belong to Southern Mabia too has a subdivision, Mid-Southern these two subgroups; Oti-Volta in turn includes Gurma Mabia (Hanga-Kamara). Hanga and Kamara, although and Mabia (Western Oti-Volta). Since the most also found in the south, are quite different from, for prominent languages of the Gur family are in the instance, Dagbane. The fifth group, Eastern Mabia, is Mabia subgroup, this subgroup is the focus of this composed of Kusaal and Talni. brief survey (see Figure 1). This attempt at classifying the Mabia languages The Mabia languages can be divided into five main does not mean to suggest that there are clear groups; the languages within most of these groups are boundaries. Most of the languages, in fact, form a more or less mutually intelligible. Major languages continuum (speakers of adjacent languages can within each group are written in capitals in the usually understand each other), and variation between diagram. Western Mabia, which includes Dagaare, the languages is rather gradual. For instance, it is Waale, Birifor, and Safaliba, is located in northwestern sometimes difficult to say whether Mampruli is more Ghana and adjacent Burkina Faso. Northern Mabia, closely related to Dagbane or to Kusaal. A similar which includes just Moore and its dialectal forms, is situation has been observed for Birifor, Waale, and found mostly in Burkina Faso. Central Mabia is Dagaare. located approximately in the middle of the Gur area. Gur languages display some rather interesting There is one overall name for this group, Gurenne, typological features. Within typological studies, which includes Frafra, Nankani, and Nabit. A linguists try to outline the characteristics of sound subdivision of this group is Mid-Central Mabia. The systems and word and sentence structures in natural two languages in this subgroup, Buli and Konni, languages and then find out how a particular language although geographically close to Gurenne, are or group of languages may be classified according to linguistically slightly distinct from Gurenne and some these types of grammatical features. of the other Mabia languages; in fact, there is some The Gur languages are marked by a preponderance controversy as to whether this group belongs in Mabia of consonants and a scarcity of vowels when compared or in Grusi. Here, it is included in Mabia because it has to Indo-European languages like English and French, more similarities in vocabulary with Mabia than with and they have some consonants with complex Grusi. Southern Mabia, comprising Dagbane, articulation that is rare in Indo-European languages. GUR Oti-Volta Grusi e.g. (Moore–Gurma) Kasem Sisaala Kabre Mabia Gurma e.g. (Western Oti-Volta) (Moore-(Mole-) Dagomba) Bassari Konkomba Moba Western Northern Mabia Central Mabia Southern Eastern Mabia Mabia GURENNE Mabia MOORE KUSAAL DAGAARE (Frafra, DAGBANE Waale Talni Nankani, MAMPRULI Birifor Nanuni Safaliba Nabit) Mid-Central Mid- Buli Southern Konni Hanga Kamara Figure 1 709 MOORE AND THE GUR LANGUAGES An example would be labio-velar stops, i.e. Bòè lá fó mààndà consonants for which the lips are pressed together as What that you doing in [b] and the tongue is simultaneously pressed against “What are you doing?” the back of the roof of the mouth as in [g]. This is actually a common African language feature, even À yáá tónd tèngá nààbá though it is conspicuously absent in languages such as (S)he saw our village chief Akan. “He saw the chief of our village.” With respect to vowels, there is the typological The language as spoken in Burkina Faso comprises feature of vowel harmony, i.e. the vowels within a four main dialects with focal points around the towns word tend to be alike in certain respects and according of Ouagadougou (the capital), Koudougou, Koupela, to specific rules. Not only does this feature distinguish and Ouahigouya, which are all major provincial between some of these languages and Indo-European centers. Glottochronological studies, which attempt to languages it also divides these languages, into determine how long ago two or more languages split harmonizing and nonharmonizing languages. For by comparing the number of very common words they instance, one difference between Western Mabia have presumably retained from their ancestral languages like Dagaare and Waale and the rest of the language, indicate centuries of divergence within this group is that these exhibit vowel harmony while the language group. They show that Moore, including all other members of the group do not seem to. It may, its dialects, is more closely related to Dagara, a variant however, be possible that vowel harmony existed in of the Dagaare language, with only two centuries of the ancestor form of all these languages. divergence between them. Gur languages are tone languages, i.e. pitch differences within a word can alone change the meaning. Most Gur languages use only two tone levels, Dagbane i.e. they distinguish between high and low tones. The Dagbane, including Nanuni, is spoken by about 1.5 former are indicated by an acute accent (á), while a million people in Ghana. It is a trade language in and grave accent (à) indicates the latter. These tones serve around Tamale, the fourth largest town in Ghana. It is to express both lexical and grammatical oppositions as a major language of education and literacy in Dagbon, in the Dagaare verbs dá “push”, dà “buy”. home of the Dagomba. It is taught in various An important typological feature for these languages undergraduate programs at universities in Ghana. is the system of noun classes––nouns belonging to Dagbane and Mampruli are mutually intelligible. different classes are inflected differently (the masculine and feminine genders of many European languages are examples of noun classes). The manifestation of noun Dagaare classes is common in Niger-Congo languages but, while Dagaare, including Waale and Birifor, is spoken in these languages usually mark noun classes with northwestern Ghana around towns including Wa, prefixes, most Gur languages use suffixes. Tuna, Jirapa, Lawra, and Nandom, and in Burkina Faso around towns including Dano and Dissin. Native Moore speakers of Dagaare number up to 2 million. It bene- fited from early missionary activities and has one of Moore, the language of the Mossi, is the most the most extensive arrays of literature in the region. It prominent Gur language in terms of the number of is taught in primary and secondary schools in Dagao, speakers and political importance. It is spoken by homeland of the Dagaaba, and in various under- about 5 million people in Burkina Faso, where it is the graduate programs at universities in Ghana and de facto national language, and about 1 million more Burkina Faso. in Côte d’Ivoire, Ghana, Togo, and elsewhere. Moore shows typical Gur language characteristics. It has two tones, it exhibits vowel harmony, it has a system of Gurenne noun classes where suffixes indicate the class a noun belongs to, and it has Subject–Verb–Object order. The Gurenne (Frafra, Nankanne, and Nabit) also has a three sentences below illustrate this structure: large number of native speakers, numbering about 1 million. It is the language of Bolgatanga, one of the M rîgdá ráágà wá cosmopolitan towns in northern Ghana. It is fast I going market the becoming a lingua franca in northeastern Ghana, “I am going to the market.” where speakers of different languages, such as Kusaal, 710  MOORE AND THE GUR LANGUAGES Kasem, Mampruli, and Talni, use it in the Bolgatanga Likpakpalnli. It is spoken by about 500,000 people. metropolitan area. Many of its speakers use Dagbane as a second language in their homeland around Saboba and Zabzugu. The language is taught in primary schools, Kusaal and there are many literacy programs involving Kusaal, language of the Kusaasi, is spoken by about Konkomba. 250,000 people in the extreme northeast of Ghana To sum up, most of the Gur languages are in use for around towns including Bawku. It is an important educational purposes in their communities. Some of language of communication in the area. them serve communities beyond their traditional areas. Glottochronological studies show that it is most There are some mass communication programs on closely related to Mampruli, with only two centuries radio and TV in these languages in Ghana, Burkina of divergence between them. Faso, and other countries. The unrelated Chadic language, Hausa, is often thought to be an effective lingua franca in the Gur-speaking area, but this is Mampruli hardly a fact. It may have been so some time ago, but Mampruli, language of the Mamprussi, is spoken by currently some of the Gur languages, including about 100,000 people around towns including Moore, Dagbane, and Dagaare, are fast replacing it as Gambaga, Nalerigu, and Walewale. Mampruli is very lingua francas in their respective areas of influence. closely related to Kusaal and Dagbane, having There are serious attempts at functional literacy and diverged from them only two centuries ago. Indeed, mass communication in all these indigenous both languages are quite intelligible to its speakers. languages. Awareness is gradually being raised regarding the importance of mother tongues as major languages of mass communication for socioeconomic Buli development in this part of West Africa. Buli, the language of the Bulsa, with about 65,000 speakers in and around Sandema, also has some literacy programs. It is closely related to the much References smaller language Konni but quite distinct from all the Barker, P.1986. Peoples. Languages and religion in Northern other major Gur languages. Ghana. Ghana Accra: Evangelical Committee. Bodomo, A.B. 1994. Language, culture, and history in Northern Ghana: an introduction to the Mabia linguistic group. Kasem Nordic Journal of African Studies 3(2). 25–43. Bodomo, A.B. 1997. The structure of Dagaare, Stanford Kasem, language of the Kasena, belongs to the Grusi Monographs in African Languages. Stanford, USA: CSLI subgroup along with languages including Sisaali, Publications. Chakali, Tampulma, Kabre, Vagla, and Mo. It is spoken Bodomo, A.B.2000. Dagaare: languages of the world materials around towns including Navrongo in Ghana and in No. 165. Munchen, Germany: Lincom Europa. adjacent settlements in Burkina Faso. It is spoken by Goody, J. 1967. The social organisation of the LoWiili. London: IAS. approximately 300,000 people. Kasem, like Dagaare, St. John-Parsons, D. 1958. Legends for Northern Ghana. was one of the first literary languages of northern Ghana, London: Longmans and Accra: SPC. having benefited from early missionary activities. Manessy, G. 1975. Les Langues Oti/Volta. Paris: SELAF. Manessy, G. 1981. Les Langues Voltaiques. Les Langues dans le Monde Ancien et Moderne, ed. by Jean Perrot. Paris: Sisaali CNRS. Naden, A.1988. Language, history, and legend in Northern Sisaali, or Isaaleng, language of the Sisaala, is spoken Ghana. Sprache und Geschichte in Afrika 9. around the towns of Tumu, Gwellu, and Lambussie in Naden, A. 1989. Gur. Niger-Congo, ed. by Bendor-Samuel. Ghana and in adjoining areas of Burkina Faso. It is Lanham, MD: Universities Press of America. spoken by about 200,000 people. Sisaali is scarcely Nikiema, E. Moore≅: Languages of the world materials No. 383. Munchen, Germany: Lincom Europa, forthcoming. intelligible with the other Gur languages, but many Swadesh, et al. 1966 . A preliminary glottochronology of the speakers of Sisaali speak Dagaare as a second language. Gur languages. Journal of West African Languages 3. Tuurey, G.1987. Introduction to the Mole-speaking community. WA: Catholic Press. Konkomba ADAMS BODOMO Konkomba, belonging to the Gurma group along with languages including Moba and Bassari, is also called See also Niger-Congo 711 MORPHEME Morpheme The word uncountable is made up of three units: un-, by overtly or covertly modifying the definition of count, and -able. Each of these units recurs elsewhere morpheme in several different ways, to such an in English, for example, in the words unkind, count- extent that there are probably few who would adopt down, and acceptable, respectively. When they recur, the characterization that has been provided here, they have the same meaning (un- means, approxi- although they might allow the definition of a mately, ‘not’ in both uncountable and unkind) and the morpheme as the minimal unit of grammatical same form or a closely related one (in the examples analysis. cited so far, the form is identical). Moreover, we Consider the plural markers in cats, oxen, mice, and cannot subdivide these units into smaller units with tempi. The regular marker on the end of cats is not a these same properties. It is true that we can see two problem and can be dealt with easily: the word can be different letters standing for two different sounds in analyzed into two morphs, a root cat and a suffix -s, v un-, / / and /n/, but although these have the same form each of which represents a morpheme. The morpheme in other words, such as sun, they do not have a for the suffix has three allomorphs, one of which is the meaning there: sun does not mean ‘s not.’ Lastly, the /s/ found on the end of cats. The first question is: is the units un-, count, and -able between them exhaustively same morpheme present in oxen? According to the analyze the word uncountable: they account for every characterization given above, it is not, because there is bit of form; there are no unexplained bits left over. no form in common between /s/ and /ən/; rather, these Let us define units that have all of these properties are two synonymous morphemes. An alternative view as ‘morphs’, although we shall see below that not all gives more weight to the meaning side of the equation, scholars require every one of these properties. along with the fact that both are members of the same When a morph can stand alone as a word, such as paradigm of number, and allows /ən/ as an allomorph count, it is called a (potentially) free morph. When, of {plural} determined by the lexeme (vocabulary like un-, it is found only attached to something else, it unit) ox. In mice, it is not clear that there are two is called an (obligatorily) bound morph. morphs at all, although some scholars have tried to Uncountable is related to the word uncountability. argue that the root of the word is the discontinuous Again, we can divide this into morphs as un·count· morph /m—s/ and the plural marker is the diphthong abil·ity, where the -ity bit recurs in words like /a I /. However, if there is just one morpheme {plural}, probability and sensitivity, with a meaning something then perhaps that morpheme is present in mice, even if like ‘quality of being’. However, a close comparison it cannot be segmented out. The term ‘portmanteau with uncountable will show that -able /əbl/ has changed morph’ is used for a morph that represents more than to -abil- /əbIl/. When the second syllable of –able is one morpheme, although some scholars restrict it to stressed, as it is whenever -ity is added, we find -abil- instances in which morphemes that are usually instead of -able. The meaning is the same, and the form represented as independent words are merged, as in is still very closely related (it still contains the sounds French du, ‘of the (masculine)’, which can be seen as /ə/, /b/, and /l/), but -able and -abil- occur in different a conflation of de, ‘of ’, and le, ‘the (masculine)’. In environments. They are in complementary distribution: the case of tempi, there is an alternative plural tempos, where you find one, you cannot find the other. We call hence, the -i marker and the -s marker are not in morphs that are in complementary distribution in this complementary distribution. However, the fact that way ‘allomorphs’ of the same morpheme. The they do not contrast is usually accepted as sufficient terminology is entirely parallel to the terminology used for them to belong to the same morpheme. in the theoretical descriptions of speech sounds, where Yet other scholars do not even insist on a constant one speaks of ‘allophones’ of the same ‘phoneme’. meaning for the morpheme. Mark Aronoff (1976) Where there are not several different allomorphs of the argues that because permit, emit, and similar words all same morpheme, as with count, whose form is always have irregular nominalizations in -mission, the count, we can still say that the single morph count consistent change from mit to mis is sufficient to instantiates the morpheme {count}, where the braces establish -mit as a morpheme, despite the fact that it show that we are treating this as a morpheme. cannot be assigned a constant meaning in modern There are many problems with this notion of a English. For other scholars, this is a matter of morpheme, and the linguistic community has responded etymology (historical development) rather than a 712  MORPHOLOGICAL TYPOLOGY matter of morphology. Joachim Mugdan (1986) References provides a survey of definitions of the morpheme. Anderson, Stephen R. 1992. A-morphous morphology. The problems with the concept of the morpheme Cambridge: Cambridge University Press. are so great that some linguists have given up on the Aronoff, Mark. 1976. Word formation in generative grammar. notion entirely. Stephen Anderson (1992) introduces Cambridge, MA: MIT Press. the notion of a-morphous morphology, where the Coates, Richard. 1999. Word structure. London and New York: Routledge. structure of words is determined by the construction of Matthews, P.H. 1993. Grammatical theory in the United States a sound structure on the basis of roots and syntactico- from Bloomfield to Chomsky. Cambridge: Cambridge semantic structure, without any recognition of an University Press. intermediate level at which there are morphs or Mugdan, Joachim. 1986. Was ist eigentlich ein Morphem? morphemes. However, such a stance is still contro- Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 39. 29–43. versial (although it is becoming more generally LAURIE BAUER accepted), and the morpheme is still recognized as one of the fundamental units of linguistic analysis. See also Affixation; Phoneme; Word Morphological Typology Morphological typology is the study of differences chúng tôi ba˘t aˆu làm bài. among the world’s languages relating to the ways in PLURAL I begin do lesson which words are formed from smaller meaningful ‘When I came to my friend’s house, we began to do units referred to as ‘morphemes’. lessons.’ If we examine the formation of plurals in English, None of these Vietnamese words shows any of the then regular plurals like cats as the plural of cat show the morphological changes that are found in the English following relation between word and morphemes. Cats is translation, such as come ~ came, begin ~ began, or I a single word, as indicated by the fact that it is a minimal ~ my; indeed, even the plural pronoun ‘we’ is free form, i.e. it is not possible to pronounce cats with a expressed by combining a separate word indicating pause within the word. In terms of its morphological plurality, chúng, with the pronoun tôi ‘I.’ The only structure, however, it clearly consists of two meaningful exception to strict isolating structure is the verb ba˘t elements, the lexical root cat and the plural suffix -s. The aˆu ‘begin,’ which is a compound word, i.e. a single single word cats thus consists of two morphemes and can word made up of two components, literally ‘seize’ and readily be divided into those two morphemes (cat + s). ‘head.’ (Note that Vietnamese orthography leaves a Not all plurals, however, follow this pattern. Thus, the space between syllables, and not only between words; plural of tooth is teeth, with the plural being marked not the Vietnamese for ‘Bolivia’ is Bô li vi a.) by attaching a suffix to the root, but rather by changing In contrast to isolating languages, we find ‘synthetic’ the vowel of the root. And occasionally in English, one languages, which permit more than one morpheme to forms the plural by means of a separate word, as in the combine to form a word. In one type of synthetic nonstandard plural of who as who all, for instance in language, namely ‘agglutinating’ languages, the Who all are coming to dinner tonight? boundaries between the individual morphemes are clear- The different ways of forming plurals in English cut. The classic example of an agglutinating language is correspond to the three major morphological types that Turkish. If we look at the way in which a Turkish noun have been identified by considering many languages changes its form to show different numbers (singular vs. around the world. ‘Isolating’ or ‘analytic’ morphology plural) and cases (e.g. nominative or dative), we can refers to a system in which each word consists of only always easily identify which sequence of sounds is the a single morpheme. Vietnamese is a language that root, which sequence indicates number, and which comes close to being an ideal isolating language, as sequence indicates case. Thus, the dative plural of adam can be seen from the following example: ‘man’ is adamlara ‘to the men’, which can be broken Khi tôi eˆn nhà ba·n tôi, down into the morphemes adam + lar + a; the root is when I come house friend I adam, and we find this in the nominative singular adam 713 MORPHOLOGICAL TYPOLOGY ‘man’ (both singular number and nominative case take ərkən ‘I have a fierce headache,’ whose components are no suffix); the plural suffix is -lar, and we find this in the the first-person singular prefix t- ‘I’, the root mejŋ- ‘big, nominative plural adamlar (adam + lar) ‘men’; the great’, the root lewt- ‘head’, the root pəγt- ‘ache’, and dative suffix is -a, and we find this in the dative singular the imperfective aspect suffix -rkən here indicating an adama (adam + a) ‘to the man’. (Note that Turkish has ongoing state. (The other instances of ə are inserted to no definite article corresponding to ‘the’.) Knowing that break up consonant clusters.) The possibility of having a the ablative suffix is -dan, we can readily form the single word corresponding to a sentence of several ablative singular adamdan (adam + dan) ‘from the man’ words in most other languages of the world is also often and the ablative plural adamlardan (adam + lar + dan) cited as a characteristic of polysynthetic languages. ‘from the men’. Incidentally, Turkish can in this way However, it is not really clear whether polysynthetic string together quite long sequences of suffixes, as in should be recognized as a distinct type. In the Chukchi avrupallas trlamyanlardansnz ‘you are one of those example just given, the basic morphological technique is who cannot be Europeanized.’ From the noun Avrupa, agglutination, and the only unusual feature is the extent the suffix -l forms the adjective ‘European’; from this, to which such agglutination is carried out, combining the suffix -las forms a verb meaning ‘become both lexical morphemes like ‘big’, ‘head,’ and ‘ache’ and European’; the suffix -tr adds the meaning ‘cause to’, grammatical morphemes like first-person singular and i.e. ‘cause to become European’ or ‘Europeanize’; -l imperfective aspect. Polysynthesis is thus arguably just makes this passive, i.e. ‘be Europeanized’; then -am agglutination (sometimes with elements of fusion) adds the element of inability, i.e. ‘unable to be carried to an extreme. Europeanized’; -yan turns this into a participle, i.e. Although some languages come close to one of the ‘being unable to be Europeanized, one who is unable to three ideal types, isolating, agglutinating, and fusional, be Europeanized’; -lar makes this plural and -dan adds such as Vietnamese, most languages combine ablative case, i.e. ‘from those who are unable to be elements of all three types to different degrees. For Europeanized’; and finally -snz adds the element ‘you’. instance, English is generally an isolating language, A second kind of synthetic language is a ‘fusional’ and it is quite possible to produce quite long sentences language, in which the various morphemes fuse that involve no word consisting of more than one together to give a single, unsegmentable whole. We morpheme, e.g. Every day I see the cat on the wall have already encountered an example of fusional near the tree. Agglutinating morphology in English is morphology in the irregular English plural teeth, in its limited to compounding (e.g. birthday, i.e. birth + relation to tooth, since in this plural the elements root day), to the genitive and plural forms of nouns (cat’s, and plural are not segmentable as they are in a regular cats), and to a handful of verb forms (third-person plural like cats. Certain Indo-European languages singular present tense walks, past tense walked, past such as Latin, Greek, and Russian provide good participle walked, present participle walking––note examples of fusional morphology. Whereas in Turkish that some verbs do have distinct past tense and past the markers of case and number are always separable participle, e.g. ate, eaten). Fusional morphology is from one another, in Russian the two are inextricably limited to a handful of irregular noun plurals (like fused. In the nominative, we have the singular stol tooth ~ teeth) and a limited number of irregular verbs ‘table’ and the plural stoly ‘tables’. In the genitive, the in the past tense and past participle (e.g. drink ~ drank singular is stola ‘of the table’ and the plural stolov ‘of ~ drunk). the tables’; it is already clear that the genitive plural The major morphological types can be given a suffix -ov cannot be segmented into a component that systematic explication in terms of two indices. The we could identify with nominative plural -y and ‘index of synthesis’ indicates how many morphemes another that we could identify with genitive singular - are combined into a single word. The ‘index of a. The picture is only further compounded as we fusion’, relevant only if the index of fusion is above 1, examine other cases, e.g. dative singular stolu ‘to the indicates how many morphemes are fused together table’ vs. plural stolam ‘to the tables’. into a single form. By calculating these indices for A fourth morphological type is sometimes, but by no representative texts in a language, one can assess that means always, recognized, namely ‘polysynthetic’. language’s index of synthesis and index of fusion, thus Characteristic of polysynthetic languages is that capturing the intuition that all languages combine the typically a large number of morphemes are combined in different types to different degrees. a word, i.e. these languages are, as one might suspect Historically, agglutination can often be shown to from their name, very synthetic. An example is provided result from separate words becoming affixes, as when by the Chukchi language spoken on the Chukotka the Modern English suffix -hood, as in childhood, peninsula at the eastern tip of Asiatic Russia, where one derives from an Old English word ha¯d meaning finds the single-word sentence tə-mejŋə-lewtə-pəγt- ‘quality, condition’. Likewise, fusion can often be 714  MORPHOLOGY shown to have developed from the effect of a lost affix; Greenberg, Joseph H., A quantitative approach to the typology e.g. a plural suffix -i, long since lost, caused the vowel of language. International Journal of American Linguistics 26. 178–94. change in nouns like tooth ~ teeth. Greenberg, Joseph H. 1974. Language Typology: A Historical and Analytic Overview, The Hague: Mouton. References Sapir, Edward. 1921. Language: An Introduction to the Study of Speech, New York: Harcourt, Brace & World. Comrie, Bernard 1989. Language Universals and Linguistic BERNARD COMRIE Typology, 2nd edition. Oxford: Blackwell and Chicago: University of Chicago Press. See also Morpheme; Typology; Word Morphology The central focus of those who study morphology is also tells us that there are restrictions on how how language users understand complex words and morphemes combine. At the same time, we can how they create new ones. Compare the two English understand novel forms such as unfaxable. It is the words secure and insecurity. The word secure cannot morphologist’s job to discover the general principles be broken down further into meaningful parts. It is that underlie our ability to form and understand some morphologically simple. By contrast, insecurity is complex forms but not others. morphologically complex because it consists of three Affixes are morphemes that appear before or after unanalyzable meaningful components (i.e. in + secure the stem morpheme (viz. prefixes and suffixes, + ity), which linguists call morphemes. The study of respectively). Affixes may vary quite widely in their the patterning of morphemes within a word and how productivity, the likelihood that they will appear in morphemes combine to form new complex words falls new words. Compare the three English suffixes -ory, - within the domain of morphology. Morphemes are ive, and -able, all of which form adjectives from verbs. meaningful elements and must be distinguished from The first suffix is almost completely unproductive in units of sound, because a simple morpheme may be Modern English; very few new words with this suffix complex in its sound structure: the simple morpheme have been added to the language in centuries. The secure is complex in terms of sound, consisting of two second occurs in such recent words as adaptive and syllables and six phonemes. The study of the sound adoptive. The third is highly productive: innovations structure in language processing is well established. such as unfaxable are common. Productivity is usually Only relatively recently, however, have defined with respect to the extent to which a psycholinguists begun to examine morphology as a morpheme is expected to appear in novel forms. For window on how we process words. example, if we search a large database for new words The first morpheme in insecurity is the prefix (in-), (words that do not appear in a large standard which means approximately ‘not’; the second or stem dictionary) containing the two morphemes in- and -ity is the adjective secure; the third and word final that we introduced at the beginning of this article, we morpheme is the suffix (-ity), which serves to form a find very few new words with in- (actual examples noun from the adjective secure. The meaning of include ineliminable and inegalitarian), but many insecurity (viz. ‘the state of not being secure’) is fully more with -ity (actual examples include avuncularity predictable from the meaning of its components. It is and deviosity). We therefore say that -ity is more semantically transparent. The same prefix occurs in productive than in-. However, sometimes researchers many other words, as does its suffix (e.g. inactivity, define productivity with respect to the total number of impartiality). Note, however, that not all adjective words in which a morpheme appears. Again, we find stems can combine with morphemes such as in or ity. that the affix able appears in many more than does the For example, the stem abashed combines with the affix ive. Analogous to affixes, stem morphemes can affixes un and ness to form the noun unabashedness differ with respect to their family size, that is, the while continue combines with dis and ity to form the number of words formed from a particular morpheme adjective discontinuity. The fact that nouns like stem. For example, many more words are formed from *incontinueness or *disambiguosity are impossible the stem sist (i.e. consist, persist, desist, insist, and 715 MORPHOLOGY approximately 31 derivations) than from the stem flect seem obscure until one realizes that in biblical times (i.e. inflect, deflect, reflect, and approximately seven one could only spy by walking around. derivations). Knowledge about words is represented in the mental Depending on whether or not they can stand alone, lexicon. A major research question for psycholinguists morphemes are classified into two basic types: free is how the mental lexicon represents morphological (e.g. secure) and bound e.g. in, ity, fect). Affixes are knowledge. One issue is whether regular forms and bound because they cannot appear in isolation, but irregular are represented differently in the mental must combine with (be bound to) another morpheme lexicon. If regularity is defined with respect to form, to form a word. Repeated additions of bound we can ask whether words with stems that undergo a morphemes allow for the formation of more complex change in spelling (and sometimes pronunciation) are words as affixes pile one onto another. The word represented differently from words whose stem is inconclusiveness contains three bound morphemes as always regular. That is, are forms such as run and ran well as a free stem [conclude] and has been built up in represented differently from forms such as turn and stages from conclude by first adding the suffix -ive to turned? Derivation tends to be semantically somewhat the verbal stem (with a sound change from d to s), so unpredictable: walker can mean either ‘one who walks’ as to produce conclusive, then the prefix -in so as to or ‘a special support that helps one to walk’. If form conclusive, and finally the suffix -ness, resulting regularity is defined with respect to meaning, a second in [[in[[conclude]V ive]A]A ness]N. Note that in cannot issue is whether inflected forms such as concluded and combine with conclude but it can combine with derived forms such as conclusive are represented in the conclusive. This reflects the strict ordering of stages same manner. Similarly, we can ask whether when complex words are formed. Sometimes, two semantically opaque (or ambiguous) forms (e.g. morphemes can be added to a stem in two different walker) as well as semantically transparent orders, yielding two different meanings. Consider the morphological relatives (e.g. talker) are represented in word unbalanced. It can mean either ‘not balanced’ or a like manner in the lexicon. Within the mental lexicon, ‘deranged’. The first meaning results from adding the some theorists express morphological relatedness in suffix -ed to balance and then adding the prefix un-, terms of representations that are decomposed and share which, when attached to adjectives means ‘not’. The a constituent morpheme. Other theorists express second meaning results from first adding the prefix un- morphological relatedness in terms of a principle of to the verb balance, giving us the verb unbalance, similarity among full forms. which means ‘derange’. When the suffix -ed is added In the psycholinguistic literature, the classic task to this complex verb, it creates the adjective meaning, for exploring morphological knowledge is the lexical hence ‘deranged’. decision task. Letter strings are presented visually. Finally, among bound morphemes, linguists Readers must decide whether each is a real word, and distinguish inflectional from derivational morphemes. reaction time to decide is measured. Words are usually Derivational morphology deals with how distinct presented in pairs, a prime and then a target. Typically, words are related to one another; inflectional prime and target occur in immediate succession. When morphology focuses on the different forms that a word both prime and target are fully visible, decision may take, depending on its role in a sentence. English latencies to the target are faster when it is preceded by is quite sparse inflectionally, as compared with many a prime that is morphologically related (e.g. turned- other languages (e.g. Classical Greek and most Slavic turn) than by a prime that is similar in form (e.g. languages), where each noun, verb, and adjective will turnip-turn) or meaning (e.g. rotate-turn) but not have a large number of inflected forms. morphologically related. Reduced decision latencies The way in which morphemes combine to form following a morphologically related prime as complex words also varies across languages. Some compared to an unrelated prime is termed languages (e.g. Chinese) have little in the way of morphological facilitation and many psycholinguists combining morphology. Others (e.g. Turkish) are interpret the effect as evidence that the same base distinctive for the manner in which multiple morpheme or stem was activated by the prime and by morphemes occur within a single word. Rules for the target. Typically, morphological facilitation arises combining morphemes also vary across languages. In both when morphological relatives of the target are English, morphemes are linked linearly as in words regular (e.g. turned-turn) and when they are irregular such as inconclusiveness. In Hebrew, by contrast, (e.g. ran-run), however. Nonetheless, the magnitude morphemes can be interleaved with one another (e.g. of morphological facilitation in lexical decision as R-G-L combines with -e-e to form ReGeL meaning well as in other experimental tasks tends to be ‘leg’ and with m-a-e?- to form mragel? meaning attenuated when the prime is irregular and involves ‘spy’). The relation between ‘legs’ and ‘spies’ may either reduced form or reduced semantic similarity 716  MURRINH-PATHA AND THE DALY LANGUAGES with its target. Finally, morphological facilitation has Booij, G., and J. van Marle. Yearbook of morphology. been documented in a variety of languages including Dordrecht: Kluwer (annual). Feldman, L. B. (ed.). 1995. Morphological aspects of language those where morphemes are not generally appended processing. Hillsdale, NJ: Lawrence Erlbaum. linearly (viz., nonconcatenative) such as American Feldman, L. B. 2000. Are morphological effects distinguishable Sign, and Hebrew, as well as in many where from the effects of shared meaning and shared form? Journal morphemes are concatenated such as Bulgarian, of Experimental Psychology: Learning, Memory and Dutch, English, French, German, Greek, Serbian, and Cognition 26. Marslen-Wilson, W., L. Tyler, R. Waksler, and L. Older. 1994. Spanish. Dimensions of morphological relatedness Morphology and meaning in the English mental lexicon. that can alter the magnitude of morphological Psychological Review 101. facilitation in variants of the lexical decision task Matthews, P. H. 1991. Morphology, 2nd edition. Cambridge: include inflection vs. derivation, semantic Cambridge University Press. transparency or opacity (or ambiguousness), and Rueckl, J.G., M. Mikolinski, M. Raveh, C.S. Miner, and F. Mars. 1997. Morphological priming, connectionist orthographic regularity vs. irregularity. networks, and masked fragment completion. Journal of Memory and Language 36. References Sandra, D., and M. Taft. (ed.) 1994. Morphological structure, lexical representation and lexical access. Hove, UK: Aronoff, M. 1994. Morphology by itself. Cambridge, MA: MIT Lawrence Erlbaum. Press. Spencer, Andrew. 1991. Morphological theory. Oxford: Baayen, H., and A. Renouf. 1996. Chronicling the Times: Blackwell. productive lexical innovations in an English newspaper. Zwicky, A., and A. Spencer. (eds.) 1997. Handbook of Language 72. morphology. Oxford: Blackwell. Bauer, L. 1983. English word-formation. Cambridge: Cambridge University Press. LAURIE BETH FELDMAN AND MARK ARONOFF Bertram, R., R.H. Baayen, and R. Schreuder. 2000. Effects of family size for complex words. Journal of Memory and See also Affixation; Inflection and Derivation; Lex- Language. 42. icon: Overview; Word Murrinh-Patha and the Daly Languages Murrinh-patha is an Australian Aboriginal language Classification spoken on the country’s far northern coast, in an area For many years, Murrinh-patha was regarded as several hundred kilometers to the southwest of something of an Australian isolate, accepted as a Darwin. The language, which now has no significant member of the Australian language family, but held to dialectal variations, is spoken by some 2,000 people, belong to no lower level subgroup. In particular, predominantly in the township of Wadeye (formerly Murrinh-patha was thought to have no close genetic Port Keats) and the numerous satellite out-stations and link with any of the dozen or so languages of the Daly dry season camps that it supplies. With this number of River region to its north and east, these being speakers, Murrinh-patha stands out as being one of the presented by Tryon (1974) as together constituting the few Australian languages not under immediate threat ‘Daly Family’. The cornerstone to this view of of extinction. Indeed, the Murrinh-patha speech Murrinh-patha’s genetic status was the lexical data; community has been growing rather than diminishing Murrinh-patha has at most an 11% shared vocabulary in size over the last few decades, with an expanding density with any other language against which it has population of Murrinh-patha children continuing to been tested. learn it as a first language, and with Marringarr-, Mati Present research is, however, overturning these Ge-, and Marrisyefin- speaking groups from the long-held assumptions. Green (2003) has now made neighboring regions showing generational shift to out a compelling case for considering Murrinh-patha Murrinh-patha as they become increasingly integrated as making up a genetic subgroup with the Ngan’gi into a greater Wadeye community. The language is languages (Ngan’gikurunggurr and Ngen’giwumirri), further supported through a bilingual program in the formerly claimed to constitute a branch of the Daly Wadeye School and language maintenance programs family. The case is based primarily on formal undertaken by the community Language Centre. 717 MURRINH-PATHA AND THE DALY LANGUAGES correspondences in the core morphological sequences Example 1. Murrinh-patha (Walsh 1993a,b) of finite verbs, which Green argues are too matching ma-nanku-ma-purl-nu-ngintha-nga-ni in their complexities and irregularities to have 1sgS.Hands.Fut-2dlO-hand-wash-Fut. Positive- plausibly come about through anything other than a 2dl.non-male.mixed sub-section-1sgS.Fut-Sit shared genetic legacy, and which he demonstrates I will keep on washing the hands of you two (who are through reconstruction of finite verb paradigms as not siblings and one or both of whom are female). being systematically derivable from an innovative common parent. Pronominal Indexing At the same time, Tryon’s construction of an overall Daly Family is now also seen as problematic. Green Verbs obligatorily index core participants such as has suggested that in place of the single family ‘subject’ (I saw you) and ‘object’ (I saw you) by bound proposed by Tryon, the formal evidence establishes pronominal prefixes. Most languages additionally five separate Australian subgroups in the region. These allow for the verbal cross-referencing of other kinds of are given in Table 1. (Identifiable branches of each participants, such as ‘goals’ (I said it to her), subgroup are listed on separate lines.) Green claims ‘benefactives’ (I did it for her), and ‘adversatives’ (My that the five subgroups cannot convincingly be related wife ran away on me). Pronominal indexing is shown together as a single genetic unit, and argues that the in example 2, from Ngen’giwumirri (< Southern similarities that Tryon considered to be diagnostic of Daly), where subject, object, and adversative are membership in the Daly Family are better accounted indexed on the one verb. Observe that the subject for either diffusionally or as genetically inherited pronoun also encodes a tense category, in this case the features shared with a wide range of northern perfective. Variation of the subject pronoun for tense is Australian languages. found in all languages, typically as the historical result of fusion with a formerly discrete tense affix. TABLE 1 Genetic Subgroups in the Daly River Example 2. Ngen’giwumirri Region Danginy-nyi-fime-ngidde-wurru Subgroup Principal Language Varieties 3sgS.Poke.Perf-2sgO-give-1sgAdv-bad She gave it away to you against my wishes. Anson Bay Batjamalh (aka Wadjiginy), Kendjerramalh (aka PunguPungu) Northern Daly MalakMalak, Tyeraty, Kuwema Verbal Classification Eastern Daly Matngele, Kamu Western Daly Marrithiyel, Marrisyefin (aka Marri Many languages of Australia’s central far north share Tjebin), Marri Ammu, Marringarr, the further characteristic of forming their verbs with not Mati Ge (aka Magati Ge) Marramaninydyi, Marranunggu one, but rather two, root-like elements. This two-part (aka Merranunggu and Warrgat), structure typically involves the pairing of a relatively Emmi-Menhdhe inert root (or ‘co-verb’), the element that provides the Southern Daly Murrinh-patha, Ngan'gikurunggurr, main lexical information for the verb, with a root that Ngen'giwumirri hosts the core grammatical affixes. The lexical roots form an open class, while the ‘grammatical’ roots belong to a closed class, normally numbering less than Areal Features 50. The two-part verbal structure is thought to be an ancient diffusional feature. And while in some other Like the majority of languages spoken in Australia’s parts of Australia’s north the grammatical root has central far north, Murrinh-patha and the Daly languages synchronically no clear semantic value, in the Southern are of the ‘polysynthetic’ structural type, and are and Western Daly subgroups, it functions as a classifier categorized within Australianist typology as belonging of the verbal action. The Southern Daly languages all in the ‘non-Pama-Nyungan’ and ‘prefixing’ groups. have more than 30 of these classifiers, while the They possess complex verbal structures built up through Western Daly varieties have around 20 or so. Verbal the addition of strings of prefixes and suffixes to the classification may simply involve specifying the relative lexical roots. Many of the affixes are suppletive in form orientation of the subject, as with the intransitive and portmanteau in nature, simultaneously encoding a classifiers in examples 3 and 4. number of grammatical categories. Equally, it is not uncommon for grammatical categories, such as person, Example 3. Marrithiyel number, and tense, to be marked discontinuously via ga-wu-fifi-nysyen-a different affixes at different points in the verb, as 3sgS.Realis-stand-smoke-then-Past illustrated in example 1. He was standing up smoking then. 718  MURRINH-PATHA AND THE DALY LANGUAGES Example 4. Marrithiyel languages have singular/dual/trial/plural systems, ga-fulh-fifi-nysjen-a while others have singular/dual/paucal/plural systems. 3sgS.Realis-lie-smoke-then-Past Secondly, some have developed systems of pronouns He was lying down smoking then. where the categories of number and person have become integrated with other categories of Verbal classification may also be relatively abstract information. For example, in Murrinh-patha features in nature, as illustrated by the transitive classifiers in of the subsection system, and a gender categorization examples 5 and 6 below. In example 5, the gram- that hinges on the minimal inclusion of at least one matical verb root functions to conceptualize the action female within a group determine pronominal choice. as performed at the end of an elongated instrument. As can be seen from Table 2 (from Walsh 1976), this Contrast this with example 6, in which the type of complexity results in nine Murrinh-patha replacement of the grammatical verb-root achieves a pronouns corresponding to English ‘we’. different schematic conceptualization, this time of the action as performed along the shaft, rather than the end, of an elongated instrument. Nominal Classification Example 5. Marrithiyel All the Daly languages have at least a few generic sjendi-gin gu-mun-git-a nouns, such as ‘meat’, ‘vegetable food’, and ‘fire’, spear-Instrumental 3sgS.Realis-jab-sever-Past which are regularly placed in front of specific nouns to He cut it with a spear (that is, by jabbing at it with encode salient cultural categories. In Murrinh-patha, the end of the spear). this has become extended into a system in which the Example 6. Marrithiyel category membership of all entities is obligatorily sjendi-gin gi-ny-git-a encoded by one of around a dozen NP initial generic spear-Instrumental 3sgS.Realis-swing-sever-Past nouns. He cut it with a spear (that is, by swinging at it with In an adjacent bloc of Daly languages, the shaft of the spear). encompassing Ngan’gikurunggurr, Ngen’giwumirri, and all but the northernmost Western Daly varieties, These rich classificatory systems are not, however, the systems have the same number of categories, but found right across the Daly region. The Northern and have become more grammaticized, displaying Eastern Daly languages, for example, have much agreement phenomena and reduction of the smaller and simpler systems, with just a single independent generic. Sometimes, agreement is marked productive transitive classifier combining with a few by bound forms attached to nouns as well as modifiers intransitives, while the Anson Bay subgroup shows such as adjectives or demonstratives. In other cases, only recent and rudimentary developments in this noun class assignment is marked by freeform generics direction. that precede specific nouns and also precede the When compared to the other languages of the Daly, modifiers. Both these possibilities are demonstrated in and to northern Australia generally, the classifying the Marrithiyel examples below. verb structures of the Southern and Western Daly languages reveal two aberrant features. Firstly, they Example 9. Marrithiyel exhibit a tight morpho-phonological binding between a-madi a-ngelfu co-verb and the inflected grammatical root. Secondly, animal-barramundi animal-many they show an innovative ordering, placing the co-verb many barramundi fish after the inflected grammatical root rather than Example 10. Marrithiyel preposed to it. Reid (2003) argues that these shared Wadi meri wadi gutlik wadi gan features result from recent diffusion rather than a Male man male blind man this shared genetic legacy, demonstrating how the This blind man Southern Daly languages Ngan’gikurunggurr and Ngen’giwumirri have acquired them only within the Noun class phenomena in Daly languages have last hundred years. proved theoretically interesting by providing a perspective on the historical development of bound class from freeform to proclitic to prefix. They have Free pronouns also contributed to theorizing about the process by The Daly languages have freeform pronoun systems which agreement phenomena develop (Reid 1997) and that are interesting for several reasons. Firstly, they to considerations of the nature of the distinction tend to be complex by virtue of grammaticizing between noun class and noun-classifying systems multiple nonsingular number categories. Some (Green 1997). 719 MURRINH-PATHA AND THE DALY LANGUAGES TABLE 2 Murrinh-Patha Freeform Pronouns 1st Inclusive 1st Exclusive 2nd Person 3rd Person Singular ngayi nyinyi nukunu Male nigunu Female Dual neki ngankuninda nankuninda peninda All male, mixed subsections ngankunginda nankunginda peninginda At least one female, mixed subsections nganku nanku piguna All same sub section Paucal nekineme ngankuneme nankuneme peneme All male, mixed subsections nekingime ngankungime nankungime peningime At least one female, mixed sub sections Plural nganki nanki pigunu All same subsection TABLE 3 Ngen’giwumirri Phonemic Inventory most linguistically complex region. Studies in Language Change 552. Pacific Linguistics: Canberra. Bilabial Alveolar Palatal Velar Green, I. 1989. Marrithiyel. Unpublished Ph.D. thesis. Voiceless p t ty k Australian National University: Canberra. stop ———. 1997. Nominal classification in Marrithiyel. In Harvey Voiced b d and Reid. stop ———. 2003. The genetic status of Murrinh-patha. In Evans. Fricative f s sy g Harvey, M. and N. Reid (eds.) 1997. Nominal classification in Nasal m n ny ng Aboriginal Australia. Amsterdam: John Benjamins. Lateral l Reid, N.J. 1989. Ngan'gityemerri. Unpublished Ph.D. thesis. Rhotics r (approximant) Australian National University: Canberra. rr (trill) ———. 2000. Complex verb collacations in Ngan'gityemerri: a Glide w y non-derivational mechanism for manipulating valency alternations. Changing valency, ed. by R.M.W. Dixon and A. Aikenvald. Cambridge: Cambridge University Press. Phonology ———. 2003. Phrasal verb to synthetic verb: recorded structural change in Ngan'gityemerri. In Evans. Australian languages generally lack phonemic fricatives, ———. 1997. Class and classifier in Ngan'gityemerri. In and have just the single series of stop. The Daly region Harvey and Reid. Tryon, D. 1974. The Daly family languages, Australia. Series shows a significant departure from this pattern. With the C, No. 32. Canberra: Pacific Linguistics. exception of Anson Bay, all the subgroups show at least Walsh, M.J. 1976. Murinypata of North-West Australia. some phonemic voicing contrast. In addition, in Unpublished Ph.D. thesis. Australian National University: Ngan’gikurunggurr, Ngen’giwumirri, and the Western Canberra. Daly varieties, there are phonemic fricatives, with a ———. 1993. Languages and their status in Aboriginal Australia. In Walsh and Yallop. resultant three-way voiced stop, voiceless stop, and ———. 1993. Classifying the world in an Aboriginal language. fricative contrast in the bilabials, alveolars, and (in some In Walsh and Yallop. languages) palatals. The phonemes of Ngen’giwumirri, ———. 1996. Vouns and nerbs: a category squish in Murrinh- showing the atypical obstruent set in an otherwise patha (northern Australia). Studies in Kimberley languages standard Australian inventory, are given in Table 3 in in honour of Howard Coate, ed. by W. McGregor, pp. 227–52. München: Lincom Europa. their practical orthography. Walsh, M., and C. Yallop (eds.) 1993. Language and Aboriginal Australia. Canberra: Aboriginal Studies Press. References NICHOLAS REID AND IAN GREEN Evans, N. (ed.) 2003. The Non-Pama-Nyungan languages of Northern Australia: comparative studies in the continent's See also Australia 720  N Nahuatl and Uto-Aztecan Languages The Uto-Aztecan (also spelled Uto-Aztekan) language Hopilavayi Project of the Hopi Cultural Preservation family is comprised of eight branches, spread across the Office has designed materials to help communities Great Basin, into the northwest, California, the south- teach the language within the homes, villages, and west of the United State, and through Mesoamerica. schools. The major groups are Numic, Tübatulabal, Hopi, Takic, Benjamin Whorf relied heavily on his study of Hopi Pimic, Taracahitic, Corachol, and Aztecan. in developing his theory of linguistic determinism. Pre-European contact Numic peoples ranged Ekkehart Malotki, working closely with Michael through the western portion of the United States into Lomatuway’ma, has published seven collections of Alberta and Saskatchewan; postcontact, their range Hopi tales, presented in bilingual format. He has also contracted. Today, groups live in Wyoming, Idaho, explored the encoding of time in the Hopi language, Oregon, Utah, California, Colorado, Nevada, Arizona, the basis for much of Whorf’s speculation. and New Mexico. Numic languages include Comanche The Takic branch, situated in Southern California, and Shoshoni (also written Shoshone), Kawaiisu, Ute, consists of Cahuilla, Cupeño, Luiseño, and Serrano. Northern and Southern Paiute, and Mono. Although Each of these languages has fewer than 50 speakers; covering a wide area, languages of this group remain only one speaker was reported for Serrano in 1994. quite similar in lexicon and morphology, suggesting a Although California has had a mentor-language-learn- relatively recent dispersal. Miller (1983) follows Lamb ing system, which pairs young Native Americans with (1958) and Fowler (1972) in surmising that the initial fluent speakers of their languages for total immersion, communities inhabited the southwestern part of the some elders prefer not teaching the language to teach- Great Basin, which began to separate into northern and ing one that they must fill with neologisms. southern groups about 2,000 years ago, moving out to The Pimic branch (also called Tepiman) includes their greatest extension around 1,000 years thereafter. Pima, Papago, and Tepehuan. Papago-Pima is also Today, Paiute and Shoshoni remain vigorous lan- known as O’odham, O’othham, Nevome, Nebome, guages. The Yerington band dialect of Paiute is expand- and Upper Piman. The term O’odham means ‘We, the ing its forms and domains of usage. Most other People’. The Pimas of the Gila River area call them- Numic-speaking communities are shifting to English selves ʔakimel ʔoʔodham ‘river people’, while the as a first or dominant language; this shift is complete Papago to the south are Tohono ʔoʔodham ‘desert peo- for Mono and Kawaiisu. ple’. Through the early 1900s, Pima elders continued The Tübatulabal branch consisted of a single family. to record their life histories on calendar sticks. The It became extinct at the end of the twentieth century. writing system appears to have been a mixture of per- Hopi remains a viable language, spoken in north- sonal symbols and ideographs. Most of these records east Arizona, Utah, and New Mexico. Eighty percent were destroyed upon the authors’ deaths. The lan- of ethnically ascribed Hopi speak their language. The guages are now recorded in the roman alphabet. The 721 NAHUATL AND UTO-AZTECAN LANGUAGES Tohono O’odham have K-12 bilingual education. Two Nahuatl scholars, produced a compendium of cultural Piman languages, both referred to as Pima Bajo, are exegeses, which he entitled General history of the spoken in Mexico, Chihuahua, and Sonora. Speakers Things of New Spain. Treatises cover precontact reli- are just becoming bilingual in Spanish. Tepehuan, gious practices, sacred songs, origin tales, divination, divided into three major dialects, is spoken in philosophy, astronomy, political structure, occupations, Durango, Chihuahua, and, until recently, in northwest- plants and animals, and the Spanish invasion. Nahuatl ern Jalisco, and has approximately 35,000 speakers. intellectuals in the early colonial period, among them Tubar, once spoken in Chihuahua, is now extinct. Chimalpahin and Tezozomoc, wrote political and his- Taracahitic consists of Guarijio, c. 5,000 speakers in torical essays, and composed poetry and songs. Major the western Sierra Madre; Tarahumara, in Chihuahua, tropes include parallelism (morphological, syntactic, with around 72,000 speakers; and the Cahitic lan- and semantic), metaphor, couplets-for-class (naming a guages: Mayo, in southern Sonora and northern class by two prime examplars, as cua¯uhtli, o¯ce¯lo¯tl Sinaloa; Ópata in the Districto Federal and the state of ‘eagle, jaguar’, the two major warrior societies stand- México; and Yaqui, with speakers in Sonora and in ing for ‘warriors’; to¯chtli, maza¯tl ‘rabbit, deer’, two Arizona. While Ópata is moribund, Mayo and Yaqui prime wild animals, standing for the ‘wild’), and inver- continue to be highly used, and are similar in structure sion. Parallelism is a common feature in indigenous lit- and lexicon. Yaqui is tonal, while Mayo is not. eratures of the Americas, as are the couplets-for-class. Corachol is made up of Cora, spoken in Nayarit Inversion is more particular to colonial Nahuatl canons. and in a pocket of western Colorado, and Huichol in In inversion, a word or phrase is used to refer to its northwestern Jalisco. All the languages of this group ‘inverse’ or antonym. Thus, cua¯uhtli, o¯ce¯lo¯tl ‘eagle, are tonal. jaguar’ may be used to refer to a commoner, a farmer, The Nahuatl branch, also called Aztec, Mexicano, rather than to the elite warriors; pilli ‘child’ is also the and Nahua, is robust, with around 1.9 million speak- term for a person of the nobility. ers. Precontact Nahuatl-speaking communities were Legal and ecclesiastical records were still kept in scattered from the southern Sonoran desert through Nahuatl through the eighteenth century. In the mid- Salvador and Honduras. The Spanish crown made twentieth century, Mexican intellectuals began a Nahuatl an official language for legal proceedings. Nahuatl revitalization, publishing poetry and prose Many non-Nahuatl communities retained Nahuatl again in Nahuatl. UNAM publishes a journal scribes to file their land claims, wills, and petitions. Tlalocan, primarily dedicated to Nahuatl studies. The Spanish priests wrote excellent dictionaries and gram- Mexican Ministry of Education sponsors bilingual mars of the language; a Nahuatl speaker trained by the education in Nahuatl communities. A full range of Jesuit priest/linguist Horacio Carochi composed a teaching materials is available for K-12. Mexican- series of ‘dialogues’ meant to serve the priests as a Americans have also taken up Nahuatl as a symbol of model for polite discourse (Karttunen and Lockhart identity. Nahuatl words, written and spoken, are used 1987). In addition, many Nahuatl town scribes record- emblematically by populations in California and, to a ed local histories in annals; although postcontact and lesser extent, Arizona. written in the roman alphabet, many of these annals Pipil is the Nahuatl variety spoken in El Salvador. record earlier events. That of Quauahtitlan covers Official estimates have long placed this language close 635–1519 CE; that of the Chichimecs spans to extinction. However, a 1979–1980 survey by the 1116–1544 CE. Some or all of these documents may Universidad Centroamericana José Simeón Cañas have been based on earlier nonalphabetically recorded revealed that there were still viable populations of texts. Precontact Nahuatl writings are predominatly speakers in the three northernmost departments. With picto- and logographic. There is evidence that some the resolution of the civil war, a bilingual education signs in late texts were being used for their phonetic project was established. Materials and teachers from value, rather than as illustrations of the words or con- sister communities in central Mexico helped jumpstart cepts they represented, perhaps indicating a shift the schools in Pipil-speaking communities. toward a more syllabic or phonetic writing system. Proto-Uto-Aztecan had a relatively simple conso- Surviving precontact writings are dynastic histories, nant inventory: p, t, , k, kw,ʔ, s, h, m, n, w, y, and five annals, maps, divinatory guides, or combinations of vowels. Classical Nahuatl had long and short i, e, a, these. and a back vowel [o~u]. Some modern languages Nahuatl schools for noble youths were often taken within the family are tonal, but tone appears to have over by priests, postcontact. Early texts produced by been developed from glottal stops and consonantal young men trained in these schools occasionally sandhi. Most languages of the group have fixed stress. retained logograph-like illustrations or marginalia. Morphologically, languages of this group tend to Fray Bernardino de Sahagún, working with a cadre of share the use of an absolutive suffix, although they 722  NAHUATL AND UTO-AZTECAN LANGUAGES vary with respect to its ubiquity. Some languages have with indefinite actor pronoun affixes on the verbs. only vestiges of this affix in citation forms. In others, Number can be marked in several ways in Uto- such as modern spoken Nahuatl, the affix appears Aztecan languages, including use of numerals, inflec- when noun stems are neither possessed nor initial ele- tion on nouns and pronouns, on verb forms, and on ments of compounds. Compounding is a productive pronominal verb affixes. Languages within the family mechanism for word formation, as are cliticization and differ with respect to the obligatoriness of number affixation. In Cora, the verb word may be quite long, marking and agreement. Classical Nahuatl did not as habitual objects, goals, and instruments may be require number agreement between pronominal affix- incorporated into the verb word. Verb–verb, es and coreferent noun phrases, although increasing verb–adverb compounds are also common. Huichol animacy brought increasing agreement. Modern spo- has productive compounds; Mayo has relatively short- ken varieties of Nahuatl more consistently mark plural er words, and Classical Nahuatl combined compound- on nouns, both animate and inanimate. There is also ing, with elaborate affixation. A verb word in Classical more agreement between pronominal verb affixes and Nahuatl could contain a negator, an incorporated their coreferent noun phrases. Classical Nahuatl had a adverbial, an antecessive marker, additional time variety of mechanisms for inflecting for plural, includ- adverbs, subject pronouns, object pronouns, move- ing reduplication. Modern Nahuatl retains examples of ment markers, reflexives, incorporated noun earlier plural classes in fossilized forms. The produc- objects/goals, reduplication (for intensity or distribu- tive morphemes are now restricted to a few forms for tion), incorporated indefinite objectmaker, further general pluralization, diminuitives, and vocatives. reduplication, a verb root, a series of derivations Hopi may mark the plurality on nouns by reduplica- affecting transitivity, a connective, one or more auxil- tion. Nouns and verbs in Hopi may also carry plural liary verbs, aspect marker, and a number marker. suffixes. Some Hopi verbs have suppletive stems for Nouns, although less complex, could be elaborate as the plural. Hopi also makes a distinction between sin- well, containing reduplication, one or two roots, gular, dual, and plural for subjects. Some speakers derivational affixes, and either an initial possessive or retain a separate dual noun suffix; others do not, but a final absolutive maker. Common postpositions the singular, dual, plural distinction is shown by marked both positive and negative affect, location, agreement on the verb. A singular noun phrase with abundance, position, or manner. singular verb agreement is interpreted as singular. A Modern Nahuatl retains most of these possibilities, plural noun phrase with a plural verb is plural, while a although the extended alternation of transitivizing and plural noun phrase with singular agreement is dual. intransitivizing verbal suffixes found in courtly colonial The Uto-Aztecan family is large and diverse; mem- Nahuatl writings does not appear in modern honorific ber languages are spoken from the Canadian border to usage. The compounding potential of Nahuatl has made El Salvador. The family retains many typological and the formation of neologisms relatively easy. This poten- structural features, with enough shared lexicon to tial is open to the other languages of this family. allow fairly strong reconstructions. Suárez (1983: 28) Classifiers for nouns or for counting are relatively gives a glottochronological estimate of 5,100 years of undeveloped within Uto-Aztecan. Most languages separation. have a few. Classical Nahuatl had seven. Modern spo- ken Nahuatl varieties seldom use more than three. This References may be related to the use of Spanish number expres- sions for most counting. Anderson, Arthur J.O., and Charles E. Dibble. 1982. Florentine Proto-Uto-Aztecan appears to have placed verbs codex: general history of the things of New Spain, Fray Bernardino de Sahagún. Santa Fe, New Mexico: The School following the complement or direct object (SOV). of American Research and the University of Utah. Many modern languages have experienced some Campbell, R. Joe, and Frances Karttunen. 1989. Foundation reordering in main clauses, while retaining the early course in nahuatl Grammar, Vols. 1 and 2. Missoula, MT: order in subordinate clauses. Most languages in this The University of Montana. family have postpositions and, at least, allow Fowler, Catherine S. 1972. Some ecological clues to proto- numic homelands. Great Basin cultural ecology: a sympo- Subject–Object–Verb order today. Preposing an argu- sium, ed. by Don D. Fowler. Desert Research Institute ment or clause is used for emphasis. Relative clauses Publications in the Social Sciences 8. 105–21. may occur without a relative pronoun. Subordination Jaquith, James R. 1970. The present status of the Uto-Aztekan is often by juxtaposition. In Proto-Uto-Aztecan, sub- languages of Mexico; an index of data bearing on their sur- ject noun phrases of subordinate clauses were inflect- vival, geographical location, and internal relationships. Greeley: Museum of Anthropology, University of Northern ed as objects. Most languages of the family retain this Colorado. trait. Nouns and adjectives may serve as predicates, Kalectaca, Milo. 1978. Lessons in Hopi, ed. by Ronald W. with a ‘zero’ copula. Generic action may be shown Langacker. Tucson, AR: The University of Arizona Press. 723 NAHUATL AND UTO-AZTECAN LANGUAGES Karttunen, Frances E. 1983. An analytical dictionary of Miller, Wick. 1983. Uto-Aztecan languages. The handbook of Nahuatl. Austin: University of Texas Press. North American Indians, 113–24. Washington: Smithsonian Karttunen, Frances E., and James Lockhart. 1987. The art of Institution. Nahuatl speech: the Bancroft dialogues. Los Angeles: UCLA Press, Margaret Lauritsen. 1979. Chemehuevi: a grammar and Latin American Center Publications, University of California. lexicon. Berkeley: University of California Press. Lamb, Sydney. 1958. Linguistic prehistory of the Great Basin. Suárez, Jorge. 1983. The Mesoamerican Indian languages. International Journal of American Linguistics 24(2). Cambridge: Cambridge University Press. 95–100. JUDITH M. MAXWELL Mason, John Alden. 1950. The language of the Papago of Arizona. Philadelphia: University Museum, University of See also Mexico; United States; Whorf, Benjamin Pennsylvania, 1950. Lee Naming Names have always been seen as a very special class do not allow for plurals. It is true that names such as of linguistic signs, and naming as a very special lin- John or Newcastle may have many different referents guistic process. Many cultures and societies associate (John Smith, John Baker, John Howard, etc.; names and naming with magic and taboo. A famous Newcastle, UK; Newcastle, Australia; etc.). But even if example is the tale of Rumpelstiltskin: a name refers to more than one single extralinguistic referent, those referents do not have anything in com- The goblin Rumpelstiltskin helps a poor maid to become mon. Therefore, sentences such as Johns make good queen but demands her first child in return. When she lovers or Newcastles are good places to live in do not gives birth and refuses to give up her child, she can only make sense. This seemingly clear-cut difference break the goblin’s power over her by discovering his secret name. When she finally does, Rumpelstiltskin between proper nouns and common nouns on the basis explodes with anger and rage and his power is broken of unique and common reference can be attacked on a forever. number of grounds, however. First, on closer inspec- tion the various referents of John and Newcastle do The relationship between person and name is so close have some things in common: Johns will be male; will that in many cultures saying the name of a person is have been born into an English-speaking environment; believed to result in serious harm. Some Australian and, given that most personal names are subject to Aboriginal groups bestow two names on each person: fashion, one can even make an educated guess about one that is commonly used, and one that is only known the age of many Johns. Second, there are a number of to the fully initiated and is only uttered on the most common nouns whose reference is also unique: sun, solemn occasions and even then only in a whisper (for moon, hell, etc., all refer to a unique extralinguistic ref- further examples of magic and taboo associated with erent, although they are not usually considered names. names, see Frazer 1993). This essay provides an Other criteria that have been suggested for distin- overview of onomastics (as the academic discipline of guishing between proper nouns and common nouns are name study is called) by reviewing the status of names related to sentence structure, translation, and context. as linguistic signs, describing the uses of naming, and The syntactic criterion that defines names (in English) discussing three categories of names in more detail, is the absence of a determiner: while common nouns namely personal names, place names, and brand names. take determiners (a/the house), proper nouns do not At first glance, the difference between names or (a/the John, a/the Newcastle are ungrammatical ‘proper nouns’ and common nouns seems straightfor- expressions). However, not all common nouns can take ward enough: the former have unique reference and the full range of determiners: a music, for example, is denote individuals, while the latter have generic (or not a possible expression in English. At the same time, ‘common’) reference and denote classes of entities. some proper nouns require a determiner (e.g. The This means that all referents of a common noun such as Hague, the Buddha, the Potomac). With brand names, house have something in common (walls, windows, the distinction is lost completely, as the determiner roofs, etc.), and a sentence such as Houses are impor- contrast can be used here to distinguish between the tant for humans to live in makes sense. However, this is name of a company and the name of a product, as in not true for proper nouns, and therefore they normally Onassis bought Ford vs. Onassis bought a Ford. 724  NAMING With respect to meaning and sentence structure, -a, as in Lydia, Moira, or Sonya. (2) Where do first proper and common nouns clearly form a gradient, with names come from? In English-speaking societies, a sig- prototypical cases at either end and many fuzzy ones in nificant source of first names has traditionally been the between. Some scholars have argued that translatability Bible (e.g. Mark, Michael, Peter). (3) What patterns of (of common nouns) and nontranslatability (of proper sex-specific names are there? For instance, the pool of nouns) can provide the only clear distinction. However, boys’ names has tended to be much more stable over the counterexamples are easy to find, even if they are rather years than the pool of girls’ names. In recent years, there rare: e.g. New York can be translated into German as is also an increasing trend toward gender-neutral names Neu York and into Spanish as Nueva Yorca, or German such as Ashley, Taylor, or Madison. (4) How has first Schwarzwald becomes Black Forest in English. name choice developed historically and what naming Ultimately, the intention and perception of language fashions are there? Less than a century ago, the pool of users is the only way to distinguish between common first names was much more limited than it is today, now nouns and proper nouns. However, this distinction is no that parents wish to express their child’s individuality as longer systematic and inherent in the language, but uniquely as possible. (5) How and why do people determined by context: whether any given noun is con- choose particular names? Reasons include family tradi- sidered a name or not would thus shift from context to tion, ‘hero worship’ (e.g. naming a child after a celebri- context. The shifting use of words as proper and com- ty), expressing a wish for the child’s future or its mon nouns is best exemplified with brand names: from personality (e.g. Victoria, Rose), or sound appeal. (6) Is the producers’ perspective, brand names have to be there a link between name and personality? Such links unique and distinctive. At the same time, they refer to undoubtedly exist but it is almost impossible to say classes of items that are more identical than any two whether people live up (or down) to their names or houses are ever likely to be. Indeed, with the wide- whether they are perceived in particular ways because spread practice of badge-engineering (diversifying your of the names they carry. product range by selling the same item under different Interest in family names, on the other hand, is often names), the name may even be the only difference that genealogical, and records of family names are used to exists across different classes of items. With respect to trace the settlement history of an area or the migratory context, there are two important uses to which names history of people. Other concerns include the etymolo- can be put. First, they can be used to identify and indi- gy of family names, the pattern of their transmission, vidualize their referents. It might be argued that this and laws regarding naming. Further areas of inquiry purpose could be more rationally achieved through into personal names are the use of pseudonyms, nam- numerical coding, but names are more memorable and ing in literature, name change, nicknames, and their much richer in connotation. Second, personal names are use in functions that are not directly related to naming. used to address people. Together with terms of address, These include generic forms such as Jane Doe or John names allow for rich combinatorial possibilities of Doe, metaphoric uses (e.g. John Bull, Uncle Sam), use expressing the social relationship between speaker and in idioms and proverbs (e.g. jack-in-the-box, all work addressee (e.g. Carrie, Carol, Carol Walker, Ms. Carol and no play makes Jack a dull boy), and metonymic Walker, Ms. Walker, Dr. Walker, Aunt Carol). Only per- uses, which are most often found in various specialized sonal names are normally used in forms of address, and registers (e.g. Mercedes, Creutzfeldt-Jakob disease). these will now be described in some more detail, fol- Place names include names for settlements, rural lowed by place names and brand names. areas, rivers, and streets. The study of place names is Personal names are names used for people, but relevant to the work of dialectologists, archeologists, names for pets are also sometimes included in this cat- and historians because place names can provide valu- egory. In many parts of the world, personal names con- able clues to the settlement history of an area, in addi- sist of a ‘given’ part and a ‘family’ part, which tend to tion to or in the absence of other linguistic and material be fundamentally different in that given names are usu- records. Britain’s Celtic (e.g. Thames), Anglo-Saxon ally a matter of choice while the surname runs in the (e.g. Winston), Latin (e.g. Chester), Viking (e.g. family. The surname is largely conventionalized in that Derby), and French (e.g. Bellevue) names reflect the it is usually the father’s family name in many parts of successive waves of conquest that have shaped the his- the world (although recent societal changes have led to tory of the island. Consequently, the main research an increase in the mother’s name or a combination of interest is in the etymologies of place names. However, parental family names also being used). Research ques- the formation of place names and their meanings are tions concerning these two categories of personal names also of interest. Place names typically refer to the form differ significantly. Questions about first names include of a place (e.g. Hillsdale), its geological quality (e.g. the following: (1) How are they formed? Female first Rocky Mountains), its situation in relation to water (e.g. names, for instance, often follow a pattern of ending in Tennant’s Creek), its position (e.g. West Ryde), its flora 725 NAMING (e.g. French Forest), or its fauna (e.g. Emu Plains). dismiss this widely held assumption as ‘primitive’ or Furthermore, place names often refer to human uses to ‘superstitious’. This stance has led to an ever-widening which the land was put (e.g. Hyde Park), they com- chasm between limited academic interest in names and memorate events (e.g. Breakfast Creek), or they are naming (as evidenced e.g. in the small number of aca- named after deities or people (e.g. San Francisco). demic programs in onomastics) and widespread popu- Brand names have often been excluded from ono- lar interest (as evidenced e.g. in the innumerable mastic research on the grounds that they are closer to internet sites devoted to names and naming). It seems common nouns than to personal and place names. that Juliet’s question remains as enigmatic as ever: However, brand names are fascinating, both in form ‘What’s in a name?’ and in meaning, because they are more consciously created than any other words. In addition, there often References exist good records of the creators’ motivation in choosing a particular brand name. As a matter of fact, Abate, Frank R. (ed.) 1994. Proper names master index: a com- some of the most creative minds are engaged in brand prehensive index of more than 200,000 proper names that appear as entries in standard reference works. Detroit: name creation. Brand names have to fulfill a number Omnigraphics. of legal and commercial requirements. The legal con- Algeo, John. 1973. On defining the proper name. Gainesville: straints are the various copyright and trademark laws, University of Florida Press. which tend to exclude offensive words as well as Arney, Roberta. 2001. Onomastics bibliography. El Paso: words of public interest (e.g. federal, national) and University of Texas at El Paso Library. <http://libraryweb. utep.edu/ref/onomast.html> disallow words that are similar to brand names already Baker, Ronald L. 1991. The study of place names. Terre Haute, in use. From a commercial point of view, brand names IN: Indiana Council of Teachers of English, Hoosier have to be distinctive and memorable: they have to aid Folklore Society. the advertising of a product and encapsulate the Dickson, Paul. 1996. What’s in a name? Reflections of an irre- brand’s image. Creating a name that fulfills all these pressible name collector. Springfield, MA: Merriam Webster. Eichler, Ernst, Gerold Hilty, Heinrich Löffler, Hugo Steger, and functions may prove to be very difficult, particularly in Ladislav Zgusta (eds.) 1995/1996. Namenforschung/Name today’s global marketplace, where a name should ide- Studies/Les Noms Propres: Ein Internationales Handbuch ally accomplish that goal in a number of languages. zur Onomastik/An international handbook of onomas- Despite the extensive market research that goes into tics/Manuel International d’Onomastique, 2 vols. Berlin, brand name creation (in the late 1990s, the creation of New York: Mouton de Gruyter. Frazer, James. 1993. The golden bough: a study in magic and an international brand name cost around $60,000), religion. Ware, UK: Wordsworth Reference; 1st edition 1922. grotesque multilingual failures are not unheard of. The Koss, Gerhard. 1990. Namenforschung: eine Einführung in die Mitsubishi Pajero, for instance, failed in Spanish- Onomastik. Tübingen: Niemeyer. speaking markets because pajero means ‘masturbator’ Matthews, C.M. 1972. Place names of the English-speaking there. Similarly, the Chevrolet Nova did not sell well World. New York: Scribners. Morgan, Jane, Christopher O’Neill, and Rom Harré. 1979. in Spanish-speaking markets because no va translates Nicknames. London: Routledge and Kegan Paul. as ‘does not move’––obviously a poor omen for a car. Piller, Ingrid. 1996. American automobile names. Essen: Blaue There can be no doubt that the relationship between Eule. a name and its referent has more psychological reality Smith, Elsdon C. 1973. The new dictionary of American fami- for language users than that pertaining between any ly names. New York: Harper & Row. A useful gateway to naming on the internet is provided at other linguistic signs and their referents. This observa- http://jerryhill.tripod.com/ tion has excited philosophers for centuries, as is evi- INGRID PILLER denced in Plato’s Cratylus. Modern linguistics, philosophy, and onomastics, however, have tended to See also Etymology; Proper Nouns; Reference Natural Classes The notion of natural class is widely used in phono- also be applied to other domains––for example, logical theory for the categorization of segments or a group of morphemes may constitute a natural class. phonemes, our focus here, although the notion can As we will see, this notion is useful for descriptive 726  NATURAL CLASSES purposes, but plays no direct role in grammatical the- of [t] ([voiced, continuant]) will be a subset of the ory. Given this derivative status, we begin with an representation of [p] ([voiced, continuant,  labi- informal example, instead of choosing one of the var- al]) or [k] ([voiced, continuant,  velar]). In such ious formalizations of the notion that are implicitly or a system, the representation of [t] is identical to the explicitly used in the literature. representation of the class [p,t,k]. Informally, the featural representation [voiceless (iv) Relation to substantive categories: Does every stop] defines a natural class of English phonology, formally definable natural class constitute a natural which includes the sounds [p, t, k] and no others. The phonetic (acoustic or articulatory) grouping of other sounds of English are either voiced (e.g. [a, m, sounds? Clearly, there is a fairly straightforward match g]) or not stops (e.g. [f s]). Because natural classes are between a natural class such as the voiceless fricatives defined in terms of a conjunction of shared properties, [voiced,  continuant] of English and a cluster of it is generally the case that using more features, yield- phonetic properties. However, it is less clear that a ing a more specific (richer) specification, will define a definable natural class such as [lateral], which class with fewer members: the class defined by [voice- would potentially include all the sounds of English less stop] has fewer members than the class defined by except for the lateral [l], forms a phonetically identifi- [stop], since the latter includes voiced and voiceless able grouping. There are several approaches to take. stops. Several issues arise when discussing natural First, it may not be the case that [lateral] describes classes: (i) language specificity, (ii) organization of the all the sounds of English besides [l]. Perhaps the vow- feature inventory, (iii) underspecification, (iv) relation els are completely unspecified for the feature [lateral]; to substantive (e.g. phonetic) categories, (v) level of perhaps the feature [lateral] is only present on the non- analysis (UR, SR), and (vi) status in grammatical the- nasal sonorants [l] and [r], the first being [ lateral] ory. We briefly consider each of these. and the second [lateral]. Second, just because a nat- (i) Language specificity: It is crucial to note that ural class is definable formally, it does not follow that natural classes can only be defined with respect to a it must play a role in the phonology of every language, given language. For example, in English, [t, n] do not or even of any language. form a natural class of sounds: using binary valued The natural classes that can be defined depend upon features, their representations both contain the set of the set of features available to the language faculty; features [continuant,  coronal]. However, this set thus, the set of natural classes is theory dependent. For of features is also contained in the representation of example, given a set of features that distinguishes the [d], which also occurs in English. Any description of vowels [i, I, e, E] on the basis of height alone, there is [t,n] that merely conjoins features cannot include [t, n] no way to group [i, e] or [I, E] into natural classes by but excludes [d]. However, in a language where there themselves. However, if the feature system represents is no [d], where say [t] and [n] are the only noncontin- the former pair as [ tense] and the latter as [tense], uant coronal consonants, [t, n] would constitute a nat- we can represent the relevant classes. Phonetic factors ural class of sounds. ([ tense] vowels share articulatory properties) and (ii) Organization of feature inventory: If phonolog- phonological factors such as the existence of harmony ical representations are assumed to consist not merely systems that treat [i, e] uniformly as opposed to [I, E] of unordered sets of features, then natural classes can- support the existence of the relevant feature. not be defined in simple set theoretical terms––mem- This brings us to an apparent circularity of reason- bers of a set are not organized in any way. For ing: rules or processes are defined as applying to natu- example, in theories of Feature geometry, which ral classes of sounds, and natural classes are defined impose hierarchical groupings of features, the more on the basis of features that are posited when a set of complex notion of subsumption must be used. sounds is treated uniformly by rules. For example, in (iii) Underspecification: Another formal problem Lamba, there is a process that palatalizes [s] to [S] arises with the use of certain kinds of underspecified before [i]. There is also a palatalization of [k] to [C] representations. If, at a given grammatical level, two before [i]. It is tempting to collapse these processes representations A and B differ in that A contains a sub- into a single rule; however, there is no feature system set of the information contained in B, then it is not that will treat [s, k] as a natural class to the exclusion possible to distinguish the representation of the natu- of [t], which is also found in the language and which ral class containing both A and B from the representa- does not undergo palatalization. Should we take the tion of just A. In other words, A and B are not distinct Lamba data as evidence of a new feature [F], such that in the sense of Chomsky and Halle (1968:328). This [s, k] are [ F] and [t] is [F]? There are several rea- issue arises in models that treat certain segments as sons not to do so. First, our identification of the two defaults. If [t] is treated as a default stop, with no fea- processes may be somewhat superficial––palataliza- tures for place of articulation, then the representation tion’ is a descriptive term that does not obviously rep- 727 NATURAL CLASSES resent a single process. It covers a wide range of syn- theoretical content, or an actual element of linguistic chronic and diachronic phenomena. Second, we can theory? The best view is probably the second––a nat- restrict the distinctive feature theory by requiring that ural class is just a convenient way to refer to the set of every featural distinction be transducible from the representations a rule applies to, but speakers’ knowl- acoustic signal. There is no possible acoustic parame- edge of phonology need not contain the notion. ter that is shared by [s, k] to the exclusion of [t]. In Natural classes of segments are affected by rules since other words, we can take a phonetic parameter to be a the Structural Description of the target of a rule may necessary (but not sufficient) correlate of a distinctive contain a subset of the information found in any par- feature. ticular token of a representation that the rule affects. (v) Level of analysis: It will have been noted that For example, if a grammar contains a rule allophonic variation was not taken into account in the [sonorant] > [voiced] / __ [voiced] discussion above. For example, it was stated that [p, t, k] constitute the natural class of voiceless stops in then any representation that is [sonorant] will under- English, without acknowledging the existence of aspi- go the rule in the right context, but this set of represen- rated stops [ph, th, kh]. Should natural classes be defined tations need not be identified within the grammar as at the level of Underlying Representation, the level of constituting a natural class, independently of this rule. Surface Representation, or at both these levels? Since, as the following paragraph will argue, the notion of natural class is a mere descriptive convenience, natural References classes can be referred to wherever it is convenient to Chomsky, Noam, and Morris Halle. 1968. The sound pattern of do so, as their existence is epiphenomenal. English. New York: Harper & Row. (vi) Status in linguistic theory: Given the dubious Clements, G.N., and Elizabeth Hume. 1995. The internal organ- status of the phoneme in generative phonology, and ization of speech sounds. The handbook of phonological the- ory, ed. by John Goldsmith, 245–306. Oxford: Blackwell. the equally difficult notion of segment, the question CHARLES REISS arises of the status of the notion natural class. Is the notion useless, a useful descriptive device without real See also Feature Theory Navajo and Athabaskan-Eyak Language The Navajo language is one of several Southern continuum, i.e. neighboring languages/dialects may be Athabaskan languages, more commonly known as mutually intelligible. Apachean languages, spoken in the Southwest. The (2) Pacific Coast Athabaskan (and extinct Athabaskan languages, along with Tlingit and Eyak, Kwalhioqua-Tlatskanai)—including Hupa and Tolowa make up the larger Na Dene language family. There in California, and Tututni in Oregon. are three major geographical groupings of Athabaskan (3) Navajo and Apache—including seven Apachean languages: languages divided into two main groups: (1) the (1) Northern Athabaskan—including about 25 lan- Western group, including Navajo in one branch and a guages spoken in Alaska and Canada, including second branch divided into the San Carlos group (San Koyukon, Holikachuk, Tanacross, Ahtna, Dena’ina, Carlos, White Mountain, Cibecue, etc.) and the Deg Hit’an, Tanana, Upper Kuskokwim in Alaska, Han Mescalero-Chiricahua group, and (2) the Eastern and Gwitch’in in Alaska and Canada, Tagish, Tutchone, group, divided into a branch containing Jicarilla and and Tahltan at the Yukon headwater, Tahltahn, Sekani, Lipan and a second branch containing Plains Apache. Tsetsaut, Babine, Carrier, and Chilcotin in the far west Of the Athabaskan languages, all but Navajo are of Canada, and Hare, Bearlake, Mountain, Kaska, moribund, with little if any intergenerational transmis- Dogrib, Slave, Beaver, Chipewyan, and Sarcee across sion. Even Navajo, with approximately 125,000 speak- the rest of western Canada. It is important to note that, ers, is threatened: while more than 95% of children rather than discrete groups, the Alaskan and Canadian entering bilingual education programs were fluent in Athabaskan languages constitute a language and dialect Navajo in the mid-1970s, today fewer than 50% of 728  NAVAJO AND ATHABASKAN-EYAK LANGUAGE kindergartners are fluent. Of the Apachean languages, composed of a root plus suffixes (or other modifica- Navajo is by far the best documented, and indeed may tion) indicating mode and aspect. Prefixes indicate be the best documented of any Native American lan- subject and objects, mode and aspect, and adverbials. guage. This is due in large measure to the pioneering The prefix complex can be analyzed as a template work of Robert Young and William Morgan. Their comprised of basic positions or zones preceding the grammars and dictionaries of Navajo over the past half stem. The prefix chart in Figure 1 gives a general idea century (e.g. 1987, 1992) have set the standard for work of the ordering of prefixes. Note that the stem is in both in Apachean and in Athabaskan more generally. position zero and subject prefixes are in positions 2 and 5, with tense, aspect, and mode in 3, and direct Phonology object in 6. Some of the prefixes noted above are obligatory in The distinctive sounds (phonemes) of Athabaskan the sense that the particular position they occupy must include obstruents, sonorants, and vowels. The ances- be filled by one of the possible variants of that prefix tor language—Proto Athabaskan—likely contained class in every derivative of the verb. These obligatory the following stop obstruents: prefixes are the person, tense, aspect, mode, and T TL TS CH Ky CHw Q voice/valence (also called the ‘classifier’) prefixes. The modern languages show a range of shifts Other, derivational, prefixes, e.g. adverbial, iterative, among the Proto-Athabaskan stops. The following list are optional—he positions these prefixes occupy may shows the modern reflexes of the Proto-Athabaskan or may not be filled depending on the meaning of the word *-tsi’ ‘head’: derivative. Sometimes, prefixes are thematic—they encode a situational participant—and they are lexical- Tanaina -tsi’ ized, which means that they are present in every deri- Tanana, Han -tthì’ Koyukon -tli’ vation of a particular verb. These thematic prefixes in Gwitch’in (AK) -kì’ combination with the voice/valence prefix and the root Gwitch’in (Can.) -chì’ of the verb are referred to as the VERB THEME. Hare -pfí’, fí’ The Athabaskan theme is the underlying skeleton or Bearlake -kwí’, kfí’ verb construction to which prefixes and suffixal ele- Navajo -tsi:’ ments are added in producing an utterance. The theme The sonorants of Proto Athabaskan were w and y, itself has a meaning and is the basic unit of the m, n, and a palato-velar nasal. Proto Athabaskan vow- Athabaskan verbal lexicon. Verb forms derived from els are reconstructed as i, e, a, u plus three reduced themes have, in addition to the stem (i.e. the aspectu- vowels (mid, low, and back). All of the modern ally suffixed root) and theme prefixes, inflectional pre- Athabaskan languages have both oral and nasal vow- fixes (person, number, etc.) and derivational prefixes els. Some of the languages have short and long vow- (aspect, adverbials, etc.). These prefixes are arranged els; others have full and reduced vowels. in relatively fixed positions preceding the stem as The phonemes of Navajo include plain, aspirated, shown in Figure 1. and glottalized stops and affricates (d, dl, dz, j, and g In Navajo, the THEME consists of the valence prefix are the symbols used for the plain stops), fricatives (s, (position 1) plus the stem (position 0). The VERB BASE z, sh, zh, h, gh), and liquids and sonorants (l, B , y, w, consists of the mode—conjugation prefixes (position n, m). Navajo has plain and nasalized short and long 3) plus the subject pronoun (position 2) plus the verb vowels, and high vs. low tone (distinctive pitch). theme. Examples of verb themes in Navajo, with verb words derived from them, are shown below: Morphology and Verb Structure Theme: na + θ + né ‘play’ Example verb: naashné ‘I’m playing’ The Verb Complex Analysis: na + θ + sh + né The Athabaskan verb is a complex polysynthetic Thematic + imperf + 1st person subject + stem structure made up of a stem plus prefixes. The stem is Theme: ha # O + B + géésh ‘cut O (object) out’ 10 9 8 7 # # 6 5 4 3 2 1 0 post derivational/ reflexive distrib./pl disjunct DO Subj. derivational/ mode/ Subj. voice/valence stem positions thematic reversionary boundary (3rd) thematic conjugation (1st & 2nd semiliterative sg & pl) iterative Figure 1. Navajo prefix chart. 729 NAVAJO AND ATHABASKAN-EYAK LANGUAGE Example verb: hadeiilgéésh ‘we’re cutting it out’ or paired objects, (3) slender, rigid, or stick-like Analysis: ha # da # θ + θ + sh + B + géésh objects, (4) animate objects, (5) plural separable Thematic # distributive plural # 3rd person direct obj. + imperf. + 1st person subj. + objects, (6) plural objects in profusion, (7) noncom- valence + stem pact matter, (8) objects in open containers, (9) flat, flexible, or cloth-like objects, (10) mushy wet objects, and (11) heavy, large, or bulky objects. The Aspectual System Aspect in Athabaskan is expressed by a complex, mul- tidimensional system. There are two major categories Word Order that combine in expressing the temporal contour of the The word order of the Navajo sentence is state or activity described by the verb. The convention Subject–Object–Verb (SOV). in Athabaskan linguistics is to call one of these two Aééchaa’í mósí yinooB chééB categories MODE and the other ASPECT (although some Dog cat it’s chasing it researchers refer to these categories as ‘aspect’ and S O V ‘aktionsart’, respectively). Both of these categories are obligatory—every verb must be marked for both mode Most transitive Navajo sentences do not name both and aspect. This morphological marking is an inter- the subject and object in noun phrases. When only one secting one, i.e. mode and aspect are marked by pre- is mentioned as a noun phrase, it is usually the direct fixal and suffixal elements, which encode both object, that is, the new information. In this case, the categories. direct object has an agreement prefix on the verb, There are seven modes in Navajo, describing activ- which is yi-. However, when the topic, or old informa- ity as incomplete, complete, ongoing, future, potential, tion, is mentioned in a noun phrase, this unusual case customary, or recurrent. Examples of verbs in the is marked by a different object agreement prefix, bi-. imperfective and perfective modes are: For example: Imperfective: yáshti’ ‘I’m talking’ Mary hayííB tî ‘He carried Mary up’ Perfective: yáB áti’ ‘I talked Mary habííltî ‘Mary carried him up’ The second basic temporal category is aspect, which describes the manner in which an activity or The Future of Athabaskan Languages event is carried out over time—whether it happens Work on language revitalization is under way in many once, or repeatedly, or at length. There are 12 aspects Athabaskan-speaking communities, and efforts in the in Navajo and they allow the expression of such dis- Navajo community are perhaps the most robust. There tinctions in meaning as: ‘I am red’ vs. “I turned red’ are language programs in many schools across the and ‘I am chewing it’ vs. ‘I bit it’. For example: Reservation, and at the two branches of Diné College. Durative: yáshti’ ‘I’m talking’ The language is also taught at the state universities of Momentaneous: ‘ayániishtééh ‘I’m starting to talk” New Mexico and Arizona. The Rock Point School on Repetitive: yádíshtih ‘I’m chattering’ the Navajo reservation is a model bilingual education program. At Rock Point, children come expecting that they will succeed and they do—the Navajo program Classificatory Verbs gives students pride in being Navajo, in their language, Athabaskan verbs in general show a wide range of clas- and in their culture. This effective program includes sificatory functions. Some verb stems are specialized lessons in Navajo by community elders on Navajo cul- according to whether the subject of the action is singu- tural matters and a strong involvement of parents and lar (or dual) or plural. Others are specialized for singu- so family both in conferences and in activities such as lar vs. plural objects. Still other actions are referred to Language Fairs and book-making nights. by different stems depending on the physical charac- teristics of the subject or object involved. These stems References are the classificatory verb stems. The Athabaskan clas- Axelrod, Melissa. 2003. The semantics of time. Lincoln: sificatory verbs categorize the material, shape, consis- University of Nebraska Press. tency, size, animacy, arrangement, quanta, and Cook, E-D., and Keren Rice (eds.) 1989. Athapaskan linguis- containment of the subjects of intransitive verbs of tics: current perspectives on a language family. Berlin: position and location, and the objects of transitive Mouton. Faltz, Leonard. 1998. The Navajo verb: a grammar for students verbs of handling. and scholars. Albuquerque: University of New Mexico Press. The 11 Navajo classificatory verbs delineate (1) Fernald, Ted, and Paul Platero (eds.) 2000. Athabaskan syntax solid, round, or compact objects, (2) slender, flexible, and semantics. Oxford: Oxford University Press. 730  NEUROLINGUISTICS Jelinek, Eloise , Sally Midgette, Keren Rice, and Leslie Saxon Rice, Keren. 2000. Morpheme order and semantic scope: word (eds.) 1996. Athabaskan language studies: essays in honor formation in the Athapaskan verb. Cambridge: Cambridge of Robert W. Young. Albuquerque: University of New University Press. Mexico Press. Sapir, Edward, and Harry Hoijer. 1967. The phonology and Kari, James. 1979. Athabaskan verb theme categories. Fairbanks, morphology of the Navaho language. Berkeley: University AK: Alaska Native Language Center Research Papers, #2. of California Press. Leer, Jeff. 1979. Proto-Athabaskan verb stem variation, part Young, Robert W. 2000. The Navajo verb system: an overview. one: phonology. Fairbanks, AK: Alaska Native Language Albuquerque: University of New Mexico Press. Center Research Papers, #1. MELISSA AXELROD Neurolinguistics Neurolinguistics is a relatively new discipline that brain-related language disorders (aphasia). Aphasia is comprises both clinical and basic research into the an acquired language disorder caused by brain dam- functional relationship between language and the age, which may induce severe impairment in both lan- brain. The research topics of neurolinguistics require a guage comprehension and production, as well as in multidisciplinary approach and share parts of tradi- reading and writing. Reasons for brain damage are tional disciplines such as linguistics, medicine, psy- mainly strokes, due to an interruption of the blood chology, and computer science, and more precisely, flow that causes the death of nerve cells in the respec- parts of the fields of psycholinguistics, aphasiology, tive area. Possible reasons for the interruption may be neurology, brain imaging, neuropsychology, and neu- a blocked vessel (arteriosclerosis), a blood clot roinformatics. Due to the combination of theories and (thrombosis), or a burst blood vessel accompanied by methods, all these fields contribute to a new discipline bleeding, e.g. caused by a ballooning expansion of a in life sciences, called cognitive neurosciences. As for weak vessel wall (aneurysm). Moreover, traumatic linguistics, this leads to an innovative research area head injuries and brain tumors in certain areas may providing empirical and neurophysiological methods cause such language deficits. Types of aphasic lan- for testing aspects of linguistic theories and models. guage disorders range from minor difficulties, which Looking back on the first emergence of common concern only naming, to major difficulties such as a research methods in the history of linguistics, five main complete loss of language. The strength of the impair- stages can be observed until now: (1) language philos- ment depends on the location and extent of the brain ophy, using introspection with an over 2000-year-old damage. Currently, aphasia affects more than tradition; (2) first empirical studies on genealogy and 1,000,000 individuals (stroke survivors) in the United roots of languages, using comparative approaches in States. By combining linguistic theory and neurologi- typology for more than 200 years; (3) empirical behav- cal knowledge on aphasia, clinical neurolinguists typ- ioral studies of psycholinguistics, using e.g. reaction ically work in stroke units and rehabilitation clinics, time experiments, with an almost 50-year-old tradition; providing mainly three occupations: (1) Diagnosis: (4) computer-aided simulation of language processes in In order to test aphasic patients on the extent and the computerlinguistics, using e.g. neural networks, for special type of language deficit in standardized exam- more than 30 years; (5) brain imaging of language pro- inations, several neurolinguistic test batteries are used: cessing and higher brain function in cognitive neuro- e.g. the Boston Diagnostic Aphasia Examination sciences during the last 20 years. Therefore, neurolinguistics is one of the newest, but also one of the most rapidly growing fields in contem- (Aphasiology, Medicine) porary linguistics. At present, neurolinguistics can be Clinical (Neuro)Linguistics divided into three major branches, depending on the research domain (see Figure 1). Experimental Simulative Clinical Neurolinguistics or Clinical Linguistics Neurolinguistics Neurolinguistics (Neurobiology, Medicine) (Computerlinguistics, Informatics) This field, formerly also called patholinguistics or aphasiology, mainly deals with clinical aspects of Figure 1 731 NEUROLINGUISTICS (BDAE), Porch Index of Communicative Ability changes within certain frequency bands, the correlat- (PICA), or the Western Aphasia Battery (WAB). (2) ed activity of different brain areas during a task can be Speech therapy: This represents the main activity of measured (= coherence analysis). This latter analysis the clinical neurolinguists, who have to perform adapt- allows to monitor the cooperation of different neu- ed speech therapy programs with the patients for sev- ronal networks during cognitive tasks such as lan- eral weeks or months. (3) Clinical research: This part guage processing. of their occupation covers the linguistic aspects of (2) Magnetoencephalography. Like EEG, magne- aphasia research in hospitals and universities and toencephalography (MEG) is the second real noninva- results in the optimization of treatment efficacy and sive technique, and provides the same advantages of effectiveness of therapy programs. direct measurement of brain processes. In contrast to EEG, MEG measures the magnetic component of electrical activity. By using superconducting quantum Experimental Neurolinguistics interference devices (SQUIDs), which are contactless- In contrast to clinical neurolinguistics, this field ly positioned around the head in a so-called Dewar, mainly covers basic research on the neurobiology of MEG records the very low magnetic fields accompa- language of healthy probands. Experimental neu- nying the neural activity. Like the EEG technique, rolinguists try to understand the nature of representa- ERP- and coherence analysis of the MEG signal tion and physiological processes that contribute allows investigating language processes in the time to normal language by investigating underlying neu- and frequency domain. ral mechanisms in general. Their main goal is to (3) Positron emission tomography. Even though the study the neurophysiological phenomena in the brain positron emission tomography (PET) technique is not during language comprehension and production. based on invasion of the skull, it is a ‘small’-invasive Combining state-of-the-art neurophysiological non- technique, since radioisotopes are injected or inhaled. invasive techniques with linguistic, neurobiological, In clinical use since the late 1970s, PET scanners and neuropsychological findings, experimental neu- visualize differences in metabolic processes of the rolinguists investigate certain aspects of language brains during a stimulus task (e.g. listening to words) processing under laboratory conditions. Since this and during a control task (e.g. listening to noise). work requires high-tech equipment and a broad Short-lived radioactive isotopes are intravenously expertise in several disciplines, this kind of research injected and can be tracked from outside the head with involves typical teamwork, often represented by more a tomographic scintillation counter after their radioac- than one laboratory. tive decay. While passing the brain tissue, the emitted Common noninvasive techniques for the study of positron of the former isotope collides after 2 to 8 mm language processes in healthy participants are the fol- with a body’s electron, producing two gamma rays. lowing techniques, which represent the most frequent- By detecting and analyzing such coincident gamma ly used empirical methods in neurolinguistics. Some rays, the spatial position of the former isotope can be invasive techniques, only used during preoperative determined. The spatial distribution of the isotopes in diagnostics of patients, are described later. the brain allows an insight into the extent of the meta- (1) Electroencephalography. By using scalp disc- bolic processes, and thus the brain’s activity during a electrodes, which are attached on to the head skin by time window of several seconds. After the computa- an electrolyte conductive gel, human electroen- tional visualization of the data, false color pictures cephalography (EEG) is a real noninvasive technique, indicate the brain areas of higher or lower neural first described in 1929 by Hans Berger. This tech- activity during a certain language task compared with nique gained its current importance in neurolinguis- a control task. tics after powerful computers became available. (4) Magnetic resonance tomography (MRT) or Recording the electrical activities of the underlying magnetic resonance imaging (MRI) are synonymous- neural substrate (at least several 10,000 neurons), the ly used for a powerful technique, which allows three- EEG signal provides information on the time course dimensional high-resolution brain imaging. After the of certain cognitive processes. Since the EEG meas- head is exposed to the strong magnetic field of a cylin- ures the brain’s activity directly, it has a very high drical superconducting magnet, the spin axes of every time resolution below 1 millisecond and this allows hydrogen atom of the head brought into alignment. assessing ‘when’ a certain event is processed in the Then, short radio waves of the hydrogen resonance brain. Analyzing the brain waves elicited of a certain frequency (125 MHz) are sent to the tissue and by event (e.g. onset of a given word), the amplitude and absorbing the energy the spin axes of the atoms are time course of the wave can be interpreted (= event- disordered for a short moment. After this radio wave related potential, ERP). Analyzing the coherence stimulation, the hydrogen spin axes return to their 732  NEUROLINGUISTICS original, randomized state and emit energy in the form Simulative Neurolinguistics of a very weak radio wave. These remitted radio waves deliver the information on the head’s structure. After In simulative neurolinguistics, the knowledge of theo- being analyzed and visualized by computer, cross-sec- retical linguistics, clinical linguistics, neuropsycho- tional images of the brain with a resolution of less than logical case studies, and results of experimental 1 millimeter can be made. Outside of clinical use, MRI neurolinguistics are used as input for a computer sim- scanners of up to 11 Tesla are used. ulation with special software environments. By using This technique provides high-resolution informa- computational techniques of computerlinguistics and tion on the anatomy of the brain and can be combined neuroinformatics (e.g. connectionist neural networks), with electrophysiological findings of EEG or MEG. In previously observed real language processes can be order to monitor cognitive processes in terms of meta- used for a computer simulation. During simulation, bolic changes, functional MRI must also be per- each condition can be modified step by step, and the formed. Functional magnetic resonance imaging results can be obtained immediately. The time required (fMRI) maps changes in oxygen concentration that depends only on the computing power of the hardware correspond to nerve cell activity. By overlaying a map used. The advantage of this technique is the enormous of oxygen concentration changes on the anatomical flexibility of the simulated paradigm. Contrary to real- data, those brain areas that show increasing activity life processes, where investigators have to wait for correlated to a given stimulus can be defined. Even patients with certain impairments, virtual impairment though the time resolution is not excellent, the spatial can be created and tested within seconds. On the other resolution is in the range of several millimeters, which hand, the reliability of the predictions of such simula- makes fMRI one of the most powerful techniques in tions depends on the implemented processes, models, neurolinguistic research. and constraints. However, even though simulative (5) Invasive techniques. During preoperative diag- neurolinguistics seem to have an enormous potential nostics, e.g. prior to epilepsy surgery, electrode arrays for future research, it did not have a remarkable influ- are implanted and monitored for several days to iden- ence on neurolinguistics or cognitive neuroscience tify the focus of abnormal electrical activity. Such until now. intracranial electrodes, placed directly onto the cortex, Based on the large progress in brain imaging and allow the recording of underlying cortical activity dur- computerized analysis of brain data during the last few ing cognitive tasks, e.g. object naming, speaking, read- years, it can be expected that neurolinguistics will ing, etc. Furthermore, the underlying cortex area can strongly contribute to linguistic theory and modeling be electrically stimulated by short pulse trains of weak during the next decades. For a few years now, brain currents. If the stimulated cortex area contributes to processes during language can be investigated with a language, speech arrest or another aphasiological resolution in time of less than 1 millisecond and in symptom will occur. In other words, by electrical stim- space of less than 1 millimeter. Thus, a powerful new ulation, temporary lesions of the brain can be made in tool is available for linguists in the at least 2000-year- order to test the possible role of a given cortex region old attempt to understand human language. All meth- in the language process. This test is required prior to ods developed and used in linguistics (see above) form the performance of a neurosurgical epilepsy operation an emergent instrument for the very different topics, to assess where certain cerebral functions are centered questions, and approaches in studying language with- such as memory and speech. in the different fields of linguistics. For most of them, Another invasive technique applied for the same an integration of neurophysiological techniques would clinical reasons as mentioned above is the Wada test or be valuable. For example, neurolinguistic studies indi- Intracarotid Sodium Amytal Test (first described by cate how and when language is learned as first or sec- Juhn A. Wada). After the injection of an anesthetic ond language in a mono- or bilingual child, how (e.g. sodium amytal) into the right or left internal different languages are learned by a juvenile or ado- carotid artery, the respective hemisphere is anes- lescent, and how it is represented in the brain as well thetized for a few minutes. The type and extent of as how learning can be facilitated or disturbed. In cerebral dominance of a patient can be assessed during addition, the contemporary theories and models of lan- the following behavioral test of language comprehen- guage may be improved by neurolinguistic findings. sion, production, object naming, and memory abilities. The enormous impact of the behavioral data of psy- In a healthy, typical right-handed volunteer, suspen- cholinguistics on linguistic modeling since the 1950s sion of the left (language-dominant) hemisphere leads did show the necessity and usefulness of empirical to a Wernicke aphasia with a neologistic jargon, evaluation. Neurolinguistic research offers an even whereas suspension of the right hemisphere leads to a stronger and more direct observation of cognitive phe- dysprosody. nomena and language processes in the brain. 733 NEUROLINGUISTICS References Grodzinsky, Yosef, Lew Shapiro, and David Swinney (eds.) 2000. Language and the brain: representation and processing. San Benson, D. Frank, and Alfredo Ardila (eds.) 1996. Aphasia: a Diego: Academic Press. clinical perspective. Oxford: Oxford University Press. Kertesz, Andrew (ed.) 1997. Localization and neuroimaging in Brown, Colin M., and Peter Hagoort (eds.) 1999. The neu- neuropsychology. San Diego: Academic Press. rocognition of language. Oxford: Oxford University Press. Rugg, Michael D. (ed.) 1997. Cognitive neuroscience. Caplan, David. 1992. Language: structure, processing, and dis- Cambridge: MIT Press. orders. Cambridge: MIT Press. Stemmer, Brigitte, and Harry A. Whitaker (eds.) 1998. Eling, Paul (ed.) 1994. Reader in the history of aphasia: from Handbook of neurolinguistics. San Diego: Academic Gall to Geschwind. Amsterdam: John Benjamins. Press. Frackowiak, Richard S.J., Karl J. Friston, Christopher Frith, Zemlin, Willard R. 1998. Speech and hearing science: anatomy and Raymond Dolan (eds.) 1997. Human brain function. and physiology, 4th edition. Boston: Allyn & Bacon. San Diego: Academic Press. Friederici, Angela D. (ed.) 1999. Language comprehension: a HORST M. MÜLLER biological perspective. Berlin: Springer-Verlag. Gazzaniga, Michael S. (ed.) 2000. The new cognitive neuro- See also Clinical Linguistics; Localization of Lin- sciences. Cambridge: MIT Press. guistic Information New Guinea New Guinea, the second largest island in the world, New Guinea was originally peopled by many dif- lies roughly 100 miles north of Australia. It contains ferent waves of migrants, whose prehistory is largely more than 1,000 languages, about one sixth of the unknown. Although human habitation of the island world’s total (or one language in every 900 square extends back some 40,000 years, recorded history is kilometers), making it linguistically one of the most very recent and in some cases goes back only a few dense and complex regions of the world. The name decades. It was the last major land area in the world to was given to the island by the Spanish navigator Ortiz be colonized by European powers; almost all regions de Retes, possibly because he saw some resemblance have a history of contact of less than a century. The between the indigenous population and the inhabitants terrain is extremely rugged, with mountains reaching of the Guinea coast of Africa. Politically, the island is altitudes of more than 15,000 feet, dense tropical rain divided between two countries: Papua New Guinea, forests, and fast-flowing rivers, which have long cut occupying the eastern half of the island, and some 600 the interior of the country off from outsiders. Many associated islands, the largest of which are New villages have no road or river links with other centers, Britain, New Ireland, and Bougainville, and Irian Jaya and some can be reached only by walking for up to (West Irian), a former Dutch colony until 1963, now two weeks. the easternmost province of Indonesia. An independ- The peoples of New Guinea are mainly settled vil- ent nation since 1975, Papua New Guinea is histori- lagers living in a subsistence economy. Their produc- cally an amalgamation of what were two separate tive activities vary according to the zone they inhabit colonies: Kaiserwilhelmsland, occupying the north- on the island’s extraordinary vertical ecology. The eastern half of the mainland and islands under German extreme highlands have an alpine climate, with wide- rule from 1884 until 1914, and subsequently mandat- spread frost. There, an intensive agriculture based on ed by the United Nations to Australian administration the sweet potato has developed over the last few hun- under the name ‘New Guinea’; and Papua, the south- dred years. This highly productive system has given ern half of the mainland, formerly British New rise to a local population boom, and the highlands Guinea, which was proclaimed the Australian territory support large, dense groups with large languages. The of Papua in 1906. Papua New Guinea has the highest ten largest indigenous languages of Papua New population (5.2 million) and land mass of the Pacific Guinea belong to the large groups of the interior island nations, with 64% of its population and 84% of Highlands; they have from 30,000 to 100,000 speak- its land mass. Irian Jaya has a population between 1.8 ers. Between them, they account for nearly one third and 4 million and occupies a land mass approximately of the population. Perhaps 80% of the languages have the size of California. fewer than 5,000 speakers, and as many as one third 734  NEW GUINEA have fewer than 500. These languages are found main- with those of Tasmania (but not mainland Australia) ly in the coastal lowlands and the intermediate areas and the Andaman Islands in the Indian Ocean, has not known as the highland fringe, consisting of pockets of been generally accepted, nor has the suggestion of a rain forest, swamps, and grassland, where the popula- link between Australian languages and the languages tion is low and thinly spread. of the Eastern Highlands of Papua New Guinea. It is Linguists generally recognize two major language possible that Australia was settled from New Guinea groupings in New Guinea: Austronesian and non- because they were one continent until around 8,000 Austronesian (or Papuan). The Austronesian lan- years ago. Only after the last Ice Age did sea levels rise guages clearly constitute a family with a common to separate them. ancestor, Proto-Austronesian, and comprise some 450 New Guinea languages are typologically diverse, descendant languages in the Pacific basin and another displaying many interesting and unusual linguistic 600 to 700 outside it, making it the largest language features. There are languages with SVO (subject– family. About one quarter of New Guinea languages verb–object), SOV, VSO, VOS, and OSV word orders; are of Austronesian origin, and most of them (with the OVS is the only unattested word order, and this order exception of the languages of western Irian Jaya) is quite rare across the world. Most Papuan languages, belong to the Oceanic subgroup established by the however, tend to be verb final. In addition, there are German linguist Otto Dempwolff. The relationships examples of noun classifier systems, such as in the among the non-Austronesian (Papuan) languages are Papuan language Nasioi, with a set of more than 100 less clear, and the label is best seen as a cover term for suffixes added to nouns, adjectives, numerals, and a number, perhaps as many as 60, of distinct families. derived nominals to classify the entity being referred Figure 1 shows the distribution of the Austronesian to. Both Papuan and Oceanic languages, however, tend and Papuan languages. Most linguists agree that the to have relatively simple sound systems, with sound coastal distribution of most of the Austronesian lan- inventories smaller than that of English. Consonant guages indicates the later arrival of their speakers. The clusters are absent or rare in Austronesian, but they are immediate ancestors of the Proto-Oceanic speakers present in many Papuan languages. migrated from eastern Indonesia through eastern Irian Almost all Oceanic languages distinguish between Jaya into the Bismarck Archipelago. The interior of inclusive (referring to the speaker and addressee or New Guinea, where the majority of Papua languages addressees, ‘I + you’) and exclusive first-person pro- are spoken, experienced no European contact until nouns (referring to the speaker and some other person shortly before (and even in some cases some time or persons, ‘I + he/she/it/they’), as well as a three-way after) World War II, and hence most of the languages distinction in number between singular (one), dual were unknown to the outside world until quite recent- (only two), and plural, or paucal (more than two). ly. A number are classified as isolates, i.e. languages Some languages have a trial, referring to only three. that seem to have no known relatives. The Indo-Pacific Very few Oceanic languages mark gender in pro- hypothesis, which attempts to link Papuan languages nouns, and most have a three-way distinction in their NEW IRIAN JAYA BRITAIN PAPUA NEW GUINEA Non-Austronesian Austronesian Figure 1. Map of New Guinea show- Unclassified ing New Britain and the distribution 0 20 40 60kms of Austronesian and non-Austronesian languages. 735 NEW GUINEA demonstratives between proximate (‘this’ near the Much less research has been done on the approxi- speaker), intermediate (‘that’ near the addressee), and mately 250 languages in Irian Jaya, only four of which distant (‘that’ away from both speaker and addressee). are spoken by 40,000 people or more, and somewhat Most Oceanic languages also distinguish two types of more on the roughly 860 languages in Papua New possessive marking, with all nouns being classified Guinea. In many cases, missionaries undertook the into alienable and inalienable. Alienable nouns refer first studies of languages in New Guinea, and mis- to objects that belong to the speaker, where the speak- sionary linguist work continues today, particularly in er typically has control over the possession and may Papua New Guinea, through the Summer Institute of reject it (e.g. food, clothes, etc.). Inalienable nouns Linguistics. refer to items inseparably connected to the speaker, Contacts between speakers of Papuan and such as relatives or body parts. Inalienable nouns are Austronesian languages have led to rapid change and marked with suffixes that vary in a minor way from diversification to the point where languages such as language to language. Magori and Maisin in Papua New Guinea are difficult Only a small number of Oceanic languages have a to classify as Austronesian or Papuan. The Motu lan- contrast between active and passive. Grammatical dis- guage spoken along the southern coast of Papua New tinctions, such as tense, are mostly expressed by inde- Guinea has the grammatical particle system typical of pendent particles rather than affixes: Austronesian languages, but it has the typical SOV Tuna yau ga gire word order of Papuan languages. ‘I’ PAST see tau ese au-na imea bogarai-na-I vada e hado ‘I saw it’. man a tree-the garden middle-its-at PERFECT he/she/it plant Papuan languages, however, often have a rich and ‘The man planted a tree in the middle of the garden.’ strongly developed affixal word structure: The small size of many Melanesian societies has Yimas ama-wa-t also permitted change to spread more rapidly than in ‘I’ go PERFECT larger societies. Significant changes have affected the ‘I went’ basic vocabulary and grammar of the Austronesian Pronoun systems vary widely in Papuan languages language Muyuw, spoken on Woodlark Island in Milne but are generally not as complex as in the Oceanic lan- Bay Province of Papua New Guinea, in just a 50-year guages. Very few Papuan languages distinguish inclu- period. Local vernaculars are seen as a unique badge sive and exclusive first-person pronouns, although a of identity and distinctiveness. Villagers in one com- number of them distinguish gender in pronouns. munity decided at a meeting that they would be differ- Articles are virtually nonexistent, and possessive con- ent from other Selepet-speaking villages by adopting a structions are less complex than in Oceanic languages. new word (bunge) for ‘no’ to replace their usual word A number of languages have more than one existential (bia) shared by all Selepet speakers. (‘be’) verb, such as Anggor in the Sepik region of Multilingualism is widespread, and because people Papua New Guinea, which has 18. The different verb marry outside of their community, husbands and wives forms depend on the shape of the object, its location, often speak different languages. Many people, espe- and posture. In Enga, ‘there are men’ is expressed as cially men, know the languages of one or two neigh- ‘the man stands’, whereas ‘there are women’ is boring communities, or perhaps a language with wider expressed as ‘the woman sits’. currency around their valley or coastline. Where lan- Switch reference, although generally rare in the guage groups were large, as in the highlands, only Pacific, is typical of complex sentences in Papuan lan- those in the border areas would be multilingual. Where guages. When two or more clauses join to form a com- groups were small, everyone was effectively in a bor- plex sentence, the last verb in the clause retains the der area, and knowledge of multiple languages was subject-tense marking of the first verb, but the other universal. In the lowland village of Gapun, whose lan- (medial) verbs do not. Instead, they incorporate a suf- guage Taiap is an isolate spoken by about 80 people, fix indicating whether the subject of the verb is the the average number of languages understood by men same as or different from the subject of the following over 40 was five: the vernacular, a lingua franca, and verb. three or so of the other local languages. Nasioi kad-o-ma nan-ant-in In addition to the indigenous languages, there are a talk-‘I’-SAME ACTOR go-‘I’-IMMEDIATE PAST number of pidgins and creoles, as well as languages of ‘While talking, I went.’ the metropolitan powers, particularly English (and for- da? po-ko nan-amp-e-ain merly also German) in Papua New Guinea and Bahasa you come-DIFFERENT ACTOR go-‘we’-FUTURE Indonesia (and formerly also Dutch) in Irian Jaya. ‘When you come, we two will go.’ Among the pidgins and creoles are those based on 736  NIDA, EUGENE ALBERT indigenous languages such as Pidgin Yimas, based on Foley, William. 1986. The Papuan languages of New Guinea. the Papuan language Yimas and spoken in Papua New Cambridge: Cambridge University Press. Kulick, Don. 1992. Language shift and cultural reproduction. Guinea’s Sepik region, and Hiri Motu (‘trade Motu’), Socialization, self and syncretism in a Papua New Guinean based on the Austronesian language Motu, spoken by village. Cambridge: Cambridge University Press. a quarter of a million people and one of Papua New Lynch, John. 1998. Pacific languages. Honolulu: University of Guinea’s three national languages, along with English Hawaii Press. and Tok Pisin, an English-based pidgin and creole. Sankoff, Gillian. 1980. Multilingualism in Papua New Guinea. The social life of language. Philadelphia: University of Papua New Guinea’s most widespread language, Tok Pennsylvania Press. Pisin, with over 2 million speakers, is the largest pid- Thurston, William. 1987. Processes of change in the languages gin/creole language in the Pacific. In many parts of of North-Western New Britain. Canberra: Pacific Linguistics Papua New Guinea, children grow up speaking Tok Series B, No. 99. Pisin and no longer acquire their local village lan- Wurm, Stephen (ed.) 1975. New Guinea area languages and language study, Vol I. Papuan languages and the New guage. In Gapun, parents began speaking mainly Tok Guinea linguistic scene. Canberra: Pacific Linguistics Series Pisin to their children, and now children over the age C, No. 38. of ten no longer use Taiap. The vernacular languages Wurm, Stephen (ed.) 1976. New Guinea area languages and have very little place in the national life of either language study, Vol II. Austronesian languages. Canberra: Papua New Guinea or Irian Jaya. In the former, Pacific Linguistics Series C, No. 39. Wurm, Stephen (ed.) 1977. New Guinea area languages and lan- English is the main language of government and edu- guage study, Vol III. Language, culture, society and the mod- cation, and in the latter, Bahasa Indonesia. ern world. Canberra: Pacific Linguistics Series C, No. 40. Wurm, Stephen (ed.) 1980. New Guinea and neighbouring areas: a sociolinguistic laboratory. The Hague: Mouton. References Wurm, Stephen (ed.) 1982. The Papuan languages of Oceania. Capell, Arthur. 1969. A survey of New Guinea languages. Tübingen: Gunther Narr Verlag. Sydney: Sydney University Press. SUZANNE ROMAINE Nida, Eugene Albert Eugene Nida’s research in linguistics and cultural ing formal with referential classes. In his works, he anthropology spans a period of 60 years, during which has amply demonstrated that, despite formal similari- he has continuously developed and refined his percep- ties, words may develop different semantic relations tion of languages and cultures. His initial studies of between themselves and the use of strictly formal cri- classical Greek and the New Testament as well as his teria of analysis can be misleading. Moreover, mean- early books on morphology and the structure of ing cannot be located in words, which have ‘fuzzy English were ideal qualifications for dealing with boundaries’, but rather on a ‘molecular’ level, in their Bible translations, a task that he was asked to fulfill by combinations with other words. Context thus becomes the American Bible Society. A large part of Nida’s a fundamental factor in Nida’s analysis of meaning, works and efforts were indeed devoted to showing and he follows Martin Joos (1972) in maximizing its translators how to better understand the Bible in order role in lexical combinations. to make it understandable, in their turn, to receptors From word and syntax levels, Nida’s research has speaking highly different languages and belonging to encompassed, in an increasingly detailed manner, the most diverse cultures. This activity brought him into structures of discourse as well as textual organization, contact with over 200 languages and cultures in Asia, texts being regarded as the basic and ultimate carriers Australia, Africa, and the Americas. His views on the of meaning. Still, his investigation of contexts goes way in which languages function are thus based on beyond strictly linguistic scrutiny. Like Edward Sapir extensive field surveys. and Bronislav Malinovski, Nida has constantly In his description of the word and sentence struc- emphasized the crucial role of cultures in describing ture of languages, Nida has resorted to the concepts of and explaining languages. Moreover, his adoption of a American structural linguistics, placing, however, the sociosemiotic stance has allowed him to explore the meaning at the center of his investigations and replac- social aspects of language as well as the ways in which 737 NIDA, EUGENE ALBERT language signs relate to other sign systems and to the communication is always possible, although there is real world of referents. no such thing as absolute communication. Eugene Nida has made a fundamental contribution In more recent publications in linguistics and trans- to the development of translation theory and practice, lation theory, Nida undertakes detailed analyses at all and his impact on Bible translators, translation schol- language levels, broadening the area of interlingual ars, and translation schools has been outstanding. He investigations. The author includes a high number of has brought rigor and scientific objectivity to the sub- case studies of Bible translations, of scientific and ject by incorporating concepts, methods, and classifi- technical texts, of various literary genres, and of cations from linguistics, pragmatics, semantics, and European Union texts, and a multitude of experience- discourse analysis as well as necessary data from cul- based examples from different languages and cultures tural anthropology, linking languages to cultures and illustrating his translation principles. broader contexts of communication. The concept of Debates around Nida’s works were mainly generat- dynamic/functional equivalence that he has introduced ed by ideological and religious speculations on his in the translational discourse has shifted the emphasis receptor-oriented position in translation as well as by in translation theory from the faithful reproduction of the principle of ‘equivalence of effect’, on which source messages to the creation of translated texts with dynamic/functional equivalence is based. Although the a strong communicative impact, focusing on the American scholar has suggested a series of tests for receivers’ needs for clarity as well as on their linguis- checking and comparing the source to target text tic and cultural expectations. This receptor orientation receivers’ understanding and response, “effect” has gives priority to the content of texts and may entail, in been regarded as too vague a notion to serve as a cri- some cases, a more radical formal restructuring of the terion for translation evaluation. source text without, however, altering its meaning. Eugene Nida’s approach to linguistics and trans- In The theory and practice of translation (co- lation is avowedly eclectic. This encyclopedic per- authored with Charles Taber, 1969) and in subsequent spective enables him to cover all major aspects of books, Nida provides a model for describing the languages and their translation. He combines translating process and analyzing meaning in more insights from structural linguistics, pragmatics, soci- detail. In his view, the translating process involves olinguistics, discourse analysis, functional grammar, analysis, transfer, and restructuring. Analysis consists transformational-generative grammar, semiotics, in a reduction of surface structures to kernels (i.e. psycholinguistics, rhetoric, stylistics, information substructures), making use of the functional classes of theory, etc. Nida openly declares his mistrust of the objects, events, abstracts, and relations. At the kernel holistic systems that ‘can stifle creativity and lock level, languages are found to ‘agree far more than on minds shut to new evidence’ (2003:140). His books the level of more elaborate structures’ (1969:39), and have aroused considerable interest among linguists, it is at this level that the transfer into the receptor lan- theologians, and translation scholars, and they have guage occurs. The transformation into a new surface certainly been of great help to all those dealing with structure through restructuring takes place according intralingual, interlingual, and intercultural commu- to the deemed expectations of the receivers. The nication, be it religious or secular. model draws on principles used in transformational- generative grammar, although the influence of this Biography linguistic direction on Nida’s work has often been overemphasized. For one thing, Nida reverses Noam Eugene Albert Nida was born in Oklahoma City, Chomsky’s model by starting from surface structures Oklahoma on November 11, 1914. He received his B.A. and moving to kernels and not the other way round. in Greek and Linguistics from the University of Secondly, Nida is interested in language-in-use California at Los Angeles an 1936, and an M.A. in (Ferdinand de Saussure’s parole or Chomsky’s per- Patristics from the University of Southern California in formance), and not in the abstract level of Chomsky’s 1939. He taught at the Summer Institute of Linguistics deep structures. Thirdly, Nida rejects the idea that lan- in 1937–1952. He did his Ph.D. on English Syntactic guages are strictly rule-governed, as this would leave Structures under the supervision of Charles C. Fries, no space for their creative use. It is nevertheless true Professor of English and Linguistics at the University that transformational-generative grammar offered him of Michigan, in 1943. He was consultant for the a theoretical perspective for explaining processes of American Bible Society and the United Bible Societies, decoding and encoding texts. From a similar stance, 1943–1981; President of the Linguistic Society of he could also account for the similarities between lan- America, 1968: and Translations Research Coordinator guages on a conceptual and even on a formal level, in for the United Bible Societies, 1970–1980. He has keeping with his conviction that effective interlingual studied languages and cultures, counseling on Bible 738  NIGER-CONGO translating in more than 90 countries, and has lectured Munday, Jeremy. 2001. Introducing translation studies. in more than 100 universities, also actively participating London, New York: Routledge. Nida, Eugene Albert. 1946. Morphology. The descriptive analy- in other scientific and academic reunions worldwide. sis of words. Ann Arbour, MI: University of Michigan Press. He also received five honorary Ph.D.s. He is author and ––––––. 1964. Toward a science of translating. Leiden: E. J. co-author of more than 40 books and 250 articles. Nida Brill. has lived in Brussels since 1995, consulting with Bible ––––––. (co-authored with Charles Taber). 1969. The theory societies and translators from the European Union. and practice of translation. Leiden: E. J. Brill. ———. 1975. Exploring semantic structures. Munich, Germany: Wilhelm Fink. ––––––. (co-authored with Johannes P. Louw). 1992. Lexical References semantics of the Greek New Testament. Atlanta: Scholars Black, Matthew, and William Smalley (eds.) 1974. On lan- Press. guage, culture and religion: in honor of Eugene A. Nida. The ––––––. 1996. The sociolinguistics of interlingual communica- Hague: Mouton. tion. Bruxelles, Belgium: Les Editions du Hazard. Fawcett, Peter. 1997. Translation and language. Manchester: St. ———. 2001. Contexts in translating. Amsterdam and Jerome. Philadelphia, PA: Benjamins. Joos, Martin. 1972. Semantic axiom number one. Language 48. ———. 2003. Fascinated by languages. Amsterdam and 257–65. Philadelphia, PA: Benjamins. Larose, Robert.1989. Théories contemporaines de la traduction. Schmitz, John Robert. 1999. The contribution of Eugene Nida to Québec: Presses universitaires de Québec. the theory and practice of translation. Translation-transition. Longacre, R.E. 1994. Eugene Nida. The encyclopedia of lan- 15th World Congress of FIT, Mons, Vol. 1, ed. by Jean-Marie guage and linguistics, Vol. 5, ed. by R.E. Ascher and J.M.Y. Vande Walle. Mons: Elma Edities. Simpson. Oxford: Pergamon Press. RODICA DIMITRIU Niger-Congo The Niger-Congo language family, previously labeled Adamawa-Ubangi Nigritic, or western Nigritic, is one of the four main language families in Africa. The other language fam- Adamawa-Ubanji, previously called Adamawa- ilies are Khoisan, Nilo-Saharan, and Afro-Asiatic. Eastern, is spoken in parts of Central African Niger-Congo languages are spoken in the basins of Republic, eastern Nigeria, northern Cameroon, and the Niger and Congo rivers and in the Nuba southwestern Chad. It has languages like Gbaya, Mountains in the Sudan. Joseph Greenberg’s (1955) Banda, and Zande. The Adamawa-Ubangi family has original classification viewed the Niger-Congo fami- two daughter languages: Adamawa and Ubangi. ly as a sister language of the Kordofanian family. He Adamawa has languages like Leko, Duru, Jen, named the ancestor language from which the two Nimbari, Mbum, Bua, Kim, Day, Waja, Daka, and emerged as Niger-Kordofanian. However, Kay Fali. Ubangi languages are spoken in the area stretch- Williamson (1989a,b) notes that the Kordofanian ing from northern Cameroon across the Central Africa family falls within the Niger-Congo family; hence, Republic to parts of southern Sudan and northern she uses the label Niger-Congo in place of Niger- Democratic Republic of Congo (formerly Zaire). Kordofanian for the whole family. The Niger-Congo Ubangi languages include Banda, Gbaya, Ngbaka group of languages is spoken by well over 80% of the (Sere, Mba), Sango, and Zande. Sango is a lingua population of Africa. Geographically, they are spoken franca of the Central African Republic. in areas stretching from Senegal (West Africa) to While Adamawa allows words to end in conso- Kenya (East Africa). They also stretch from Sudan nants, none of the Ubanji languages, with the excep- (North Africa) to the south in the Republic of South tion of Gbaya, allows final consonants. Whereas other Africa. Niger-Congo has ten daughter languages, language families may have nasal sounds, Adamawa- namely Adamawa-Ubangi (formerly Adamawa Ubanji has nasal morphemes (root words). Word order Eastern), Atlantic (formerly West Atlantic), Benue- is Subject–Verb–Object, and verb reduplication (dou- Congo, Dogon, Gur, Ijoid, Kordofanian, Kru, Kwa, bling of verbs in certain grammatical constructions) and Mande. is common to all the languages in the subfamily. The 739 NIGER-CONGO languages have between two and four tonal phonemes, it may also be classified as an independent member of i.e. pitch patterns that may alone distinguish meaning. the Volta Congo group. Dogon has six major dialects: Donnɔ Sɔɔ, Tombɔ Sɔɔ, Torɔ Sɔɔ, Jamsay, Togo Kan, Atlantic and TomO Kan. Important linguistic features include independent nasalized vowels (phonemic nasalization Atlantic languages, also called West Atlantic languages, of vowels), vowel harmony (co-occurrence restrictions are spoken in West Africa in the area stretching from the in the distribution of vowels), a two-tone system, and Senegal river down into Liberia. Most languages in this Subject–Object–Verb word order. In Dogon noun family are spoken in Senegal, The Gambia, Guinea, phrases, the noun always comes first. Guinea-Bissau, Liberia, and Sierra Leone. Atlantic has about 45 languages and over 30 million speakers. Gur Languages in this family with large numbers of speak- ers include Fula, Wolof, Diola, Serer, Manjaku, Balanta, Geographically, the Gur group of languages is distrib- Basari, Limba, Kisi, Sua, and Temne. Important lin- uted over a wide area stretching from the Ivory Coast, guistic features include consonant mutation/alternation through Ghana, Togo, Benin to Burkina Faso. Gur lan- (specific consonants may change into another in partic- guages are also spoken on the fringes of Niger, Benin, ular sound patterns), noun class, and concord systems in and Mali. The Gur phylum subdivides into Oti-Volta, which the choice of specific parts of speech (especially Bwamu, Kurumfe, Grusi, Kirma, Dyan, Gan, Dogoso, nouns) requires other elements (verbs, adjectives, among others. There are nearly a hundred languages in adverbs, etc.) of the sentence to occur with particular the Gur family. Some of the languages in this group affixes, depending on the grammatical construction. include Moore (spoken in Burkina Faso by nearly seven million people), Grusi, Gurenne, Wali (Dagaari), (New) Benue-Congo Dagbani, Buli, Kabre, Kotokoli, Kasem, Konni, Tayiri, Kusaal, Bassari, Ntrobo, Sisaala, Waali, Mampruli, and The Benue-Congo language family is the largest sub- Nafaanra. Over twenty million people speak the Gur family of the Niger-Congo group. It is spoken in the languages. Some of the phonological features found in areas in and around the Benue and Congo river basins. Gur languages are tone, vowel harmony, uncommon It is made up of the former Eastern Kwa––Yoruba, consonants such as implosives, coarticulated sounds Edo, Nupe, Idoma, Igbo––and Bantoid (subclassified (e.g. /kp/ /gb/), and a syllable-timed rhythm. into non-Bantu and Wide Bantu). The New Benue- Congo language family is subclassified into eight sub- Ijoid families: Defoid (Yoruba, Akoko), Edoid (Edo), Nupoid (Nupe, Ebira, Asu, and Gbagyi), Idomoid Ijo, the smallest branch of the Niger-Congo language (Idoma), Igboid (Igbo), Kainji and Platoid (Kainji, family, is spoken by the Ijo of Nigeria and covers both Eloyi, Kagoma, Jukun), Cross River (Obolo, Ogoni, Ijo and Defaka (Afakani). It is spoken in the Niger Legbo, Ogbia), and Bantoid (Tiv, Swahili, Kikuyu, River delta region and adjacent riverine areas within Kirundi, Kinyarwanda, Shona, Zulu, and Xhosa). The the Rivers, Bendel, and Ondo states of Nigeria Bantoid group is by far the largest group in the Benue (Jenewari 1989). The Ijo language family is made up of Congo subfamily. Important linguistic characteristics seven languages: Biseni (Amegi), Okodia, Oruma, of this large subfamily include cross-height harmony Nkoroo, Eastern Ijo (which comprises Kalabari, (where either only tense or only lax vowels occur in Okrika, and Ibani), Brass Ijo (Nember-Akassa), and words) and vowel coalescence (where two or more Izon (Bumo, Kolokuma, Mein, and Arogbo). Speakers vowels merge to become one). Syllable structure of Ijo are a little over one million. The Ijo were among ranges from CV (consonant–vowel) to CCV and CVC. the first West Africans to have contact with Europeans, Consonant types include such uncommon sounds as and the Kalabari, an Ijo language, is believed to be one breathy voiced plosives /bh dh/ and labio velar sounds of the first Nigerian languages to be written. An impor- /kp gb/. Other important features include noun class tant unique structural feature is a consonant harmony and concord systems and an SVO word order. Some of also called ‘implosive harmony’ (which requires that the Bantu languages have a fixed stress any given word either contains implosives or plosives, pattern––words in Swahili, for example, are always but never both), and a noun class system drawing stressed on the penultimate (last but one) syllable. grammatical distinctions based on animateness and biological gender. The basic word order is Subject– Object–Verb, although different word orders are possi- Dogon ble (Object–Subject–Verb and Subject– Verb–Object) Dogon, which is spoken on the mountains of Mali, if object or subject need to be marked as the topic of was classified as a Gur language until 1989. However, conversation. 740  NIGER-CONGO Kordofanian (and other Gbe languages), Ga, Dangme, Gwa, Avikam, Anyi, Baule, Chakosi, Nzema, Santrokofi, The Kordofanian language family is located in the Likpe, Adele, Logba, and Kposo. Important linguistic Nuba Mountains of Sudan. Its status in the Niger- features associated with some of the New Kwa lan- Congo family has been widely debated. As mentioned guages are the occurrence of voiceless and voiced labi- earlier, it was first viewed as a sister family of Niger- al and velar fricatives, double articulated sounds Congo in a larger grouping called Niger-Kordofanian, (single sounds with two different places of articula- but later it came to be seen as a sub-branch of Niger- tions e.g. /kp, tp, gb/ ), vowel harmony, and tone ter- Congo. Kordofanian has over half a million speakers racing, where pitch is lowered toward the word end. and consists of about 20 languages. Moro, Mudo, Talla, Miri, Krongo, Talasa (Tumtum), Tiro, Utoro, Rere, Ngile, Tocho, Goy, Gom (Tegali), and Kalak are some Mande of the languages of this subfamily. Kordofanian lan- The Mande group of languages is spoken in West guages have dental /d, D, t/ (tongue touches teeth) and African countries: Senegal, The Gambia, Mali, the Ivory retroflex consonants /t ¢, d¢ / (tongue curls back). Plosives Coast, Guinea, Sierra Leone, Liberia, Guinea Bissau, like /t/ occurring between vowels change to fricatives Burkina Faso, Mauritania, Benin, Niger, Nigeria, and such as /s/, and the noun class system is reminiscent of Ghana. Bambara, Dyula, Susu, Mende, Kpelle, Vai, those found in other Niger-Congo. Lorma, Loko, Soninke, Kweni, Dan, Maninka, Kpelle, Busa, Bisa, Ligbi, Togo and Bobo are Mande languages. Kru Over 40 languages are found in this family. The Mande The Kru language family is spoken in southwestern languages are spoken by a little over 20 million people. Ivory Coast and southern Liberia. There are 24 lan- Phonologically, many Mande languages have between guages in this subfamily and the total number of speak- seven and nine vowel phonemes, labiovelar stops, con- ers ranges between three and four million. Some of the sonant mutation, two tones, and tone sandhi (adjacent Kru languages include Grebo, Klao, Dida, Godie, Bete, tones influence one another). Unlike most African lan- Nyabwa, Konobo, Bassa, Gbii, Bakwe, Kuwaa, Aizi, guages, Mande languages do not have serial verb con- Wobe, Dewoin, SEmE, Guere (Krahn, WEE), Tepo, structions. Word order is Subject–Object–Verb. In noun Chedepo, and Neyo. Most Kru languages have eight phrase constructions, definite determiners, articles, and vowel phonemes and a vowel harmony where only plurals tend to follow the noun. However, possessive vowels from a particular set may occur in any given pronouns precede nouns. Some of the languages such as morpheme. Kru syllables tend to end in a vowel. Central Vai, Mende, Loma, and Kpelle in this group have com- vowels such as /i, u, a,/ are found in some of the Kru binations of Arabic script, Latin writing systems, and languages. Uncommon consonant sounds such as unique African writing systems. implosives, double articulated sounds, and velar frica- tives are found in the Kru languages. Kru languages References have subject–verb–object word order with indirect Bendor-Samuel, John. 1989. Niger-Congo. The Niger-Congo objects preceding direct objects. When there is an aux- languages. A classification and description of Africa’s iliary, then both the direct and indirect objects precede largest language family (N-CL). New York: University Press the verb. Sentences may be negated with the help of an of America. auxiliary or a particle as well as changes in tone pattern. Bendor-Samuel, Olsen, and White. 1989. Dogon. In John Bendor-Samuel, pp. 169–77. Dwyer, David. 1989. Mande. In John Bendor-Samuel, pp. 47–66. Kwa Greenberg, Joseph. 1955. Studies in African linguistic classifi- cation. New Haven: Compass Publishing Company. Kwa has undergone tremendous reclassification with Heine, Bernd, and Derek Nurse, 2000. Niger-Congo, N-CL. several languages and language groups such as Ijo, New York: University Press of America. Kru, Yoruba, and others moved to other major lan- Jenewari, Charles. 1989. Ijoid. In John Bendor-Samuel, pp. guage families. The New Kwa, as it is now called, is 105–18. Köhler, Oswin. 1975. Geschichte und Probleme der Gliederung spoken in Côte d’Ivoire, Ghana, Togo, Benin, and der Sprachen Afrikas. Die Völker Afrikas und ihre tradi- Nigeria by about 30 million people. It subdivides into tionellen Kulturen, ed.by H. Baumann, 137-373. two main subfamilies, namely Nyo and Left bank. Wiesbaden: Franz Steiner. Important subfamilies under Nyo include Avikam- Marchese, Lynell. 1989. Kru. In John Bendor-Samuel, pp. Alladian, Agnéby, Potou-Tano, and Ga-Dangme. Gbe, 119–40. Mukarovsky. 1976. A study of Western Nigritic. Vienna: Institut Avatime-Nyangbo, Kposo, and Keby-Animere are für Aegyptologie und Afrikanistik. identified as members of the Left Bank subfamily. Naden, Tony. 1989. Gur. In John Bendor-Samuel, pp. 141–68. Important New Kwa languages include Akan, Ewe Samarin, William. 1971. Adamawa-Eastern. Samarin, William. 741 NIGER-CONGO 1971. Adamawa-Eastern. Current trends in linguistics, Vol. Stewart, John. 1989. Kwa. In John Bendor-Samuel, pp. 217–46. VII: Linguistics in Sub-Saharan Africa, ed. by T. A. Sebeok, Williamson, Kay. 1989a. Niger-Congo overview. In John 213–22. The Hague, Netherlands: Mouton. Bendor-Samuel, pp. 3–46. Sapir, John. 1971. West Atlantic: an inventory of the languages, Williamson, Kay. 1989b. Benue-Congo. In John Bendor- their noun class systems and consonantal alternations. CTL 7. Samuel, pp. 247–74. 45–112. Wilson, W. A. A. 1989. Atlantic. In John Bendor-Samuel, pp. Schadeberg, Thilo. 1989. Kordofanian. In John Bendor-Samuel, 81–104. pp. 67–80. SAMUEL GYASI OBENG Nigeria The Federal Republic of Nigeria is the most populous guages, Achebe utilizes Nigerian English, which country in Africa. This ‘giant of Africa’ gained inde- echoes the oral tradition the text celebrates. The text is pendence from the British on October 1, 1960. Nigeria rich in proverbs, metaphor, simile, and folklore, as is located in West Africa. In the south, it is washed by exemplified by the following excerpts: the Atlantic Ocean. Its four neighbors are Nigér Republic in the north, Chad in the northeast, a. ‘Okoye said the next half a dozen sentences in proverbs. Among the Ibo, the art of conversation is Cameroon in the east, and Benin Republic in the west. regarded very highly, and proverbs are the palmoil with Presently, it has a presidential form of government. which words are eaten’ (7). The centrally located city of Abuja is Nigeria’s new b. ‘A clan was like a lizard; if it lost its tail, it soon grew capital. The Niger river and its tributary, the Benue, another’ (171) divide Nigeria into three parts: the northern region or Hausaland, the southwestern region or Yorubaland, Stories of Igboland are recounted and significant and the southeastern region or I(g)boland. The names beliefs are celebrated, such as the Week of Peace, the of these regions mirror the linguistic basis of their New Yam Festival, and the precolonial conception of nomenclature; Hausa is the primary language of time in relation to market weeks and the position of Hausaland, Yoruba is the main language of the moon. Yorubaland, and Igbo (or Ibo) is the primary language In Nigeria, English serves multifarious func- of Igboland. The importance of River Niger is evident tions––in administration, education, commerce, and in the name of the country, which is exemplary of a more. The national anthem and the national pledge, creative linguistic process, which combines the words the two most widely recited verses of patriotism, are ‘Niger’ and ‘area’. ‘Nig- area’, spelled NIG-ER-IA, both in English, and illustrate how Nigeria is concep- literally refers to the area around the River Niger. tualized as both a fatherland and a motherland. As a Nigeria has the largest number of languages found result of the dynamics of English usage in different in any African nation, accounting for over one quarter strata of Nigerian society, at least three varieties of of Africa’s languages. Nigeria’s linguistic landscape is English are discernable: the acrolect, the mesolect, and as variable as its vegetation, which ranges from the basilect (see Bamiro 1991; Pandey 1997). Many swampland on the southern coast, to lush tropical rain- Nigerians now use Americanisms. forests, and a continuum of savannah land––guinea Multilingualism, with its accompanying problems savannah, through sudan savannah, sahel savannah, as regards the selection of a national language, has and finally desert, studded with the mighty baobab. ‘forced’ English––a neutral code––to function, in The flora and fauna are distinctive to the region; many many ways, as the quasinational language (Emenyonu plants and trees are not known to the western world; 1989:83). English is now a permanent resident in hence, they have become part of the lexicon of Nigeria, a primary Nigerian language in all senses of Nigerian English. Prime examples are kola (nut), cam the term. In Achebe’s words, ‘It has been bent and wood, the udala tree, bitter leaf (from which bitter leaf twisted to bear the burden of the African experience’ soup is made), and the acacia tree. In Things fall (1975). It is a Nigerian brand of English (Ubahakwe apart, a classic bestseller that has sold over two mil- 1979; Akere 1982; Bamgbose 1982; Odumuh 1987), lion copies and has been translated into over 50 lan- with a distinctive local flavor. As Akere (1982) puts it, 742  NIGERIA What has happened here in Nigeria… is that the come as no surprise that NPE is the most widely spo- resources of a second language are superimposed on an ken pidgin in the world, with over 40 million estimat- intricate system of social and kinship relationships, and ed speakers, not all of whom are Nigerians (Faraclas on a completely different pattern of cultural outlook and 1996:1). Some speakers are from neighboring social expectations. …The English forms of address and Cameroon and Ghana––both of which witnessed a greetings have become modified to suit local commu- ‘brain drain’ (the Nigerian English term for intellectu- nicative needs. (97) al loss) or an exodus of immigrants to Nigeria––during The educated Nigerian generally has access to the Nigeria’s oil boom. For the most part, these and other largest language repertoire and even initiates switches expatriates have acquired NPE. In the absence of a to the stigmatized Nigerian Pidgin (English) in con- standard orthography, NPE is written the way it versations with peers and lesser-educated Nigerians. sounds (to the speaker). Faraclas (1996) observes that This is a marked difference between the ‘Outer Circle’ the name NPE is actually a misnomer, as it has devel- (see Kachru 1982) to which Nigeria belongs on oped trilectal varieties: a pidginized basilectal variant account of the variety of English it has grown, and the often referred to as ‘Pidgin Propa’ (i.e. ‘Pidgin Proper’ ‘Inner circle’, where it is rather unusual to find col- or PP) and frequently used by less educated or unedu- lege-educated individuals being the most multilingual cated Nigerians; a creolized ‘mesolect’; and a decre- and multidialectal citizens who readily and willingly olized variant that is often used by educated Nigerians. switch to nonstandard varieties in exchanges with less- Eze (1980) refers to this variety as ‘Hyperanglised er-educated citizens (see Pandey 1999). Not surpris- Pidgin’ (HPP) and notes that it differs extensively from ingly, code-mixing, code-switching, and speech ‘Pidgin Propa’ in its more-Standard-English-oriented accommodation are frequently used conversational use of prepositions, as opposed to the use of ‘for’ to strategies, even by Nigerian creative writers whose represent all prepositional functions in ‘Pidgin Propa’. language use mirrors the conversational realities of There are, of course, other differences, such as the Nigerian society (see Pandey 1997). more economical lexicon of ‘Pidgin Proper’. Occupying approximately three fifths of Nigeria’s Examples include: land area, the Niger-Benue river basin is a major source of sustenance. Its waters have also nourished (1) a. Im mother been come meet am in the office and transported a variety of water-borne languages, (HPP) namely, Pidgins and Creoles, and specifically Nigerian b. Im moda been come see am for (im) office Pidgin English (NPE hereafter), the nation’s linguistic (PP) lifeline or primary lingua franca. NPE still bears lin- [Gloss: His/Her mother came to meet or see him/her in guistic evidence of the earliest European contact in the office OR His/Her mother DID (indeed) come and Nigeria. The Portuguese are known to have arrived on meet him/her at the/his/her office] the coast of West Africa in the fifteenth century. Exemplary Portuguese words that constitute part an (2) a. ’The thing pained me bad bad because I parcel of the core vocabulary of this linguistic medley wanted to be big man like lawyer or doctor rid- (NPE) include sabi (comprehend, understand), pickin ing car and talking big big English’ (HPP, (child or children), palava (trouble), and dash (a gift excerpted from Saro-Wiwa’s Sozaboy, p. 2) or bribe, including the act of giving). The excerpt that follows, an Efik Chief’s diary entry, provides some of [Gloss: The thing/It hurt me very badly because I want- the earliest evidence of the development of a ed to be an accomplished and successful person, like a pidginized interlanguage in Calabar: ‘about 6 am in doctor or a lawyer, driving a nice car and impressing aqua Landing with small Rain morning so I walk up to people with my command of the English language] see Esim and Egbo Young so I see Jimmy Henshaw b. E/De ting vex me bad bad. I wan be big man come to see wee and wee tell him for go on bord …’ like lawya or docto wey get car and tok big big (Forde 1956). Around Lokoja, a confluence town, dif- English (PP) ferent varieties of pidgin are spoken in a relatively (3) ‘Everyting scatta’ (from ’Zombie’) [Gloss: small area. The same has been recorded for the Delta Everything is scattered/is in disarray] and Rivers regions, where several varieties of NPE (4) ‘I look the D.O. well well. Then I tell am say no have been documented (Marchese and Shnukal 1982; be tallness go fight the war…’ (from Saro- Faraclas 1985). Opinions differ regarding the conjec- Wiwa’s Sozaboy, p. 27) tured expansion of NPE into a Creole in some parts of Nigeria (Donwa-Ifode 1983). [Gloss: I looked at/stared at the D.O. for a long time. Even though English is Nigeria’s official language, Then I told him that when it came to fighting the war, it NPE is the preferred lingua franca; hence, it should didn’t matter how tall you were] 743 NIGERIA (5) E get one pikin, I/Ai get four (pikin ) [Gloss: rotten English written by the recently hanged Nigerian She/he has one child; I have four (children)] writer Ken Saro-Wiwa, who, although not a linguist, (6) ‘If you neva ready to carry the load, O’, why made the following sociolinguistic observation: put am for another person head?… Dis kain Both ‘High Life’ and Sozaboy are the result of my life, na so so wahala’ (lyrics excerpted from fascination with the adaptability of the English ‘Wait for Me’ composed by King Sunny Ade Language and of my closely observing the speech and and Shina Peters) [Gloss: If you aren’t ready to writings of a certain segment of Nigerian society. For, shoulder your responsibility, why burden others as Platt, Weber, and Ho accurately observe in their with it? That approach is full of troubles] book The new Englishes (1984), ‘In some nations … (7) ‘No money, woman no go follow you, even if the New Englishes have developed a noticeable range you fine fine pass everybody’ (from Saro- of different varieties. … ’ Sozaboy’s language is what Wiwa’s Sozaboy, p. 21) I call ‘rotten English’, a mixture of Nigerian pidgin English, broken English, and occasional flashes of [Gloss: If you have no money, women won’t come after good, even idiomatic English. This language is disor- you, even if you are the most handsome man in the world] dered and disorderly. … It thrives on lawlessness, and (8) Wettin you de talk? (What are you talking is part of the dislocated and discordant society in about? OR What thing are you referring to?) which Sozaboy must live, move, and have not his being (‘Author’s Note’). Nonetheless, some core features that different vari- Works by other Nigerian artists that are entirely or eties of NPE share include an anglicized vocabulary: predominantly in NPE include ‘Zombie’, lyrics by Fela African grammatical and discourse structures, the Kuti, one of Africa’s foremost musicians, and No nonuse of tense markers, as in (3) and (4), the zero longer at ease in which the characters’ use of NPE both copula and zero distinctions in gender and number mirrors and intensifies the neocolonial discord Achebe (see (1b)), zero genitive forms (see (6)), reduplica- depicts through his artful use of varieties of Nigerian tion––a feature of many African languages (see (4) English. Some of Africa’s best-known writers, play- and (7)), and multiple meanings associated with a sin- wrights, and musicians come from Nigeria, and their gle word (e.g. the word chop, which means con- works have been instrumental in the development of sume(d), eat(en), ate, digest(ed), food, a snack, and a Nigerian English. Several are world-renowned and bribe). Other similarities include a mixed vocabulary have contributed substantially to both African literature that contains many words from Nigerian languages, and world literature. In 1986, Wole Soyinka became reduced or zero articles, and demonstratives; commu- the first African to receive the Nobel Prize in literature. nicative contractions such as ‘tori’ for ‘story’, ‘gree’ Chinua Achebe was awarded the Nigerian National for ‘agree’, and ‘kain’ for kind”; communicatively Merit Award, Nigeria’s highest award for intellectual expressive compounds like ‘sofahead’ (a creative term achievement. Arrow of God won the New Statesman- used to refer to one who worries too much), and coa- Jock Campbell Award, and Anthills of the Savannah lesced creations, such as ‘likam’ (like him/her, likes was the finalist for the 1987 Booker Prize. him/her), ‘wey’ (who, whose, where), and ‘wettin’, a Nigerian artists such as Chinua Achebe,Wole combination of ‘what’ and ‘thing’, as in (8). Soyinka, Ola Rotimi, Amos Tutuola, John Pepper Like Nigerian English, NPE is particularly creative, Clark, Segun Oyewole, Tunde Fatunde, Christopher and echoes the rich Nigerian oral tradition (over 2,000 Okigbo, Gabriel Okara, and Femi Osofisan have been years old). Proverbs are a vital ingredient and color this known to utilize NPE for literary effect––to color their code in a refreshing manner. Examples include: Who characters or the plot (e.g. Amos Tutuola’s The Palm- go dash monkey banana? (Gloss: Who gives monkeys Wine Drinkard), to add humor (e.g. Soyinka’s Jero’s bananas out of sheer goodwill?/There are no free Plays, Segun Oyewole’s Katakata for Sofahead ), sar- lunches) and Kaki no be leda (Gloss: Khaki is not casm, local flavor, or to give the text a linguistic facelift leather; in short, it’s like comparing apples and (see Pandey 1997). NPE is also frequently used in oranges). Sayings in NPE enable their speakers to suc- Onitsha Market literature, including well-known plays cinctly communicate a whole lot. In a culture where like Ogali Ogali’s Veronica my daughter that has sold economy, metaphor, and wit are prized in conversation, over 250,000 copies since 1956, and short stories writ- this code is valuable. Because of its function as the lan- ten by a host of Nigerian pamphleteers (Sander 1980). guage of the people, many Nigerian literary artists uti- Ogali uses NPE or ‘uncooked English’, as Veronica lize NPE in their works (Lindfors 1974). In fact, for a calls it, to portray her traditional father (Chief Jombo) long time, NPE served as a safe outlet for political and Chief Bassey, and Standard English to portray the expression, particularly during military regimes (Eka other characters. Bomber Billy speaks a bombastic 1999). Exemplary works include Sozaboy: A novel in variety of English, and is ridiculed by the writer and by 744  NIGERIA one of the characters in his play, Paulina. She describes The vast majority of Nigeria’s languages––includ- him as ‘a negligible pocket radio that utters useless ing Yoruba and Igbo, two of the three major lan- words’. In Sander’s (1980) opinion, guages––belong to the Niger-Congo family. Estimates vary regarding the exact number of indigenous lan- Ogali is making fun of those people in Nigerian society guages thriving in Nigeria. While Bamgbose (1991:2) who, eager to demonstrate their mastery of the English and McArthur (1992) put the number at 400, Katzner language, go to any extremes in the use of polysyllabic words or the creation of new words. … Through the use (1995) estimates that there are ‘about 250’ (352). of various levels of English, Ogali manipulates the While we cannot rule out the threat of language response of the audience (x). endangerment arising from the central place that English occupies in Nigeria’s linguistic horizon, the In short, NPE is frequently utilized in Nigerian lit- discrepancy in estimates has a lot to do with differ- erature, and its continued use contributes to its popu- ences in opinions regarding what constitutes a lan- larity. As far back as 1979, NPE was observed to have guage and what constitutes a dialect. For instance, replaced Efik as the lingua franca of the Cross River Ibibio, Efik, and Annang, being mutually intelligible, area (Dunstan 1969:35). are viewed by many Nigerians as one and the same Being a West African pidgin, NPE is of vital impor- language (Dunstan 1969:35), namely Efik/Calabar (or tance to the study of creolistics, because of its histori- Ibibio to some Nigerians), which has acquired the sta- cal role in the development of anglophone pidgins and tus of a Standard dialect, on account of its use in creoles. In the words of Nicholas Faraclas (1996), broadcasts and textbooks. Some Nigerians do, howev- ‘Nigerian Pidgin can be considered to be one link in a er, regard them as separate languages. According to chain of English-lexifier pidgins and creoles spoken some, Ikwer(r)e/Ikwerri has 12 dialects. Many non- along the coast of West Africa and in African diaspora Ibibio speakers regard Qua, Oron, Eket, Okobo, and communities throughout the Atlantic Basin’ (1). Ibuno as dialects of Ibibio, while others view them as Advocates of the Afrogenesis hypothesis, which traces lesser-used languages. While some view Kolokuma the roots of pidgins and creoles to Africa, look to West and Ijo as dialects of the same language, others might African pidgins and creoles for answers (Romaine not, as Nigerian codes of communication have distinct 1994; Dalphinis 1985). According to one view, names. Most have several names. Isoko, for instance, ‘Created in West Africa, … work pidgins were trans- is also called Biotu, Sobo, and Igaba; yet, some con- ported across oceans where they took on new roles as sider the last two names to be offensive. lingua francas among the enslaved. …’ (McWhorter To most Nigerians, dialects tend to coincide with 1984:240). In the case of NPE, as well as Nigerian geographical boundaries or specific urban locations. English, many of the ingredients are African. These Chief dialects identified for Eksako, for instance, are include the names of food (akara balls, fried bean said to be those of Auchi, Aviele, Ekperi, South Ibie, cakes; moimoi, steamed bean cakes; dodo, fried plan- Uzairue, and Weppe Wano––names of towns and tains; gari, powdered cassava; Jollof rice, rice flavored cities, and of specific ‘clans’ (Dunstan 1969:47). The with Nigerian peppers; masa; millet cakes eaten in three main dialects of Fula/Fulani found in Nigeria Hausaland; pounded yam, egusi soup, etc.); clothing also coincide with the names of major cities or geo- (agbada, the flowing robes that Yoruba men wear; graphical terrains, namely Sokoto, Adamawa, and the buba and wrapper, the women’s traditional outfit); central dialect of Kano, Bauchi, Gombe, Katagum local customs and festivals (naming ceremonies for Division, and Jos (Dunstan 1969:57). newborns); regional celebrations (e.g. Masquerade Most educated Nigerians speak at least three lan- Day, a tribute to the ancestors in the Ekiti region); and guages, English (the official language), one or more of expressions of linguistic pride such as ajibotta, the the major regional languages, a local language, and well-known derogative Yoruba word for a butter eater NPE, not to mention the other dialect(s) they might or westernized Nigerian (the Nigerian English equiva- switch to in the middle of a conversation. For most lent is been-to). Exemplary Yoruba words in NPE Qua speakers, for instance, Efik is a second language, include Oga (a respectful term, roughly equivalent to Igbo the third, NPE the fourth, and (Nigerian) English ‘Sir’), katakata (major disturbance or problem), and the fifth. the honorific O’. Some Hausa words that have entered In Nigeria, ethnicity and linguistic identities are NPE include wahala (trouble) and haba (Goodness! or inextricably intertwined. McArthur (1992) recognizes Listen up!). Igbo words in NPE include sef (a dis- eight major ethnic groups, namely 21% Hausa, 21% course marker), yanga (boasting or boastfulness), and Yoruba, 18% Igbo, 10% Fulani, 6% Tiv, 5% Kanuri, na (a particle with multiple meanings, such as ‘it is’, 5% Ibibio, 4% Edo, and 10% minorities. Because of and the attention-getter Na wa, O’!, a popular expres- the close ties between the Hausa and the Fulani, it is sion of shock). not unusual to find these groups conjoined under one 745 NIGERIA label, namely, the Hausa-Fulani. The minority ethnic Hausa enjoyed official language status alongside groups would include, among others, the Ogoni, English in what was then the Northern Region. At that Isoko, Izon, Kwale, Degema, Itsekiri, Ibuno, Izi, Ewe, time, Hausa was written in the Arabic script known as Andoni, Amo, Angas, Birom, Chip, Chella, Efik, Ajami, but now the Roman script is used. Mada, Nabor, Nembe, Tera, Yergam, Gwari, and Yoruba, a paradigm Kwa language and a subgroup Gonja peoples of Nigeria. The names of their lan- of the Niger-Congo family, is spoken in southwestern guages tend to coincide with their ethnic descriptors, Nigerian, from Illorin (Kwara State) down to Akure, further emphasizing the linguistic basis of ethnic iden- Ado-Ekiti, and Igede-Ekiti (Ekiti State), Ibadan (Oyo tity in Nigeria. Each ethnic group has its own unique State), and Lagos. It is Nigeria’s second major lan- customs and language, making for a diverse group of guage. This language and the rich mythology (of Nigerians; yet, the minor languages tend not to be as Sango and Ogun) constitute seminal ingredients in carefully studied. Some (Fula, Tiv, Igala, Efik, and many of Wole Soyinka’s works. The mythology is also Edo) enjoy a more functional status than others, resonant in the Afro-Caribbean beliefs transported to because of their use in the media. The nation’s lin- the Caribbean (see Burnett 1986). Some 20 dialects of guistic and cultural diversity has a lot to do with its Yoruba can be heard in Nigeria, and one, ‘based large- historical location at the heart of transcontinental ly on the Oyo dialect’ (Dunstan 1969:80), is the migration routes. The greatest concentration of lan- Standard. Yoruba is also spoken in some other West guages and ethnic groups is found in Central Nigeria. African nations, including parts of Ghana, Togo, and One has only to travel to Kwara State or Illorin to get Benin. Tonality is mirrored in its orthography. Three a feel for the linguistic diversity. letters and their corresponding sounds are noteworthy: The most widely spoken indigenous language is o, pronounced as in ‘sought’ (e.g. oba); é pronouced as Hausa (Katzner 1985; McArthur 1992). This is fol- in ‘let’ (e.g. éjé which means blood); and s, pro- lowed closely by Yoruba (18% of the population or nounced /sh/ as in ‘ekuse’ (i.e. Well done!). In orthog- some 20 million speakers), then Igbo (11% or some 15 raphy, the acute and grave accents indicate tone (not million speakers), Fulani (approximately 8 million stress). speakers), Kanuri (3 million), Efik and Ibibio (3 mil- Igbo, another Kwa language like Yoruba and Ewe, is lion), Tiv (spoken around Gboko by some 2 million), Nigeria’s third major language. It is spoken in south- Ijo or Ijaw (2 million), Edo (one million), Urhobo eastern Nigeria, in Port Harcourt, Enugu, Calabar, (half a million), Nupe (in the Gur subfamily of Niger- Onitsha, Owerri, and surrounding areas. The three Congo languages, with roughly 500,000 speakers), main dialects that are recognized are the Central, Idoma (some 250,000 speakers), and Eksako (also Owerri, and Umuahia dialects. Prior to 1961, its spelled Esako and Etsako(r), an Edo language with orthography was disputed (Dunstan 1969:85). The offi- some 120,000 speakers). These numbers are more esti- cial orthography, adopted by the Onwu Committtee, is mates than up-to-date survey findings of ‘native now compulsory in the School Certificate Igbo exami- speakers’ of these languages, as representative lan- nation. Igbo features regularly in the works of Chinua guage surveys are hard to administer in this complex Achebe. In Things fall apart, some 37 Ibo words and multilingual environment, and such surveys have phrases are woven into the fabric of the text. Many rarely been administered (see Bamgbose 1995:34). have no translation equivalents, including chi (person- Hausa, a Chadic language that contains many al god), ogbanje (which Achebe describes as a words borrowed from Arabic, is spoken by roughly ‘changeling’ that will keep dying until its ‘iyi-uwa’ is 27% of the population or some 40 million Nigerians. first dug up and destroyed), uri (a betrothal day, after It is the primary language or mother tongue of at least the bride-price has been paid), Nne (mother, including 25 million Nigerians (Katzner 1985:288). The Kano one’s nonbiological mothers in a polygamous house- dialect is regarded as the Standard variety. The number hold), umunna (kinsmen), and ndichie (one’s elders). of Hausa speakers has clearly grown; in 1979, an esti- These indigenous words bring the richness of precolo- mated 12 million Nigerians were native Hausa speak- nial Igbo society to life. From a sociolinguistic stand- ers (Dunstan 1969:73). Hausa is the lingua franca for point, they constitute prime examples of the the bulk of northern Nigeria and also plays the role of Sapir–Whorfian hypothesis, which illustrates how spe- a regional language in West Africa. It is, in fact, the cific cultural concepts are embodied in words. most widely spoken African language in West Africa, Fula(ni) is generally termed Fulfulde in Nigeria. with speakers found in neighboring countries, includ- This Niger-Congo language (the West Atlantic branch) ing Nigér (as many as five million speakers), Ghana, of the Fulani/Fulbe people is of great historical signif- Togo, Benin, Mali, and Senegal. The Hausa Language icance, as the Fulani are ‘a people of great antiquity’ Board, established in 1955, has assisted in the stan- (Katzner 1995:289) and have heavily influenced the dardization of the language. Prior to independence, sociolinguistics of northern Nigeria, where over eight 746  NIGERIA million of the estimated 15 million Fulani people independence, then, several universities were estab- reside. Many are cattle-herders, and follow the lished in Nigeria, most notable among them The nomadic and seminomadic lifestyle they are accus- University of Nigeria in Nssuka, eastern Nigeria, tomed to. Areas in which Fulani is spoken in Nigeria Ahmadu Bello University, located at the heart of include Sokoto, Kano, Kaduna, Plateau, Bauchi, and Hausaland, and the University of Ile-Ife, located in Ife, Kwara States. In Mauritania, Gambia, Benin, Ghana, a very important city in Yorubaland, and home of the Mali, Burkina Faso, Chad, Nigér, and Cameroon, Oba of Ife, one of the most highly regarded Chiefs in where it is also spoken, it is called Fula. In Senegal, it the Yoruba chieftaincy. Ahmadu Bello University (bet- is called by its other name, Pulaar, and in Guinea- ter known as ABU) was named after (Sir) Ahmadu Bissau, it is termed Pulle. The root morpheme in all Bello who was northern Nigeria’s first Premier. A resi- these names is {Ful-} or {Pul-} from which the dent of Hausaland, and a well-known political figure, Germanic name Fula(ni) and the Frenchified ‘Peul’ northerners (called ‘Gambaris’ by some southerners) originated. Its regional spread makes it a frequently named the University after him, as a tribute to his used lingua franca in West Africa. accomplishments. The University is located in Zaria, a The national language question has received quite a famous northern city that played a pivotal role in the bit of attention since the mid-1970s. In Language and intricate emirate system that was in place prior to the the nation, Ayo Bamgbose observes that: colonial period. The language question in Sub-Saharan Africa arises from English and NPE usage declines steadily as one the fact that not only are most of the countries multilin- moves north of the town of Illorin. Instead, one tends to gual, the colonial experience has led to the importation hear more Hausa and Fulani in everyday conversations of foreign official languages which have taken on the and even in service-counter exchanges. Koranic schools roles of national communication, unification, adminis- and after-school Arabic classes are commonplace in the tration and medium of education from early or late pri- north, where Arabic is a language of great (religious) mary to university level. Thus, the existing multilingual importance, although rarely used in conversations. In situation is compounded by the addition of imported lan- contrast, in the open markets in the Yoruba-speaking guages whose strength does not lie in numbers of speak- Ekiti area or in Onitsha, which is famous for its market, ers, but rather in the superior roles assigned to them (1). NPE is the preferred transactional code, ideal for bar- During the colonial period, western education was gaining. So important is NPE in the markets of restricted to only a small segment of the population that Yorubaland and Igboland that mastery of NPE gives played a role in the ‘indirect rule’ in place; hence, even buyers a competitive advantage over non-NPE speakers, today, only the elite uses Standard English. British as haggling is a culturally acceptable and expected dis- administrative policy in Nigeria relied on empowered course practice. One who does not speak the language Nigerians––primarily Chiefs and Emirs––to serve as of the seller is usually at a disadvantage. Some vendors intermediary rulers. English usage and English-medium can be heard speaking what is called ‘Broken English’ western education were most widespread in parts of the in Nigeria, and mainly in their limited-but-meaningful country where Christian missions were established exchanges (e.g. ‘Oga, buy tomato! Good Good!’) with (Ajayi and Crowder 1985:68). Islam (roughly 47% of the Oy˜ιbo(s), the Yoruba term used to identify outsiders the population) and Christianity are the two primary or non-Nigerians. Some traders and children use the religions, with an estimated 19% of the populace being compound term ‘Oy˜ιbo man’. This term literally means ‘traditionalists’ (McArthur 1992). As Christianity was ‘white man’ or ‘white person’, but its use is not limited resisted in Islamic northern Nigeria, where the Jihad or to Europeans alone. Also, it is not intended as an insult, Holy War led by Uthman dan Fodio in the early nine- particularly when it is followed by the respect marker teenth century had left its religious mark, the discrepan- ‘O(h)’, although it could easily be misconstrued as cy in English usage between northern and southern such, particularly when one is surrounded by a crowd Nigeria has persisted to this day. So pronounced were chorusing ‘Oy˜ιbo O’’. The term should be interpreted the differences on the eve of independence that there as more of a congregational announcement of the pres- was even a fear that those with an English-based ence of a visitor whom one would like to honor by Western education ‘would dominate the economic and acknowledging. ‘O’’ is an honorific term in Yoruba, political life’ (Ajayi and Crowder 1985:68) of the new which means roughly ‘respectful Sir or Ma’am’, and nation. The establishment of major regional universities has been adopted by NPE. The equivalent term for can be attributed, in part, to this regional-linguistic com- expatriates in Hausa is ‘Baturé’ (masculine) or petition. As the University of Ibadan, Nigeria’s oldest ‘Baturiya’ (feminine). institution, was garnering a lot of (inter) national atten- It would be a mistake to view Nigeria as an ESL tion, many Nigerians saw the need for a university in environment, because it is an English-as-an-indige- Igboland and another in Hausaland. At the wake of nized-language context (Bamgbose 1995). Even 747 NIGERIA though English is the medium of instruction from of a national language have advanced the idea of an third grade onward, most Nigerians speak Nigerian indigenous national language. English, which differs from Standard British or Special programs have been designed to reduce American English. In the past 10 years, a matter of tribalism and foster national unity. One such measure grave concern to Nigerian educators has been the is the Nigerian Youth Corp Service, better known as ‘mass failure syndrome’ (Bamgbose 1995:130–52) or the NYSC, which is a year-long commitment required the increasing number of failing grades in English of all graduates. It requires them to serve in a region obtained in the main college-entry examinations, that is linguistically different from their own, so that namely, the WAEC or the West African School they learn to be more tolerant of other ethnic and lin- Certificate Examination, and the JAMB or the Joint guistic groups. This year of service almost always Admissions and Matriculation Board Examination. ensures a process of assimilation. Most ‘youth cor- These examinations are patterned along the colonial, pers’ end up acquiring yet another language––usually British variety, yet the students hear and learn a a major regional language––during their NYSC year. Nigerian variety. Many of the idioms that students are The prop on which this program rests is essentially tested on are, in fact, foreign to the culture and envi- sociolinguistic. Another multilingual measure is the ronment. Examples include the following italicized attempt to revive the local languages, particularly the items: a finger in every pie, comparing apples and three major languages that constitute the cornerstones oranges, looking for a needle in a haystack, over the of Nigeria’s linguistic ‘stool’ (a symbol of the seat of grape vine, and penny wise, pound foolish. Students government in the precolonial Chieftaincy era). This are often tested on noncount nouns like information, includes making them instructionally viable (Rufa’i even though most Nigerians pluralize these. The end 1991; Arohunmolase 1998, 1999) through the devel- result is a mismatch between the variety of English opment of textbooks, teacher preparation, and so on. that is learned, spoken, heard, and taught, on the one The WAEC now offers certification examinations in hand, and what is tested. Admission into college is some Nigerian languages. contingent upon a credit pass in English, and yet a curricular change is unlikely, as administrators do not References recognize Nigerian English. Achebe, Chinua. 1958. Things fall apart. London: Heinemann. In a report submitted to the Federal Ministry of Achebe, Chinua. 1964. Arrow of God. London: Heinemann. Information, the Nigerian Public Service Review Achebe, Chinua. 1987. Anthills of the Savannah. London: Commission drew attention to the need for a national Heinemann; 2nd edition, 1989. language: Achebe, Chinua.1975. Morning yet on creation day. New York: Doubleday Press. An overriding problem, which affects the public service Ajayi, Ade, and Michael Crowder (general eds.) (E. Dunstan, as it does all aspects of society, is that of language. linguistics ed.) 1985. Historical atlas of Africa. Cambridge: Nigeria shares with many developing, and some devel- Cambridge University Press. oped nations, the lack of an indigenous lingua franca. Akere, Funso. 1982. Sociocultural constraints and the emer- What this means for efficiency in the conduct of gov- gence of Nigerian English. New Englishes, ed. by J.B. Pride, ernment business is rarely even thought about perhaps 85–9. Rowley, MA: Newbury House. Altback, Phillip, G. (eds.) 1999. Publishing in African lan- because there seems to be no immediate answer. But it guages: challenges and prospects. Asmara; Trenton, NJ: is perfectly clear to the careful observer that below the Africa World Press, Inc. top-most levels in the various sectors of society, people Amkpa, Awam. 1998. Nigerian dramatic literature: the exam- are conducting their business in a language which, in ples of Femi Osofisan and Tunde Fatunde. The growth of varying degrees, they have not in fact mastered (6). African literature: 25 years after Dakar and Fourah Bay, ed. by Edris Makward, et al. Asmara: Africa World Press, Inc. Since that time, some of the major motions that Arohunmolase, Oyewole (ed.) 1998. Nigerian languages for have been tabled include the idea of a trilectal nation- national development and unity. Trenton, NJ: Africa World al language policy, a rotational language policy (which Press, Inc. also draws on the three regional languages, one by Avery-Coger, and Greta M.K. McCormick (eds.) 1988. Index of one), and the creation of a single conglomerate lan- subjects, proverbs, and themes in the writing of Wole Soyinka. New York: Greenwood Press. guage, like Esperanto, which would be called WAZO- Bamgbose, Ayo. 1982. Standard Nigerian English: issues of BIA, and which would combine elements of Hausa, identification, In Kachru, pp. 148–61. Yoruba, and Igbo. Another idea that has received con- Bamgbose, Ayo. 1991. Language and the nation: the language siderable attention is the pan-Africanist idea of adopt- question in sub-Saharan Africa. Edinburgh: Edinburgh ing the most widely spoken African language, namely, University Press. Bamgbose, Ayo, et al. (eds.) 1995. New Englishes: a West Swahili. This proposal has been led by Wole Soyinka. African perspective. Ibadan: Mosuro. In the last few years, French has been added to the list Burnett, Paula. 1986. The Penguin book of Caribbean verse in of contested languages. In general, Nigerians in favor English. Harmondsworth: Penguin Publishers. 748  NILO-SAHARAN LANGUAGES Dalphinis, Morgan. 1985. Caribbean and African languages: McWhorter, John, H. 1984. The missing Spanish Creoles: social history, language, literature, and education. London: recovering the birth of plantation contact. Berkeley: Karia Press. University of California Press; 2nd edition, 2000. Donwa-Ifode, S. 1983. Is Nigerian Pidgin English Creolizing? Odumuh, E. 1987. Nigerian English. Zaria: Ahmadu Bello Journal of the Linguistic Society of Nigeria 2. 199–203. University Press. Dunstan, Elizabeth (ed.) 1969. Twelve Nigerian languages. Osofisan, Femi. 1975. A Restless run of locusts. Ibadan: New York: Africana Publishing Corporation. Onibonoje Press. Eka, David. 2000. Aspects of language in Ken Saro-Wiwa’s Oyekunle, Segun. 1983. Katakata for Sofahead. London: Sozaboy: a novel in rotten English. Before I am hanged, ed. Macmillan Education. by O. Okome. Trenton, NJ: Africa World Press, Inc. Pandey, Anita. 1997. The Pragmatics of Code Alteration in Emenyonu, Ernest. 1989. National language policy in Nigeria: Nigerian English. Studies in the Linguistic Sciences 25(1). implications for English teaching. Language planning and 75–117. English language teaching, ed. by C. Kennedy, 82–91. Pandey, Anita. 1999. Code alteration and Englishization across Oxford: Oxford University Press. cultures: cyclic differences. Paper invited for presentation at Eze,Smart, N. 1980. Nigerian Pidgin sentence complexity, the International Association for World Englishes (IAWE) Band 8. Wien, Germany: Afro-Publishers. Convention, Singapore, and invited for inclusion in The Faraclas, Nicholas. 1985. Rivers Pidgin English: tone, stress, or Three Circles of English by the Conference Organizer, pitch-accent language? Language in African culture and Edwin Thumboo, University of Singapore. society. Studies in the Linguistic Sciences 14(2), ed. by E. Romaine, Suzanne. 1994. Language in society: an introduction Bokamba. 67–76. to sociolinguistics. Cambridge: Cambridge University Press. ––––––. 1996. Nigerian Pidgin. New York: Routledge. Rufa’i, Abba (ed.) 1991. Twentieth Year Commemorative Fatunde, Tunde. 1985. No food, no country. Benin City, Symposium of the Center for the Study of Nigerian Nigeria: Adena Publishers. Languages (1989: Bayero University), Nigerian Languages Forde, C. D. 1956. Efik traders of old Calabar. London: Yesterday, Today, and tomorrow: Proceedings of the International African Institute. Symposium. Kano, Nigeria: Bayero University Press. Kachru, Braj (ed.) 1982. The other tongue: English across cul- Sander, Reinhard. 1980. Veronica my daughter and other tures. Urbana: University of Illinois Press; 2nd edition, Onitsha plays and stories by Ogali Ogali. Port Harcourt: 1992. Three Continents Press, Inc. Katzner, Kenneth. 1995. The languages of the World. New Saro-Wiwa, Ken. 1985. Sozaboy: a novel in rotten English. Port York: Routledge. Harcourt: Saros Publishers. Lindfors, Bernth (ed.) 1974. Dem-say: interviews with eight Ubahakwe, Ebo (ed.) 1979. Varieties and functions of English Nigerian writers. Austin, TX: University of Texas at Austin in Nigeria. Ibadan: African Universities Press. Press. ANITA PANDEY ———. 1997. African textualities: texts, pre-texts, and contexts of African literature. Trenton, NJ: Africa World Press. See also Bilingualism; Code-Switching; Hausa and Maja-Pearce, Adewale (ed.) 1994. Wole Soyinka: an appraisal. Chadic Languages; Igbo and Igboid Languages; Ibadan: Heinemann. Marchese, L., and A. Shnukal. 1982. Nigerian Pidgin English of Niger-Congo; Official Language Selection; Pidgins Warri. Journal of the Linguistic Society of Nigeria 1. and Creoles; Sapir–Whorf Hypothesis; Yoruba and 213–19. Yoruboid Languages Nilo-Saharan Languages The idea of a Nilo-Saharan (NiSa) phylum among significant changes, most of which have been accepted African languages alongside Afro-Asiatic, Niger- since and partly also substantiated in detailed studies: Congo, and Khoisan was brought into existence by (a) The inclusion of Mimi into the Maban group. Joseph H. Greenberg (1963). On the basis of 161 lex- (b) The combination of Tucker and Bryan’s ‘Nilotic’ ical and 29 grammatical sound–meaning similarities, and ‘Nilo-Hamitic’ into one group with three he suggested genetic unity among 82 languages, branches: Western, Eastern, and Southern Nilotic. which had been treated as 22 separate language units (c) The combination of Moru-Mangbetu and before by Tucker and Bryan (1956, 1966) and for Bongo-Bagirmi into a Central Sudanic group. some of which genetic unity had never been suggested (Table 1). Greenberg’s fusion of Gule and Koma into one group, On the level of language units as presented in however, has been revised. Bender (1996) suggests again Tucker and Bryan (1956), Greenberg proposed some two units although differently subdivided, i.e. 749 NILO-SAHARAN LANGUAGES TABLE 1 Language Groups of Tucker and Bryan (1956) and Greenberg (1963) Tucker and Bryan (1956) Greenberg (1963) 1. Songhai 2. Eastern Sudanic 2. Saharan: (a) Kanuri, Kanembu; (b) Teda, Daza; (c) Zaghawa [today also: Beria], Berti 10. Mimi 3. Maban: Maba, Runga, Mimi (of Nachtigal), Mime (of Gaudefroy-Demombynes) 11. Maba group 13. Fur 4. Fur 20. Nubian group 1. Nubian group 27. Didinga-Murle group 2. Murle, Logarim, etc. [today: Surmic] 21.Barea 5. Chari-Nile 3. Barea [today: Nara ~ Nera] 24. Tabi A. 4. Ingassana (Tabi) [today also: Gaam/Jebel] 15. Nyimang group Eastern 5. Nyima, Afitti 16. Temein group Sudanic 6. Temein, Teis-um-Danab 12. Tama group 7. Merarit, Tama, Sungor 14. Daju group 8. Daju group 30. Nilotic 9. Nilotic 31. Nilo-Hamitic 29. Teuso 10. Nyangiya, Teuso [today: Kuliak] 1. Moru-Mangbetu B. Central Sudanic 2. Bongo-Bagirimi 23. Berta C. Berta 22. Kunama D. Kunama 25. ‘Gule’ [extinct] 6. Coman: Koma, Ganza, Uduk, Gule, Gumuz, Mao 26. Koma group I. Koman: T’wampa (Uduk), Gule, Komo, Kwama presented in these studies highlight significant fea- and Opo (Shita) and tures of the current state of discussion in the field of J. Gumuz (dialect cluster). Nilo-Saharan genetic studies: Several morphological sound–meaning similarities —The idea of a Nilo-Saharan phylum remains used by Greenberg (1963) to define Nilo-Saharan are widely accepted, although the external still major arguments regarding Nilo-Saharan unity, boundaries of Nilo-Saharan as well as among them: significant parts of its internal subdivision remain highly controversial. —independent subject pronouns, especially in the —Greenberg’s proposal of a Chari-Nile group has singular, show the vowels a (1st person), i or u been rejected as a valid genetic unit, while the (2nd person) and e (3rd person), genetic unity of Eastern Sudanic and of Central —ma and ko as relative and adjective formant, Sudanic remains unchallenged. —n/k singular–plural alternation, —The lower-level units (Tucker and Bryan’s —t/k singular–plural alternation, language units as revised by Greenberg) are —nominal derivational prefix k- (‘movable’ k-), also largely uncontroversial. —verbal dative affix -kV-, —The most substantial progress since Greenberg —causative in t has been made regarding language documentation, genetic classification, and Several morphological sound–meaning correspon- reconstruction within these lower-level units. dences proposed by Greenberg to support Nilo- Saharan unity, however, are also found in neighboring Afro-Asiatic (e.g. negation in b or m) or in Niger- External boundaries Congo languages, thus weakening the argument and Discussions regarding the external boundaries of Nilo- pointing to the general difficulty of separating areal Saharan relate particularly to the following topics: from genetic features. Songhay cluster: Its geographic distance from the Most recently, only two comprehensive studies other Nilo-Saharan languages and long-standing contact on the genetic classification of Nilo-Saharan since with Mande (Niger-Congo) and Berber (Afroasiatic) Greenberg’s have been published, i.e. Bender (1996) languages continue to cast doubt on the Nilo-Saharan and Ehret (2001). The comparison of the results affiliation suggested by Greenberg (1963) and supported 750  NILO-SAHARAN LANGUAGES by Bender (1996) and Ehret (2001). Nicolai (1990) this derives not so much from the still highly fragmen- reconstructs it as a post-Creole with a Berber base, while tary documentation of numerous Nilo-Saharan lan- some other scholars follow an old argument by guages, but from their generally great internal lexical Delafosse and discuss Mande affiliation. and grammatical heterogeneity. This feature may point Kadugli-Krongo (Kadu): This group appears in to a history of thousands of years for the whole Greenberg (1963) under the name of Tumtum as one phylum. Detailed studies of lower-level language of five Kordofanian branches despite divergent proper- units, however, also reveal continuous episodes of ties. Following suggestions made by Dimmendaal heavy language contact, thus complicating lexical and (1987) among others and supported by the fact that morphological reconstruction and hence higher-level this group shows several of the typical Nilo-Saharan grouping considerably. features (1sg pronoun a'a, 2sg pronoun
’
, movable Based on morphological as well as on lexical com- k, n/k plural formation, etc.), Bender (1996) includes parison, Bender (1996) proposes four branches of Nilo- Kadugli-Krongo in his Core of Nilo- Saharan, while Saharan: (A) Songhay, (B) Saharan, (K) Kuliak, called Ehret (2001) rejects any Nilo- Saharan affiliation. ‘Outliers’, and the rest, called ‘Satellite–Core’, which Shabo (formerly: Mikeyir): The genetic position of he further subdivides as indicated in Table 2. this rudimentarily documented language of SW Ehret’s (2001:88f, 70f) alternative genetic classifi- Ethiopia remains unclear. Bender proposes a Nilo- cation of Nilo-Saharan uses phonological and lexical Saharan affiliation, which Ehret rejects. isoglosses derived from an extensive comparative Meroitic: The extinct language of the ancient lexical database, but also considers grammatical and Meroitic Empire (Sudan) preserved by a number of derivational morphemes as far as they are accessible written records has been linked to various genetic through language descriptions. His result is present- groups and phyla, among them Afro-Asiatic, Tokharian, ed in a mainly bilaterally branching tree, with one Saharan, and Eastern Sudanic. This latter affiliation was specific small language group or language branching also supported by Greenberg, and has most recently off at each level and set against the rest, which again been substantiated by Claude Rilly, among others. receives novel geography-based labels. In the follow- Relation to Niger-Congo: Due to a number of lexi- ing condensed representation of his result, previous- cal and morphological similarities between Nilo- ly known group and language labels are underlined Saharan and Niger-Congo languages, Gregersen (see graph). (1972) proposed to combine these two phyla into a The extreme lexical heterogeneity of Nilo-Saharan single one, i.e. Kongo-Saharan. His line of argument languages, even in the most basic vocabulary, is paral- has been taken up by Roger Blench in recent years, leled by a large grammatical diversity, and hardly who attempts to establish Niger-Congo as a branch of any of these grammatical features is specific to the Nilo-Saharan. phylum. Nilo-Saharan languages are tone languages with two to four underlying level tones, which are used for Internal subgrouping grammatical as well as lexical purposes. The number General disagreement also characterizes higher-level of tones and their main functional domain correlate to units within Nilo-Saharan. Greenberg’s subgrouping a considerable degree with the morphological type of (Table 1) has only partly been accepted. The reason for the respective language. Five-vowel systems with TABLE 2 Bender’s (1996) Genetic Classification of Nilo-Saharan A. Songhay Outliers B. Saharan K. Kuliak C. Maban Nilo-Saharan Satellites D. Fur F. Central Sudanic G. Berta Satellite-Core H. Kunama Ek. Nubian, Nera, Nyima, Tama E. Eastern Sudanic En. Surmic, Jebel (Gaam); Temein, Core I. Koman Daju, Nilotic J. Gumuz L. Kadugli-Krongo 751 NILO-SAHARAN LANGUAGES Nilo-Saharan Koman Sudanic Central Sudanic Northern Sudanic Kunama Saharo-Sahelian Saharan Sahelian Songhay For[Fur] Eastern Sahelian Maban Eastern Sudanic Astaboran Kir-Abbaian Kuliak Nara W-Astaboran Jebel Kir Nubian Tama W-Jebel Bertha Nuba-M. Daju Surma-Nilotic Temein Nyimang Surmic Nilotic Graph. Ehret’s (2001) genetic classification of Nilo-Saharan. advanced tongue root (ATR) vowel harmony, often TABLE 3 Nilo-Saharan Languages: Languages and with contrastive length, are frequently synchronically Speakers Per Country or diachronically reconstructable. The consonant sys- Country No. of NiSa No. of NiSa % of tems show either four or five (e.g. Gaam, Berta, Koma, Languages Speakers Population Western Nilotic) places of articulation, which partly Kenya ~6 7.5 Mio. 29.2 also extends to the nasal. Ehret (2001:16) claims five Sudan ~ 64 7.2 Mio. 29 places of articulation for Proto-Nilo-Saharan in the Uganda ~ 20 5.6 Mio. 25.5 case of plosives and four in the case of nasals. He also Nigeria 6 3.5 Mio. 3.5 DR Congo 22 3.0 Mio. 7 suggests a complex set of plosives for Proto-Nilo- Chad ~ 33 2.6 Mio 40.6 Saharan, with voiced implosives and explosives, Niger ~7 2.7 Mio. 32.5 voiceless aspirated and unaspirated as well as glottal- Tanzania ~4 858 000 0.3 ized plosives. Although all of these sets do occur in Mali 1 640 000 6 Nilo-Saharan languages synchronically, only Twampa Ethiopia 16 376 000 0.7 CAR 13 223 000 6.6 (Uduk) comes close to such a system. Eritrea 2 210 000 5.7 The morphological type of Nilo-Saharan languages Cameroon 2 150 000 1.2 ranges from largely isolating (e.g. Central Sudanic Burkina Faso 2 125 000 0.8 languages), to agglutinative (e.g. Southern Nilotic) Egypt 1 100 000 0.1 and highly fusional (e.g. Kanuri, partly Western Benin 2 50 000 0.9 Libya 2 9 000 0.1 Nilotic). Languages of the same genetic unit often belong to different morphological types, which possi- bly has to be interpreted as an outcome of substantial Syntactically, all the common basic word orders language contact. Nonisolating Nilo-Saharan lan- with its concomitant features occur, namely SOV guages are mostly characterized by complex verbal (e.g. Kanuri, Nubian), VSO (e.g. East and South derivational morphology, while isolating Nilo-Saharan Nilotic, Kuliak, partly Kadugli-Krongo), and SVO. languages tend to make use of serial verb construction Morphological and partly also syntactic ergativity has for derivational and morphosyntactic purposes. been reported for Jur-Luwo, Anywa, Päri, Shilluk Nominal morphology is equally heterogeneous, with (Western Nilotic), and Toposa (Eastern Nilotic). some languages having gender (e.g. Eastern Nilotic), others not. A morphologically tripartite number sys- Demography tem, in which either a singular/singulative or a plural form or both are morphologically derived from a nom- Numbers given for Nilo-Saharan languages range inal root, is widespread in Nilo-Saharan and may even from 80–90 (lumpers) up to 199 (splitters). The area in be a potential isogloss. which they are spoken stretches from the river Nile in the 752  NOOTKA AND WAKASHAN LANGUAGES TABLE 4 Numerically Dominant Nilo-Saharan Languages Speakers Language Sub-division Country 3–4 Mio. Songhai Songhay Niger, Mali, Nigeria, Burkina, Faso, Benin Kanuri Saharan Nigeria, Niger, Cameroon, Chad, Sudan Dholuo Western Nilotic Kenya, Tanzania Dinka Western Nilotic Sudan > 2 Mio. Lwo (Acholi + Lango) Western Nilotic Uganda, Sudan Kalenjin Southern Nilotic Kenya > 1 Mio. Lugbara Central Sudanic DR Congo, Uganda Alur Western Nilotic DR Congo, Uganda Teso Eastern Nilotic Uganda, Kenya Maa Eastern Nilotic Kenya, Tanzania Nuer Western Nilotic Sudan, Ethiopia > 500,000 Ngambay Central Sudanic Chad, Cameroon, Nigeria Lendu Badha/Lendu DR Congo, Uganda Mangbetu Mangbetu-Balese DR Congo Fur Fur Sudan, Chad > 250,000 Maba Maban Chad Teda-Daza Saharan Chad, Niger, Libyen Masalit Maban Chad, Sudan Bari Eastern Nilotic Sudan, Uganda, DR Congo Karimojong Eastern Nilotic Uganda Nile Nubian (Nobiin) Nubian Sudan, Egypt northeast up to Lake Chad and––in the case of Kanuri References and Songhay––further west to northeastern Nigeria and Bender, L.M. 1996. The Nilo-Saharan languages––a compara- to the Mali-Niger-Burkina Faso border area, respectively. tive essay. (LINCOM Handbooks in linguistics, Vol. 6.) Nilo-Saharan languages are spoken as a first language in München & Newcastle: Lincom Europe. 17 African countries by over 30 Mio. speakers (Table 3). Dimmendaal, Gerrit J. 1987. Krongo: between universal, areal The highest number of speakers are found in Kenya and genetic norms. Journal of African Languages and Linguistics 9. 161–77. (7.5 Mio.), Sudan (7.2 Mio.), Uganda (5.6 Mio.), Ehret, C. 2001. A historical-comparative reconstruction of Nigeria (3.5 Mio.), DR Kongo (3.0 Mio.), Niger (2.7 Nilo-Saharan. (Sprache und Geschichte in Afrika. Beiheft Mio.), Chad (2.6 Mio.), but these speakers represent sig- 12.) Köln: Rüdiger Köppe. nificant portions of the population only in four countries, Greenberg, J.H. 1963. The languages of Africa. International namely Chad (41%), Niger (32%), Sudan (29%), Kenya Journal of American Linguistics 29, 1 (Part 2). [Repr. 1966 by Bloomington: Indiana University; The Hague: Mouton.] (29%), and Uganda (25%). With respect to the number Gregersen, E.A. 1972. Kongo-Saharan. Journal of African of Nilo-Saharan languages per country, Sudan comes Languages 11. 69–89. first (~64), followed by Chad, DR Congo, Uganda Nicolai, R. 1990. Parentés linguistiques (à propos du songhay). and––despite the overall low number of NiSa speak- Paris: Éditions du CNRS. ers––Ethiopia (16) and Central African Republic (13). Tucker, A.N., and M.A. Bryan. 1956. The non-Bantu languages of north-eastern Africa. London, New York, Cape Town: Table 4 presents major Nilo-Saharan languages in Oxford University Press for International African Institute. terms of number of speakers and geographic distribu- Tucker A.N., and M.A. Bryan. 1966. Linguistic analyses: the tion. As indicated, most of them are cross-border lan- non-Bantu languages of north eastern Africa. London, New guages, thus making language-planning activities York, Cape Town: Oxford University Press for International comparatively difficult. African Institute. MECHTHILD REH Nootka and Wakashan Languages Wakashan is one of a dozen or so language families diversity. The term is a variation of ‘Wakashian’, indigenous to North America’s Pacific Northwest, a which Captain James Cook proposed to call the people lush and mountainous region of tremendous linguistic he visited in Nootka Sound in 1778: 753 NOOTKA AND WAKASHAN LANGUAGES The word wakash ... was frequently in their mouths. It proto-Wakashan vocabulary on local maritime culture seemed to express applause, approbation, and friend- also suggests that the Wakashan Urheimat lies in its ship. For when they appeared to be satisfied, or well present area. pleased with anything they saw, or any incident that Historically, the Wakashans were vigorous peoples. happened, they would, with one voice, call out wakash! For instance, they used giant cedar trees to construct wakash! expansive houses for their extended families, to carve The study of Wakashan languages has contributed beautiful totem poles, and to build huge canoes which significantly to the development of linguistics. For they used to fish for halibut and to hunt sea mammals instance, the lasting and profound influence of such ––including colossal humpback and gray whales. But famous linguists as Franz Boas, Edward Sapir, Morris they were almost decimated by the arrival of the Swadesh, and Mary Haas is due in large part to their Europeans. Epidemics (especially smallpox), alcohol, seminal work on these languages. Besides refuting the and firearms used in intertribal wars reduced their popular misconception that native languages are population from tens of thousands before contact (the somehow primitive, their pioneering work spawned 1700s) to a few thousand in 1929. Fortunately, valuable research in linguistic affiliation relating to the Wakashan peoples are now recovering rapidly from ethnohistory of North America, and it also uncovered these historical disasters; for instance, at present, the important structural and semantic phenomena that had Nootka (Nuu-chah-nulth) and the descendants of not been found in more widely studied languages. Kwakw’ala-speaking tribes (Kwakwaka’wakw) num- Well-known contemporary linguists such as Emmon ber almost 8,000 and 6,000, respectively. Bach and Stephen Anderson also credit their fieldwork Unfortunately, Wakashan languages remain in a on Wakashan languages for improving their under- grave state of decline. The Canadian government not standing of language. only prohibited Wakashan peoples’ main social events Wakashan languages fall neatly into two branches, (‘potlatches’) from 1884 to 1951 but also relocated each comprising three languages. The southern branch their children to church-run residential schools from (‘Nootkan’) includes Nootka (Nuu-chah-nulth) and 1888 to 1983. These measures, which were taken delib- Nitinat (Ditidaht) spoken along the west coast of erately to break down the transmission of culture and Vancouver Island, as well as Makah spoken on language, were all too effective. More passively, too, Washington’s Olympic Peninsula. The northern Wakashan languages have become obsolete under the branch (‘Kwakiutlan’) consists of Haisla-Henaksiala influence of English, which has become the primary and Heiltsuk-Oowekyala spoken on the north and cen- language at school and at home in every community tral coasts of British Columbia, and Kwakw’ala spo- thanks in part to the influence of the mass media. Thus, ken on northern Vancouver Island and the adjacent currently there are little more than a couple hundred mainland coast. The genetic relation between the two speakers each of Nootka (Nuu-chah-nulth), branches was discovered by Franz Boas in 1889. Their Kwakw’ala, Haisla-Henaksiala, and Heiltsuk- kinship is most clearly evidenced by the locational lex- Oowekyala. Makah has just 20 to 30 speakers, while ical suffixes (-i ‘indoors’, -as ‘outdoors on the Nitinat (Ditidaht) has less than ten. Recently, however, ground’, -is ‘on the beach’, -a ‘on a rock’, etc.), which communities have sought to counteract the loss of their are of great frequency in all Wakashan languages. ancestral languages by integrating them in schools. To Edward Sapir and Morris Swadesh hypothesized this end, they have adopted standardized orthographies, that Wakashan belongs with Salish, Chimakuan, and compiled dictionaries, and developed multimedia Kutenai under a larger ‘stock’ called Mosan, and with materials on their languages. The long-term effect of Algic under a still larger ‘phylum’ called Algonkin- these efforts on the survival of Wakashan languages is Wakashan (or Almosan). At present, these primeval unknown but hopeful. affiliations are not widely accepted by specialists, Let us now focus on Nootka. This term is actually an most of whom have chosen to recoil from such large- error dating to Cook’s visit. On one popular account, scale classification until the histories of the smaller the legendary captain mistook it as the name of the groupings are better understood. The lexical differ- natives who told him nootka (‘circle about!’) to come ences between Nootkan and Kwakiutlan alone are to their village on the other side of an island. The term such that Swadesh estimated the two to have separat- continues to be used widely in linguistics, but nowa- ed around three millenia ago, a time depth that is con- days the people themselves prefer to be called Nuu- sidered plausible by some (e.g. Jacobsen 1979) but too chah-nulth (‘along the mountains’). There are a dozen brief by others (e.g. Embleton 1985). For both branch- distinct Nuu-chah-nulth tribes, with many different es, dialectal differentiation is greatest on Vancouver dialects. Grammars exist for three of these: the south- Island, which is therefore assumed to be the original ern Tseshaht, the central Ahousaht, and the northern home of the ancient Wakashans. The specialized Ka:’yu:k’t’h’ (formerly Kyuquot). The following is a 754  NOOTKA AND WAKASHAN LANGUAGES brief survey of some structural characteristics of nak tʃ ’apats (big-have canoe) ‘to have a big canoe’, or Nootka that the reader might find strikingly different to a ‘dummy’ root as in  u-nak  i tʃ ’apats (it-have and interesting. big canoe) ‘to have a big canoe’. It can also combine Nootka shares 12 consonant phonemes with with other lexical suffixes as in tupk-nak-maiqt English, and has at least 27 others. Fifteen of these (black-have-want) ‘to want to have a black one’. On involve various kinds of glottalization. For instance, the the other hand, much of Nootka morphology is non- above-mentioned dialect name Tseshaht is pronounced concatenative in that many lexical and aspectual dis- with a postglottalized ‘ejective’ consonant and a glottal tinctions are expressed by reduplicating the root or by stop [ts ’iʃaʔat], while Nuu-chah-nulth is pronounced lenghtening its vowel(s). For instance, from the verb with a preglottalized ‘creaky’ consonant [nutʃanu]. form mitx-ʃit ‘to make a turn’, we get the ITERATIVE Nootka consonants draw on all known places of articu- mitxmitx-ʃit ‘to start in on turns at intervals’, lation except dental. For example, the dialect names the GRADUATIVE mitx-ʃit ‘making a turn’, the REPETI- Ka:’yu:k’t’h’ and Ahousaht involve uvular, pharyngeal, TIVE mitxmitx-ʃit ‘to start turning around and and epiglottal consonants: [qajuk’at, ausat]. around’, and (tenuously) the DISTRIBUTIVE REPETITIVE Nootka’s consonantal inventory would have been even mimitxmitx-ʃit ‘to start turning around and around larger had it not historically abandoned the lateral here and there’. In sum, the Nootka word consists of a approximants [| |] and the voiced obstruents [b d dz d root, which itself may be modified by reduplication or g g G G], which are still in use in other Wakashan lengthening, and to which may be added a large num- languages. In contrast to its manifold use of consonants, ber of suffixes with various lexical and grammatical Nootka has just three basic vowels, in long [i a u] meanings. and short [i a u] varieties, plus two marginal vowels Turning to syntax, it is often obscured in Nootka by [ε, ɔ] found mainly in foreign borrowings. Nootka the complex morphology that allows words to convey lacks schwa [ə], which is widely used in Kwakiutlan. sentences, as in Ka:’yu:k’t’h’ ʔu-ki-tʃip-’itʃ-is-im Nootka phonemes are implicated in many active (it-make-for-IMPV-you(PL)-me-will) ‘you’ll make it for phonological processes, including lenition, glottaliza- me!’, ʔu-sup-intiʃ (it-kill-PST.IND) ‘he killed it’, and tion, assimilation, shortening, and deletion. For muk-sup-intiʃ (deer-kill-PST.IND) ‘he killed the deer’. instance, all these processes are involved in the fol- When individual words are used instead of affixes, the lowing utterance (Tseshaht): /hi-as-uk-ap-ma/; following basic order is apparent: predicate–sub- [hijasuk’ama] ‘I put mine in a container’ (LOC-‘in ject–object, as in qasap-intiʃ tʃ akup muwitʃ (kill- vessel’-POSS-CAUS-LS). Every Nootka syllable consists PST.IND man deer) ‘a man killed a deer’. Interestingly, of a vowel obligatorily preceded by a single conso- any category of word can be a predicate in Nootka, nant, and optionally followed by up to three conso- including adverbs as in ʔati-intiʃ qasap muwitʃ nants, as in [.int.tin.ʔi.] ‘the one made of snot’. (nightly-PST.IND kill deer) ‘it was at night he killed However, in the Ka:’yu:k’t’h dialect, vowels are regu- the deer’, nouns as in tʃ akup-intiʃ qasap muwitʃ larly dropped inside and at the ends of words, such that (man-PST.IND kill deer) ‘it was a man that killed the any number of consonants is possible in sequence; for deer’, and demonstratives as in ʔu-intiʃ tʃ akupi example, [t’ut’uʃinkʃtts’] (…-inuk ‘at hands’ qasap muwitʃ (that-PST.IND man kill deer) ‘it was that + -ʃit mom. + -ts’i ‘at fire’) ‘he was drying his man that killed the deer’. Of particular interest in this hands at the fire’. Incidentally, an extreme form of context is the oft-repeated claim that Nootka lacks cat- consonant sequencing occurs in Oowekyala: this egory distinctions. This claim is controversial, but Kwakiutlan dialect is remarkable in permitting conso- most linguists agree that this kind of distinction is nant-only words such as [t’xt’k’s] ‘fish hawk’ and weak in Nootkan syntax. For example, in Tseshaht even consonant-only utterances such as [tspstkts] quʔas ‘man’ and mamuk ‘work’ act as noun and verb, ‘this (not visible) will be a nice thwart’. Words in respectively, in mamuk-ma quʔas-ʔi ‘the man is Nootka show regular stress patterns (e.g. working’ (work-he man-the), but these grammatical [haja akʃi at] ‘did not know now’), as in other categories appear to be reversed in quʔas-ma mamuk- Wakashan languages (but not all; Haisla has a pitch ʔi ‘the working one is a man’ (man-he work-the). accent system and Heiltsuk is tonal). Nootka morphology is also very complex. On the References one hand, it is polysynthetic in that hundreds of lexi- cal suffixes can combine with well over a thousand Boas, Franz. 1947. Kwakiutl grammar. Philadelphia: American roots to form large stems with composite meanings. Philosophical Society. Cook, Captain James. 1784. A voyage to the Pacific Ocean: For example, a verbal suffix like -nak ‘to have’ can be undertaken by the Command of His Majesty, for making added to a nominal root as in tʃ ’apats-nak (canoe- discoveries in the Northern Hemisphere; performed under have) ‘to have a canoe’, to an adjectival root as in  i- the direction of Captain Cook, Clerke, and Gore, in His 755 NOOTKA AND WAKASHAN LANGUAGES Majesty’s Ships the Resolution and Discovery, in the years Lincoln, Neville J., and John C. Rath. 1980. North Wakashan 1776, 1777, 1778 and 1780. London: W. and A. Strahan. comparative root list. Ottawa: National Museums of Canada. Embleton, Sheila M. 1985. Lexicostatistics applied to the McMillan, Alan D. 1999. Since the time of the transformers: the Germanic, Romance, and Wakashan families Word 36. ancient heritage of the Nuu-chah-nulth, Ditidaht, and 37–60. Makah. Vancouver: University of British Columbia Press. Goddard, Ives (ed.) 1996. Languages. Handbook of North Mithun, Marianne. 2001. The languages of Native North American Indians, Vol. 17. Washington: Smithsonian America. Cambridge, UK: Cambridge University Press. Institution. Nakayama, Toshihide. 1999. Nuu-chah-nulth (Nootka). Haas, Mary R. 1972. The structure of stems and roots in Munich: LINCOM EUROPA. Nootka–Nitinat. International Journal of American Sapir, Edward, and Morris Swadesh. 1939. Nootka texts: tales and Linguistics 38. 83–92. ethnological narratives with grammatical notes and Lexical Jacobsen, William H. Jr. 1979. Wakashan comparative studies. materials. Philadelphia: American Philosophical Society. The languages of North America: historical and comparative Stonham, John T. 1999. Aspects of Tsishaath Nootka phonetics assessment, ed. by Lyle Campbell and Marianne Mithun. and phonology. Munich: LINCOM EUROPA. Austin, TX: University of Texas Press. DARIN HOWE Northwest Caucasian Languages The languages of the Northwest Caucasian fami- proper, the different ethnic groups have generally pre- ly––Abaza, Abkhaz, Adyghe, Kabardian, and served their own languages, although for some Ubykh––were originally spoken in a contiguous Russian has become the dominant language. region stretching from Abkhazia on the Black Sea in Many Caucasian linguists believe that the the south to the Kuban River in the north. As a result Northwest Caucasian family is genetically related to of the relocations following the Russo-Caucasian War the neighboring Northeast Caucasian family, but the (1817–1864), substantial communities of each now linguistic arguments presented in support of this con- exist in Turkey, as well as smaller enclaves in Syria, nection are not convincing, resting on typological sim- Jordan, Israel, former Yugoslavia, Western Europe, ilarity rather than the comparative method. All of the and New Jersey. The languages of this family fall into Northwest Caucasian languages have literary forms, three groups: Abkhaz/Abaza, Ubykh, and Circassian, with the exception of Ubykh, which is now extinct. the latter containing the languages Adyghe and Their current orthographic forms are based on the Kabardian. Cyrillic script, although there is currently movement In 1850, there were approximately one million to develop a new orthography based on the Roman speakers of Northwest Caucasian languages, most of script (most likely its Turkish manifestation). Abkhaz them Circassians. The Russo-Caucasian war and the and Abaza are quite similar to one another and are Russian colonization of the Caucasus drastically considered by some scholars to be dialects of the same changed this situation: many Abkhaz and Circassians language; the same holds for Adyghe with respect to were forced to flee westward to the Ottoman Empire, Kabardian. Ubykh is linguistically intermediate and the Ubykh and Sadz Abkhaz were entirely dis- between these two groups. placed from their homelands near the Kwydypsta The Northwest Caucasian languages are most River just north of present-day Abkhazia. At present, famous for pairing unusually rich consonant invento- some 800,000 Northwest Caucasians still reside in the ries (containing as many as 83 consonants, in Ubykh) Caucasus, most of whom retain their ancestral lan- with unusually small vowel inventories, typically con- guages, as well as Russian. In addition, more than a sisting (in native words) of only a low vowel /a/ and a million Northwest Caucasians now live in the diaspo- non-low vowel //. The richness of the consonant sys- ra, primarily in Turkey. Current estimates of the num- tem derives for the most part from the addition of labi- ber of North Caucasians in Turkey range from one to al and palatal secondary articulations. It appears that six million, depending on the sources used. It is diffi- the enormity and smallness of the consonantal and cult to generate reliable figures, because many vocalic systems, respectively, are related: the ancestral Northwest Caucasians now have complicated ethnic (proto) language most likely had more conventional identities resulting from cultural and linguistic assim- vowel and consonant systems, and subsequently rein- ilation and intermarriage. In the Northwest Caucasus terpreted the [back] and [round] features of the vowels 756  NORTHWEST CAUCASIAN LANGUAGES as resulting from secondary7 articulations on neigh- Turning to the individual languages, Abaza had boring consonants. (Turkish, Arabic, and Persian but 34,800 speakers in the Karachay-Cherkess autonomous not Russian loans underwent this reanalysis as well: region of Russia at the time of the 1989 census; as of cf. Turkish dükkân ‘store’→Abkhaz /a-dwk’yan/ 1995, there were 10,000 more in Turkey, 80 (out of 150 [ɑ´d buk’jɑ´n].) One can see in this example that what members of the ethnic group) in Germany, and about are˘now phonologically consonantal secondary articu- 15 in the United States. There are three main dialects: lations still surface on the vowels in most contexts. All Tapanta, Ashkhar, and Bezshagh. The literary language of the Northwest Caucasian languages distinguish is based on the Tapanta dialect. Abaza is reported to be three types of laryngeal consonants, opposing plain mutual with Abkhaz. voiced, voiceless aspirated, and ejective stops. Many Abkhaz had 101,000 speakers in Abkhazia in 1993, also possess extremely complicated intonational stress as well as 4,000 speakers out of approximately systems, most famously Abkhaz. 15–30,000 ethnic Abkhaz in Turkey, and smaller num- The system of agreement marking in Northwest bers scattered in other countries. Abkhazia was under Caucasian verbs is also uncommonly intricate. Abkhaz, Georgian rule during the Soviet period, but seceded for example, overtly marks verbal agreement with sub- shortly after the dissolution of the Soviet Union in the jects, direct objects, indirect objects, and various other mid-1990s. The continued refusal by foreign nations oblique arguments, as in the following Abzhuy exam- to recognize Abkhazia’s sovereign status, together ple from George Hewitt (transcribed in the IPA): with the forced immigration into Abkhazia of Slavs, Armenians, and Mingrelians during the 1930s, and sɑɾɑ´ ɑ-pɥ ´s ɑ-sɑp’´n Georgia’s continuing attempts to eliminate its ethnic I the-woman the-soap minorities, have placed the Abkhaz language in danger s-χɑɾph (∅-)ɑ-lɑ-l s´-ɾ-dɥ dɥɑ´-(∅-)jt’ of disappearing. The assignment of literary status to my-shirt (it-)it-by her-I-cause-wash-(PAST-)FINITE Abkhaz (including TV broadcasts since 1978) has helped counteract this tendency within Abkhazia, but ‘I got the woman to wash my shirt with (the) soap’ few members of the younger generation are learning It is possible to string together as many as nine the language in the diaspora. There are three main overt morphemes in a row in a sequence of verbal pre- dialect groups: Bzyp (north of Sukhumi), Abzhui, the fixes. literary dialect (south of Sukhumi), and Sadz (now The nominal morphology, on the other hand, is rel- spoken only by a handful of Abkhaz in Turkey). atively simple, containing only two cases (direct The Circassians live in the Adyghe, Kabardino- object and oblique or adverbial). Possession is Balkar, and Karachay-Cherkess republics in the expressed via prefixes, as in Abkhaz s-nap’-kwɑ´ ‘my Russian Federation. The Circassians have two main hands’ (I-hand-nonhuman.plural). Abkhaz (and languages, Adyghe (West Circassian) and Kabardian Abaza) distinguish human vs. nonhuman in certain (East Circassian), the former of which has four main morphological classes, as can be seen in the plural in dialects (Shapsugh, Bzhedug, Temirgoy (the literary this example. All of the Northwest Caucasian lan- dialect), and Abzakh) and the latter of which has six guages distinguish two genders, two numbers, and (Besleney, Malka, Bakhsan, Terek, Mozdok, and three persons in their system of agreement markers. Lesser Kabardian). As of 1993, there were some Finding lexical items shared by all of the Northwest 125,000 Adyghe and 46,000 Kabardians in Russia, Caucasian languages is difficult, but not impossible: 100,000 Adyghe and 202,000 Kabardians in Turkey, and a total of approximately 750,000 Circassians in Proto- Common Ubykh Common the world as a whole. Both languages have literary sta- Northwest Circassian Abkhaz tus and are taught in the schools in their respective Caucasian republics, but their survival in the diaspora is less *pxja ‘back’ pxa pšja *(p)χja secure following the closing of the Circassian school *bza ‘tongue’ *bza bzja *bzə in Turkey by Atatürk in 1920 and the outlawing of *zwja ‘cook, boil’ *zwja zwja *zwjə publication in minority languages in Turkey in 1983. *gwjə ‘heart’ *gwə gj ə *gwə The Ubykh community originally lived on the Black Sea just north of Abkhazia, but all 50,000 of its The Northwest Caucasian numeral system is members were deported to the Ottoman Empire fol- interesting in that it preserves vestiges of an orig- lowing the Russian conquest of the Caucasus, ending inal vigesimal system in the numbers 30–99. For up in Hac Osman, a town near Istanbul. Although the instance, Abkhaz (Cwyzhy dialect) eɥzɥejzɥɑbɑ ‘thir- community retains a distinct identity, its language is ty’ is literally ‘twenty (eɥzɥɑ) and (-j) ten (zɥɑ-)’; now extinct. Isolated individuals can still be found ɥneɥzɥɑ ‘forty’ is ‘two twenties’, and so on. who know pieces of the grammar and vocabulary, but 757 NORTHWEST CAUCASIAN LANGUAGES most of the remaining Ubykhs speak Adyghe and/or Hewitt, George. 1979. Abkhaz. Amsterdam: North-Holland. Turkish. Hewitt, George (ed.) 1989. The indigenous languages of the Caucasus, Vol. 2: The North West Caucasian languages. Delmar, NY: Caravan Books. Kuipers, Aert. 1960. Phoneme and morpheme in Kabardian. References The Hague: Mouton. Allen, Sidney. 1956. Structure and system in the Abaza verbal Nikolaev, S., and S. Starostin. 1994. A North Caucasian etymo- complex. Transactions of the Philological Society. 127–76. logical dictionary. Moscow: Asterisk Publishers. Chirikba, Viacheslav. 1996. Common West Caucasian. Leiden: Paris, Catherine. 1974. Système phonologique et phénomènes CNWS Publications. phonétiques dans le parler besney de Zennen Köyü. Paris: Colarusso, John. 1988. The Northwest Caucasian languages: a Klincksieck. phonological survey. New York: Garland Press. Smeets, Rieks. 1984. Studies in West Circassian phonology and Colarusso, John. 1992. A grammar of the Kabardian language. morphology. Leiden: Hakuchi Press. Calgary: University of Calgary Press. Sommerfelt, Alf. 1934. Études comparatives sur le caucasique Dumézil, Georges. 1931. La langue des oubykhs. Paris: du Nord-ouest. Norsk Tidsskrift for Sprogvidenskap 7. Champion. 178–210. Dumézil, Georges. 1932. Études comparatives sur les langues von Mészáros, Julius. 1934. Die Päkhy-Sprache. Chicago: caucasiennes du nord-ouest (Morphologie). Paris: Adrien- University of Chicago Press. Maisonneuve. BERT VAUX Noun Incorporation The term Noun Incorporation refers to the linguistic (Examples (a) and (b) are taken from Baker phenomenon according to which a noun and a verb 1988:76.) may combine into a single word. The two elements Incorporating languages thus have two ways of form one morphological unit that is often described as expressing the verb and its object: either the two ele- a complex verb (or as a compound). Usually, the incor- ments are realized as a single word after incorporation porated noun is interpreted as the object of the verb. of the noun into the verb, as shown in example (a), or For example, in Onondaga, an American Indian lan- they appear as two separate words. In the latter case, guage of the Iroquian language family, the noun (stem) the noun does not incorporate into the verb, but it fol- hwist ‘money’, which is in bold below, may appear as lows the verb, as shown in example (b). The two con- part of the verb (stem) ahtu ‘lose’. In particular, the structions may differ slightly in their interpretation. noun (stem) immediately precedes the verb (stem), as For example, Incorporation sometimes has the func- we can see in the following example: tion of narrowing the semantic range of the verb. This (a). Pet wa?-ha-hwist-ahtu-?t-a? implies that the speakers of incorporating languages Pat he-money-lost chose between the nonanalytic configuration (where ‘Pat lost money’ the object is incorporated into the verb) and the ana- lytic configuration (where the object is not incorporat- In this case, the noun (stem) and the verb (stem) ed into the verb) on the basis of the particulars of the form a single word, namely the complex verb wa?-ha- information they want to communicate. hwist-ahtu-?t-a? ‘he-money-lost’. The incorporated Noun Incorporation is a productive process mainly noun (stem) hwist ‘money’ is interpreted as the object in non-Indo-European languages. It is attested in some of the verb ahtu ‘lose’. North American languages (Athapaskan, Caddoan, In the same language, the object of the verb can be Chimariko, Iroquian, Siouan, Kiowa-Tanoan, Natchez, an independent word in the sentence. In this case, the Takelma, Uto-Aztecan, Yana, Zuni), in Australian lan- object is one word: ne?o-hwist-a? ‘the money’, while guages (Rembarnga), in Paleo-Siberian languages the verb is another word, namely, wa?-ha-htu-?t-a? (Chukchee), in Oceanic languages, in Turkish, and in ‘he-lost’. The noun now follows the verb, as illustrat- other languages. ed below: There has been a long debate in the literature of (b) Pet wa?-ha-htu-?t-a? ne?o-hwist-a? Generative Grammar regarding Noun Incorporation. Pat he-lost the-money The debate concerns the locus where these complex ‘Pat lost the money’ words are formed. On the one hand, it has been argued 758  NOUN INCORPORATION that Noun Incorporation is an instance of compounding truck, which is the object of the verb, cannot become in the Lexicon (Mithun 1984; Di Sciullo and Williams part of the verb drive, but rather the two elements have 1987; Rosen 1989). Compounding refers to the combi- to remain two separate words. The noun truck can be nation of two (or more) stems or words into one single combined with the noun driver, giving rise to a complex word. For example, the English word bedroom is a noun: truck driver. However, it can be argued that these compound consisting of two elements: bed and room. cases are different from genuine instances of Noun Hence, it has been argued that Noun Incorporation is a Incorporation in that the derived word is a noun rather case of a noun and a verb that are put together in the than a verb. There are some complex words in English same way that bed and room are combined into bed- that resemble cases of Noun Incorporation. It could be room. According to this view, the noun–verb complex argued that the formation of verbs like baby sit possibly is not related structurally to its analytic counterpart. involves an operation of syntactic movement. However, That is, there is no association between the noun–verb such examples lack a parallel analytic configuration. compound and the structure that involves the verb and That is, to sit a baby cannot be considered as the paral- its object, despite the fact that the two configurations lel analytic configuration of the complex verb to baby may have very similar interpretation. The noun and the sit. The latter has a very unexpected meaning. It verb are put together morphologically and behave as expresses the notion ‘to look after a baby’. If baby sit is one syntactic unit. This hypothesis is usually known as not an instance of Noun Incorporation, it could be the lexicalist hypothesis. viewed as an instance of compounding. This would Baker (1988), on the other hand, was the first to mean that the verb baby sit is derived by a lexical or a argue that Noun Incorporation is a syntactic phenom- morphological process and not by a syntactic process. enon. His main idea is that Noun Incorporation is a Lastly, it has been argued that Incorporation is a syntactic process that derives complex verbs from sen- widely spread phenomenon that is not necessarily tences where the verb takes a noun as its object. restricted to nouns. It is a process that allows any two Specifically, Baker argues that the incorporated noun elements to combine into a single word. One of the starts out as an independent word, in a position fol- two parts of the derived word is usually a verb. It has lowing the verb. It ends up in a position preceding the been proposed, for example, that certain kinds of pro- verb after an operation of syntactic movement. That is, nouns incorporate into the verb. These are called clitic syntactic movement changes the position of the noun pronouns and are set apart from other pronouns by cer- and places it within the verbal complex. An empty slot tain special characteristics. While ordinary pronouns is left behind, in the original postverbal position of the do not appear as parts of the verb, clitic pronouns may noun, which is called the trace of the moved element. do so. This process is often named cliticization. The Baker provides a straightforward explanation of the process of cliticization can be distinguished, though, fact that the incorporated noun is interpreted as the from Noun Incorporation in that the former possibly object of the verb. It starts out as the complement of the involves incorporation of a functional element (a clitic verb, i.e. the two elements are adjacent and in a specif- pronoun) whereas the latter involves incorporation of ic order: the noun follows the verb. This is considered as a lexical element (a noun). the structural position where the internal theta-role of the verb is assigned (following Baker’s Universal Theta- role Assignment Hypothesis––the UTAH––according References to which each thematic role of the verb is assigned in a Baker, Mark. 1988. Incorporation. Chicago: University of specific syntactic configuration). That is, any element Chicago Press. Baker, Mark. 1996. The polysynthesis parameter. Oxford: that is base-generated in that particular syntactic posi- Oxford University Press. tion is interpreted as the direct object of the verb, name- Di Sciullo, Anna Maria, and Edwin Williams. 1987. On the def- ly as the theme (i.e. the argument that is affected by the inition of word. Cambridge, MA: MIT Press. action described by the verb). Hence, if the incorporat- Mithun, Marianne. 1999. The languages of native North ed noun starts out as the complement of the verb, it is America. Cambridge: Cambridge University Press. Mithun, Marianne. 1984. The evolution of Noun Incorporation. assigned the internal theta-role of the verb and is thus Language 60. 847–94. interpreted as the object of the verb, even after incorpo- Rosen, Sara Thomas. 1989. Two types of Noun Incorporation: rating (by syntactic movement) into the verb. a lexical analysis. Language 65. 294–317. Incorporation is predicted to be possible with any noun Spencer, Andrew. 1991. Morphological theory. Oxford: that starts out from the verbal complement position. Blackwell. Spencer, Andrew, and Arnold Zwicky (eds.) 1998. The hand- This is actually true: Noun Incorporation is a productive book of morphology. Oxford: Blackwell Publishers. process in incorporating languages. DIMITRA PAPANGELI In nonincorporating languages, like English, the object cannot form a single word with the verb. Words See also Generative Grammar; Mohawk and Iro- like truck drive are not in use. This means that the noun quoian Languages 759 NUMBER MARKING Number Marking Number is the morphological category that expresses Both inflection and derivation are traditionally contrasts involving countable quantities. The most explained exclusively in terms of affixation. As widespread number opposition is likely that between regards inflection, affixation consists in the addition of singular (one) and plural (more than one). Several lan- an inflectional affix (an obligatorily bound morph that guages, however, make further distinctions. Some of contains grammatical information) to another morph them (e.g. Greek or Sanskrit) exhibit dual forms that with lexical content, as in house-s. This approach to apply to two individuals, and a few languages (e.g. inflection, however, is said to face important difficul- some Southwest Pacific languages) additionally dis- ties. One of them is illustrated by irregular English tinguish a trial form that refers to three individuals. In plurals such as men, in which plurality is marked not some systems, there are also paucal forms for indicat- by the addition of an allomorph, but rather by means ing a small number, as in Arabic. of a process that causes a vocalic change. The so- Although number implies a basic linguistic con- called portmanteau morphs (morphs that realize more trast, languages deal with this category in different than one morpheme, as the ending -us in the Latin ways. Most languages mark number in nouns; yet, noun annus ‘year’, which indicates simultaneously there are languages (e.g. Maori, a Polynesian lan- singular number and nominative case) are also prob- guage) in which it is signaled by determiners, and oth- lematic, since this approach assumes a one-to-one cor- ers that lack number distinctions (e.g. Nancowry, respondence between form and meaning. In view of spoken in India’s Nicobarese Islands). The explicit such shortcomings, recent morphological theories of expression of number, on the other hand, may be inflection have abandoned this classical treatment and obligatory, as in most English nouns; limited to certain have adopted a model in which words are modified nominal categories, as in most Algonquian languages by means of different morphological processes that (spoken by American Indians); or optional, as in apply to the word itself, which is considered the base Halkomelen (another American Indian language). or stem. Number marking differs greatly among languages, The most simple device to express number is, as in depending on the morphological system they display. traditional accounts, affixation. Affixes can be divided In languages traditionally classified as analytic, such into three main classes: prefixes, which attach to the as Chinese, words consist of only one syllable and are front of the stem; suffixes, which come after the stem; invariable. Thus, plural nouns do not exhibit any for- and infixes, which occur within the stem. Most Indo- mal marker and the notion of plurality must be European languages make use of suffixes, such as the inferred from the context. Only when this information English or Spanish regular plural ending -s, to express is not contextually available or the idea of plurality is number contrasts. Number marking by means of pre- to be emphasized, separate words are added: ki fixation can be found in Bantu languages such as ‘some’, šu ‘number’. Synthetic languages, on the con- Kikuyu (mu˜-ru˜u˜thi ‘lion’, pl. m˜- ι ru˜u˜thi), whereas the trary, rely on inflection, i.e. the modification of a Uto-Aztecan language Oaxaca Chontal illustrates the word’s form, to express number distinctions. use of infixes to distinguish number (kwepoʔ ‘lizard’, Modifications of words may equally result in deriva- pl. kwe-B-poʔ ). tion, which involves the creation of new words with Other processes by which number oppositions may new meanings. In traditional morphological theory, be expressed involve the modification of the stem. which has its origins in American structuralism, This shift may affect the quantity or quality of the the analysis of word structure was based on the internal vowel, as in the English pair foot/feet. Vowel morpheme—the smallest meaningful unit of language. change is generally referred to as ‘ablaut’, although, Morphemes are, however, abstract elements that are because of its historical origin, the alternation just physically realized by morphs. The word houses, for mentioned receives the distinct name ‘umlaut’. Not example, consists of two morphs: house and -s. It is only vowels but also consonants may be modified to very frequent that a particular morpheme is represent- mark number. In the West African language, Fula, the ed by different morphs in different environments. For plural of the word yiite ‘fire’, is formed by changing instance, the plural morpheme in the English noun the initial consonant, giite. Such consonant alterna- oxen is not realized by the regular morph -s, but by the tions are frequently the result of a phonological irregular -en. These alternants are called allomorphs. change induced by an affix, which is then lost. 760  NUMBER MARKING In many languages, information about number is device: in Navajo, the form for ‘one is standing’ is dif- conveyed by means of modifications of the supraseg- ferent from that for ‘several are standing’. mental features tone and stress. In Somali (an East African language), for example, number is indicated References by a change in the tone pattern: the singular form èy Anderson, Stephen R. 1985. Inflectional morphology. ‘dog’ takes a falling tone, whereas the plural éy takes Language typology and syntactic description, Vol. III, ed. by a high tone. Russian illustrates the use of stress shift to Timothy Shopen. Cambridge: Cambridge University Press. distinguish singular from plural: in a particular group Anderson, Stephen R. 1988. Morphological theory. Linguistics: of neuter nouns, the singular is ending-stressed, while the Cambridge survey, Vol. I, ed. by Frederick J. Newmeyer. the plural is stem-stressed: oknó ‘window’ (nominative Cambridge: Cambridge University Press. Anderson, Stephen R. 1992. A-morphous morphology. singular), ókna (nominative plural). Cambridge: Cambridge University Press. Another common morphological process in certain Bauer, Laurie. 1988. Introducing linguistic morphology. languages is reduplication, which consists in the Edinburgh: Edinburgh University Press. copying of the whole stem––total reduplication––or Bloomfield, Leonard. 1933. Language. New York: Holt, only of a part of it––partial reduplication, typically Rinehart, and Winston; Chicago and London: The University of Chicago Press, 1984. the leftmost portion. Reduplication is said to be fre- Jespersen, Otto. 1922. Language. Its nature, development and quently used iconally, i.e. the form of the word origin. London: George Allen & Unwind Ltd. reflects its meaning. For this reason, reduplication is Matthews, P. H. 1972. Inflectional morphology. A theoretical a common marker of plurality. Total reduplication to study based on aspects of Latin verb conjugation. Cambridge: form the plural is found in Indonesian (babi ‘pig’, pl. Cambridge University Press. Matthews, P. H. 1974. Morphology. An introduction to the babibabi), whereas partial reduplication is used in theory of word-structure. Cambridge: Cambridge University Motu, a language of Papua New Guinea (tau ‘man’, Press. pl. tatau). O’Grady, William, Michael Dobrovolsky, and Francis Katamba The last morphological operation associated with (eds.) 1987. Contemporary linguistics. An introduction. number marking is suppletion. This process replaces London and New York: Longman; 3rd edition, 1996. Spencer, Andrew, and Arnold M. Zwicky. 1998. The handbook one form by a phonologically unrelated form. A clear of morphology. Oxford, Malden, MA: Blackwell. example of this phenomenon is the use of went as the LUISA GONZÁLEZ-ROMERO past tense of go. Suppletion appears in many Western North American languages as a number marking See also Affixation; Inflection and Derivation 761  O Official Language Selection Official language selection may be simply defined as Pakistan, where Urdu was declared the official lan- the political choice of which language, or languages, guage on independence from Britain in 1947. to use in the legislative, executive, and judicial busi- However, because Urdu was not an indigenous lan- ness of government. Given that all nations in the twen- guage of East Pakistan, the overwhelmingly Bengali- ty-first century comprise populations speaking speaking majority there felt marginalized; the different languages, the selection of official languages unrequited demand for language rights led to civil war, is much more complex than the definition. This is after which Bangladesh achieved independence from because granting official status to a particular lan- Pakistan in 1971. Therefore, some newly independent guage will enhance its prestige, extend its use to edu- nations select the language of the previous imperial cational and nonofficial domains, privilege its power; this was the pattern among most ex-colonies of speakers, and impinge on the linguistic rights of Britain in the 1960s and 1970s, such as Ghana, speakers of other languages within the community. Nigeria, and Kenya. Although this choice means that a Both newly independent and well-established coun- new nation may relatively easily build on the existing tries face difficulties when selecting official lan- communications infrastructure, the choice may be guages, although the contributing factors differ. inappropriate. There is often a residue of ill-feeling Examples of each will now be considered. toward the culture and language of the erstwhile Countries that are colonized or subjected to imperi- rulers; such a decision may also leave in positions of alism in one form or another usually have the language influence those who collaborated with the colonial of the colonial state imposed on them. Upon reaching power and who may be out of sympathy with a new independence, the selection of the official language is regime. Because of this, some countries adopt another a matter of considerable importance––and controver- international language. The example of East Timor, sy––because there are often several contending lan- which became independent in 2000, is a case in point. guages. One of them may be the first language of a None of the indigenous languages was considered majority of, or dominant group within, the population, appropriate as the official language. The language of and this would seem the obvious choice for nation the erstwhile dominant power––Indonesia––was con- building. This was the case with the nations that re sidered unacceptable on political and cultural grounds. emerged after the demise of the Soviet Union in the Portuguese––the language of the earlier colonizer––was 1990s (Ukraine, Latvia, Belarus, and so on). However, eventually chosen because it was widely understood in such cases, speakers of other indigenous languages and was used both within East Timor and internation- may feel that they are regarded as second-class citi- ally. Since 1994, South Africa has presented a very zens for whom access to public services is restricted, different example of language policy––one of national if not denied completely, by the lack of official recog- pluralism. At the national level, 11 indigenous nition of their mother tongue. This situation arose in languages have official status, and each provincial 763 OFFICIAL LANGUAGE SELECTION legislature can determine which of the languages will United States have been unsuccessful, although since be used for internal official purposes. the 1980s many state legislatures have passed laws pro- Longer established nations may also have difficulty claiming, with varying degrees of specificity, English as in selecting official languages, as examples from major the official language within their own borders. English-speaking countries illustrate. For many years, Language is not merely a means of communication: the francophone population of Canada struggled for it is the most important cultural symbol of a commu- legal recognition of their language until eventually the nity. For many countries––Japan, Somalia, Malta, and 1969 Official Languages Act declared that English and many others––the national language is a core value French had equality of status with equal rights and priv- that bonds the people together and indeed defines that ileges. The case of the United Kingdom is different very community. Increasingly, however, all nation because although English is the national language, it has states are host to diverse linguistic minority communi- no official status: the only legally recognized official ties in their midst, whether this is caused by temporary language in Great Britain is, since 1967, Welsh. business activity, permanent migration, urgent asylum- Similarly, since 1987, the only official language seeking, or the revitalization of indigenous languages. throughout New Zealand is Maori. The reason for this It is, therefore, necessary for national policy-makers to apparent legal anomaly is that the indigenous speakers consider the extent to which multilingualism (and of those two minority ethnic groups campaigned to have multiculturalism) presents either a threat or a resource their languages officially sanctioned for use in public to their communities, and thereafter to select official services such as the legal system and education, health, languages and frame national language policies in the and social services. With regard to English, there was no light of this understanding. need to formalize such rights because they existed in fact, although not in law. Some English-speaking people in these countries opposed the granting of prestige and References rights to these minority languages; they also opposed the Bourdieu, P. 1991. Language and symbolic power. Cambridge: allocation of public resources for their development and Harvard University Press. diffusion. None of these countries has an explicit over- Eggington, William, and Helen Wren. 1997. Language policy: dominant English, pluralist challenges. Amsterdam, all national policy on languages, and there is consider- Philadelphia: John Benjamins Publishing Company. able scope for uncertainty and even anxiety regarding Fairclough, N. 1989. Language and power. London: Longman. the language rights of speakers of nonofficial lan- Herriman, Michael, and Barbara Burnaby (eds.) 1996. guages:for example, immigrants and minority ethnic Language policies in English-dominant countries. communities in all three countries and the indigenous Clevedon, Philadelphia, Adelaide: Multilingual Matters Ltd. Phillipson, Robert. 1992. Linguistic imperialism. Oxford: First Nation peoples of Canada. By contrast, Australia Oxford University Press. has been developing a comprehensive National Schiffman, H.F. 1995. Linguistic culture and language policy. Languages Policy since 1987; although this now London: Routledge. declares that English is the official language of Tollefson, James, W. 1991. Planning language, planning Australia, it recognizes the aboriginal languages and inequality. London, New York. Longman. Varennes, Fernand de. 1996. Language, minorities and human also specifies the provision of services in languages rights. The Hague: M. Nijhoff. other than English. The United States has never selected ROGER BARNARD an official language, despite the large number of indige- nous, colonial, and immigrant language communities. See also Australia; Canada; Language Planning; Recent attempts to change the federal constitution to Nigeria; South Africa; Soviet Union: Successor adopt English as the official language throughout the States Ojibwe and Algonquian Languages Ojibwe (sometimes Ojibwa, Chippewa) is a Native Although widely dispersed, these languages share American language and belongs to the Algonquian many of the same grammatical features. Other mem- family, one of the largest of precontact North America. bers in this group include Blackfoot, Cheyenne and 764  OJIBWE AND ALGONQUIAN LANGUAGES Arapaho, Cree, Potawotami, Menominee, Fox, As a spoken language, Ojibwe is fairly robust, Illinois, Shawnee, Micmac, Passamaquoddy, and whereas several other Algonquian languages have Abenaki. declined or died out altogether. Most speakers live on Algonquian speakers have inhabited North America reservations in Minnesota, Wisconsin, and Michigan for thousands of years. Their homeland was suggested (United States), as well as in the province of Ontario to be somewhere in the region of Lake Huron or (Canada). One recent estimate puts the total number of Georgian Bay, but given the ease of movement Ojibwe speakers at 51,000, but most other estimates are throughout the eastern woodlands area, most lower. In any case, only a small percentage represents researchers are noncommittal in this regard. At the unilingual Ojibwe speakers. Many efforts are under time of European contact, Algonquian languages were way to preserve and/or revive the language. These usu- spoken along the eastern seaboard north of Virginia, as ally take the form of bilingual education programs in well as to the east of the lower Mississippi River. the schools, language centers for adults, or the estab- Today they can be found from the Atlantic Ocean to lishment of institutional or personal websites. Recently the Rocky Mountains along what is now the United published word lists and dictionaries abound, and States/Canadian border. Place names such as Ojibwe has received a relatively good deal of attention Connecticut, Manitoba, Massachusetts, Michigan, from linguists. Otherwise, the model of loss is similar Minnesota, Mississippi, Ontario, Saskatchewan, and to endangered languages worldwide: younger speakers Wisconsin are all Algonquian in origin, as are the are heavily influenced by the dominant culture (in this words moose, chipmunk, and toboggan. case, Anglo-America) and use their first language pri- Linguists often speak of Algonquian in terms of marily in family or tribal situations. Borrowed speech regional subdivisions, but only Eastern Algonquian habits continually erode native sound and sentence pat- constitutes a family in the true sense; i.e. it can be terns until only the elders speak the language fluently. shown that these languages have a common ancestor. The sounds of Ojibwe are typical of other Members of this group include Abenaki (spoken in Algonquian languages. There are five vowels of dif- Québec and Maine), Passamaquoddy-Maliseet ferent quality, and vowel length (indicated by a colon) (Maine, New Brunswick), and Micmac (Nova Scotia, distinguishes meaning. The vowels are [i, e, (schwa), New Brunswick). Arapaho, Cheyenne (Wyoming), i:, a:, and o:]. The consonants are [p, b, g, k, m, n, ch, and Blackfoot (Montana, Alberta) fall into the western j, w, y ,h, and ʔ]. Note the absence of [t, d]. group. ‘Central Algonquian’—of which Ojibwe is a Ojibwe (henceforth Algonquian) is a polysynthetic member—is spoken around the area of the Great language, a term that carries with it many implica- Lakes. Ottawa (Odawa), Algonquin, and Salteaux (all tions: single, complex words typically express the spoken in Canada) are essentially the same as Ojibwe; meaning of whole sentences, pronouns are ‘under- Cree, Fox, and Shawnee are close relatives. Divisions stood’ rather than overt, and expressions such as ‘all’ along geographical lines are not completely arbitrary: or ‘every’ may be separated from the nouns they mod- members of each region share certain sounds, vocabu- ify by other words. In addition, polysynthetic lan- lary, and sentence patterns not found in others. The guages are rich in agreement: subjects, direct objects, differences between one language and the next shift and even indirect objects cause the verb to inflect for gradually from west to east, with western languages person, number, and gender—and sometimes case. being the oldest. Noun phrases, meanwhile, do not appear in fixed posi- Algonquian was one of the first language families tions within the sentence, as in English; instead they to be reconstructed by using methods developed in the are considered ‘appositional’ (add-ons), only loosely nineteenth century for Indo-European. By comparing linked to the verbs. Many meanings that are conveyed words in various languages, the American structuralist in full phrases in other languages are expressed via Leonard Bloomfield was able to determine the forms particles that attach to the verb. of ‘proto’ Algonquian, at the same time making signif- Verbs in Ojibwe (and all Algonquian languages) icant contributions to our understanding of linguistic agree with subjects as well as direct objects, some- change itself. Since then, many other researchers have times in very intricate ways. In the sentence [ni-ki:- followed in Bloomfield’s footsteps, to the point where wa:pem-a:-n] kene:peko:n ‘[I saw] the snake’, the Algonquian is often looked on as a textbook case of prefix ni- indicates that the subject is first person, structuralist analysis. Outside of taxonomy and analy- whereas the absence of a plural marker ensures that it ses of sound change, not much systematic work has is singular. The third-person feature of kene:peko:n been done on Algonquian. Compared with other lan- ‘snake’ is encoded by the suffix -a:, whereas -n signals guages, our understanding of word and sentence struc- that the (singular) object is different from the subject ture is incomplete, and research on meaning and (cf. obviation below). Because of such rich agreement, discourse is still in its infancy. there is no need for separate pronouns other than 765 OJIBWE AND ALGONQUIAN LANGUAGES emphatic ones. In the absence of an object, this sen- The seemingly many names for snow in Eskimo tence would simply mean ‘I saw him/her/it.’ When (Yup’ik, Inuktitut) is a classic example. Similarly, plural morphemes are added to the verb, agreement is Algonquian languages (like French) seem to contain ‘skewed’, yielding the order an inordinate number of verbal paradigms. The differ- ence, for example, between ‘if he comes’ and ‘when he person prefix - person - number - number comes’ depends on the change or deletion of a single (verb stem) suffix suffix suffix vowel. The third major order in Algonquian is the SUBJECT OBJECT SUBJECT OBJECT imperative. The important thing to realize is that In other words, subject as well as object agreement speakers do not memorize these myriad forms, but is split between two nonadjacent positions. Various rather freely produce and understand them by internal- linguistic theories have attempted to deal with facts izing a finite set of rules. like these, but so far very few have succeeded in pro- Nouns in Ojibwe (Algonquian) are classified on the viding a natural account. Finally, nouns inflect for per- basis of animacy, roughly ‘being like a conscious liv- son, number, and obviation too, as in ni-gwis-ag [first ing thing’. As one might expect, humans and animals person-son-plural] ‘my sons’. Note that often the same are [animate], whereas rocks and similar types of affixes involved in verbal agreement also show up in objects are [animate]. The division is not always this context. based on clear-cut properties, however. For example, A large class of elements called preverbs appear in trees are [animate], along with containers used for fixed positions before the verb stem. These typically liquid (cups, spoons, pens, etc.). In Passamaquoddy modify the state or action denoted by verb, much like (Eastern Algonquian), the words for rope, milk, and modifier phrases do in English. Examples from fallen snow are all [animate], as are those for finger- Ojibwe are ojaanimi ‘busy, noisy’, wani ‘mistakenly’, nail and knee. Other body parts (heart, tongue) are and aano ‘in vain, without result’. Some preverbs are [animate], however. The animacy of a given noun is themselves verbal in character, such as giizhii ‘finish, overtly indicated by the verb-stem final, but only get through doing’, gwiinawii ‘don’t know, not able under certain grammatical conditions. to’, and goji ‘try, attempt’. Still others pertain to the Ojibwe speakers have a complex system of referring way that a speaker perceives an event: bi ‘here, toward to things or people in a sentence or discourse. English the speaker’, madwe ‘audible from a distance, can be normally uses pronouns like ‘he’ or phrases like ‘that heard’, ani/ni ‘going away from the speaker, guy’ to refer to someone after first introducing him by approaching the time of doing something’. In short, name. In Ojibwe, however, pronouns are not expressed preverbs have different functions that parallel phrasal overtly: one usually infers their presence through sub- expressions in nonpolysynthetic languages. Although ject/object agreement on the verb or possessor agree- it is possible for several preverbs to appear in a string, ment on the noun. The ‘obviation’ system guarantees their order is not random. that certain noun phrases (inaudible pronouns as well One of the characteristics of a polysynthetic lan- as audible noun phrases) are not confused with others. guage is the richness of word formation processes. In Within a complex noun phrase like niw wday-an ‘his Ojibwe/Algonquian, many words are made up of mul- dog’, for example, the possessive pronoun ‘his’ (which tiple roots, not unlike English compounds, e.g. is only implied) would not be indicated by inflection, ‘killdeer’ (a kind of bird), ‘snow removal’, etc. Roots whereas niw ‘dog’ is marked as obviative (the suffix - can be divided into initials, medials, and finals, an). This ensures that ‘his’ and ‘dog’ do not refer to the depending on their position within the word. Initials same entity. The second domain of obviation is the carry most of the meaning, often describing processes clause. This can be seen in uki:-necci:we’a:-n [eniw or resulting states. Medials are usually nounlike, kwi:wesse:ns-an] ‘He scolded [that boy]’, where the denoting body parts or instruments, and finals typical- subject is unmarked and the object triggers obviation ly indicate simple states or actions. The complex (underlined) on the verb as well as the noun phrase. As Passamaquoddy stem napici koli-hqeh-mon ‘She or he in possessed noun phrases, obviation within a clause is sticks it onto something’ could be translated as ‘onto obligatory. In the third domain of obviation, subjects of stuck-surface-make’ [preverb initial-medial-final]. main clauses are unmarked, but those in subordinate Final roots in Ojibwe/Algonquian also specify the clauses are obviative (underlined): gender class of intransitive subjects and transitive gii-boonii-w dash maa dVdibew [mitigoonsikaa-ini-g] objects (cf. the discussion of gender below). A com- ‘Then she landed on the shore [where there were bushes]’ mon pitfall in understanding polysynthetic languages has been to ascribe variations in word form to the Further restrictions determine whether subjects and speaker’s viewpoint of the world, when in fact they objects are overtly expressed or implied. Although it is probably result from regular grammatical processes. possible to say ‘I wrote them’ or ‘You hurt me’, for 766  OKANAGAN AND SALISHAN LANGUAGES instance, the opposite ‘They wrote me’ or ‘I hurt you’ Fund Publications in Anthropology, Vol. 6. New York: cannot be expressed without changing the affixes of Viking. ———.1957. Eastern Ojibwa. Ann Arbor: University of the verb. In the first two sentences, a ‘direct’ mor- Michigan Press. pheme is attached to the verb stem, and in the latter Campana, Mark. 1996. The conjunct order in Algonquian. two an ‘inverse’ one. The choice of direct or inverse is Canadian Journal of Linguistics 41(3). 201–34. determined by a ‘participant hierarchy’ that ranks *Campbell, Lyle. 1997. American Indian languages: the histor- noun phrases according to person and obviation. In ical linguistics of Native America. New York: Oxford University Press. direct sentences, subjects outrank objects on the scale: *Dahlstrom, Amy. 1991. Plains Cree morphosyntax. New York: 2 > 1 > 3 > 3—that is, second persons (‘you’) are Garland. ranked higher than first persons (‘I/we’), which in turn Goddard, Ives. 1978. Central Algonquian languages. Handbook are higher than third persons (‘she or he’). of North American Indians 15. 583–7. Nonobviative third persons (3) are ranked higher than ———.1990. Primary and secondary stem derivation in Algonquian. International Journal of American Linguistics obviative (3) ones. In inverse sentences, exactly the 56(4). 449–83. opposite holds: objects must outrank subjects on the ———. 1994. The east-west cline in Algonquian dialectology. same scale. Strikingly, subjects in the inverse appear Papers of the 25th Algonquian Conference, ed. by W. as objects in the direct, and vice versa: Cowan, Carleton University, 187–211. Ottawa. Hockett, Charles. 1966. What Algonquian is really like. Direct Inverse International Journal of American Linguistics 32. 59–73. ni-wa:pam-a:-k ni-wa:pam-ik-o:k Leavitt, Robert. 1992. Lexical exploration and educational 1-see-third person/direct-pl 1-see-third person/ insight. Papers of the 23rd Algonquian Conference, ed. by W. Cowan. Carleton University, Ottawa. inverse-pl ———. 1996. Passmaquoddy-Maliseet. Languages of the ‘I see them’ ‘They see me’ world/materials, Vol. 27. Newcastle: Lincom Europa. Lyovin, Anatole. 1997. An introduction to the languages of the Hierarchical effects such as those exhibited by world. Oxford: Oxford University Press. Ojibwe/Algonquian pose a challenge to linguists Mithun, Marianne. 1999. The languages of native North attempting to explain them. America. Cambridge: Cambridge University Press. Nichols, John. 1980. Ojibwe morphology. Doctoral Dissertation, Harvard University. References *Rhodes, Richard. 1985. Eastern Ojibwa-Chippewa-Ottawa Aanii: an introduction to the central Ojibwe language. Video dictionary. Berlin: Mouton de Gruyter. Series. Lessons in conversational Ojibwe as spoken at ———. 1990. Obviation, inversion, and topic rank in Ojibwa. Wikwemikong, Ontario. Videotapes with Ojibwe and Berkeley Linguistic Society 16. 101–15. English subtitles, pauses for repetition, and additional Siebert, Frank. 1967. The original home of the proto-Algonquian vocabulary. Native Vision Productions, Box 23, Ohsweken, people. ed. by A.D. DeBlois, Contributions to anthropology: Ontario N0A 1M0, Canada (tel.: 519/445-0639; fax: linguistics I (Algonquian), 13–47. Anthropological Series 78, 519/445-0639; e-mail: mwwwturner@yahoo.ca; website: National Museum of Canada Bulletin 214. www. www.nativevision.homestead.com) ncs4.net/Ojibwe. The definitive link to Ojibwe linguistic and Baker, Mark. 1996. The polysynthesis parameter. Oxford: anthropological material. www.citilink.com. Nancy Vogt’s Oxford University Press. website, with brief descriptions of language and culture; bib- Bloomfield, Leonard. 1946. Algonquian. Linguistic structures liography of further references. of native America, ed. by Hoijer Harry, 85–129. Viking MARK CAMPANA Okanagan and Salishan Languages Okanagan is one of four languages that belong to the to early speculations that the Salishan languages had Southern Interior branch of the Salishan linguistic spread from an inland location (Boas 1905), recently family of North America. All but two languages of the scholars have proposed that speakers of the original family occupy contiguous territory that extends longi- language occupied a maritime area, and migrated out- tudinally from about 123°W to about 113°W, and lati- ward from there, following routes to the interior along tudinally from about 52°N to about 45°N. In contrast such major rivers as the Fraser and Thompson (Suttles 767 OKANAGAN AND SALISHAN LANGUAGES and Elmendorf 1963; Suttles 1987). The linguistic evi- Washington, and westward in the Similkameen and dence adduced focuses on terms of flora and fauna Methow valleys; and along the north–south expanses (Kinkade 1991). of the Sanpoil and Kettle rivers, and the area west of Subgroupings: The family divides into five branch- the Columbia river as far as the bend around Wilbur, es, the result of migrations. A single group is thought Washington. Dialectal differences are minor but to have headed northwest and settled in the area where remain to be described. (Figure 1 here) Bella Coola is now found, surrounded by non-Salishan Phonology: In spite of the size of the family, and the languages, Wakashan seaside, and Athapascan inland- number of different languages represented, it is possi- side. Another group is thought to have settled along ble to provide a generalized phonology of Salish (see the coast, and then spread further, one subgroup going Figure 2). southward to where Tillamook, now extinct, was spo- The angled brackets notation should be read as ken, and formed the third branch of the linguistic fam- either/or: with few exceptions, a Salishan language has ily. This language, too, is surrounded by non-Salishan one or the other series, the velar being the conservative languages, clockwise from the north: Chinookan, one. Okanagan has the velar series. The northern Athapascan, Takelman, and Maidu. A fourth group, dialects of Okanagan include voiced postpalatal reso- forming the Tsamosan branch, also moved southward, nants (γ', γ) that correspond to y' y in the other and a fifth group, forming the Interior branch, moved dialects. In the same northern dialects, the labialized eastward (see Figure 1). One now extinct Athapascan pharyngeals have merged with their unrounded coun- language, Nicola, was spoken in the approximate geo- terparts. Voiced stops are rare in Salishan languages, graphic center of the Salish area. but occur in Coeur d’Alene, Twana, Lushootseed, and The area where Okanagan is spoken by perhaps as Comox, where they do not derive historically from many as 1,000 speakers of mostly mature age spans proto-stops. Pharyngeals are found only in the Salish along the north–south expanse of the Okanagan valley languages of the Interior; and no language, except from what is now Enderby to the south of Okanogan, Comox, has a phonemic nonejective lateral affricate. Bella Coola Comox (Pentlatch) Sechelt Squamish Lillooet Halkomelem SALISHAN Shuswap (Nooksack) Northern N.Straits Thompson Central Coast Interior Straits Columbian Klallam Okanagan Southern Lushootseed Kalispel (Twana) Tsamosan Coeur d'Alene Quinault (L. Chehalis) (U. Chehalis) (Cowlitz) (Tillamook) Figure 1 768  OKANAGAN AND SALISHAN LANGUAGES Figure 2. Generalized Salish consonant inventory. Okanagan has a symmetric three -vowel system, /i u (prefixes and suffixes; parentheses abbreviate vari- a/, while the vowel systems of Salishan languages ants): often include four vowels, sometimes five, with some i(n)- 1sg -tt 1pl asymmetries, notably the occasional absence of a high a(n)- 2sg -mp 2pl back vowel. Schwas are for the most part epenthetic. -s/-c 3sg -s-lx / -c-lx 3pl Salishan languages are wellknown for their propensity for consonant clusters; Bella Coola for several lexical which yield such forms as items that consist entirely of voiceless obstruents. an-lʔíw your father Morphosyntax: Salish languages have large num- in-χmínk I like/want it bers of affixes, predominantly suffixes. All Salish lan- guages have developed lexical affixes, bound forms which, in turn, may combine with members of the kn with lexical content, with a function that resembles set (kwu subset) to yield forms such as that of incorporated nouns, and all Salish languages kwu an-1ʔíw I am your father. have several reduplicative patterns that mark primarily kw in-χmínk I like/want you (you are my augmentative, diminutive, and inchoative forms. A wanting). nominalizer with the shape s- is nearly universally kw i-ks-ʔam-t-ím an-lʔíw I am going to feed found, but homophonous aspectual prefixes also your father. occur, notably in Coeur d’Alene and Okanagan. Person marking: Six Interior languages have differ- The last is the nominalization of a future (ks-) posses- ent sets of person markers for transitive, intransitive, sor applicative (-t) verb form (root ʔam, feed), in and subordinate predicates; the paradigms are less dif- which the suffix -(i)m, sometimes referred to as the ferentiated in Lillooet, closest to the languages of the antipassive, is required. coast, and in the noninterior languages. Possessive The transitive subject set, often called the ergative paradigms include prefixes and suffixes in all the lan- set, consists of the following suffixes (parentheses guages. Okanagan has four main sets of person refer- abbreviate stressed and unstressed variants): ence markers: the kn set (intransitive), the i(n)- set -(í)n 1sg -(í)m /-t 1pl (possessive), the -(í)n set (transitive subject), and the -(í)xw 2sg -(í)p 2pl (transitive) object set. -(í)s 3sg -(í)s-lx 3pl The kn set consists of clitics (marked with the liga- These markers follow the object markers, which, in ture) and a suffix: turn, follow one of several obligatory transitive mark- kn 1sg kwu 1pl ers (see below). kw 2sg p 2pl The (transitive) object set consists of the following Ø 3sg Ø ...-lx 3pl markers (one proclitic and suffixes): These markers accompany stems that in English kwu 1sg kwu ...-m 1pl translate as intransitive verbs, nouns, and adjectives. -s / -m 2sg -(úl)m 2pl -Ø 3sg -Ø... -lx 3pl kn ʔitx. I slept. kw sqilxw. You are an Indian/a person. Because third -person object markers and thirdperson ʔayχwt (axáʔ). This one is tired. intransitive subject markers are Ø, Salishan languages are often characterized as split ergative systems. The A subset of these markers, identical in all persons allomorphy of the second singular object is transitivizer except for 1sg kwu , co-occurs with the possessive set dependent. The disambiguation of number in the first- of person markers, and is reserved for double posses- person object is accomplished by the suffix -m and such sives and verb nominalizations. forms are interpreted as 3rd indef subject - 1pl object: The possessive set, used with nouns, psych verbs, kwu sp’-nt-is He whipped me (-nt transitivizer). and verb nominalizations, consists of these markers kwu sp’-nt-im They whipped us/We were whipped. 769 OKANAGAN AND SALISHAN LANGUAGES -(í)m occurs also with Ø, and the interpretation of sql’tmixw yaʔχís That one over there is a man; that is, these forms can be indefinite subject, or passive: by juxtaposing (in either order) the stem sql’tmixw sp’-nt-is 3rd person whipped 3rd person. and a deictic stem (ixíʔ, yaʔχís). In traditional terms, sp’-nt-im 3rd person indef whipped 3rd per- these sentences would be analyzed as exocentric equa- son/3rd person was whipped. tional constructions consisting of a subject and a pred- icate. The participant persons kn and kw are Word classes: Aspectual criteria can be established to pronominal subjects; third-person forms can be ana- distinguish word classes, and, as expected, these may lyzed as having a nominal subject of the classes men- derive forms of other classes––nouns can derive verbs tioned, which, in context, can be deleted. Another and verbs can be nominalized (for example, N. Mattina complication for the interpretation of all full words as as reported in Kroeber 1999). A prototypical noun like predicative is presented by the different markings for kWilstn sweat lodge, culturally relevant and categorially morphological and syntactic plurals: the morphologi- marked (-tn instrumental), derives a verb with -m: cal plural of citxw house is the reduplicated form kn kwilstn-m. I sweat bathed. ct-citxw houses, while the syntactic plural of the same form is citxw-lx (ixíʔ) (Those) are houses. Similarly, qwacqn hat derives qwacqn-m wear a In recent times, when scholars are preferring to hat (intransitive); ntχwχwin noon derives ntχwχwin-m view all constructions to have heads (or centers, in the do lunch (intransitive). old terminology), the question is raised as to what con- Analogously, qiʔs dream (intransitive) derives stitutes the head of such a sentence as kn sql’tmíxw. s-qiʔs dream, and the latter form can be inflected with Most common is the hypothesis that the verb is the possessive markers and interpreted as a possessive head of the sentence (here it would be the predicate noun form, or as a nominalized verb form. nominal), but because the identification of head with Most Okanagan stems can also be transitivized (see lexical head can be dispensed with, just as abstract below). features within the Inflection or Agreement nodes Nominal and pronominal arguments: Scholars have have been proposed to head sentences, and just as the argued that Salishan languages are pronominal argu- determiner has been proposed to head Determiner ment languages: a form like wik-nt-xw You saw it is a Phrases, so can kn be proposed to head the sentence full sentence with a third-person object (Ø) and kn sql’tmíxw. An utterance like xwuy He went, then, second-person subject (-xw). In this interpretation, any can be viewed as the abbreviation of xwuy ixíʔ That object expressed in nominal form is an adjunct, not a one went, and analyzed either as having a null subject, (nominal) argument. The claim is countered with the or as requiring a third-person nominal subject which suggestion that in applicative sentences like undergoes deletion in the appropriate circumstances. kwu tq-t-is in-kílx. He touched my hand Intransitive, possessive, and transitive paradigms: Beside the intransitive constructions already dis- the noun phrase in-kílx my hand functions as one of cussed, Okanagan uses kn inflection in a number of the arguments of the possessor applicative verb form forms derived by means of prefixes, suffixes, and cir- kwu tq-t-is He touched my ... and this argument is not, cumfixes. Among these forms are: To-Be nouns (kn and cannot be, referenced in pronominal form on the k-noun). k-ilmíxwm snk’lip Coyote will be chief/is verb. chief-to-be; k- have forms (kn k-noun). kn k-qwac- Intransitive forms are also analyzed as fully pred- qn I have a hat; inchoatives (kn verb+-ʔ- before icative. stressed vowel). kn c’ʔax I got ashamed. (root c’ax); kn xwuy I went. patient forms (kn verb+VC2). kn t’k’w-ak’w I fell. (cf. kW ilmíxwm You are a boss. t’k’w-nt put something down); get patient forms (kn kW χast You are fine. c+verb). u ilíʔ kn c-lak’ I was in jail a long time (cf. lk’-nt tie something); habitual/durative forms In these sentences, the clitics kn and kw are the sub- (kn c+verb). ilíʔkn c-wix I live there; imperfective jects, and the word to which the clitics are attached are forms (kn s-c+verb-(mi)x I have been X-ing); inceptive the predicates. Third-person forms have Ø subject per- forms (kn ks+verb-(mí)xaʔx I am about to X); and son marking, and forms like sql’tmixw have been ana- past perfect forms (kn ksc+verb). kn ksc-nik’ I have lyzed as full predications that should be translated as cut some, I have some cut. something like ‘He is a man’ or ‘It’s a man.’ In the Beside the possessive, double possessive, and psych stream of discourse, such words can function as pred- forms discussed, other forms, some intransitive and icative elements. The normal way to express either of others transitive (the latter always marked by the suffix the isolated propositions ‘He’s a man’ and ‘It’s a man’ -m), take possessive inflection: durative/intent forms is with utterances like ixíʔ sql’tmixw That’s a man, or (i+s+verb) s-q’sápiʔ-s ilíʔ i-s-ilíʔ I lived there a long 770  OLD CHINESE time (root ilíʔ there, lit. long-time there I-there); per- Conferences, Vol I: The North Pacific to 1600. Portland: fective forms (i+sc+verb). in-χást i-sc-ʔítx I slept well Oregon Historical Society Press. Kinkade, M. Dale, William W. Elmendorf, Bruce Rigsby, and (my-good my-having-slept); future forms (i+ks+verb) Haruo Aoki. 1998. Languages. Plateau Handbook of North lut a-ks-xwúy Don’t go; kw i-ks-(s)íw-m I’ll ask you; American Indians, Vol. 12, ed. by Deward E. Walker future imperative forms (i+kc+verb) lut a-kc-náq’w (William C. Sturtevant, general editor), 49–72. Washington, You will not steal. χast a-kc-k’wúl’-m You will work DC: Smithonian Institution. well; and future applicative forms (a-ks-verb-t-m) kw i- Kroeber, Paul D. 1999. The Salish language family. Reconstructing syntax. Lincoln, NE: University of Nebraska ks-may’-xít-m ... I am going to tell you ... Press. Finally, all transitive forms take transitive person Suttles, Wayne. 1987. Northwest coast linguistic history. A markers. Okanagan has two transitivizers, -nt and -st; view from the coast. Coast Salish Essays, ed. by Wayne, sut- a causative -st; three applicatives -t, -x(í)t, -tút; and tles. Vancouver: Talonbooks. three suffixes that prepare stems for transitivization: Suttles, Wayne, and William W. Elmendorf. 1963. linguistic evi- dence for Salish prehistory. Proceedings of the 1962 Annual -nun, -min, -xix. Customary transitive forms are Spring Meeting of the American Ethnological Society, ed. by marked with the circumfix c-...-st, as in c-wik-st-n I Viola E. Garfield, and Wallace L. Chafe. Seattle. always see it. Thompson, Laurence C. 1973. The Northwest. Current trends in linguistics, Vol. 10. Linguistics in North America, ed. by Thomas A. Sebeok, 979–1045. The Hague: Mouton. References –––––– 1979 Salishan and the Northwest. The languages of Boas, Franz. 1905. The Jesup North Pacific expedition Native North America: historical and comparative assess- Proceedings of the 13th International Congress of ment, ed. by Lyle Campbell, and Marianne Mithun, Americanists, 91–100. New York, Easton, PA: Eschenbach 692–765. Austin: University of Texas Press. Printing Co. Thompson, Laurence C., and M. Dale Kinkade. 1998. Czaykowska-Higgins, Ewa, and M. Dale Kindade (eds.) 1998. Languages. Northwest Coast of Handbook of North Salish languages and linguistics. Theoretical and descriptive American Indians, Vol. 17, ed. by Wayne Suttles. (William perspectives. Berlin: Mouton De Gruyter. C. Sturtevant, general editor), 30–51. Washington, DC: Kinkade, M. Dale. 1991. Prehistory of the native languages of Smithonian Institution. the northwest coast. Proceedings of the Great Ocean ANTHONY MATTINA Old Chinese Old Chinese may be subperiodized into Early Old bordered on the Chinese-speaking area in Antiquity, Chinese, Middle Old Chinese, and Late Old Chinese. the affiliations of Chinese must be considered uncer- It is generally thought that most Early Old Chinese tain and the relationship with Tibeto-Burman likely to polysyllabic words became monosyllabic by or during be due to convergence. Even if the Sino-Tibetan theory the Middle Old Chinese period, so the Late Old is correct, Proto-Chinese and the Tibeto-Burman lan- Chinese language was overwhelmingly monosyllabic. guages must have diverged long before the Oracle Proto-Chinese, the ancestor of Old Chinese, is prac- Bone Inscriptions, the first documents written in Old tically unknown. Although many scholars believe it to Chinese. These texts appear in the mid-second milleni- be related to the Tibeto-Burman family of languages in um BCE and already contain distinctively Chinese a ‘Sino-Tibetan’ genetic family, the theory remains phonological and syntactic characteristics that had controversial due to the lack of regular correspondenc- developed in Proto-Chinese, before the language was es in phonology (sound structure), morphology (word first written down. For example, only one negative par- structure), and syntax (sentence structure) between ticle, *ma, is reconstructable for Proto-Tibeto-Burman, Chinese and Tibeto-Burman. Moreover, from Early but the Oracle Bone Inscriptions have two negative Old Chinese times to the present, the Chinese language roots, *pa- and *ma-. This distinction is preserved in all has spread into territory inhabited by peoples who orig- later forms of Chinese, including modern Mandarin, inally spoke other languages, and for centuries it has where it is still found in numerous bound forms and in been spoken by far more people than any other lan- the free negative forms bù [pu] and méi [mej]. guage in the world. Without a thorough comparative The Old Chinese writing system contains important study of loanwords in languages known to have information for the reconstruction of the language. 771 OLD CHINESE Although it includes many characters with no phonet- mately on the Qieyun. However, unlike earlier works, ic elements, such as guî ‘tortoise, turtle’ (originally the Yunjing organizes the material in tabular form. a pictograph), some characters are phonetically ‘bor- Historic Sinological Reconstruction of Old Chinese rowed’. (N.B.: Pronunciation of characters is given in depends heavily on the projection back in time of the Mandarin unless otherwise noted.) For example, the categories established by the tables in the Yunjing, originally pictographic character used to write the adjusted according to the rhyme categories implied by word for ‘wheat’ was borrowed to write the then the Shijing poems and to phonological information homonymous word lái ‘to come’. Most characters derivable from the characters themselves. Change in are actually constructed of two or more parts, of which one part of the system necessarily entails change in one part is a ‘phonetic’ element and another is an often many other parts in order to maintain consistent corre- simplified semantic or ‘significant’ element usually spondences between the categories of the tables and the called the ‘radical’, such as zhuî ‘short-tailed bird’ phonological categories of Middle Chinese and the (originally a pictograph) and *zhuî (now usually modern Chinese dialects, hence the reconstructions of pronounced, irregularly, huán) ‘grass used for making one scholar’s system often look radically different from mats’, written with the ‘grass’ radical as the significant those of another. Unfortunately, Historic Sinological and zhuî as the phonetic. Although other exam- Reconstruction does not allow subperiodization of Old ples––such as lu˘ ‘blunt, stupid; Lu, the home state Chinese––contemporary proponents of the method of Confucius’ and its phonetic yú ‘fish’––have dif- explicitly claim that they are not actually trying to ferences that are important for reconstruction (see reconstruct a real language spoken by real people in dif- below), the phonetic information contained in the ferent places and at different times but only a theoreti- script is not by itself sufficient to allow reconstruction cal construct from which all later forms of Chinese can of Old Chinese phonological structure. In fact, be derived––so it is impossible to systematically use Chinese is not well attested phonetically until the sev- contemporaneous data to check the reconstructions enth century CE, when Middle Chinese forms begin to themselves. This results in a tendency to ignore attested be recorded in Old Tibetan alphabetic script. The data in favor of the system. For example, Sergei phonology of Old Chinese can therefore only be Starostin (1989) reconstructs ‘woman; you (in the recovered through reconstruction. latter sense also written )’ as Old Chinese *nraʔ even Modern reconstructions of Old Chinese nearly all though overwhelming Oracle Bone Inscription and belong to a tradition that may be called Historic other evidence indicates that the word initial must have Sinological Reconstruction. It derives ultimately from been a bilabial in Early Old Chinese and apparently the early Chinese grammarians, on whose work the pio- well into the Middle Old Chinese period. Another result neering Sinologist Bernhard Karlgren based his system is the production of many unlikely forms, such as Old of reconstruction. Although the appearance of most Old Chinese * srjʔ(s), the reconstruction of ‘affair, mat- Chinese reconstructions has changed radically since the ter’ by William Baxter (1992). Although scholars work- publication of Karlgren’s etymological dictionary ing in this tradition have made progress toward the Grammata Serica recensa (1957; A Chinese Grammar, reconstruction of Old Chinese, recently their method Revised), Historic Sinological Reconstruction is still has been challenged. Reconstructions have been pro- based on the method pioneered by Karlgren. posed that are not based on the categories of the rhyme The most important sources used in the traditional tables but on contemporaneous data, emphasizing Old Historic Sinological Reconstruction approach to Old Chinese character variations, loanwords found in neigh- Chinese phonology are: the phonological information boring languages, and Middle Chinese reconstructions derivable from analysis of the characters themselves grounded in foreign transcriptions and the earliest (mostly mid-second millenium BCE to second century attested text of the Qieyun (706 CE). BCE) and their variants or substitutions; the rhymes of Early Old Chinese (called by some scholars ‘pre- the Shijing (late first millenium BCE; Book of Songs); Archaic Chinese’), the language of the Yin, or late Chinese transcriptions of known foreign words (c. first Shang dynasty period (fourteenth to eleventh centuries century CE onward); books by Chinese grammarians on BCE), is preserved mainly in the Oracle Bone dialects and rhymes (first century BCE onward), espe- Inscriptions, which are divination texts inscribed on cially the lost Middle Chinese Qieyun (601 CE; Cut bones and turtle shells. These texts have been found in Rhymes); the rhyme books, particularly the Guangyun great numbers in the area of the lower Yellow River (Extensive Rhymes) and Jiyun (Collected Rhymes), valley. The syntax of Early Old Chinese, although both from the eleventh century CE; and, above all, the slightly different from that of the earliest classical rhyme tables, especially the Yunjing (Mirror of texts from a millenium later in the late Middle Old Rhymes), from the twelfth century CE. The latter three Chinese and Late Old Chinese periods, is already SVO works, although compiled much later, are based ulti- (Subject–Verb–Object) in basic sentence word-order, 772  OLD CHINESE as in modern Chinese. Several scholars have argued synonyms must be related. Since *lŋâ, one dialect that Early Old Chinese must have had many disyllab- form of the Middle Old Chinese first person pronoun, ic words or roots. Reconstructable examples include may be reconstructed as *laŋa, the other dialect form, Early Old Chinese *Cwena (where ‘C’ stands for an *lâ, may be reconstructed as *laγa, both deriving unknown consonant) ‘woman; you’, and Early Old from an Early Old Chinese *laga or *lege. This word is Chinese *kara ‘bitter’. However, Early Old Chinese evidently related in turn to another early first person phonology has yet to be studied in depth. pronoun, Early Old Chinese *aga or *ege. Middle Old Chinese is essentially the language of the During the Middle Old Chinese period, most previ- post-Shang bronze and stone inscriptions and the earli- ously disyllabic morphemes became monosyllabic. est transmitted Classical texts. The inscriptions on This change produced many homonyms by late bronze vessels are often dated, and cover the period Middle Old Chinese––such as *lŋâ ‘first person pro- from the early Zhou dynasty (c. 1145 BCE onward) into noun’ ~ ‘five’ ~ ‘fish’ ~ ‘speak’—and resulted in other the Warring States period (475–221 BCE). Some of drastic phonological changes, including widespread them contain rhymed passages and variant characters metathesis (segments changing place) in one or more from which phonological information can be derived. dialects. For example, middle Middle Old Chinese The Middle Old Chinese period may be divided into *lŋâ ‘fish’ (from Proto-Chinese *laka) became late three subperiods: the early Middle Old Chinese of the Middle Old Chinese *ŋlâ or *ŋrâ. The development of Western Zhou dynasty, when the capital was located in monosyllabism was also probably a major motivation the area of present-day Shaanxi Province and the pres- for the eventual development of phonemic tone (into- tige language was influenced by the local dialect; the national pitch distinguishing meanings). The shift to a middle Middle Old Chinese of the Eastern Zhou, when monosyllabic morpheme structure was largely com- the capital moved eastward into another dialect area; plete by the time the Shijing was recorded. and late Middle Old Chinese, the language of the earli- Late Old Chinese is characterized by a number of est transmitted ‘Confucian’ classical texts, including the phonological changes, including that of the Middle Shijing. These periods are marked by phonological Old Chinese syllable-initial *l to either *y [j] or *d, of changes, partly motivated by prestige-dialect shift when Middle Old Chinese medial *l or *r to *, and of the the capital moved to a new dialect region, and partly due Middle Old Chinese syllable-final *r to *n or *y [j]. to change over time. One example must suffice here. (The rules for these changes are still unclear.) The The usual first person singular pronoun (‘I’) in early change of final *r to *n took place in the ancient cen- Middle Old Chinese is (also written ) yú, but in tral dialect and is found in all modern Chinese dialects, middle Middle Old Chinese (beginning with the Stone but it was not completed in some peripheral dialects, Drum inscriptions, c. fifth to sixth centuries BCE) this notably in the northeast, where ancient final *r was word is replaced by wú, which by Late Old Chinese retained as a syllable coda at least into the Middle displaces yú except in deliberately archaizing texts. Chinese period. Another significant change includes The character wú ‘I’ has as its phonetic element the the beginning of what has been called the Great character wu˘ ‘five’, while in other middle Middle Chinese Vowel Shift, in which the Middle Old Chinese Old Chinese texts it is generally written instead with vowel *â (or [a˘]) became Middle Chinese *ɔ and yú ‘fish’ as its phonetic. Because the Old Tibetan Mandarin u. Thus, late Middle Old Chinese *ŋlâ (or numerals are known to be cognate with (derived from *ŋrâ) ‘fish’ became Late Old Chinese *ŋâ, Middle the same historical source as) Chinese, Old Tibetan Iŋa Chinese *ŋɔ1, Late Middle Chinese *ŋü1, and ‘five’ is cognate with the Chinese word for ‘five’. Mandarin yú [ü]. Some traces of Late Old Chinese (Whether the Tibetan numerals are borrowed from forms are found in the modern Southern Min dialect. Chinese or inherited from a common ‘Sino-Tibetan’ (Chinese is otherwise internally reconstructable only ancestor is irrelevant for present purposes.) Since yú back to Middle Chinese.) The written Late Old (from Middle Chinese *ŋɔ1) ‘fish’ is also the phonetic Chinese of the Former Han dynasty (206 BCE to 9 in the character lu˘ (from Middle Chinese *lɔ2) ‘Lu’, CE) became the standard Classical Chinese literary it is clear that not only must the Middle Old Chinese language, which continued in use into the twentieth form of the word for ‘five’ have been close to *lŋâ, but century and is still used for some purposes even today. it and the word for ‘fish’, the first person pronoun, ‘Lu’, and other words, such as yu˘ ‘speak, speech’, were References pronounced the same in one or more of the Middle Old Chinese dialects. Because the character used to write Baxter, William H. 1992. A handbook of Old Chinese phonolo- gy. Berlin: Mouton de Gruyter. the earlier first person pronoun yú is also used as a Beckwith, Christopher I. 2002. The Sino-Tibetan problem. phonetic in other characters, it is reconstructable as *lâ, Medieval Tibeto-Burman Languages, ed. by Christopher I. which derives from an earlier *laCa. It is clear that these Beckwith. Leiden: E. J. Brill. 773 OLD CHINESE Beckwith, Christopher I. 1999. Review of Laurent Sagart, The ––––––. 1996. Early contacts between Indo-Europeans roots of Old Chinese. Amsterdam: John Benjamins; in and Chinese. International Review of Chinese Linguistics 1.1. Anthropological Linguistics 44.2 (2002). Sagart, Laurent. 1999. The roots of Old Chinese. Amsterdam: Benedict, Paul. 1972. Sino-Tibetan: a conspectus. Cambridge: John Benjamins. Cambridge University Press. Schuessler, Axel. 1987. A dictionary of early Zhou Chinese. Karlgren, Bernhard. 1957. Grammata Serica recensa. Stock- Honolulu: University of Hawaii Press. holm: Museum of Far Eastern Antiquities; reprinted Starostin, Sergei A. 1989. Rekonstrukcija drevnekitajskoj fono- 1972. logieskoj sistemy (Reconstruction of the Old Chinese Pulleyblank, Edwin G. 1962. The consonantal system of Old phonological system). Moscow: Nauka. Chinese. Asia Major 9. CHRISTOPHER I. BECKWITH Old Church Slavonic Old Church Slav(on)ic (OCS) is the language of a Orthodox Slavs and some other peoples, such as the group of Slavic texts containing copies of religious Romanians. works, almost all originally translated from Greek. Old Church Slavonic was written in two alphabets: The corpus includes eight parchment manuscripts hav- Glagolitic, invented (or, less likely, edited and brought ing 100 or more folia, and an approximately equal into its final form) by St. Cyril, perhaps with the par- number of fragments. A half-dozen inscriptions sur- ticipation of his brother Methodius, for their mission vive from the same period and area. No manuscript to the Slavs in the 860s, and Cyrillic, invented later, bears an explicit date or locale, but paleographic and probably in Eastern Bulgaria, and named in honor of linguistic features indicate that most of the surviving Cyril. Cyrillic was based upon the Greek uncial letters; manuscripts were copied in the eleventh century CE, Glagolitic has no immediately obvious resemblance to with a very small number possibly from the late tenth. any other alphabet, although some features compare The original translations are generally assumed to go suggestively to the Greek cursive of the time. back to the second half of the ninth or early tenth cen- Glagolitic was used in some parts of Croatia until the turies, and were connected with the mission of Cyril early twentieth century, but elsewhere (Kievan Rus’, and Methodius to the Slavs and the activities of their Bulgaria, Serbia, etc.) was quickly displaced by pupils in Bulgaria following the death of Methodius in Cyrillic, which has remained in use among the 885 CE. Every manuscript deviates in some ways from Orthodox Slavs until the present day. normalized Old Church Slavonic, but as a group they Late Common Slavic introduced far-reaching show an Eastern South Slavic dialect which has often changes to the sound system of late Indo-European been called Old Bulgarian, and it is in fact very close (IE). As in some other dialects of IE, *o and *a con- to the reconstructed Late Common Slavic (LCS), flated to *å, with the long vowel giving Slavic a and the which is usually considered to have lasted until around short vowel giving Slavic o. IE * e¯ became LCS * e (in the end of the first millennium CE. With the passage of OCS, a low front vowel [æ]), and IE * ˘e went to LCS time, this liturgical and literary language took on cer- *e. IE * ¯ gave LCS *i, IE * ˘ gave LCS *ь, a high short tain local characteristics in each area where it (lax) front vowel. IE *u¯ eventually became *y, a high remained in use, producing Russian/Bulgarian/ unrounded nonfront vowel (*su¯nus > *synъ ‘son’), and Serbian, etc. recensions of Church Slav(on)ic, which IE * ˘u became LCS *ъ, a high short (lax) back vowel. remained mutually comprehensible and collectively This gave a vowel system with binary oppositions served as a common literary language for the in height, fronting, and length (long/short): front nonfront front nonfront high ¯ / ˘ u¯ / ˘u realized as OCS high i/ь y/ъ nonhigh e¯ / ˘e a¯ / ˘a nonhigh e / e a/o 774  OLD CHURCH SLAVONIC plus diphthongs. Those diphthongs in *w and *y were LCS syllable structure had undergone a number of converted to monophthongs: *ow > *u; *ew > *ju; *oy significant changes. Syllables in LCS were always > *e or *i; *ey > *i. Diphthongs in *r and *l were trans- ‘open’ Syllable-closing consonants were dropped: *sup- posed (*gord- > *grad- ‘enclosed area,’ *xold- > xlad- nos > *sъnъ (su ˘ nu ˘ ). The basic syllable formula was ‘cold’, and diphthongs in *m or *n became nasal vow- strV, where s stands for s or z, t for almost any conso- els (*ronka > *roka ‘hand,’ *se¯men > *se me ‘seed’). nant, and r for a sonorant, although the details are quite As a result, the vowel system of OCS contained i, complex. The consonant clusters that actually occur can ь, e , e, a, o, y, ъ, u, e, o ,with oppositions based upon be quite formidable: OCS umrъštvl¸o ‘I will kill.’ front(i, ь, e , e, e)/back, high(i, ь, y, ъ, u)/low, round- Syllables also showed a type of ‘synharmony’, ed(ъ,u, o, o )/unrounded, tense/lax, and oral/nasal. where consonants and vowels were subject to mutual Slavic is a satem language, as shown by such OCS accommodation, such as fronting of vowels after j or words as sъto ‘hundred’ and zlato ‘gold.’ The consonant palatal consonants, the ‘palatalization’ of the velars k g system shows only a two-way contrast, between voiced x to J D š or c  s before a front vowel, or ‘jotization’ of and voiceless, with a division into labial, dental, palatal combinations of consonants with j: sj > š, kj > J , etc. (not palatalized), and velar. Stops were p b t d k g, con- This produced paired endings and consonant alterna- tinuants included s z š D x, and the affricates were c (ts), tions in several instances. For examples of the changes,  (dz), J, and in some dialects,  . Sonorants were m n l r see the o/jo and a/ja declensions given below, including n¸ (=nj), ¸l(=lj), ¸r (=rj), glides contained w and j. the forms of vlьkъ and ro ka given in the same tables. TABLE 1 (a) Singular Gloss stem nom. voc. acc. gen. loc. dat. instr. ‘city’ o gradъ grade gradъ grada grade gradu gradomь ‘man’ jo moDь moDu moDa moDa moDi moDu moDemь ‘wolf’ o vlьkъ vlьe vlьkъ vlьka vlьce vlьku vlьkomь ‘village’ o selo selo selo sela sele selu selomь ‘field’ jo pol¸e pol¸e pol¸e pol¸a pol¸i pol¸u pol¸emь ‘woman’ a Dena Deno Deno Deny Dene Dene Denojo ‘soul’ ja duša duše dušo duše duši duši dušejo ‘hand’ a roka roko roko roky roce roce rokojo ‘bone’ ˘ kostь kosti kostь kosti kosti kosti kostьjo ‘son’ ˘u synъ synu synъ/-a synu synu synovi synъmь (b) Plural meaning nom./voc. acc. gen. loc. dat. instr. ‘city’ gradi grady gradъ grade xъ gradomъ grady ‘man’ moDi moDe moDь moDixъ moDemъ moDi ‘wolf’ vlьci vlьky vlьkъ vlьce xъ vlьkomъ vlьky ‘village’ sela sela selъ sele xъ selomъ sely ‘field’ pol¸a pol¸a pol¸ь pol¸ixъ pol¸emъ pol¸i ‘woman’ z eny Deny Denъ Denaxъ Denamъ Denami ‘soul’ duše duše dušь dušaxъ dušamъ dušami ‘hand’ roky roky rokъ rokaxъ rokamъ rokami ‘bone’ kosti kosti kostьjь kostьxъ kostьmъ kostьmi ‘son’ synove syny synovъ synъxъ synъmъ synъmi (c) Dual meaning nom./voc./acc. gen./loc. dat./instr. ‘city’ grada gradu gradoma ‘man’ moDa moDu moDema ‘wolf’ vlьka vlьku vlьkoma ‘village’ sele selu seloma ‘field’ pol¸i pol¸u pol¸ema ‘woman’ Dene Denu Denama ‘soul’ duši dušu dušama ‘hand’ r oce roku rokama ‘bone’ kosti kostьju kostьma ‘son’ syny synovu synъma 775 OLD CHURCH SLAVONIC TABLE 2 present root aorist s-aorist extended a. imperfect imperative 1 sg. vedo vedъ ve sъ vedoxъ vede axъ – 2 sg. vedeši vede vede vede vede aše vedi 3 sg. vedetъ vede vede vede vede aše vedi 1 dual vedeve vedove ve sove vedoxove vede axove vede ve 2 dual vedeta vedeta ve sta vedosta vede ašeta vede ta 3 dual vedete vedete ve ste vedoste vede ašete – 1 pl. vedemъ vedomъ ve somъ vedoxomъ vede axomъ vede mъ 2 pl. vedete vedete ve ste vedoste vede ašete vede te 3 pl. vedotъ vedo ve se vedoše vede axo – Inflected words in OCS may be divided into nomi- between definite and indefinite; definite adjectives are nal and verbal groups, with the nominals subdivided made from indefinite adjectives by appending into nouns, adjectives, and pronouns. Numerals pat- pronominal suffixes to the indefinite forms; thus, nova tern with nouns or pronouns; ‘one’ through ‘four’ and re ka means ‘a new river,’ but novaja re ka means ‘the their compounds also show gender agreement. new river’; novo selo means ‘a new village’ and novo- Both verbs and nouns show number (singular, plu- je selo (o > e after j by ‘synharmony’) means ‘the new ral, and dual). Nouns distinguish up to seven gram- village.’ The comparative (more...) is formed by suf- matical cases marking subject (nominative), direct fixation; the superlative (most ...) is made syntactical- object (accusative), possession (genitive), location ly from the comparative. (locative), indirect object (dative), instruments/means Although no single OCS verb is attested in all pos- (instrumental), and also a form of address(vocative). sible forms, the conjugation of ved- ‘lead’ must have Nouns have inherent gender (masculine, feminine, and been as shown in Table 2. neuter). Verbs distinguish six tenses, and have indica- Word order in OCS is most often S(ubject) V(erb) tive, imperative, and conditional forms; they can be O(bject), but can be quite free. It is often quite difficult active or passive and indicate completed vs. incom- to know when OCS word order is following that of the pleted actions. Greek from which the text was originally translated Noun markings in OCS fall into five patterns and when it represents truly Slavic usage. All gram- (declensions); in Indo-European terms they are o/jo, matical cases except the nominative occur with prepo- a/ja, ˘ , u˘, and consonant stems. The difference sitions, but all, including the locative, occur without between the o and jo stems and between the a and ja prepositions. Under the influence of Greek, OCS syn- stems are due largely to the intrasyllabic changes tax often becomes rather convoluted. referred to above as ‘synharmony’. The o and the u˘ OCS vocabulary is primarily of Indo-European, or stem endings have intermixed so much by the time at least Balto-Slavic origin, although it also contains the OCS texts were copied that one can no longer borrowings from several other sources, including really set up u˘-stems as a separate declension. Iranian (e.g. bogъ ‘god,’ rajь ‘paradise’) and Consonant stems come in several varieties, e.g. r- Germanic (e.g. xle bъ ‘bread,’ kupiti ‘buy’). As one stems such as mati, gen. sg. matere ‘mother’; s-stems would expect, most of the religious terminology and such as nebo, gen. sg. nebese ‘heaven’; en-stems an overwhelming majority of the abstract, legal, philo- such as vre me, gen. sg. vre˘mene ‘time’ and kamy, sophical, administrative, and didactic vocabulary is gen. sg. kamene ‘stone’. The productive declensions either direct borrowings or loan translations (calques) are the o/jo, a/ja, and ˘ -stems. Some of the u˘-stem from Greek. endings occur frequently and even become produc- tive in the o/jo declension. References Sample Declensions Huntley, David. 1993. Old Church Slavonic. ed. by Bernard Comrie and Greville G. Corbett. The Slavonic Languages, See Table 1. London, New York: Routledge. Adjectives are inflected and show agreement in Lunt, Horace G. 2001. Old Church Slavonic Grammar, 7th revised edition. Berlin, New York: Mouton de Gruyter. gender, number, and case. Masculine and neuter adjec- Schenker, Alexander M. 1995. The Dawn of Slavic. An tives decline in principle like o/jo-stem nouns, while Introduction to Slavic Philology. New Haven: Yale feminine adjectival forms are essentially like those of University Press. the a/ja-stem nouns. Adjectives also distinguish CHARLES GRIBBLE 776  OLD ENGLISH Old English Old English (OE) is the collective name given to the toward Scandinavia: between 1016 and 1042, England varieties of Germanic brought to Britain by the Jutes, was ruled by the Danish king Cnut and his sons. Saxons, and Angles in the fifth and sixth centuries. The event separating Old from Middle English is the Orthography Norman Conquest of 1066. The root Engl- in the names Engl-ish and England < Engla-land ‘land of the The earliest OE written records are runic inscriptions Angles’ represents a form of the Latin and Common dated c. 650–700 CE. They were either records of indi- Germanic name of the tribe Angli. The OE language is vidual names, or brief decorative messages. The runic also sometimes referred to as Anglo-Saxon (AS), alphabet, originally linked to Germanic pagan rituals, although more frequently a distinction is made between was abandoned after the adoption and spread of the OE language and AS history, law, literature, culture, Christianity during the sixth and seventh centuries. The etc. The closest linguistic relative of OE is Old Frisian. alphabet used subsequently was a modified form of the OE and Old Frisian belong to the West Germanic sub- Roman alphabet, known as the Insular hand in which group of the Germanic branch of Indo-European. the letter <s> was written in its long form <ʃ>, and <g> The date usually associated with the beginning of appeared as <>, known as ‘yogh’. The three specifical- OE is 449 CE, when Germanic-speaking warriors, led ly OE letters, not used in the Roman alphabet, are <æ> by the legendary brothers Hengist and Horsa, sailed to ‘ash’, for the vowel [æ], and <þ> ‘thorn’ and <ð> ‘eth’ Britain to fight against the Picts at the invitation of the (or ‘ðæt’), used interchangeably for the consonant Celtic king Vortigern. From the middle of the fifth cen- sounds [ð], as in that, or [θ], as in thump. Consonant let- tury onward, Germanic settlers arrived in Britain in ter combinations specific to OE writing are: <sc>, which considerable numbers. The Jutes remained mostly in represented [sk] for most of the period, but probably Kent, parts of Hampshire, and the Isle of Wight, and also [ʃ] after c. 1000, and <c>, which stood for []. the Saxons occupied the lands south of the Thames, as well as Middlesex and Essex. The Angles spread Dialects westward and as far north as the Scottish Lowlands. The three original groups of settlers, the Jutes, the The Germanic invasions and settlements resulted in Saxons, and the Angles, maintained their dialect differ- the partial displacement of the indigenous Celtic- ences throughout the AS period. The main dialects of speaking populations from central England to the Old English are Kentish, West Saxon, East Saxon, more southern, western, and northern parts of the Mercian, and Northumbrian. Mercian and Northumbrian country. The most common periodization of OE is into were two varieties of Anglian, spoken to the south and early OE, from the beginnings to c. 800 CE, classical north of the river Humber. Prior to the Scandinavian OE, c. 800–950, and late OE, c. 950–1100. invasions, Northumbrian was a dialect of great prestige Two notable historical events that had a profound due to the religious, artistic, and intellectual achieve- effect on AS culture and the vocabulary of OE are the ments of the Northumbrian kingdom. After the end of conversion to Christianity, initiated by the arrival of St. the ninth century, the kingdom of Wessex unified and Augustine in Kent in 597, and the Viking invasions and dominated the rest of the country, and the new political settlements. The Vikings, who were Scandinavian sea- and cultural situation led to the spread of the West Saxon faring warriors, began their raids on Northumbria in the literary norms to the neighboring dialects. West Saxon is late eighth century. In the ninth century, their attacks the tenth century literary koiné of Anglo-Saxon England. culminated in the establishment of a separate adminis- Classical West Saxon is the variety of OE described in trative unit, the Danelaw (< Dena lagu ‘law of the the standard reference works, and it is also the dialect on Danes’), which included large parts of the northern, which the following descriptions are based. central, and eastern regions of England, roughly north- east of a line linking London and Chester. The treaty Phonology which established the Danelaw, negotiated by the West Saxon king Alfred the Great, who ruled from 871 to The system of OE stressed vowels consisted of seven 889, provided for a relatively peaceful period during short vowels, seven long vowels, and two diphthongs. the tenth century. At the beginning of the eleventh The colon sign in parentheses is the phonetic symbol century, the balance of political power shifted again for vowel length. Vowel length was not marked in the 777 OLD ENGLISH OE manuscripts, although many modern printings use The palatal fricative ʃ and the affricate  are in paren- the convention of a macron (-) over a vowel letter to theses because their existence as independent units indicate that it is long. before the end of the OE period has not been proven. The voiced fricatives v, ð, and z could appear only Simple vowels Diphthongs word-medially when flanked by vowels or nasals and i(:) y(:) u(:) eo or eə liquids. In all other positions, the fricatives were e(:) o(:) æa or æə voiceless. The distribution of these sounds in OE is æ(:) ɑ(:) responsible for present-day alternations of the type leaf-leaves (OE leaf, sg.-leafa gen. pl, leafum, dat. pl. The most important phonological change, which ‘leaves’), bath-bathe (OE bæþ, n.-bæþian ‘to bathe’), occurred prior to the first OE written records and glass-glaze (OE glæs,n.-glæsen, adj. ‘made of glass, changed the quality of many vowels in some positions, vitreous, glazed’). All OE consonants except the was I-Mutation, also known as I-Umlaut. This is a approximants could be doubled in the word-medial right-to-left (regressive) vowel harmony phenomenon: position. back vowels became front and low front vowels were Word stress in OE was regularly placed on the first raised before /i, j/ in the same word. (The dashed line syllable of the root. This correlates with verse compo- indicates the direction of the sound change from early sition in OE, where alliteration on the first stressed OE to late OE and ME.) syllable in each half-line is one of the basic structural i(:)<------ y(:) <___________ u(:) characteristics of the line. The most frequent prefixes, ↑ ge- and be-, were never stressed, but some noun and e(:) <------ ø(:) <___________ o(:) . . . before /i,j/ ↑ adjective prefixes that retained their lexical meaning æ(:) <__________________ ɑ(:) could be accented: óferfyllo, n. ‘repletion’ but oferfýl- lan, v. ‘feed to excess’, andswárian, v. ‘answer’, The change created stem alternations, depending on ándswaru, n. ‘answer’. In compounds, both roots were whether the stem was used by itself or whether it was stressed on the first syllable, with the first stress followed by an affix containing /i,j/. I-Mutation is stronger than the second: níed-fàru ‘enforced journey, responsible for the different vowels in pairs such as death’, túngol-wìtega ‘astrologer’, scéarp-ècged full-fill (OE full - fyllan < *fulljan ‘to fill’), foot-feet ‘sharp edged’. (OE fo- t – fe- t < *fo- tiz ‘feet’, pl. ), man-men (OE mann – menn < *manni- ‘men’, pl. ). An important quantitative change during the OE Morphosyntax period was the lengthening of short vowels before OE was a predominantly synthetic language, which some consonant groups, most notably -ld, -mb, -nd. means that the grammatical relations between the This development, which started during the ninth cen- words in the clause were expressed by means of tury, accounts for the historically long vowels in words inflexional affixes. The nouns had special markings such as child, comb, hound, kind. A third consonant for the following categories: prevented the lengthening; this is the reason for the survival of the short vowels in words such as children, (a) Number: singular, plural, e.g. stan ‘stone’, sg. hundred, kindred. stan-as ‘stones’, pl. Only short vowels could appear in fully unstressed (b) Case: nominative (for the subject), accusative syllables. The most frequent spellings of unstressed (for the direct object), genitive (noun modi- vowels are <e>, <o>, or <a>. In late OE, the spelling fiers, possessives), dative (indirect and prepo- distinctions for unstressed vowels were disappearing, sitional objects), e.g. stan-es, ‘of stone, suggesting that by the end of the period most of the stone’s’, gen, stan -um ‘with stones’, dat. pl. unstressed vowels were neutralized to [ə]. (c) Gender: masculine, feminine, neuter. The Classical OE had the following consonants: grammatical gender of a noun was determined by agreement with the demonstrative pro- nouns and adjectives accompanying it, not by Labial Dental Alveolar Palatal Velar reference to biological sex, e.g. se stan ‘this Voiceless stops p t k stone’, masculine, þæt wif ‘this wife’, neuter, Voiced stops b d g and seo rod ‘this cross’, feminine. Fricatives f/v θ/ð s/z (ʃ) x,F Affricates (), The adjectives in OE had separate markings for the Nasals m n same categories as the nouns they modified (number, Liquids l, r case, gender). They also varied according to their syn- Approximants j w tactic placement. The strong, or definite declension 778  OLD ENGLISH was used for adjectives which were part of a definite by one or more adverbs, and even by an object, noun group (se brada sæ ‘this broad sea’), and the although noun objects normally followed the verb; sta- weak, or indefinite declension was selected when the tistically the order V–O was dominant. The subject adjective was used predicatively (wæs þæt rice brad could precede or follow the verb, depending on ‘that kingdom was broad’). Like modern adjectives, whether the first constituent position in the clause was OE adjectives could be marked for degrees of compar- filled and on whether the subject was a full noun or a ison: comparative (brad -re ‘broader’) and superlative pronoun. (brad-ost ‘broadest’). On þam twelftan dæge cristes acennednysse comon þa The personal pronouns of OE were: ic ‘I’, þu ‘thou, þry tungelwitegan… you, sg.’, heo ‘she’, he ‘he’, hit ‘it’, we ‘we’, e ‘you, pl’., hie/hy ‘they’. They agree with the main verb in On the twelfth day of Christ’s birth came the three person (first, second, third) and number (singular, plu- astrologers… ral). In addition to the familiar singular–plural number distinction, OE had two special ‘dual ‘pronouns: wit With complex verb forms, the first verb occupied the ‘the two of us’ and it ‘the two of you’. Like the second position, while the second part of the verb, an nouns, the pronouns had case forms depending on infinitive or a past participle, appeared at the end of the their syntactic function. Personal pronouns and clause. Thus, the two parts of the verb formed a ‘brace’ demonstratives also had three genders: feminine, mas- enclosing all but the very first clause constituent. culine, and neuter. There were no gender distinctions We sceolon… mid halgum mægnum þone eard ofgan… in the plural. We must…with holy virtues this country obtain… The OE verbs had inflexions for the following In subordinate clauses, the normal position of the grammatical categories: inflected verb was at the end: (a) Tense: present and preterit, e.g. fylle- fylde ‘(I) Ac heo weron synfulle, þeah þe heo swa ne wendon… fill-filled’; riseþ-ras ‘(she) rises-rose’. But they were sinful, though they so (did) not think. (b) Person: first, second, third, e.g. fylle-fyllest- fylleþ ‘(I) fill-(you) fill-(he) fills’. The verb do was not used as an auxiliary verb in (c) Number: singular and plural, e.g. fylle -fyllaþ OE. Negative sentences were formed by attaching the ‘(I )fill-(we) fill’. particle ne ‘not’ to the left of the verb, and questions (d) Mood: indicative, imperative, subjunctive, required inversion of the verb and the subject. Multiple e.g. fyllest-fyll(a)-fylle ‘(you) fill-fill!-(should, negation was the norm in OE. would) fill’. Ne genealæceþ him nænig yldo ne sorh ne sar ne deaþ. OE had two types of verbs. The ‘weak’ verbs used Not approaches him no old age, nor sorrow, nor pain, nor death a dental suffix: -de, -te, -ed, -od, etc., to form the preterit and the past participle; fylde ‘filled’, (Ζe)- fylled ‘filled’. Strong verbs changed their root vowel, Vocabulary and Word Formation depending on tense and number: rise, ras, rison ‘rise- (it) rose-(they) rose’. Modern English survivals of this The recorded vocabulary of OE is estimated at approx- group are called ‘irregular’ verbs. Verbs like bind, imately 30,000 words. Only about 3% of these were of choose, drink, eat, ride, run, see belong here. non-Germanic origin. The number of Celtic borrow- Unlike Modern English, which has a very elaborate ings such as binn ‘bin’, torr ‘rock’, is limited. The set of complex verb forms, OE used such forms very influence of Latin is most noticeable in religious and sparingly. The verbs willan ‘want’ and scullan ‘must’ learned texts; abbod ‘abbot’, cantere ‘cantor’, fers were used to express the future tense only occasional- ‘verse’, and Læden ‘Latin’ are some examples. ly, and the verbs have and be had not developed their The most significant outside addition to the OE vocab- auxiliary functions, as in we have arrived, we had been ulary comes from Scandinavian. Although many traveling. The predecessors of the Modern English Scandinavian loanwords were not recorded until after modal verbs, e.g. cunnan ‘know, be able’, magan ‘be the Norman Conquest, OE is arguably the period dur- able’, motan, p.t. moste ‘be allowed to’, were more like ing which about 1,000 words were borrowed from the rest of the verbs in the language in that they had Scandinavian. This component of the vocabulary tense, person, number, and mood inflexions. includes everyday words: Modern English words such In terms of word-order, OE was, to a large extent, a as bank, call, fellow, guess, law, leg, loan, score, skill, ‘verb-second’ language: typically, the inflected verb in sky, skin, take belong in this group. Scandinavian is the main clauses appeared in the second constituent posi- source of the pronouns they, them, their, and the tion. The first position could be filled by the subject, function words till and though. There are also about 779 OLD ENGLISH 1,400 Scandinavian place names in the northern and Alistair Campbell. London, New York: Oxford University eastern parts of England. These are place names end- Press, c. 1921. Campbell, Alistair. 1959. Old English grammar, Oxford: ing in -by ‘settlement’ (Carnaby, Rugby, Thirtleby), - Clarendon Press. thorpe ‘hamlet’ (Barleythorpe, Grimsthorpe, Fulk, Robert D. 1992. A history of Old English meter. Fridaythorpe), -thwaite ‘clearing’ (Hampsthwaite, Philadelphia: University of Pennsylvania Press. Hunderthwaite, Husthwaite). Healey, Antonette di Paolo (ed.) 1998. Dictionary of Old English OE was extremely resourceful in enriching its vocab- Old English Corpus. Ann Arbor: University of Michigan Press. Electronic access: http://ets.umdl.umich.edu/o/oec/. ulary through derivation and compounding. New items Hogg, Richard, M. 1992. A grammar of Old English. Oxford, could be derived by prefixation as in a-sendan ‘send UK; Cambridge, MA., USA: B. Blackwell. away, dispatch’, in-genga ‘invader’, un-wisdom ‘stupid- –––––– (ed.) 1992. The Cambridge history of the English lan- ity’. Frequently used OE suffixes are: -ere, as in fisc-ere guage: Vol. I, The beginnings to 1066. Cambridge: ‘fisher’, -estre, as in spinn-estre ‘female spinner’, -ing, Cambridge University Press. Lass, Roger. 1994. Old English: a historical linguistic compan- as in cyn-ing ‘head of the kin, king’. Other highly pro- ion. Cambridge: Cambridge University Press. ductive OE suffixes were: -dom, -had, -scipe, -nes(s) for Mitchell, Bruce. 1985. Old English syntax, 2 vols. Oxford: nouns, and -ig, -isc, -sum, -en, -fæst, -lic, -leas, -ful for Clarendon. adjectives. Compounding was another prominent vocab- Mitchell, Bruce, and Fred C. Robinson. 1992. A guide to Old ulary enrichment strategy. Some examples are: heafod- English, 5th edition. Oxford, UK; Cambridge, MA., USA: Blackwell. mann ‘head-man, leader’, eorð-cræft ‘earth-craft, Nielsen, Hans Frede. 1998. The continental backgrounds of geography’, wid-sæ ‘wide-sea, ocean’, mild-heort ‘mild- English and its insular development until 1154. Odense: hearted, merciful’. Metaphoric compounds, or kennings, Odense University Press. are characteristic of the OE poetic tradition, e.g. ban- Quirk, Randolph, and C.L. Wrenn. 1994. An Old English gram- cofa ‘chamber for bones, body’, gar-berend ‘sword car- mar, with a supplemental bibliography by Susan E. Deskis. DeKalb: Northern Illinois University Press. rier, warrior’, guðwudu ‘battle wooden piece, spear’. Roberts, Jane Annette, and Christian Kay with Lynne Grundy. 1995. A thesaurus of Old English, 2 vols. London: King’s References College London Medieval Studies 11. Bosworth, Joseph. 1992. An Anglo-Saxon dictionary: based on DONKA MINKOVA the manuscript collections of Joseph Bosworth. Supplement See also Great Britain; Indo-European 2: Germanic by T. Northcote Toller, with rev. and enl. addenda, by Languages Old French ‘Old French’ refers to the set of dialects spoken in the conclusion that a new language has emerged northern and central France from the ninth to the four- (although any firm date is clearly artificial––the teenth centuries. The term is most commonly identi- changes had to have been in effect for some time prior fied, especially in the later stages, with francien, the to their appearance in the Serments). The early OF speech of the Île-de-France region surrounding Paris, period extends from the ninth to the early twelfth cen- because of that variety’s central geographic position tury, and constitutes a period of rapid linguistic change and its association with the royal court and the bur- and considerable regional diversity. The later OF peri- geoning political, economic, and cultural importance od (mid-twelfth to fourteenth century) sees the estab- of the capital. lishment of the University of Paris and the law courts, The birth of Old French (henceforth OF) is tradi- the flourishing of literary works (by no means limited tionally linked with the Serments de Strasbourg (AD to the dialect of the Ile-de-France), and the increasing 842), a document containing oaths sworn by the troops prestige and expansion of francien. of two of Charlemagne’s grandsons concerning the How might one characterize linguistically this new partition of his empire. This document contains struc- language? (We must bear in mind that any summary of tures (particularly a new future form) sufficiently dif- five centuries of evolution is inevitably far from com- ferent from Classical Latin (henceforth CL) to allow plete––we concentrate here on one representation of 780  OLD FRENCH the ‘classic’ forms of the twelfth century and on the Declensional classes for nouns are illustrated below major differences leading from Latin through to for murs ‘wall’, rose ‘rose’, ber ‘baron’, and none Modern French [MF].) In its pronunciation, OF lost ‘nun’, respectively. the distinction between long and short vowels charac- OF declensional classes (major types only, includ- teristic of Latin; often modified stressed vowels (BRE- ing definite articles; imparisyllabic forms have a dif- VIS /bewis/ > OF brief /bjf/ > MF bref 1 ‘brief’, ferent number of syllables in the nominative singular HORA /hoRa/ > OF eure /ewr/ > MF heure ‘hour’); as compared to the remaining forms; ‘oblique’ refers deleted many unstressed or final vowels (except /a/ to an amalgam of all nonnominative cases): and those vowels following a final group of conso- nants [final /a/ becomes //]: MALUM > mal ‘bad’, ARBOREM > arbre ‘tree’, GUTTA > goute ‘drop’); Masculine Feminine Double Stems introduced the affricates [t d ts dz] through palatal- (Impari- syllabic) ization, as in OF chalt ‘hot’, jorn ‘day’, cent ‘hun- dred’, onze ‘eleven’, respectively; began a long-term Nominative li murs la rose li ber la none process of vowel nasalization culminating in forms singular Nominative le mur la rose le baron la nonain such as MF un [œ] ‘one’ < UNUM, chien [j] ‘dog’ < plural CANEM, pont [p] ‘bridge’ < PONTEM; and deleted Oblique li mur les roses li baron les nonains or weakened consonants between vowels, as in VITA singular > vie ‘life’, SAPONEM > savon ‘soap’. Oblique les murs les roses les barons les nonains In word structure, the changes were no less dramat- plural ic. The loss of final vowels had profound effects on (Thus, La none li baron veit. ‘The nun sees the baron.’) many grammatically relevant suffixes. Among the nouns and adjectives, the original CL system used suf- fixes to distinguish five separate classes of nouns, In the verb system, equally striking losses occurred. three genders (masculine, feminine, and neuter), two Many distinctions (passive forms, for example) disap- numbers (singular and plural), and six cases (nomina- peared completely. A new future tense arose, built with tive, genitive, dative, accusative, vocative, ablative, the infinitive and suffixes based on HABERE ‘to have’: each with various syntactic roles indicating, among CANTABO (the older form) > CANTARE HABEO ‘I other things, subject [nominative case], direct object have to/will sing’ > OF chanterai ‘I will sing.’ Verb [accusative] and indirect object [dative], possession endings also eroded, so that person/number distinctions [genitive], terms of address [vocative], and so on). (I, you, he/she/it [sg.]; we, you, they [pl.]), formerly These were reduced in OF to three classes of nouns, indicated by suffixes, were also threatened. As a result, two genders, two numbers, and two cases. With the pronoun use gradually increased, especially in the sin- disappearance of the final vowels, the suffixes disap- gular as a way of preserving the former differences. peared in large part, so that the syntactic distinctions The form of OF words was significantly more vari- originally indicated by these suffixes were also direct- able than in the modern language. The effects of earli- ly affected. As a result, we see compensating develop- er sound change often resulted in different forms of ments in the increased use of prepositions (e.g. le livre noun and verb stems that were regularized subsequent de Cicéron ‘the book of Cicero = Cicero’s book’ rather to the OF period (OF truef––trovons > MF trouve–– than LIBER CICERONIS), and OF word order pat- trouvons ‘find’ [1 sg.––1 pl.]; serf––sers > MF serfs terns become more constrained than in CL (where, ––serfs ‘serf’ [sg.––pl.]; OF larc––large [masc.–– because suffixes indicated the words’ function, words fem.] > MF large ‘wide’). Fusion of unstressed forms could appear in virtually any order depending on sty- is also widespread: a le > au ‘to + the (sg.)’, en les > listic factors [CANEM [‘dog–accusative’] HOMO es ‘in + the (pl.)’, je le > jel ‘I + him/it’, ne les > nes [man––nominative] VIDET > l’homme voit le chien ‘neg + them’, qui me > quim ‘who + me’, si me > sim ‘The man sees the dog’]). The determiner system also ‘if + me’, and numerous others. underwent radical change, with great increases in the In its sentence structure, OF still permitted relative- use of demonstratives (e.g. OF cest ‘this’, cel ‘that’) ly free word order compared to the modern language, and particularly articles (e.g. OF li, le, la, les ‘the (sg. but certain patterns, particularly those where the verb is and pl.)’, uns, un, une ‘a (masc. and fem.)’). in the second position in the sentence, began to pre- dominate: subject–verb–object: Li vilains apele son fil. ‘The peasant calls his son.’ [Fabliaux 3, l. 39]; the fre- 1 The symbol ‘>’ indicates the direction of development: quent object–verb–subject (– X): Ses barons fist li rois Latin BREVIS becomes (>) OF brief. Conversely, ‘<’ indi- venir. ‘The king had his barons come.’ [Le roman de cates the source: OF chien comes from (<) Latin CANEM. Renart, l. 1807]; De venoison ont grant plente. ‘They 781 OLD FRENCH have a lot of venison.’ [Tristran, l. 1773]; and ny’); and importations from a variety of other sources subject–object–verb: Li rois Tristran menace. ‘The (Celtic chemin ‘road’, if ‘yew tree’; Arabic alchimie king threatens Tristran.’ [Tristran, l. 770]. Given the ‘alchemy’; Greek eglise ‘church’; Occitan amour disappearance of the CL passive verb forms, it is not ‘love’ [notably via the influence of the Troubadour surprising that replacements of these CL constructions poets], and many others). Nonetheless, the unmistak- appeared based on estre ‘to be’ plus the past participle ably major lexical source is Latin. (La pucele fut donc pendue. ‘They hanged the young But despite this parentage, OF is clearly French, no girl’ [Vie de sainte Marguerite, l. 1]); on the increased longer a ‘corrupted’ Latin, and has distanced itself use of pronouns such as on/l’en (On me desrobe en from its origins much more than such well-known vostre terre. ‘I am being robbed in your land.’ [Fabliaux Romance relatives as Italian, Occitan, Catalan, or 11, l. 191–2]); or on the expansion of constructions Spanish. While not easily comprehensible to modern with the reflexive pronoun se (Carles se dort. ‘Charles readers, it leaves a stong impression of familiarity, and falls asleep.’ [Chanson de Roland, l. 724]). OF sentence should inspire us as a source of great cultural and lin- patterns, in other words, remind us much more of the guistic richness. modern language than of CL. In vocabuary, the Latin origins of OF are clear ––the great majority of lexical items descend directly References from Latin, although often from informal or popular Buridant, Claude. 2000. Grammaire nouvelle de l’ancien rather than classical speech. In addition to popular français. Paris: SEDES. items, which have become the norm (TESTA ‘jug’ > Einhorn, E. 1974. Old French. A concise handbook. OF teste ‘head’; CABALLUS ‘nag, packhorse’ > OF Cambridge, New York: Cambridge University Press. Harris, Martin. 1978. The evolution of French syntax. A com- cheval ‘horse’; GENUS ‘knee’ — GENUCULUM parative approach. London, New York: Longman. [vernacular diminutive form] > OF genoil [Modern Kibler, William. 1984. An introduction to Old French. New French genou] ‘knee’), we see innovative use of many York: The Modern Language Association of America. suffixes (e.g. OF -age used to indicate taxes of various Pope, Mildred. 1934. From Latin to Old French with especial types: abeillage, arivage, cheminage, melage [on bees, consideration of Anglo-Norman. Manchester: Manchester University Press. docking, roads and apples, respectively]; Germanic Price, Glanville. 1971. The French language: present and past. loans from the early Frankish conquerors (biere ‘cof- London: Edward Arnold. fin’, helme ‘helmet’, honte ‘shame’); learned words DOUGLAS C. WALKER entering from religious or legal texts (avaricieux ‘avaricious’, crestiien ‘Christian’, testimonie ‘testimo- See also France; Indo-European 4: Romance; Latin Old High German The German language is estimated to be about 1,500 fixing of main intonational stress on the initial syllable years old based on the earliest written sources, though of the word, and a variety of vowel changes. as a spoken language it is almost surely older. Old The origins of the Germanic peoples––the speakers High German is the traditional designation of the of the ancestor language Proto-Germanic––are cloaked German language in its earliest stage, roughly between in mystery, as indeed are the origins of the Indo- the years 750 and 1050 CE (some scholars use the Europeans. We are relatively certain that sometime period 600–1100). prior to 1000 BCE Germanic-speaking people occu- German is a member of the Germanic subfamily of pied their primeval home (Urheimat) in what is today the ‘centum’ branch of the Indo-European language the area comprising southern Sweden and Norway, family. The linguistic features that set Germanic off Denmark, and northern Germany. Between 1000 BCE against other Indo-European languages are several, and 500 BCE, at least some of the Germanic tribes principal among which are the First Sound Shift (also began to move away from their original home, migrat- called the Germanic Sound Shift, the First or Germanic ing farther north or striking out to the south and the Consonant Shift, or Grimm’s Law), Verner’s Law, the east. The reason for this migration was doubtless due in 782  OLD HIGH GERMAN part to inundation of the Urheimat, much of which is While there are competing representations for illus- covered today by the relatively shallow North Sea. trating the relationships among the Germanic family Out of this Völkerwanderung (‘migration of peo- depending on which criteria are given priority, the most ples’) arose the traditional classification of the widely used and traditional genetic classification of the Germanic languages into East, North, and West Germanic languages is a tripartite one into East, North, Germanic. German belongs to the latter group, as do and West. The family tree (stammbaum) of Proto- English, Dutch, Frisian, and Low German. We know Germanic can be represented (and simplified) as fol- nothing for certain about the people who occupied this lows. (The terms ‘Ingvaeonic’ and ‘Istvaeonic’ refer to territory before the Germanic tribes arrived. We know the names of Germanic tribes; ‘North Sea Germanic’ nothing about the early contacts between the Germanic and ‘Weser-Rhine Germanic’ are used almost inter- intruders and the autochthonous inhabitants of the changeably with ‘Ingvaeonic’ and ‘Istvaeonic’. Old Urheimat, nor do we know what language(s) the latter Low Franconian gave us Dutch, and Old Saxon was the spoke. The general rule of thumb in historical linguis- ancestral language of modern Low German, which is tics is that the greater the degree of contact between spoken today in northern Germany.) (See Figure 1.) languages, the greater the amount of language change. The principal linguistic change that set Old High The changes in pronunciation and vocabulary that German apart from its other West Germanic sibling Germanic languages underwent (the First Sound Shift languages was the Old High German Consonant Shift, and others) were extensive compared with more con- also called the Second Consonant Shift or the Second servative Indo-European languages. Moreover, approx- Sound Shift. This, along with the First Sound Shift dif- imately one-third of the vocabulary of the Germanic ferentiating Germanic from Indo-European, is one of languages is from other than Indo-European stock. the most illustrious sound changes in Germanic lin- Therefore, we assume that contacts with the autochtho- guistics. It affected the voiceless stops /p t k/, changing nous inhabitants were deep and extensive––and that the them according to the phonetic environment into languages of these speakers had many fricative conso- affricates /pf ts kx/ or fricatives /f s x/, the latter of nants (the source of English f, th, s, sh). which could occur as both long (geminated) /ff ss xx/ East Germanic Gothic (†) Icelandic West Norse Old Norwegian North Germanic Old Danish East Norse Old Swedish Proto-Germanic Old English Old Frisian Ingvaeonic West Germamic Old Low Franconian Old Saxon Istvaeonic Old High German Figure 1 783 OLD HIGH GERMAN and short /f s x/. In general outlines, the Old High are spoken in an east–west band lying between Low German Consonant Shift went as follows: German in the north and Upper German in the south. The principal Central German dialects are Franconian Pre-Old High Old High Phonetic (Middle Franconian, Rhine Franconian, East German German Environment Franconian) and East Central German, which overlaps ptk > pf ts kx Word-initially and with Thuringian, Upper Saxon, and Silesian. following m, n, l, r Upper German dialects are characterized by greater > ff ss xx Medially following completeness in the effects of the Old High German a short vowel Consonant Shift. It is assumed therefore that the Old > fsx Medially following High German Consonant Shift began in the far south a long vowel and and spread toward the north: the underlying idea is that word-finally sound change goes further toward completion in the We can see the effects of the Old High German area it has been around in longest. Most Upper Consonant Shift from a comparison of modern German dialects have all of the changes stated above German words with their English cognates: as comprising the Old High German consonant shift. Central German dialects show more mixed results. Modern German English Central German dialects are most consistent in the Pfund pound shift of k to x and less consistent in the shifts of t to s Pfeffer pepper and p to f. Thus, where Upper German dialects will Wasser water consistently have fricatives in ich ‘I’, dorf ‘village’, machen make and das ‘that’, Central German dialects can be found Zeit tide with ich, dorf, dat and ich, dorp, dat. helfen help The distinctive sounds of Old High German are: auf up kochen cook Vowels An additional sound change of Pre-Old High German Short Long Diphthongs d to Old High German t is considered by some lin- i u à u¯ ei ou guists to be part of the Old High German Consonant e o e¯ o¯ ie uo Shift. Its effects are seen in cognates such as English a a¯ do, day, ride vs. German tun, Tag, reiten. The designation ‘High’ has two meanings in the Consonants context of High vs. Low German. On the one hand, it Labial Dental Palatal Velar Glottal is a purely linguistic designation for those dialects of Stops vl. p t k West Germanic that underwent any part of the Old vd. b d g High German Consonant Shift. Its original meaning in Fricatives vl. f s s. x this context, however, referred to altitude: those Affricates vl. pf ts kx dialects of German that did not undergo the Old High Nasals m n German Consonant Shift are located in northern Liquids l r Germany on the flatlands bordering the North and Semivowels w y h Baltic Seas––‘low lying’in other words, hence ‘Low’ German––and the dialects of German that did undergo ([s] and [s.] are respectively dorsal––formed with the the Old High German Consonant Shift are located in blade of the tongue against the alveolar ridge––and api- the higher regions of southern and central Germany, cal––formed with the tip of the tongue against the alve- Switzerland, and Austria, thus ‘High German’. One olar ridge. They have different historical sources: [s] is occasionally encounters the use of ‘High’ to mean created by the shift of t to s in the Old High German ‘good, correct, elevated, superior, grammatical’ and Consonant Shift; [s.] continues Indo-European s). ‘Low’ to mean the opposite, linguistically, of those The back vowels [u o a u¯ o¯ a¯] had front variants qualities. One does not encourage this usage. [ü ö e æ ü¯ ö¯ æ¯ ] before syllables containing a high front Old High German is divided into Central German and vowel or semivowel [i à y], thus gesti ‘guests’, skôni Upper German dialects. Upper German consists of those ‘beautiful’, and ubil ‘evil’ were phonetically [gesti dialects spoken in southern Germany, Switzerland, and skoni übil]. These are the so-called ‘umlaut’ sounds of Austria. The principal Upper German dialects of Old German (Gäste, schön, übel). High German are Alemannic, spoken in present-day The German language is traditionally divided into Switzerland, and Bavarian, spoken today in the German four periods: Old High German (750–1050); Middle state of Bavaria and in Austria. Central German dialects High German (1050–1350); Early New High German 784  OLD HIGH GERMAN (1350–1650); and New High German (1650–present). between the kings Louis and Charles of the Franks, The division between Old High German and Middle sons of Charlemagne. Because the eastern Franks High German is dated from the reduction of the spoke Old Low German and the western Franks Old unstressed vowels [i e a o u] into a single sound schwa French, the oaths were composed in both languages, [ə]. Since this did not occur in a single day, the transi- thus giving us linguistic information as useful for a tion from Old High German to Middle High German student of the history of the French language as for a is variously placed between 1050 and 1100, at which student of the history of German. time the reduction of unstressed vowels to schwa was There is in Old High German a certain amount of complete except in remote and isolated dialects. From original composition not derived from Latin transla- the beginning of the Common Era, Old High German tions and more substantial than ‘give me my horse’. was in contact with Romans during the expansionist The Muspilli is a religious poem of some length that is phase of the Roman Empire. Lexical borrowings were strikingly original, what we have of it, and in the later heavy, and their semantic domains reveal much about Old High German period there are Memento Mori and the nature of these early contacts: ‘wine’ (Old High The song of Ezzo. German wîn, Latin vinum), ‘arrow’ (Old High German It must be admitted that Old High German literature pfîl, Latin pilum). is inferior in comparison with the contemporaneous lit- The earliest written records of the Old High erature of Old English (Beowulf) or even its close rela- German language are words and fragments written in tive, Old Saxon (Heliand). It is vastly inferior to the runes. Somewhat later, after the Germanic tribes had rich medieval literature of Middle High German with been Christianized, we find Old High German glosses its courtly epics, its Nibelungenlied, its poetry. Old interspersed in Latin texts. In the Vocabularius Scti High German literature is not equal to the slightly later Galli, slightly later than 765 CE, we find sapiens literature of Old Icelandic. There is very little original wizzo (in normalized Old High German spelling), sci- in Old High German literature, much that is derivative. tus wiser, fortis stark. Lengthier texts are found begin- However, its literary deficiencies should not detract ning in the ninth century, many of them translations from its linguistic importance. It was the language of into Old High German of religious writings originally the Second Sound Shift, the Old High German in Latin: Isidor’s diatribe Contra Judaeos, the Consonant Shift, which along with the First Sound Weissenburg Catechism, the gospel harmonies Tatian Shift (Grimm’s Law) and Verner’s Law is one of the and Otfrid, and the translation of Boethius made by great defining phonological events in the Germanic Notker, an especially talented monk from the family of languages. The Old High German Consonant monastery of Saint Gall in Switzerland. Shift—its causality, its spread, how it happened—has Typical of the genre of most early Old High provided and continues to provide historical linguistics German writing is the Lord’s Prayer from St. Gall: with grist for its theoretical mill. The importance of a Fater unsêr, dû bist in himile, wîhi namun dînan, kweme language must never be judged by its literature alone. rîhhi dîn, werde willo dîn, sô in himile sôsa in erdu. brôd unsêr emezzihig gib uns hiutu, oblâz uns skuldi References unsêro, sô wir oblâzêm uns skuldîgêm, enti ni unsih fir- Barber, C. Clyde. 1951. An Old High German reader. Oxford, leiti in khorunka, ûzzer lôsi unsih fona ubile. Blackwell. Braune, Wilhelm. 1987. Althochdeutsche Grammatik. Typical too were ‘How to say it in Old High Tübingen: Max Niemeyer (numerous editions). German’ booklets such as the ‘Conversations from Hawkins, John A. 1987. ‘German’ The world’s major languages, Paris’, written presumably for monks literate in Latin ed. by Bernard Comrie. New York: Oxford University Press. for traveling in Germany (the unnormalized spelling of Lockwood, W.R. 1968. Historical German syntax. Oxford: Clarendon Press. Old High German points clearly and amusingly to the Murdoch, Brian O. 1983. Old High German literature. Boston: French background of the monk who wrote the thing): Twayne Publishers. Gueliche lande cumen ger? (de qua patria?) ‘What land Priebsch, R., and W.E. Collinson. 1962. The German language. London: Faber and Faber. do you come from?’ Waterman, John T. 1976. A history of the German language. Guer is tin erro? (ubi est senior tuus?) ‘Where is your Seattle: University of Washington Press. master?’ Wells, C.J. 1985. German: a linguistic history to 1945. Oxford: Ne guez. (nescio.) ‘I don’t know.’ Clarendon Press. Gimer min ros. (da mihi meum equum.) ‘Give me my Wright, Joseph. 1906. An Old High German primer. Oxford: horse.’ Clarendon Press. ROBERT D. KING Of particular cultural and historical interest are the ‘Oaths of Strassburg’, which date from 843. These are See also Germany; Grimm, Jacob; Indo-European oaths of allegiance intended to end a fratricidal rivalry 2: Germanic Languages 785 OLD IRISH Old Irish Old Irish is a now extinct Indo-European language glosses, dating in manuscript from around 845 CE but belonging to the Insular Celtic branch of the Celtic containing copies of earlier material. To these texts language family. Insular Celtic is that branch of Celtic can be added a number of further sources consisting of that is (relatively speaking) indigenous to the British additional glosses as well as a small number of con- Isles. Old Irish is the direct ancestor of present-day tinuous texts such as poems. In addition to the Irish, as well as of Scots Gaelic and Manx. Classical Old Irish texts, there also exist a small num- Within earlier Insular Celtic, it is possible to distin- ber of older texts from the 600s CE. guish two major dialects, known respectively as Q- It is generally agreed that further Old Irish material Celtic (sometimes called Goidelic or Gaelic) and is also preserved in later manuscripts, and, indeed, it P-Celtic (sometimes called Brythonic or Britannic). has been argued that some of this material illustrates a These labels have been retained by linguists to classify period of Old Irish that pre-dates the earliest contem- the later distinct Celtic languages. Q-Celtic is so called porary manuscripts. Certainly, some of these texts do because it shows /k/ as the descendent of an original show differences in comparison with other texts of the Indo-European /*kw/. (The name ‘Q-’, rather than ‘K-’, same date, notably in word order. However, this is a Celtic originates in Latin spelling.) In contrast, P-Celtic disputed area, and it has been argued convincingly that shows /p/ as the descendent of /*kw/. Irish, along with at least some of this variation may be due more to Scots Gaelic and Manx, belongs to the Q-Celtic group, genre than to date. The texts may also have come while the P-Celtic group contains Welsh and Cornish, under the influence of Latin models, and some of their as well as Breton. We may contrast the development of distinctive characteristics may in part reflect interfer- the inherited word for ‘four’, which in Old Irish ence from Latin. Thus, the contemporary Old Irish became cethair but in Old Welsh became petguar. manuscripts still form the main, and safest, basis for The earliest surviving evidence for Q-Celtic comes the linguistic description of Old Irish, although schol- from inscriptions written in the native Ogham alpha- ars vary as to how far they will admit evidence from bet, which date from the fourth to seventh centuries the later manuscripts. CE. These consist primarily of proper names, and so The Old Irish texts that have come down to us are the amount of grammatical information that they can linguistically very homogeneous and show rather little provide is limited; however, they do provide informa- evidence of dialect variation. The classical view has tion about the sound system and word structure of Q- been that what variation does exist is due to differ- Celtic in the immediately pre-Old-Irish period. ences in date, although it has been tentatively suggest- The Old Irish period proper is normally considered ed that these differences do, in fact, reflect dialectal to fall between 600 and 900 CE, and it is from this variation. Certainly, the relative homogeneity of the period that the earliest literary evidence for Irish Old Irish sources should not be taken to indicate that comes. Within the Old Irish period are sometimes fur- some degree of dialect variation did not exist within ther distinguished an Archaic Old Irish period (prior to Old Irish; indeed, it would be unusual if this were the the 700s CE) and a Later (or Classical) Old Irish peri- case, since the language was spoken in the whole of od (the 700s and 800s CE). Ogham writing, inciden- Ireland and in a large part of Scotland. We do not know tally, continued to be taught until around the how small or substantial these dialect differences may seventeenth century, but its role after the inscriptional have been in the spoken language. As regards the period of Q-Celtic was greatly restricted and most Old homogeneity of the literary sources, it should be Irish material is thus written using a form of the remembered that they were written by scholars in Roman alphabet. monastic communities, and some degree of dialect The corpus of contemporary texts that represents mixing, and possibly informal standardization, will Old Irish is relatively limited in both size and genre. undoubtedly have taken place. The most important texts, dating from the Classical The vocabulary of Old Irish is predominantly of Q- Old Irish period, consist of marginal notes (or ‘gloss- Celtic origin, but also contains substantial loan mate- es’) on Latin manuscripts. Of these, the most impor- rial, primarily from P-Celtic and from Latin. Some of tant are the Würzburg glosses, dated mainly to around the Latin loans have undergone sound changes that 750 CE, the Milan glosses, normally considered to suggest they were borrowed indirectly via P-Celtic, date from sometime before 825 CE, and the St. Gall while others seem to have been borrowed directly. 786  OLD IRISH In structural terms, Old Irish has a number of fea- absolute and conjunct endings and the near disappear- tures that are interesting from a historical and compar- ance of the prototonic and deuterotonic distinction. ative perspective. On the other hand, Old Irish also has features that On the one hand, the evolution of Old Irish to mod- are characteristic of itself and of the wider Celtic lan- ern Irish shows trends toward the breakdown of com- guage family, in contrast with other Indo-European plex systems of functional endings (inflexions), languages. simplification, and the atomization of meaning-carrying In common with the other Celtic languages, Old elements that are paralleled in other language Irish has Verb–Subject–Object as its basic word order. groups––for instance, in the development from Latin to The Celtic languages are the only Indo-European lan- modern Romance languages such as Italian and French. guages to have this as their basic word order. There has In the noun, for instance, Old Irish maintained a been some discussion about the emergence of this complex set of inflexions: for example, it had special word order in Old Irish. One suggestion has been that dual forms for pairs of things, as well as a set of five it is the consequence of generalizing the order that case endings indicating grammatical relations such as resulted from suffixed pronouns (see below). subject, object, and possession. The same was also In Old Irish, as in other Celtic languages, the true of the adjective, which agreed with the noun that boundaries between the different language levels of it modified. Although already largely obsolete in Old sentence structure, word structure, and sound structure Irish, owing to the emergence of a fairly fixed word are not discrete. For instance, the initial sounds of Old order, this system of endings was to be retained well Irish words can change in certain contexts (so-called into the modern Irish period (up to the 1600s CE). ‘mutations’). These changes were originally triggered However, it has since broken down substantially, and by the final sound in the preceding word, but a number must already have been breaking down at an earlier of these sounds later disappeared and the pattern of date in the spoken language. Some modern Irish mutation was also extended by analogy to other con- dialects have practically abandoned noun inflexion, texts. Some mutations thus came to be used to mark and even the more conservative standard written lan- grammatical relationships, rather than being the result guage uses a reduced set of endings. of local phonetic influence. This is the case, for exam- The Old Irish verb, too, was exceptionally complex, ple, in certain kinds of relative clause, where the main and has consequently received much attention from verb of the relative clause undergoes mutation. Thus, linguists. The basic distinction was between simple sentence function triggers a change in wordform, verbs and complex verbs, that is, those that were which was originally phonetically motivated. formed from simple verbs by the addition of various A further peculiarity of Old Irish is the frequency preverbal elements. For instance, from beirid (‘bear’) with which certain grammatically interrelated words could be derived compound verbs such as do·beir within a sentence or clause are fused into a single (‘give’), for·beir (‘grow’), as·beir (‘say’), and word unit, where, in contrast, other languages (such as tremi·beir (‘transfer’). In addition to these compound- English) would normally use two or more free-stand- ing preverbal elements, other preverbal affixes existed: ing words. One example of this is the situation where for example, ro was used to distinguish, among other the object of a verb is a pronoun. In Old Irish, pronoun things, between the perfect and simple past (contrast objects were often suffixed to the verb, rather than English ‘I have run’ and ‘I ran’). Each of the two class- standing on their own as separate words–for example, es of verb had, in turn, two full sets of personal end- beirthi (‘bears it’) or iurrus (‘will wound them’). ings, which were used in parallel to one another, Following one of the preverbal elements referred to depending on the grammatical context: in the simple earlier, a pronoun could actually become infixed with- verb, they were known as absolute and conjunct end- in the verb—for example, ro-m·gab (‘has taken me’). ings, and in the compound verb as prototonic and In contrast to this Old Irish tendency, pronouns in deuterotonic endings. In addition, there were special modern Irish may also be free-standing, and the use of endings in the simple verb to indicate a verb’s partici- this option is becoming increasingly frequent, at the pation in a relative clause. Furthermore, as with other expense of the suffixed pronoun construction. A relat- languages such as Latin, pronoun subjects were not ed phenomenon, which modern Irish has retained, is expressed, and the meaning was carried by endings the case of the so-called ‘conjugated prepositions’, in (contrast the unchanging bear in English ‘I bear’ and which forms of the pronoun are also suffixed to prepo- ‘you bear’ with Old Irish biru [‘I bear’] and biri [‘you sitions, resulting in ‘one-word’ expressions such as lat bear’]). One of the main changes distinguishing Old (‘with you’, from la = with and -t = you). Irish from later periods of Irish is the breakdown, and Aside from its importance as a crucial chapter in the consequent simplification, of this complex verbal sys- history of modern Irish and the other Q-Celtic lan- tem, notably the disappearance of the two sets of guages, Old Irish is linguistically important more 787 OLD IRISH generally for two main reasons. Firstly, from a histor- Mac Eoin, Gearóid. 1993 Irish. The Celtic languages, ed. by ical perspective, it provides further evidence for the Martin J. Ball and James Fife. London, New York: Routledge. McCone, Kim. 1996. Prehistoric Old and Middle Irish. universality of certain principles of language change. Progress in Medieval Irish studies, ed. by Kim McCone and Secondly, it is interesting typologically, in that it Katharine Simms. Maynooth: Department of Old Irish, St. exhibits a number of features not normally found in Patrick’s College. Indo-European languages outside the Celtic language ––––––. 1997. The Early Irish verb. Maynooth: An Sagart. family. Sims-Williams, Patrick. 1984. The double system of verbal inflex- ion in Old Irish. Transactions of the Philological Society 82. Strachan, John. 1904. Old Irish paradigms and selections from References the Old Irish glosses. Dublin: School of Irish Learning ; 4th edition, revised by O. Bergin, Dublin: Royal Irish Academy, Ahlqvist, Anders. 1988. Remarks on the question of dialects in 1949. Old Irish. Historical dialectology, regional and social, ed. by Thurneysen, Rudolf. 1946. A grammar of Old Irish. Dublin: Jacek Fisiak. Berlin, New York: Mouton de Gruyter. Dublin Institute for Advanced Studies. Lewis, Henry, and Holger Pedersen. 1937. A concise compara- ANDREW WILSON tive Celtic grammar. Göttingen: Vandenhoeck and Ruprecht.; 3rd edition with supplement, 1989. See also Celtic Languages Old Japanese Old Japanese is the oldest attested representative of the (712 CE), the ‘Nihonshoki (Annals of Japan)’ (720 Japonic (Japanese-Ryukyuan) language family. There CE), and the ‘Fudoki (Gazeteers)’ (713–737 CE); are two major varieties of Old Japanese: (1) Western poetic anthology ‘Man’yôshû (Collection of myriad Old Japanese (seventh to eighth century CE), a lan- leaves [of words])’ (c. 759 CE), prose texts ‘Senmyô guage based on the dialect of Asuka and Nara regions (Imperial Edicts)’ (seventh to eighth century), and (roughly corresponding to modern Nara prefecture); ‘Norito (Liturgies)’ (seventh to ninth century). There and (2) Eastern Old Japanese (eighth century CE), a are several important inscriptions as well, the largest dialectal continuum located roughly in the southern of them being a poetic one: ‘Bussoku seki uta (Songs part of the modern Chûbu and Kantô regions. Western of the Buddha’s footprint)’ (c. 755 CE). Old Japanese was the basis of the literary language of Writing system: Western Old Japanese uses the the time, and consequently there are many more extant writing system known as ‘man’yôgana’ (the Western Old Japanese texts than Eastern Old Japanese ‘Man’yôshû’ writing). Man’yôgana is a syllabic sys- texts. Furthermore, it is quite apparent that the west- tem of writing, where Chinese characters are used ernmost dialects in the Eastern Old Japanese dialect phonetically (to represent syllables), although they can continuum were considerably influenced by the also be used logographically (to represent words or Western Old Japanese standard, and this influence concepts). When Chinese characters are used both gradually diminishes from west to east. There is a phonetically and logographically, there is a clear ten- widespread but mistaken opinion that Western Old dency to write roots or stems logographically and suf- Japanese is practically identical to Middle (Classical) fixes and particles phonetically; however, due to the Japanese. On the contrary, both languages are very dif- syllabic nature of writing, the exact boundaries ferent, with a number of important distinctions found between stems (especially verbal stems) and suffixes in pronunciation, vocabulary, and grammar. may not be reflected in writing. Man’yôgana can be Furthermore, Western Old Japanese is not even a subdivided into two major types according to the his- direct predecessor of Middle Japanese: both are based toric stage of the underlying Chinese pronunciation of on geographically close but not identical dialects. the characters on which man’yôgana is based. The Sources: All Eastern Old Japanese and most later type is based on the Late Middle Chinese charac- Western Old Japanese texts are poetry, although ter readings of the eighth century, and it is used exclu- Western Old Japanese also has two texts that are writ- sively in the Nihonshoki. The earlier type is based on ten in prose. The major noninscriptional texts are: the Early Middle Chinese character readings of the poetry from the ‘Kojiki (Records of Ancient Matters)’ sixth to seventh century; most likely it was not coined 788  OLD JAPANESE directly in Japan on the basis of these readings, but nativized loanwords, such as e.g. uma ‘horse’, umey was borrowed from Korea. All other Western Old ‘plum blossom’ (both from Old Chinese *mra and Japanese texts use this earlier type of the man’yôgana. *mi), and tera ‘Buddhist temple’ (from Old Korean Eastern Old Japanese also uses this type of Western *tyerV). A few loanwords can be seen in prose texts, Old Japanese writing system, and it is necessary to most from Chinese or Sanskrit (via Chinese interme- note that all Eastern Old Japanese texts are essentially diary). written in Western Old Japanese orthography. Morphology: The description of word structure and Phonology: Western Old Japanese has the following word formation in the present article is based on consonants: p, b [mb], t, d [nd], k, g [ŋg], m, n, s, z [nz], Western Old Japanese, because of its prestigious w, y, r. No consonant is possible in the syllable-final nature and due to the less complete data from Eastern position; thus, only syllables of C(onsonant)V(owel) Old Japanese. Both Western Old Japanese nominal and V structure are found in the language. Consonants and verbal morphology are significantly different from b, d, g, z, and r occur only in word-medial position in any later stages of the language, including Middle native vocabulary. There are no sequences *wu and Japanese (morphological elements and their functions *yi in Western Old Japanese, but the rare exceptions to typical only for Western Old Japanese and not found in the general syllable template of [C]VCVCV…, like later stages of the language are indicated in bold kai ‘oar’, suggest that at least the sequence *yi existed typeface). once at the pre-Western Old Japanese stage. The Nominal morphology: Most nominals (with the Eastern Old Japanese consonant system seems to be exception of some pronouns) in Western Old Japanese identical to Western Old Japanese. have no formal markers, distinguishing them from Western Old Japanese vowels include the following other parts of speech, e.g. yama ‘mountain’, pyito vowels (Yale notation is given with the most likely ‘person’, yukyi ‘snow’, myidu ‘water’, piy ‘fire’, puta phonetic values in brackets): a [a], yi [i], u [u], iy [ï], ‘2’, towo ‘10’. ye [e], o [ə], wo [o], and one diphthong ey [əy], which The following is the list of the most frequent WOJ was previously considered to be a unit vowel [ε]. The affixes. Nominal prefixes: honorific myi-, intensive contrast between vowels yi/iy, ye/ey, and o/wo that ma-, dimunitive wo- and kwo-, locative sa-. Plural suf- later merged as /i/, /e/, and /o/ is not found in every fixes: -ra (neutral plural marker), -domo, -tati, -na. possible position even in Western Old Japanese. Thus, Case markers: active-i, possessive -ga, genitive -no, there is no contrast in initial position for any of the genitive-locative -tu, dative-locatives -ni and -ra, three pairs; there is no contrast between yi/iy and ye/ey accusative-absolutive -wo, comitative -to, ablative - after coronal consonants (such as t, d, s, z), and a sug- ywo/-ywori/-yu/-yuri, directives -gari and pye, com- gested contrast between po and pwo is questioned by parative -no/-ni/nasu/-nosu, and terminative -made. some linguists, although it is likely that there is some Dimunitive suffixes: -ra, rø -kwo. supportive evidence for it in the earliest texts. In all Some Western Old Japanese pronouns have two cases where there is no contrast, it is customary to stems: unextended, and extended with suffix -re, write just i, e, o. which may be treated as a formal marker of these pro- The set of Eastern Old Japanese vowels is marked- nouns, distinguishing them from other parts of speech, ly different: it includes only five vowels: a, i, u, e, o, and also from other pronouns that do not have the spe- although there is the possibility that one or two addi- cial stem in -re. Personal pronouns: wa-/ware 1ps and tional vowels can be hidden behind the Western Old 1pp (rare), a-/are 1ps, na 1ps, na-/nare 2ps and 2pp Japanese spelling system. In any case, vocalic corre- (rare), [myi]masi 2ps, si/so- 3ps. Reflexive pronoun: spondences between Western Old Japanese and ono-/onore. It is necessary to note that Middle Eastern Old Japanese are not straightforward, indicat- Japanese has a unified system of personal reflexive ing that a parent language of both used to have a sys- pronouns (...self), which does not exist in Western Old tem somewhat different from its descendants. Japanese or Eastern Old Japanese. Demonstrative pro- Data on Western Old Japanese accent system are nouns: proximal ko/kore ‘this’, mesial so-/sore ‘that’, controversial: it has been suggested that in the distal ka/kare ‘that over there’. Demonstrative pro- ‘Nihonshoki’, Chinese characters with level tone were nouns indicating place or direction: proximal: koko, used to transcribe Western Old Japanese low pitch and koti, konata ‘here’; mesial: soko ‘there’; distal: kanata Chinese characters with other tones to transcribe ‘over there’. The distal forms in Western Old Japanese Western Old Japanese high pitch, but it appears to be and Eastern Old Japanese may be an innovation, as true only statistically. they are rather rare. In Western Old Japanese there is Lexicon: Japanese poetic texts did not use words another distal demonstrative pronoun woti/woto/wote that, at the time, were obvious loanwords until quite that is used more frequently than ka/kare, but its usage late, so poetic texts contain only native vocabulary and is predominantly limited to indicating distal place or 789 OLD JAPANESE time. Interrogative pronouns: ta-/tare ‘who’, nani slot, whereas suffixal positions are multiple. There are ‘what’, ika ‘how’, idu/iduku ‘where’, iduti ‘where to’, a sizeable number of verbal prefixes, most of which idure ‘which’, itu ‘when’, iku/ikuda/ikura ‘how occur only in Old Japanese—i-, ka-, ari-, ta-, etc.— many’, nado/nazo ‘why’. Collective pronouns: miyna, with mostly unclear or poorly investigated functions. moro ‘all’. Verbal suffixes can be divided into word-final and The numerical system of Western Old Japanese word-nonfinal, the main difference between these two sharply contrasts with later stages of the language; groups being that the second group cannot conclude numerals of Chinese origin are not yet used, or at least the verbal form and is always followed by another suf- they are not present in the texts. The main problem with fix. The rules governing the combination of verbal attestation of numerals is that they are frequently writ- roots and suffixes are quite complex. However, a gen- ten logographically, and the phonetic attestations are eral rule of thumb is that suffixes with initial vowel lacking. The following cardinal numerals are attested keep the vowel after consonantal roots, and lose it after phonetically: pyito ‘1’, puta ‘2’, myi ‘3’, yo ‘4’, itu ‘5’, vowel roots, while consonant-initial suffixes keep ini- mu ‘6’, nana ‘7’, ya ‘8’, kokono ‘9, towo ‘10’. After tial consonant after vowel roots, but lose it after con- towo ‘10’, attested tens include pata ‘20’, myi-swo ‘30’, sonantal roots. Since some vowel-initial suffixes do yo-swo ‘40’, and ya-swo ‘80’. There is only one pho- not lose their initial vowels after vowel roots, but netic attestation of a numeral including both tens and rather cause the last vowel of the verbal root to be digits: myiswo-ti amar-i puta-tu ‘32’ (30-class exceed- dropped, it is convenient to subdivide vowel verbs into inf 2-class) that, alongside later glosses to numerals strong vowel verbs (that never lose their final vowels) written logographically in Western Old Japanese texts, and weak vowel verbs (that lose their final vowels provides us with the way in which the digits were before certain suffixes). Due to these complexities, it added to tens: tens + classifier + infinitive of the verb is necessary to list verbal affixes in two forms: after amar- ‘to exceed’ + digit + classifier. Among higher consonantal stems and after vowel stems (with the numerals, only the following are attested phonetically: possible differentiation between forms found after mwomwo ‘100’, i-pwo ‘500’, ya-pwo ‘800’, ti ‘1,000’, weak and strong vowel verbs). Thus, for example the yorodu ‘10,000’. The system of classifiers is in its attributive suffix has the following forms: -u (<*-uru) infancy, and only the following classifiers are attested: - after consonantal verb, -uru after weak vowel verb, tu (objects used with digits), -ti (objects used with tens and -ru after strong vowel verb: yuk-u pyito ‘person and hundreds), -ri (people), -moto (grassy plants), -pye who goes’, kwop-uru pyito ‘person who loves’, and (layers and folds), -ka (days). myi-ru pyito “person who sees”. Verbal morphology: One of the greatest differences It is impossible to provide a list of all verbal affix- between Western Old Japanese and all later stages of es here, so only the most important are listed: infinitive the language is that quality verbs like modern -i/-Ø, negative infinitive -azu/-zu, finite -u, attributive Japanese siro-i ‘white’ or Middle Japanese siro-ki ‘id’ -u/-uru/-ru, imperative -ye/-yo, negative -an-/-az-/-n-/- are still in the process of formation. This is because in z-, tentative -am-/-m-, volunative -ana/-na, iterative - Western Old Japanese stems of quality verbs can still ap, passive -ye-/?-raye-, causative -asimey-/-simey-, behave as regular adjectives: e.g. sira namyi ‘white debitive -ubey-, negative debitive -umasizi-, nominal- waves’, opo kapa ‘big river’, cf. Middle Japanese siro- izer -aku, honorific -as-. ki nami ‘white waves’ and opo-ki kafa ‘big river’ with Besides suffixation, Western Old Japanese widely the obligatory attributive suffix -ki, which is not oblig- uses some auxiliaries after the infinitives: past tense atory in Western Old Japanese. Therefore, Western -kyi, retrospective -kyer-, perfectives -te-/-t- and -n-, Old Japanese adjectives are in transition from nomi- perfective-progressive -tar-. In the later stages of lan- nals to verbs. guage, these grammaticalized auxiliaries became Western Old Japanese verbs are divided into sever- inseparable from the infinitive marker -i, and thus al classes: consonantal verbs, with roots ending in a formed new secondary suffixes. consonant (yuk- ‘go’, kyir- ‘cut’, omop- ‘love, think’, Syntax. Old Japanese syntax remains the most poor- nokos- ‘leave’, etc.), vowel verbs, with roots ending in ly studied area of Old Japanese grammar. Nevertheless, a vowel (kwopiy- ‘love’, tasukey- ‘save’, myi- ‘see’, it appears that there were some significant differences miy- ‘turn’, etc.), and irregular verbs (ko- ‘come’, se- with later periods of the language, particularly con- ‘do’, ar- ‘exist’, sin- ‘die’, in- ‘go [away]’). In addi- cerning embedded clauses. tion, there are defective verbs—n-, to ‘be’, to ‘say’— that have only a limited number of forms and are irregular as well. References Western Old Japanese verbs can take both prefixes Bentley, John R. 2001. A descriptive grammar of Early Old and suffixes. The prefixal position is limited to one Japanese prose. Leiden, Boston, Köln: Brill. 790  OLD NORSE Hendriks, Peter. 1992. Adverbial and adnominal modification Vovin, Alexander. 1997. On the syntactic typology of in Old Japanese: copular particles and predications. New Old Japanese. Journal of East Asian Linguistics 6(3). Haven, CT:Yale University Ph.D. Dissertation. 273–90. Martin, Samuel E. 1987. The Japanese language through time. ——— 2005. A descriptive and comparative grammar of New Haven, London: Yale University Press. Western Old Japanese, Vol. 1. London: Global Oriental. Miyake, Marc H. 2003. Old Japanese: a phonetic reconstruc- Whitman, John B. 1985. The phonological basis for the com- tion. London: Routledge Curzon. parison of Japanese and Korean. New Haven, CT: Yale Takeuchi, Lone. 1999. The structure and history of Japanese: from University Ph.D. Dissertation. Yamatokotoba to Nihongo. London, New York: Longman. ALEXANDER VOVIN Old Norse Old Norse (ON), in the widest sense of the term, refers it might be regarded as Proto-North-West Germanic to the varieties of the North Germanic family used in rather than specifically Norse. Thus, PrGmc *ae Scandinavia and its colonies from the breakup of its became NGmc and WGmc *a but EGmc *e e.g. ON parent Germanic until around the fourteenth century. mani OS mano vs. Got mena ‘moon’; likewise, PrGmc A narrower usage would restrict ON to Old Norwegian *z developed into NGmc and WGmc *r (in NGmc, per- (ONo) and Old Icelandic (OIc) between the seventh haps via some kind of palatalized fricative, usually rep- and fifteenth centuries; still more restrictively, resented as R) but remained in EGmc (or was Classical ON (ClON) refers to the literary language of devoiced,) e.g. ON eyra OHG ora vs. Got auso ‘ear’. Iceland from around 1150 to 1350, a period well However, some have regarded parallels between NGmc attested in saga literature, and it is this variety from and EGmc as more significant for grouping the lan- which examples will generally be drawn. guages, e.g. PrGmc *ww gave NGmc and EGmc *ggw As a member of the Germanic branch of the Indo- but WGmc *uw (then becoming a diphthong), e.g. ON European family, ON preserves a number of archaic tryggr (acc. tryggvan) Got triggws vs. OE triewe ‘true’. features, which make it useful for comparison with The ‘Norse’ of this period is attested in a number of other early Indo-European languages. For nouns, two inscriptions in the runic alphabet (the older futhark, con- numbers, three genders, and four cases are distin- taining 24 runes), some of which predate the fourth-cen- guished; around 14 declension patterns occur, based tury Gothic biblical translation, and are thus the earliest on original stem type. Adjectives and pronouns agree direct evidence for Germanic. They are found mainly in with nouns in number, gender, and case. Many verbs Denmark and southern Norway and Sweden, and reflect retain the ancient vowel ablaut to mark tense, e.g. a time when the phonology and morphology of PrGmc bita—beit—bitinn ‘to bite—bit (sg.)—bitten’ (com- were still well preserved. Thus, the Gallehus horn pare Greek leipo—leloipa—elipon). From the Proto- inscription from Denmark from around 400 CE reads ek Germanic (PrGmc) period, evidence of the Grimm’s hlewagastiR holtijaR horna tawido—‘I, Hlegest son of Law consonantal shift remains, e.g. fiskr ‘fish’, kinn Holti, made (the) horn’, which would produce the short- ‘cheek’ (compare Latin piscis, gena), and more indi- er ClON equivalent ek Hlegestr Holtir horn tadha. rectly, of Verner’s Law, e.g. finna—fundu ‘to find— Between roughly the sixth and tenth centuries, vari- they found’, vas—varu ‘was—were’. New dual first ous important changes further distinguished Norse and second person pronouns, a double declension of from the rest of Germanic, while internal distinctions adjectives, and a past tense suffix containing a dental were minimal. This period might be termed Common consonant are also PrGmc innovations retained in ON. or Viking Norse, since from around 800 to 1000 viking The earliest period of Norse history covers the grad- expansions carried Norse speakers outside Scandinavia ual separation of the language from PrGmc, approxi- itself, sometimes resulting in lasting settlements: from mately between the first and sixth centuries CE. This Norway, west to Scotland, Ireland, the Faroes, Iceland, era is often termed urnordische/urnordisk (Proto- and Greenland; from Denmark, to England, northern Norse/Ancient Scandinavian, etc.), although since France, and eastern Prussia; from Sweden, east to many changes were still shared with WGmc (ancestor Gotland and Russia. The language is still mainly of Old High German (OHG), Old English (OE), Old known from runic inscriptions, largely in the younger Saxon (OS), etc.) but not with EGmc (i.e. Gothic (Got)), 16-rune futhark, although much poetry preserved in 791 OLD NORSE later manuscripts, in the court (skaldic) and heroic From the tenth century onward, dialectal differ- (eddaic) traditions, was composed orally in this period. ences began to develop between West Norse (WN), During Common Norse (CN), unstressed vowels spoken in Norway and its colonies (chiefly Iceland), were reduced to only three qualities, ceased to distin- and East Norse (EN), spoken in Sweden, Denmark and guish length, and were completely lost in many pho- their colonies (although Gutnish might be regarded as netic environments. This caused the loss of a number intermediate between WN and EN). A notable distinc- of morphological markers and hence several mergers tion was the EN monophthongization of CN [Ei] > [e:] and analogical reformations in the nominal system. and [ou], [Ey] > [o:], e.g. Old Danish (ODa) enn ‘one’, However, many distinctions were retained through a dothaer ‘dead’ vs. Old Icelandic (OIc) einn, daudhr. shift of information from (often lost) final vowels to Mutation by [i] and [u] was carried out more com- stressed root vowels: by the process of mutation or pletely in WN than EN, and palatal mutation was also umlaut, root vowels underwent partial assimilation to caused by -R and by [g]/[k] before high front vowels following sounds, and these new pronunciations were in WN, e.g. OIc nofnum ‘names (dat.)’, thaer ‘they then phonologized, expanding the stressed vowel (fem.)’ vs. Old Swedish (OSw) nafnum, thar. By con- inventory. Thus, through palatal mutation, a following trast, breaking occurred more widely in EN. unstressed [i] (or [j]) changed [a] to [E] (later written During the late tenth and early eleventh centuries, <ae> or <e>), [o] to [o] (written <oe> when long) and the Scandinavian countries were converted to [u] to [y] (i.e. fronting/raising), while labial mutation Christianity, and Norse began to be written in a modi- (by [u] and [w]) rounded [a] to [o] (written <o>), [i] to fied form of the Roman alphabet (borrowing the rune [y], and new [E] to [oe] (written <o>), i.e. five new thorn <th> and the letter eth <dh> from Old English vowel qualities were created. A following unstressed usage in some regions to represent the dental frica- [a] also caused lowering of preceding high vowels, and tives, and adding diacritics to some vowels). The old- breaking/diphthongization of preceding [e] to [ia]. est surviving manuscripts in Norse are, however, from Morphological alternations such as fotr—foetr (from the late twelfth century. From this era until around *fotiR) ‘foot—feet’, land—lond (from *landu) 1350 (when the Middle periods of the Scandinavian ‘land—lands’, and geldr—gjalda (from *geldan) languages might be said to begin), dialectal distinc- ‘pays—to pay’ were thus established. tions within EN and WN became increasingly marked, An apparent irregularity in the operation of i-muta- and one can refer to OIc, ONo, etc. tion has attracted considerable interest among Norse EN, and in particular ODa, was generally more inno- researchers. The vowel of a short syllable does not seem vative than WN. ODa merged the unstressed vowels as to have been mutated by lost [i], e.g. katlar (from schwa (written <ae> or <e>), e.g. ODa brothaer ‘broth- *katiloR) ‘kettles’ as opposed to the mutated singular er’, dottaerson ‘daughter’s son’ vs. OIc brodhir, dottur- ketill (from *katilaR). The traditional theory of Axel son. Intervocalic consonants underwent lenition, with Kock explains this as the result of i-mutation having voiceless stops being voiced, voiced stops sometimes taken place in two different periods: first, only long syl- becoming fricatives, and voiced fricatives becoming lables were affected, and the following [i] was then lost; semivowels or being lost, e.g. ODa nogaer ‘some, a cer- later [i] was generally lost after short syllables; finally, tain’, sauthe ‘said’ vs. OIc nokkurr, sagdhi. The case all instances of remaining [i] caused mutation. However, system was substantially reduced, and there was level- this model does not appear entirely satisfactory, and a ing among the personal endings of verbs. The pitch range of other explanations have been presented. accent distinguishing original monosyllables from bisyl- Several notable grammatical developments also lables was changed into a reduced glottal stop, termed occurred during CN. A new set of verbal forms with the stod, in stressed syllables with long vowels or final reflexive/medio-passive significance arose from the voiced consonants; this accent became lexically distinc- suffixation of the reflexive pronoun sik to the verb, e.g. tive where old monosyllables became bisyllabic through berja ‘to beat, strike’ vs. berjask ‘to fight’. The combi- the insertion of epenthetic vowels to break up consonant nation of ‘to have’ with the neuter past participle pro- clusters, or through suffixation of the definite article. duced new perfect and pluperfect constructions, e.g. OSw underwent many of the same developments as fadhir hans hafdhi haldit skottum fyrir Hakoni ‘his ODa, but to a lesser degree, and later. A phenomenon father had withheld taxes from Hakon’. A definite arti- of vowel balance affected unstressed vowels: after cle was developed from the demonstrative hinn, which long syllables, they were pronounced more weakly, could be suffixed to nouns, e.g. medh ollum farminum being written as <ae e o> in some manuscripts, and ‘with all the cargo’; in arma jotna systir ‘the wretched eventually merged and underwent apocope; after short giants’ sister’. A pitch accent distinguishing monosyl- syllables, however, they retained their qualities longer, lables and bisyllables may also have come into exis- being written as <a i u>, and were more resistant to tence at this time, although it was not shown in reduction and loss, e.g. mother ‘mother’ vs. fathir writing. ‘father’, gangae ‘to go, walk’ vs. fara ‘to go, travel’. 792  OLD TIBETAN ONo, while generally similar to OIc, shared some orrosta ok fell mart af hvarumtveggjum… ‘Then they developments with EN, e.g. initial clusters of [hl] and brought the fleets together; the fiercest battle begins [hr] were simplified to [l] and [r]: ONo lut ‘share there and many fell on both (sides)…’. (acc.)’, ratt ‘pushed’ vs. OIc hlut, hratt. Some parts of ON is worthy of study not only for its intrinsic lin- ONo exhibited vowel harmony of unstressed vowels: guistic interest, but also for the access it gives to a following nonhigh vowels, [i u] were lowered to [e o], wealth of literature, reflecting a time when pagan and e.g. honom ‘him (dat.)’ vs. sinum ‘their (dat.)’, Christian worlds met, when explorers and conquerors kononge ‘king (dat.)’ vs. bili ‘moment (dat.)’. expanded out of their original homelands, and when OIc (i.e. ClON) was generally a very conservative the modern nations of Scandinavia were born. variety; however, it innovated in some ways. Around the start of the literary OIc period, the language had a seven- vowel basic ‘triangle’: three front vowels [i e E] and four References back vowels [u o o a], with two front rounded vowels [y Antonsen, Elmer. 1975. A concise grammar of the Older Runic o] that resulted from the CN mutations; all could be inscriptions. Tubingen: Max Niemeyer. either long or short, and nasalized long vowels were also Benediktsson, Hreinn. 1982. Nordic umlaut and breaking: thir- ty years of research. Nordic Journal of Linguistics 5. contrastive. There were three falling diphthongs: [Ei ou Faarlund, Jan. 1994. Old and Middle Scandinavian. The Ey]. This system is described in the remarkable twelfth- Germanic languages, ed. by Ekkehard Konig and Johan van century native work known as the First Grammatical der Auwera. London, New York: Routledge. Treatise. During the literary period, the low–mid vowels Gordon, Eric. 1927. An introduction to Old Norse, revised by underwent mergers: [E] merged with [e], [E:] with [o:], Arnold Taylor. Oxford, New York: Oxford University Press; 2nd revised edition, 1957. [o] with [o] (later written <o>), and [o:] with [a:], e.g. Haugen, Einar. 1950. The first grammatical treatise: the earliest OIc segidh ‘say!’, fell ‘fell’ vs. ONo saeghit, fell. Germanic phonology. An edition, translation and commen- Nasalization was also lost; [y] merged with [i]; [e:] diph- tary. Baltimore: Linguistic Society of America; 2nd revised thongized to [jE] and [o:] to [ai]. edition, London: Longman, 1972. The consonantal system of ClON underwent limited ———. 1976. The Scandinavian languages: an introduction to their history. London: Faber and Faber. change from CN. Before [l], [r] and most rounded vow- Kristjansson, Jonas. 1988. Eddas and sagas. Reykjavik: Hidh els, [w] was lost; it survived as [v] in some environ- islenska bokmenntafelag; 3rd edition, 1997. ments, e.g. OIc reka ‘to drive’, ordhinn ‘having Nielsen, Hans. 1979. De germanske sprog: baggrund og grup- become’, vidh ‘against’ vs. OSw vraekae, wordhin, Old pering. Odense: Odense universitetsforlag; as The Germanic Gutnish withr. Velar stops palatalized before front vow- languages: origins and early dialectal interrelations. Tuscaloosa, London: University of Alabama Press, 1989. els and [j] during the literary period, and consonant clus- Noreen, Adolf. 1884. Altislandische und altnorwegische ters ending in sonorants were broken up with epenthetic Grammatik (Laut- und Flexionslehre) unter berucksichtigung [u], e.g. Modern Icelandic madhur ‘man’ vs. OIc madhr. des urnordischen (Old Icelandic and Old Norwegian grammar If a tonal accent existed in CN, it was lost in Icelandic. (phonology and morphology) in view of Old Norse). Halle Grammatically, the morphology of CN was very (Saale): Max Niemeyer; 4th revised edition, 1923. ———. 1904. Altschwedische grammatik, mit einschluss des well preserved in ClON. There were a number of loan- altgutnischen (Old Swedish grammar, with the inclusion of words from Latin, English, or German associated with Old Gutnish). Halle: Max Niemeyer. the new Church, e.g. kirkja ‘church’, paskar ‘Easter’, Riad, Tomas. 1998. The origin of Scandinavian tone accents, biskup ‘bishop’. Word order was fairly free, although Diachronica 15. with a tendency to subject–verb–object, in prose at Skautrup, Peter. 1944. Det danske sprogs historie (History of the Danish language), Vol. 1. Copenhagen: Gyldendal. least. A notable feature of narrative style was the fre- MARISA LOHR quent use of the historical present, with tense often switching casually within sentences, e.g. Sidhan See also Gothic; Old High German; Old English; logdhu their saman flotann; teksk thar in grimmasta Swedish and Scandinavian Languages Old Tibetan Old Tibetan is the ancestor of all modern Tibetan area (if we compare with historical Tibet’s huge size). dialects. This language is attested from the seventh cen- The origin of the Tibetan people was most probably the tury onward. It was originally spoken in a relatively tiny Lhoka in southern Tibet, where remain the Yum-bung 793 OLD TIBETAN bla-sgang, the oldest fortress in Tibet, and the royal Middle Chinese are restricted to cultural vocabulary tombs in ‘Phyong-rgyas. From Emperor Srong-btsan (lcog tse ‘table’, dong tse ‘coin’). The loanwords from sGam-po at the beginning of the seventh century up to Archaic Chinese have never been studied yet, as it is the middle of the ninth century, the Tibetan Empire difficult to tell apart loanwords from words with the became one of the major powers in Central Asia. After same ancestry (cognates) (one such example is lcags this brief period of expansion, the Tibetan-speaking area ‘iron’ : it must be a loanword, since iron appeared only has not changed much up to now. in the fifth century BC in China). Besides, Tibetan Old Tibetan should not be confused with Classical absorbed many non-Tibetan languages during the Tibetan, which is still the standard written language expansion of the empire, and these have certainly for most Tibetans. Classical Tibetan is an old dialect influenced its vocabulary. This might be one cause for that is not the direct ancestor of all modern dialects, the existence of redundantly duplicated words with but it is the basis for the orthography of Modern similar meaning, such as rmi and rmang ‘dream’. Literary Tibetan, also called ‘Written Tibetan’. We Tibetan has several striking lexical innovations, consider only texts older than the tenth century as gen- such as bdun ‘seven’ (Burmese khu hnac), khyed ‘you’ uine ‘Old Tibetan’. (Burmese nang), khrag ‘blood’ (Burmese sweh), not The Buddhist conversion of Tibet entailed the found in any other language of the family. coinage of new terms, and the creation of a dictionary, Clear traces of the derivational morphology inherit- the Mahatvutpatti (bye brag rtog ched) at the beginning ed from Proto-Sino-Tibetan are found in Tibetan. of the eighth century. Besides, the orthography under- However, it probably was not in active use anymore went several reforms by the end of the Imperial Period. even by the time of the first Buddhist translations, Few documents exist in genuine Old Tibetan. There are since new means were created to coin new terms: the famous Sino-Tibetan bilingual treaties inscribed on instead of using the -s suffix deriving an action noun pillars (rdo ring), the extensive Dunhuang documents from a verb (za ‘to eat’ zas ‘food’), translators chose to that have been partially translated into French, English, add the verb byed ‘to do’ in its future form bya after a and Chinese, and only two transmitted texts that pre- verb to translate the Sanskrit past participle. serve the most archaic features: the bKa-chems ka- khol-ma (testament of Srong-btsan sGam-po) and the Writing System and Phonology sBa-bzhed (chronicle of bSam-yas). Many texts from the Buddhist canonical scriptures (bKa-‘gyur and The phonology of Old Tibetan is obscured by the bsTan-‘gyur) date from the Imperial Period, but they irregularities in spellings found in most Dunhuang were rewritten so many times that most archaic features documents, which often make the translation quite dif- have been weeded out. No grammar of Old Tibetan ficult if not impossible. The exact pronunciation of exists yet and the extent of the difference between the Tibetan in the imperial period is mostly uncontrover- Classical and the Old language is not fully understood. sial, except for a few letters. The prefixed ‘-, incor- rectly called ‘a chung’ even by Tibetan specialists (‘a Language Classification chung designates in fact the little ‘a used to indicate vowel length in Sanskrit words), has been the object of Tibetan is usually classified as a Bodic language. The some debate, as it seems to have three distinct func- Tamang-Gurung-Thakali languages of Nepal are gen- tions in Tibetan orthography: when used as an initial, erally considered its closest relatives. However, no it represents a voiced glottal fricative (as in ‘od reconstruction of the ancestor language Proto-Tibetan ‘light’). This sound is still heard in some dialects. As a based on the comparison with these languages has preinitial, it indicates prenasalization (as in ‘khor ‘to been undertaken so far; hence, our knowledge of the turn’, most likely pronounced as *ŋkhor). Finally, prehistory of Tibetan is mostly based on internal since the vowel a is not written by any special symbol reconstruction and on comparison with more remote in the indic alphabet, this letter also serves to disam- cousins like Archaic Chinese and Burmese. Some biguate some combinations of letters (mda’ ‘arrow’, scholars have proposed that Tibetan is especially close written M D ‘ as opposed to mad ‘under’ written M D). to Chinese within the Tibeto-Burman family, but fur- The final stops -b -d -g are written with the symbols ther investigations are needed. for voiced consonants, although these were most prob- ably pronounced as unreleased stops as in most lan- Vocabulary guages in the area. Tibetan has few loanwords from either Sanskrit or Old Tibetan boasts of a great quantity of complex Chinese. Indic words are limited to animals and plants consonant clusters that are preserved in modern not found in Tibet (seng-ge ‘lion’, pad-ma ‘lotus’, orthography but only partially in dialects. These clus- ping kyur ma ‘a kind of bird’). Loanwords from ters are extremely interesting typologically, since 794  OLD TIBETAN some of them are highly unusual: clusters such as rt- subject. Transitive verbs have four forms (present, are permitted, but not the cross-linguistically much past, future, and imperative), and intransitive verbs more common tr-. have two forms. In most cases, one form is not suffi- Tibetan admits up to four initial consonants (brgyad cient to derive the four forms, so that the verbal con- ‘eight’), but no more than two finals (dmangs ‘peo- jugation is far from simple. Here are examples of ple’). One of the most conspicuous differences verbal patterns found in Old Tibetan (for some forms between Old and Classical Tibetan is the da drag: unattested in Old Tibetan, we provide Classical while in Classical only -s can stand as a second final, in Tibetan in italics): Old Tibetan some clusters exist with -d as a second Present Past Future Imperative final, such as in mnard ‘he oppressed’ (Classical mnar). In the Tibetan writing system, there are three series byed byas bya byos of initial consonants : voiced, aspirated, and plain ‘dzin / zin bzung gzung zung to take unvoiced. The former can appear in all contexts, but the ‘dogs btags gdags thogs to tie latter two seem to be variants of a single type in za zos bza zos to eat Classical Tibetan: except for some rare exceptions, only sbyor sbyard / sbyar sbyor to join aspirated consonants can begin a word or stand after spyard nasal preinitials (m- and ‘-), while when preceded by Some elements of the verbal conjugation are quite any other preinitial (stop, liquid or s-), only plain voice- old: the past tense -s suffix is also found in rGyalrong, less stops are permitted. This reminds one of English Qiang, and Kachin, and the change of a to o in the stops (p is aspirated in pit and nonaspirated in spit). In imperative is related to the imperative -o suffix found Old Tibetan, the distribution of these graphemes is in Tamang. However, many peculiarities of the verbal much less predictable (gchig systematically for system (the change a to o in the present tense of some Classical gcig ‘one’, for instance), but it is not probable verbs, the past tense b- prefix) do not seem to be relat- that these two series are truly distinctive in Old Tibetan. able to anything in other Tibeto-Burman languages. Finally, there is a vowel in Old Tibetan never found An interesting irregularity in conjugation that is in the Classical language: the inverted I, transliterated common to Tibetan and Kiranti languages is the con- as ï. This sign does not seem to represent a phoneme jugation of the verb ‘to eat’. In Tibetan, it presents an distinct from i, though. The rules governing their dis- exceptional vowel alternation in the past tense a à o tribution are still unknown. unrelated to the more common alternations found in the present and imperative. In Hayu, ‘to eat’ is one of Verbal System the four verbs with a -o stem alternation: dza-ng (I eat Like modern Tibetan dialects, Old Tibetan was an it), dzo (You eat it). The same irregularity is found in ergative language, which means that the object of a a dialect of Limbu. This must be an archaism going transitive verb bears the same mark as the subject of an back far in the prehistory of the language. intransitive verb (in Old Tibetan, this case has no overt marking), while the subject of a transitive verb is marked with ergative case (variously marked as ‘is, References kyis, gyis, or gis). Bacot, J.F.W. Thomas, and C. Toussaint. 1940–1946. Documents de Touen-houang relatifs à l’histoire du Tibet. klu-gong glo-ba nye’o (Zhol inscription, South:20) Paris. Klu-gong (no suffix)—intention––near—finite verb Huang, Bufan, and De Ma. 2000. Translation and annotations ending on Dunhuang Tibetan manuscripts about the history of Klu-gong was loyal. Tufan period. Beijing. Imaeda, Y. and T. Takeuchi. 1990. Choix de documents gzhan ‘is bdag bsad sha zos dka’ myi dka’ (Pelliot Tibétains, tome III: corpus syllabique. Paris. Tibétain 126: 20) Imaeda, Y. et al. 2001. Choix de documents Tibétains, Tome IV: other—ergative suffix—yourself (no corpus syllabique. Tokyo. Li Fang Kuei, and W. South Coblin. 1987. A study in the Old suffix)—kill––flesh (no Tibetan inscriptions. Taipei. suffix)—eat––hard—not—hard McDonald, A. 1971. Une lecture des Pelliot Tibétain 128/6, If others killed you and ate (your) flesh, wouldn’t it 1287, 1038, 1047 et 1290. Etudes tibétains dédiées à la be hard? mémoire de Marcelle Lalou, 190–391. Paris. Richardson, H.E. 1985. A corpus of early Tibetan inscriptions. The ergative suffix, however, serves as an intru- London. mental case and can mark emphasis. JACQUES GUILLAUME Unlike Kiranti or rGyalrongic languages, Tibetan has no trace of agreement between verb and e.g. the See also Sino-Tibetan Languages 795 ONOMATOPOEIA Onomatopoeia Onomatopoeia is better understood as a semantic phe- meaning based on phonetic similarity. Mandarin and nomenon if language is likened as a diagram. A referent Cantonese speakers believe that the sound of 8, i.e. ba, is codified through the phonetic-symbolic means of lan- is sound symbolic to fa ‘prosper’. As such, car regis- guage into a word, i.e. the linguistic depiction of its rep- tration numbers that contain many 8’s are more expen- resentation. All languages show some kind of imitative sive in Malaysia and Singapore. Since 4 sì is sound linguistic codification known as onomatopoeia. symbolically related to sì ‘die’ in these languages, Onomatopoeia represents a referent based on a dia- karaokes in Singapore and Malaysia will add lucky grammatic sound–meaning identification. This direct phrases to tags with the number 4 before offering this sound–meaning relationship is nonarbitrary or isomor- queue number to customers. One can imagine how phic. Examples are easily found in many languages, confident a typical Chinese needs to be in order to especially the use of animal sounds like the Malay drive in a car with 4444 as the registration number. kokok ‘the crow of a rooster’, embek ‘the bleat of a Secondary sound symbolism plays an important goat’, the Mandarin gaga ‘the quack of a duck’, and role in Chinese speech communities when one choos- miao miao ‘the meow of a cat’. Onomatopoeia as a uni- es a name for a child. Lóng ‘dragon’, a symbol of versal language feature can be divided into three types: divinity and good fortune, is a favourite character used sound imitation, secondary sound symbolism, and echo. for boys. An excellent example would be Chen Long Onomatopoeia of the first type occurs when lin- (Jackie Chan), who became the dragon of Chinese cin- guistic sound imitates sounds of the real world. Each ematography. Bruce Lee’s middle name is also the linguistic imitation is taken as the diagrammatic sign character of a dragon. Other favorite characters used of the signified referent. Many terms are derived from as Chinese personal names are fa ‘prosperous’ as in sound imitation, for example, cuckoo, cockatoo, boom, Chow Yuen Fatt, an ‘safe’ as in Ang Lee, and kuang buzz, rumble, twang, chirp, slam, thump, clap, flap, ‘bright’ the first name of the former Singapore Prime and rattle. More common onomatopoeic words like ha Minister. Onomatopoeia in the form of secondary ha ha that signifies laughter in English and Cantonese sound symbolism is culturally significant as it influ- opera is used as a sarcastic tag as in ‘How funny, ha ha ences one’s belief. ha’ in specific contexts of communication. Onomatopoeic words are salient in Malay vocabu- TABLE 1 lary as illustrated in Table 1. Sound symbolism occurs when there is a recurrence Onomatopoeic Lexicon Meaning of sound–meaning correlation. An example would be Debuk The sound of slapping or boxing the high front unrounded vowel taken to symbolize a Debum The sound of heavy things falling sense of small, insignificant, slight, or weak. The vowel Debung The sound of drumming Debur The sound of waves breaking on contrast in pinch vs. punch, gripe vs. grope, and sip vs. the shore suck points to this sound–meaning symbolism in /i/. Debus The sound of flapping of wings Research shows that the distribution of sounds in men’s Debut The sound of squeezing out of a and women’s names is not random. It is common for small hole Caucasian women’s names to end with /i/ and men’s Debak The sound of boxing each other Debar The sound of heart pounding with consonants like fricative, i.e. /s, f/, and stop, i.e. /t, Debik The sound of hitting with one’s k/. The psychology experiment concludes that the gen- palm dered sound difference in personal names is best under- Degar The sound of blowing with a stick stood as reflecting emotional differences. Men’s names Deguk The sound of gulping water are associated with active, cheerful, nasty, and unpleas- Dekah The sound of hearty laughter Deham The sound of a little cough ant emotions while women’s names are associated with Dengkung The wailing of a dog passive, sad, soft, and pleasant emotions (see Whissell Dengung The sound of a plane 2001). If language is diagrammatic, personal names are Dengkur Snoring depictions of human beings whose respective names Dentang The sound of hitting the iron become the aspirations intended for them. Denting The sound of falling coins Dentung The sound of thunder Secondary sound symbolism in the onomatopoeia Dering The ringing of the phone context is a linguistic sound taken to connote another 796  OTO-MANGUEAN LANGUAGES TABLE 2 meaning, i.e. quantity iconicity where more forms Semantic Criterion Examples entail more meaning. Despite the skepticism about the relevance of ono- Motion huff-puff, flip-flap, flipperty-flopperty matopoeia, it remains resilient in language. The fact Tumult rumble-jumble, hubble-shubble, wringle-wrangle that onomatopoeia exists through time and manifests Inferiority shally-wally, phoney-baloney, whim- creatively in various forms testifies to the significance sy-whamsy of this component either as pragmatic tool, cultural Trickery hooky-crooky, hanky-panky, cuddle- routine, or semantic paradigm that prevails in lan- muddle guage use. Mental instability randy-dandy, shilly-shally, hazy-dazy Mutuality equal-aqual, hob-nob, tiggle-taggle High degree tip-top, wee-jee, itsy-bitsy References Hypocoristic mousey-pousey, tootsy-wootsy, lovey- dovey Abbi, Anvita. 1992. Reduplication in South Asian languages: an areal, typological and historical study. New Delhi: Allied Publishers. Bolinger, Dwight. 1992. Sound symbolism. International Echo is a form of reduplication that occurs in the Encyclopedia of Linguistics 4. form of freezes, binomials, echo formation, and other Haiman, John. 1985. Natural syntax: iconicity and erosion. forms of bipartites. Echoic morphology is language Cambridge: Cambridge University Press. universal. Table 2 illustrates some English echos. Hinton, Leanne, Johanna Nichols, and John Ohala (ed.) 1994. Malay echoic reduplications like gunung-ganang Sound symbolism. Cambridge: Cambridge University Press. Jakobson, Roman, and Linda R.Waugh. 1979, 1987. The sound ‘ranges of moutain’, batu-batan ‘rocks’, sayur-mayur shape of language. Berlin: Mouton De Gruyter. ‘all kinds of vegetables’, bengkang-bengkok ‘very Jespersen, Otto. 1922. Symbolic value of the vowel i. crooked’, and cucu-cicit ‘descendants’ are commonly Philologica 1. used in daily conversation. Malay partial reduplication Sew, Jyh Wee. 1995. Sound meanings in Malay. M.A. Thesis. is important in word fomation, especially in naming Singapore: The National University of Singapore. ––––––. 1996. Symbolisation in Malay: evidence in genres and things: tikus ‘mouse’ > tetikus ‘computer mouse’, kunci lexicon. Proceedings of the 4th International Pan-Asiatic ‘key’ > papan kekunci ‘keyboard’, and pejal ‘solid Symposium of Language and Linguistics, Vol. I. Salaya: (adjective)’ > pepejal ‘solid (noun)’. South Asian Mahidol University Press. echos, on the other hand, are loaded with pragmatic Thun, Nils. 1963. Reduplicative words in English: a study of functions. Bhojpuri echos like deslai-oslai ‘matchbox formations of the types tick-tick, hurly-burly and shilly- shally. Sweden: Carl Bloms. etc.’ is used as an indirect request for cigarettes by Tiwary, K.M. 1968. The echo-word construction in Bhojpuri. youngsters or subordinates because smoking in the Anthropological Linguistics 10. presence of parents or elderly persons is bad manners. Waugh, Linda R. 1993. Against arbitrariness: imitation and In Tamil, vendu ‘come’ is echoed as vendu gindu ‘come motivation revived, with consequences for textual meaning. etc.’. The echo is used for the addressee to come (and Diacritics 23. Whissell, Cynthia. 2001. Sound and emotion in given names, run an errand for him). Cantonese echos like yùhn lu¯k Names 49. lu¯k ‘rounded’, ch¯ lahp lahp ‘sticky/goey’, dung b¯ ng JYH WEE SEW b¯ ng ‘freezing cold’, and baahk syu¯t syu¯t ‘snow-white’ come in tripartites. These echos illustrate more intense See also Iconicity Oto-Manguean Languages Oto-Manguean languages are distributed throughout agricultural products among the peoples of Archaic Mesoamerica and form the language stock with the Mexico beginning about 4400 BCE. In this account, deepest time depth in the Americas. The emergence of early Oto-Manguean diversified on a landscape where its eight linguistic families patterns geographically with smaller, more mobile bands came to rely increasingly the ecological zones of the domestication of maize, on domesticated plants and center their production in chili, squash, and beans, which emerged as primary villages on the mesas and highland valleys of the 797 OTO-MANGUEAN LANGUAGES Central Altiplano, the Mixteca Alta, the Valley of Over 100 years of Oto-Manguean scholarship pro- Oaxaca, and the Tehuacán Valley. These regions help vides us with evidence for several statements about the define four branches of Oto-Manguean, with grammatical structure of Proto-Oto-Manguean, the Otopamean-Chinantecan and Tlapanecan-Manguean assumed ancestor language. Simple words typically forming the Western division and Amuzgo-Mixtecan had only one syllable, and complex words could be and Popolocan-Zapotecan the Eastern division. formed by compounding simple word roots or by Most of the language families of West Oto- attaching prefixes. The language most likely had tone Manguean are centered in the highlands and water- distinctions; i.e. melodic pitch differences between sheds from the Central Altiplano to the Pacific in otherwise identical words could alone indicate a Guerrero, and East Oto-Manguean families in high- meaning difference. The vocalic system included min- lands and watersheds from the Mixteca Alta to the imally *i, *e, *a, *u, several diphthongs, the semi- Pacific in Oaxaca. Some consonant differences vowels *w and *y, and the consonants *t, *ts, *k, *kw, between the separate reconstructions of the ancestor *s, *x, *xw, *l, *n, *m (the asterisks indicate that these languages for East and West suggest that the oldest sounds have been reconstructed on the basis of docu- accessible form of Oto-Manguean may already show mented languages). dialect differentiation, possibly resulting from disper- While parts of the Oto-Manguean family as pre- sal among these watersheds. The connections between sented here were separately grouped by researchers of the Chiapanecan and Manguean languages in the far the nineteenth and early twentieth centuries, it was east of Mesoamerica and Tlapaneco of West Oto- Edward Sapir’s 1929 classification that was the first to Manguean are accounted for as the result of popula- include the earlier groupings of Mixtecan-Zapotecan, tion migrations in prehistory. Otomian, Chiapanecan-Manguean, and Chinantecan West Oto-Manguean includes the two branches of in a single group. From the 1940s until the mid-1960s, Oto-Pamean-Chinantecan and Tlapanecan-Manguean. Oto-Manguean research had a blossoming period dur- Northern Oto-Pamean consists of Chichimeco of ing which several dedicated scholars worked out the Guanajuato and the Pame languages of San Luis relations of diversity within the Oto-Manguean Potosí. Southern Oto-Pamean consists of Matlatzinca branches. A student of Sapir, Morris Swadesh, com- and Ocuilteco around the Valley of Toluca, and an piled the first comparative study of the sound corre- Otomian group comprising the Mazahua language and spondences of the Zapotecan languages. In the 1950s, six branches of Otomí in the states of Mexico, Sarah Gudschinsky worked on the internal diversity of Tlaxcala, and Michoacan. The Chinanteco language Popolocan and supported a larger grouping that family represents the extension of the stock east of the included the East Oto-Manguean families of Mixtecan Sierra Madre and contains at least six languages. The and Popolocan, which she called ‘Proto-Popotecan’. Tlapanecan-Manguean branch consists of Tlapanecan- Eric Hamp refined and revised both Gudschinsky’s Subtiaban and Chiapanecan-Manguean. Tlapaneco is earlier subgrouping of Proto-Popolocan and the analy- spoken in several varieties in the state of Guerrero, and sis of the sound system of Chiapanecan-Manguean. Subtiaba was historically spoken near the Pacific coast María Teresa Fernández de Miranda also worked of Nicaragua. Chiapanecan-Manguean consisted of intensively with the Popolocan language family and Chiapaneco in Chiapas and Mangue in coastal the Manguean family and was working out a recon- Honduras and Nicaragua. struction of Proto-Zapotecan before her untimely East Oto-Manguean includes the Popolocan- death. Robert Longacre’s extensive work on the inter- Zapotecan and Amuzgo-Mixtecan branches. nal relationships of Amuzgo-Mixtecan also began with Popolocan comprises four Mazateco languages of publications in the 1950s. In the same decade, our Oaxaca, Puebla, and Veracruz; four Chocho-Popolocan knowledge of West Oto-Manguean developed with languages of Oaxaca and Puebla; and Ixcateco of Stanley Newman and Robert Weitlaner’s Proto-Otomí Oaxaca. Zapotecan includes three Chatino languages in reconstructions. In the 1960s, Doris Bartholomew southwestern Oaxaca and five branches of Zapoteco in contributed work on Oto-Pamean and revised the the mountains of southern, western, southwestern, and Proto-Otomí reconstructions. Along with Michael northern Oaxaca, and in the Valley of Oaxaca east to Piper, she prepared Fernández de Miranda’s unfin- the Isthmus of Tehuantepec. Each branch of Zapoteco ished Proto-Zapoteco reconstructions for publication. further divides into several distinct languages. A virtual dearth of publications spans over a decade Amuzgo-Mixtecan is made up of Amuzgo in Guerrero, before the publication of the most influential work on two Trique languages in Oaxaca, and Mixteco- Proto-Oto-Manguean. Calvin Rensch began systematic Cuicateco, which itself includes the Cuicateco lan- work on Proto-Oto-Manguean in the early 1960s. A guage of Oaxaca and several divergent Mixtecan brief period of new interest developed after the 1976 language areas in Guerrero, Puebla, and Oaxaca. publication of Rensch’s 1966 thesis, Comparative Oto- 798  OTO-MANGUEAN LANGUAGES Manguean phonology, in which he focuses on recon- language-internal changes and regionally defined pat- structing the tone-carrying syllables of words in Proto- terns of linguistic diffusion. From Archaic through Mixtecan. Based largely on Rensch (1976), J. Kathryn Post-Classic Mesoamerica, webs of political and eco- Josserand, Marcus Winter, and Nicholas Hopkins’s nomic interaction mediated the divergence of related (1984) Essays in Oto-Manguean culture history treated languages as villages allied themselves to larger cen- the unique opportunity that Mesoamerica provides for ters emerging with the origins of the states of Pre- integrating historical linguistic and archeological Columbian Mexico. approaches to culture history. In the 1980s, a student of Oto-Manguean language history also involves con- Eric Hamp, Terrence Kaufman, critically assessed the tacts with languages outside the family but within a word and sound correspondences contained in the work greater Mesoamerican culture area. Such relations of Rensch, Longacre, and other historical sources, and include speakers of Huave, Totonac, Mixe-Zoque, in 1994 presented the classification adopted here. Maya, and later arriving Uto-Aztecan speakers includ- The work of the handful of scholars mentioned ing the Aztecs. Loanwords and structural similarities above only begins to point to the interest of this lan- shared among many of the languages of Mesoamerica guage stock for linguistic and historical linguistic the- support a view that networks of contact have been ory. The bulk of the above works focus mostly on the most important in shaping the patterns of linguistic sound system and the vocabulary. However, the lan- features in this culture area. The Spanish Conquest guages of Oto-Manguean also present challenges for brought a common language of the empire that must research in intonation, sentence grammar, word struc- also be counted in the histories of the divergence of ture, and meaning, for which the literature is almost Oto-Manguean languages as well as in shaping our entirely lacking. knowledge of those that were chronicled in Colonial The Oto-Manguean sound system exhibits some and Post-Colonial Mesoamerica but are no longer spo- unique characteristics. For example, many of the lan- ken at the present time. guages lack [p] and [b], and most consonants are Today the status of Oto-Manguean languages varies voiceless, i.e. they are produced, like [t, k, s], without from village to village. Some families have well over vocal cord vibration. Voiced consonants such as [d, g, 100,000 speakers, but the languages of that family may z], on the other hand, are rare. range from villages of several thousand, where all the Oto-Manguean is the world’s oldest language stock children learn an Oto-Manguean language as a first that provides sufficient evidence for tone distinctions language, to villages where less than a dozen elders in the ancestor language. The complicated tone sys- have any memory of their language. In most communi- tems distinguish between two and six pitch levels; i.e. ties, bilingualism is the norm, with Spanish being in the most complex scenario, one could theoretically learned either at home or in school. Many villages that have six different one-syllable words differing only in once spoke an Oto-Manguean language have entirely the pitch or the pitch pattern of the vowel. shifted to Spanish. This trend is occurring in some ‘Whistled speech’ has been reported in Mazateco, regions at an alarming rate. Several key languages and Mixteco, and Zapoteco. This is a form of speech that language families are already extinct. Chiapaneco and is basically stripped of individual segments and con- all of Manguean have long since passed. While the doc- sists almost exclusively of melody. It provides speak- uments on Manguean are sufficient to demonstrate its ers with the ability to communicate a wide range of membership in the Oto-Manguean stock, the data were utterances across a distance. not recorded with attention to tone and other complex The grammar of Oto-Manguean languages makes issues of pronunciation. Most of the Tlapanecan- heavy use of particles. In Oto-Manguean languages, Manguean languages have not been adequately docu- they can carry meanings associated with intonational mented at all. For Oto-Manguean overall, there are patterns in languages like English or the Romance lan- very few dictionaries and grammars, and few practical guages. For example, yes/no questions, topicalization, orthographies have been developed. The next couple of exclamation, and emphasis are indicated by particles. generations will decide whether Oto-Manguean voices However, at this time, very little work has been done will still be heard in Mexico’s future. on Oto-Manguean sentence structure. Oto-Manguean continues to have great import in References the development of historical linguistic theory. It is a family with an internal complexity comparable to that Bartholomew, Doris A. 1960. Some revisions of proto-Otomí of Indo-European but has diversified within a restrict- consonants. International Journal of American Linguistics 26. 317–29. ed cultural-ecological zone associated with highland Campbell, Lyle, Terrence Kaufman, and Thomas C. Smith- agriculture. Thus, Oto-Manguean languages provide a Stark. 1986. Meso-America as a linguistic area. Language natural testing ground for addressing the interaction of 62(3). 530–70. 799 OTO-MANGUEAN LANGUAGES de Miranda, Fernández, and María Teresa. 1995. El Longacre, Robert E. 1966. On the linguistic affinities of Amuzgo. ProtoZapoteco, ed. by Michael J. Piper and Doris International Journal of American Linguistics 32. 46–9. Bartholomew. Mexico D.F.: El Colegio de Mexico. 1995. Newman, Stanley, and Robert Weitlaner, 1950. Central Gudschinsky, Sarah C. 1995. Lexico-statistical skewing from Otomian I: Proto-Otomi reconstructions. International dialect borrowing. International Journal of American Journal of American Linguistics 16. 1–19. Linguistics 21. 138–49. Rensch, Calvin R. 1976. Comparative Oto-Manguean phonolo- Hamp, Eric P. 1958. Protopopoloca Internal Relationships. gy. Bloomington: Indiana University Press. International Journal of American Linguistics 24. Swadesh, Morris. 1947. The phonemic structure of Proto- 150–53. Zapotec. International Journal of American Linguistics 13. Josserand, J. Kathryn, Marcus Winter, and Nicholas Hopkins 220–30. (eds.) 1984. Essays in Otomanguean culture history, ———. 1960. The Oto-Manguean hypothesis and macro Nashville: Vanderbilt University Publications in Mixtecan. International Journal of American Linguistics 26. Anthropology No. 31. 79–111. Kaufman, Terrence. 1994. The native languages of MARK SICOLI Mesoamerica. Atlas of the World’s languages, ed. by Moseley and Asher. London, New York: Routledge. See also Mexico; Tone Languages 800  P Pacific The area covered by the Pacific ocean occupies about a Chile; and Northern Marianas—a Commonwealth of third of the Earth’s surface and is by far the world’s the United States), those that are self-governing in largest geographic region. It contains more than 1,000 free association with other countries (e.g. Cook indigenous languages, Asian mainland languages, half Islands, Tokelau, and Niue in relation to New a dozen European languages, and a number of pidgins Zealand, and the Federated States of Micronesia, and creoles. Geographically, it includes the island Chuuk, Pohnpei, Yap, and Kosrae, in relation to the nations and territories of Micronesia, Polynesia, and United States), and those that are colonies (e.g. Melanesia, collectively known as Oceania. The islands Pitcairn Island—a British colony), to those that have occupy just less than 500,000 square miles of land area. been incorporated into other nations (e.g. Hawaii— New Guinea, the second largest island in the world formerly a territory of the United States, now the after Greenland, represents 64% of this total, and New fiftieth state). Zealand represents another 20%. The remaining 10% is The Pacific islands were peopled by influxes from divided among more than 10,000 scattered islands. both mainland and island Southeast Asia, with the ear- Micronesia comprises the island groups east of the liest migrations to Melanesia. From Melanesia, gener- Philippines and north of the equator, including, for ations of voyagers headed northward into eastern instance, the Mariana, Marshall, Caroline, and Gilbert Micronesia and eastward into Polynesia. European islands. Melanesia encompasses the island groups in exploration of the Pacific began with the voyages of the southwestern Pacific extending southeastward from Ferdinand Magellan in the 1520s, which inaugurated a the Admiralty islands to Fiji, including such nations as period of Spanish and Portuguese influence, followed Papua New Guinea, Solomon Islands, Vanuatu, and by Dutch, English, French, and German influences. New Caledonia. Polynesia consists of a group of Most of the major island groups were annexed by islands extending from New Zealand north to Hawaii Britain, the United States, France, and Germany in the and east to Easter Island. nineteenth century. Many Micronesian territories Politically, the status of the islands ranges from changed hands a number of times, beginning with those that are independent (e.g. Papua New Guinea, Spanish colonial administration that was followed by Fiji, Kiribati, Palau, Samoa—formerly Western German, Japanese, and American regimes. Samoa, Tuvalu, and Tonga), those that are quasi- The historical partition of the islands among the var- colonies or territories (e.g. New Caledonia, French ious European colonizers has left the islands’ popula- Polynesia, and Wallis and Futuna—French overseas tions divided primarily into a Francophone and territories; Guam and American Samoa—United Anglophone set of nations, each with different political, States territories; Norfolk Island—an Australian economic, linguistic, and cultural orientations. More territory; Rapa Nui/Easter Island—a dependency of recently, however, notions of a Pacific islander identity 801 PACIFIC have begun to emerge, fueled partially by recognition of Austronesian language containing glottalized conso- a common history of colonization, as well as by nants, (i.e. pronounced with simultaneous contraction demands from nonindependent territories and nations of the vocal folds). Most of the Micronesian languages for increased political autonomy. have relatively small numbers of speakers, with a few, The indigenous languages of the Pacific belong to such as Sonsorolese on Sonsoral island in Palau, hav- two language families: Austronesian and non- ing only several hundred speakers. Austronesian (Papuan). The Austronesian language Historical relationships among the Papuan lan- family is the largest and most widespread family in the guages, found primarily in the New Guinea interior, world. The Oceanic or easternmost subgroup of the an area of immense linguistic complexity, are less Austronesian family is the largest and best defined of clear than in the much better studied Austronesian all subgroups, containing all Polynesian languages, all language family, and the label is best seen as a cover Micronesian languages (except Palauan and term for many, perhaps as many as 60, distinct fam- Chamorro), and all Austronesian languages of ilies. The Papuan-speaking peoples were the earliest Melanesia east of the Mamberamo River in Irian Jaya people in the region, occupying the Sahul continent (the western half of the island of New Guinea, (which later partially submerged to become the Indonesia’s easternmost province). island of New Guinea) at least 40,000 years ago. By The Polynesian languages are a subgroup of 30,000 years ago, the Bismarck Archipelago east of Oceanic containing approximately 30 languages New Guinea was occupied by speakers of Papuan organized into two major subgroupings: Tongic languages. Speakers of Austronesian languages, who (including Tongan and Niuean) and Nuclear arrived much later, established coastal communities Polynesian (all the rest). The latter is divided into two in New Guinea, the southeastern Solomons, further groups: Samoic Outlier (consisting of Samoan Vanuatu, New Caledonia, and Fiji. Contacts between and Tuvaluan) and Eastern Polynesian (consisting of speakers of Papuan and Austronesian languages Hawaiian, Tahitian, Maori, Rapa Nui, and have made it difficult to classify some languages as Marquesan). Speakers number fewer than a million Austronesian or Papuan, such as Maisin in Papua people spread across a large area of the Pacific, rang- New Guinea, which displays features typical of both ing from Hawaii in the north to New Zealand in the language families. south and Easter Island. This so-called Polynesian The European languages spoken in the Pacific Triangle encloses an area approximately twice the region are primarily those of the former colonizers and size of the continental United States. Despite their still have official status in most areas. Although French great geographic spread, the Polynesian languages is important, no European language has had the impact show a relative homogeneity, which indicates that that English has had. In addition to the standard vari- they have dispersed only in the last 2,500 years from ety of British or American English that serves as the an original center in the Tonga-Samoa area. In addi- official language in Anglophone territories, English tion, there are some 18 further Polynesian-speaking exists in a number of locally distinct varieties in New communities known as ‘Polynesian Outliers’ in Zealand and Hawaii, and it also exists in the form of Micronesia (e.g. Kapingamarangi and Nukuoro, two numerous pidgins and creoles. French has left its lega- atolls in the southwestern portion of Pohnpei state) cy in the form of a number of local varieties of colo- and Melanesia (e.g. Ontong Java atoll and Rennell nial French spoken, for example, in New Caledonia island in the Solomon Islands). Although some and Tahiti, as well as in a French creole, Tayo, in New Polynesian languages have a relatively large number Caledonia. Standard French still has official status in of speakers, such as Samoan with about 200,000, French territories of the Pacific, as well as in inde- many others have very few, and some, such as Maori pendent Vanuatu, formerly a condominium ruled joint- and Hawaiian, face possible extinction as their speak- ly by Britain and France, where it shares co-official ers have shifted to English. status with English. Because Germany lost most of its Most of the Oceanic Micronesian languages (with colonies after World War I, German never had a the exception of Nauruan, which has not been suffi- chance to spread as widely as other colonial lan- ciently studied) form a closely related subgroup called guages. The language left some remnants, nonethe- Nuclear Micronesian. Palauan and Chamorro, howev- less, in the form of a German creole, Unserdeutsch, as er, appear to be the result of distinct migrations from well as borrowings, both into forms of pidgin English, Indonesia or the Philippines. Although Yapese is prob- such as Tok Pisin in Papua New Guinea, as well as into ably Oceanic, it does not appear closely related to any indigenous languages. Spanish has only a limited pres- of the other Micronesian languages, because of its ence in the Pacific as the official language of Rapa complex history of borrowing. It is also the only Nui/Easter Island. 802  PACIFIC In addition to European languages, other languages many as 2 million speakers. Most business in the were introduced as a result of colonization, such as House of Assembly, the country’s main legislative Japanese, Korean, varieties of Chinese (for example, body, is conducted in Tok Pisin, which is the most in Papua New Guinea, Nauru, Tahiti, and elsewhere), widely shared language among the members. In and Vietnamese in New Caledonia. Varieties of Hawaii, a creole English developed on plantations and Southeast Asian languages, such as Hindi, were intro- is still in use, primarily among working-class speak- duced into Fiji, and Javanese was introduced into New ers. There is some dispute over the status of the vari- Caledonia. Philippine languages such as Tagalog, ety of English spoken on Pitcairn and Norfolk Islands, Ilocano, and Visayan were introduced into Hawaii spoken by descendants of the British mutineers from through the importation of migrant labor and inden- HMS Bounty. tured plantation workers. Only Tok Pisin and Bislama have received some The Pacific pidgins and creoles originated primari- official recognition. Tok Pisin is a de facto official ly in the late eighteenth and early nineteenth centuries. language in Papua New Guinea and is spoken by By the latter part of the nineteenth century, trade con- more than half the population; however, English is tacts between English speakers and Pacific islanders the official medium of education. Although Bislama had led to the formation of English-based pidgins spo- is recognized by the constitution of Vanuatu as the ken in various forms and with differing degrees of sta- national language of the country, it is forbidden in bility in almost the entire Pacific basin from New the schools. English and French, the languages of the Guinea to Pitcairn island, from the Marshall Islands former colonial powers, are still the official lan- and Hawaii to New Caledonia and New Zealand. Not guages of education. all of these have survived. In Micronesia and most of The only example of a German creole is Rabaul Polynesia (with the exception of Hawaii), forms of Creole German, called Unserdeutsch, ‘our German’, pidgin English disappeared because their role in inter- which arose at the turn of the century as a lingua franca nal communication was quite limited. Chinese Pidgin of the Catholic mixed-race community in Vunapope near English, the oldest form of Pidgin English in the Rabaul in what is now East New Britain Province of Pacific, which developed around the port of Canton in Papua New Guinea. This was formerly part of the 1700, has subsequently died out, although a mixed German colony Kaiserwilhelmsland from 1884 to 1914. form of Melanesian-Chinese Pidgin English is still This language was apparently used by children at a mis- used on Nauru in a variety of commercial contexts, sion boarding school and creolized in one generation. It including Chinese trade stores and restaurants, and in is now extinct. the phosphate mining industry, which began in 1906. The only example of a French creole is Tayo, spoken In addition, there are a number of pidgins and cre- in the southern part of New Caledonia around St. Louis, oles based on indigenous languages such as Hiri Motu the site of a Catholic mission and plantation, where lin- in Papua New Guinea, one of the three official lan- guistic diversity among the surrounding tribes led to the guages of Papua New Guinea (along with English and emergence of a creolized French as a lingua franca. Tok Pisin). A number of indigenous pidgins such as Pidgin Maori, Pidgin Fijian, and Pidgin Hawaiian are no longer in use. References Pidgin and creole languages based on English are Bellwood, Peter. 1991. The Austronesian dispersal and the ori- more numerous than those related to any other lan- gin of languages. Scientific American July. 70. guage, attesting to the greater spread of English than Foley, William. 1986. The Papuan languages of New Guinea. any other metropolitan language. Because the French Cambridge: Cambridge University Press. and Germans were relative latecomers to the Pacific, Hollyman, K.L. 1971. French in the Pacific. Current trends in linguistics, Vol. 8, Linguistics in Oceania, ed. by Thomas A. they found varieties of Pidgin English well established Sebeok. The Hague: Mouton. in many areas. Melanesian Pidgin English emerged on Krupa, Viktor. 1982. The Polynesian languages: a guide. plantations in Queensland, Australia, to which workers London: Routledge & Kegan Paul. from all over Melanesia were recruited, and later Lynch, John. 1998. Pacific languages. Honolulu: University of evolved into a number of distinctly named varieties, Hawaii Press. Romaine, Suzanne. 1991. The Pacific. English around the including Tok Pisin, spoken in Papua New Guinea; World: sociolinguistic perspectives, ed. by Jenny Cheshire. Bislama, spoken in Vanuatu; and Pijin, spoken in the Cambridge: Cambridge University Press. Solomon Islands. In these highly multilingual coun- SUZANNE ROMAINE tries, these languages are the normal everyday medium of communication for millions of people. Tok Pisin is See also Austronesian; New Guinea; Pidgins and the largest language in the South Pacific today, with as Creoles 803 PAMA-NYUNGAN Pama-Nyungan The term Pama-Nyungan, coined by Ken Hale in the phrases). There are a few exceptions: the Tangkic lan- early 1960s after the words for ‘man’ in the northeast- guages are suffixing and dependent marking but non- ern and southwestern extremities of Australia, refers to Pama-Nyungan; Yanyuwa is Pama-Nyungan but a group of 160 or so languages spoken over about seven prefixing and head marking. The classification has eighths of the continent (shown in gray on the map). been the subject of considerable controversy in recent This group was initially proposed on the basis of lexi- years. There is, however, no widely accepted alterna- costatistical comparisons of 100 basic words, and was tive, and most Australianists accept it in broad outline. believed to represent a single family of genetically relat- ed languages. The remaining eighth of the continent is Current State of the Languages considerably more diverse, and the languages were assigned to some 28 additional families, referred to col- Most Australian languages (Pama-Nyungan and non- lectively as non-Pama-Nyungan. It was considered that Pama-Nyungan) are endangered or moribund. Many almost all Australian languages, Pama-Nyungan and languages of the east and southeast, where conflict non-Pama-Nyungan, are ultimately genetically related. with Europeans was particularly intense and violent, The lexicostatistical classification correlates well are no longer used as vehicles of everyday communi- with a typological classification (the essence of which cation, and are effectively dead; some disappeared was proposed in 1940 by Arthur Capell). Most Pama- before adequate records could be made of them. On Nyungan languages are ‘suffixing’ (by definition, have the other hand, some of the most viable languages, only suffixes, and no prefixes) and dependent marking with the best chance of future survival, are Pama- (grammatical relations signalled on noun phrases), Nyungan. These include the Western Desert language, while non-Pama-Nyungan languages are overwhelm- with some 4,000–5,000 speakers; Arrernte and ingly ‘prefixing’ (by definition, have both prefixes Warlpiri, with some 3,000 speakers each; Yolngu vari- and suffixes) and head marking (grammatical relations eties with some 2,000; and Kala Lagaw Ya (Torres signaled on the verb, and sometimes also on noun Strait Islands) with some 3,000–4,000 speakers. 1 Kalibamic 12 Gumbaynggir Mabuiagic Kala Lagaw Ya 2 Kukatyic? 13 Marawaric 3 Karantyic? 14 Yotayotic Yolngic 4 Walangama? 15 Yaithmathangic Arnhem Land Gulf of 5 Waramungic Daly Carpentaria Lamalamic 6 Kalkatungic River il ie s Warluwarric fam 7 Yalarnngic ngan -Nyu Yalanjic Mbarbaram 8 Yandic -P ama Tangkic Non Yidinyic 9 Durubulic Kimberley Garrwan 1 23 4 Dyirbalic 10 Giabalic? 5 11 Bandjalanic Warluwarra Mayapic Walmajarri 6 Warlpiri 7 8 Pama-Maric Pilbara Arandic Arrernte Western Waka-Kabic Desert South West varieties Nganyaywana Karnic Nyungic 9 10 13 11 12 Wiradjuric ic Darling ur -K in Kaurna Yu Narrinyeric 14 Kulunic 15 Boundary of Pama-Nyungan and non-Pama-Nyungan Kurnic Group boundaries Questionable groups 804  PAMA-NYUNGAN Efforts are being made by speakers and their proto-non-Pama-Nyungan. Evans (1988) proposed descendants to reintroduce some dead Pama-Nyungan further that Pama-Nyungan could be identified as a languages (e.g. Kaurna in the Adelaide region), and to subgroup (within Australian) on the basis of specific strengthen some weakening languages (e.g. phonetic and grammatical characteristics, for example, Walmajarri in the northwest). But few such efforts the wide occurrence of the variants -lu
-nggu as a have achieved appreciable results, and considerably marker for ergative case. more needs to be done before the process of attrition Dixon has always been a virulent critic of the can be considered to have been arrested generally, or notion that Pama-Nyungan languages constitute a for any language. genetic family of their own, lambasting the idea on methodological and other grounds. Dixon has made some valid criticisms of lexicostatistical methodology. Status of Pama-Nyungan and Relation to non- However, the lexicostatistical classification was Pama-Nyungan always regarded as a tentative means of providing a The notion of a Pama-Nyungan family is widely preliminary genetic classification, one that must be accepted by Australianists; most believe that a genet- refined and supported by the comparative method. ic family of roughly the size shown on the map will Most Australianists who accept the classification do so ultimately prove valid, although we are a long way tentatively (an imperfect or faulty classification is from providing convincing arguments. It has not been preferable to none, provided that its provisional status established as a family by rigorous application of the is borne in mind). comparative method: little of (ancestral) proto-Pama- Dixon also argues against Pama-Nyungan on the Nyungan phonology, lexicon, or grammar has been grounds that the putative innovations in the pronomi- reconstructed, and few subgroupings have been nal and case-marking systems—e.g. the first-person argued on the evidence of shared innovations. One dual inclusive ngali ‘we two (including you)’ and the reason for this situation is that investigators have ergative variants -lu
-nggu—are not present in all assumed the genetic relatedness of Australian lan- Pama-Nyungan languages. Dixon avers that areal dif- guages, and have focused on the search for shared fusion rather than common ancestry provides the sim- innovations, rather than attempting to reconstruct plest explanation for the distribution of these features. proto-Pama-Nyungan. He suggests that the -nggu allomorph was independ- With improved descriptions of more languages, and ently innovated (via natural phonological condition- use of the comparative method, the membership and ing) in four different places, and subsequently diffused scope of Pama-Nyungan have been refined somewhat areally. Others consider that the distribution of -nggu since the 1960s, as have details of the subclassifica- is better explained as a retention from a proto-lan- tion. Thus, Mbabaram (Queensland rain forest region) guage that was lost in some modern languages or and Nganyaywana (northeast New South Wales), orig- groups. And as Evans (2003) pertinently observes, inally excluded from the Australian superfamily, were although instances of diffusion of specific morpheme shown by Robert Dixon and Terry Crowley, respec- variants are well attested in Australia, diffusion of tively, in the 1970s to be closely related to their neigh- variant sets (with complex conditioning factors) such bors; both had undergone radical sound changes that as -lu
-nggu are not. In a similar way, many obscured cognates (related words). Australianists consider it to be probable that the Perhaps the major revision was proposed in the late absence of the first-person dual inclusive ngali in a 1980s by Barry Blake (1988) and Nicholas Evans few Pama-Nyungan languages represents loss of the (1988). They reclassified the Tangkic languages, pre- form in these languages, rather than incomplete diffu- viously considered to be Pama-Nyungan (largely on sion over the Pama-Nyungan region. typological grounds, since these languages show low Both Barry Alpher (e.g. 1990) and Harold Koch vocabulary sharing with Pama-Nyungan) as non- (e.g. 2003) argue the advantages, indeed necessity, of Pama-Nyungan, and Yanyuwa, previously taken to be taking inflectional paradigms into account in recon- a non-Pama-Nyungan isolate, as Pama-Nyungan. They structing proto-Pama-Nyungan, and not merely the assigned Yanyuwa to a subgroup with Warluwarra, forms of roots and affixes separately. Both show that spoken to the south and separated from Yanyuwa by sets of paradigmatic forms—verb inflection paradigms the Garrwan languages (whose classification has also by Alpher (1990), and case-paradigms for singular proved problematic). (These are indicated by vertical free pronominals and interrogative particles by Koch hatching on the map.) Their evidence mainly con- (2003)—can be identified in a diverse selection of cerned forms of pronominals and case-markers. Pama-Nyungan languages that permit reconstruction Blake (1988) reconstructed a pronominal system of inflectional paradigms, and thus support the genet- for proto-Pama-Nyungan, and a different one for ic unity of Pama-Nyungan. 805 PAMA-NYUNGAN Subgrouping innovations in personal pronouns, nominal inflections, and interrogatives. In Karnic languages, they include The map gives an approximate idea of the groups lexical and morphological innovations. within Pama-Nyungan. Some 30 groups are identified, which are further divided into almost seventy sub- Origin and Dispersal of Pama-Nyungan groups. The groups and subgroups differ considerably Languages in distribution and size. One group, the Nyungic group, takes up nearly half of the Pama-Nyungan The uneven distribution of language families across region, and is divided into about a dozen subgroups. Australia, and of groups within the Pama-Nyungan On the other hand, Warumingic, Kalkatungic, and region is striking, and invites explanation. Some pro- Yandic consist of single languages. posals have been put forward. These accept that proto- The establishment of the groups and subgroups of Pama-Nyungan was originally spoken in a much Pama-Nyungan is in a very preliminary state, and smaller homeland region, and that the continent as a there are doubts about the status of many. Various whole once showed linguistic diversity comparable to changes to the group and subgroup structure have been that of the non-Pama-Nyungan region today. Around proposed since 1966. Attempts have also been made to 3,000–5,000 years ago (the dating is a guess, based on motivate groups or subgroups by more reliable means the apparent diversity within the family), it began than the original lexicostatistical criteria. In a number spreading out to ultimately cover the large region it of cases, however, the genetic relatedness of a set of occupies today. languages has been established, although not subgroup Stephen Wurm (1972) suggests a link with the status by identification of shared innovations. emergence of an advanced technology (ground axes In a careful study of the Pilbara situation, Alan and hafted stone tools) that also appears on the arche- Dench (2001) has pointed out a variety of difficulties in ological horizon some 5,000 years ago. He proposes determining whether shared morphological and syntac- that users of this technology, speakers of proto-Pama- tic features are innovations indicative of common inher- Nyungan, expanded out into areas formerly occupied itance or the result of diffusion. He concludes that by speakers of different languages, their advanced although the languages are probably genetically related, technology presumably giving them an edge over the there is no evidence that they form a subgroup of any local inhabitants. larger group; the languages form, in his view, a linguis- A more elaborate model is suggested by Nicolas tic area within which much lexical and grammatical dif- Evans and Rhys Jones (1997). These authors propose fusion has occurred. Although Dench takes the extreme (on the basis of the diversity of subgroups of Pama- position that diffusional explanations should be privi- Nyungan and the number of potential sister families) a leged over genetic retention, he correctly observes the homeland region in the vicinity of the Gulf of difficulties inherent in distinguishing genetically inher- Carpenteria. In contrast to Wurm, they do not link the ited features from areally diffused features. spread of Pama-Nyungan to major migrations and Dixon’s current position (2002) is that the putative conquests by the wielders of a new technology, but to Pama-Nyungan family is made up of some 30 low- a cluster of social changes, including the spread of cer- level genetic ‘subgroups’, which are effectively small emonial activity associated with the new highly valued language families on their own (the prefix ‘sub-’, he technology, and the exploitation of new vegetable explains, does not imply that there is necessarily any foods that could support large intergroup ceremonial higher order containing group, only that there may be). activities. They suggest that initiation into the rituals Dixon further proposes that some of the accepted associated with the new technology might attract pay- groups within Pama-Nyungan are areal groups, not ment in the form of wives for the sons of the initiators, genetic subgroups. These include the Arandic and adja- thus giving rise to larger networks of exogamous cent Karnic groups of Central Australia. However, social interactions than had existed previously. In such Koch (2004) and Claire Bowern (2001) provide care- circumstances, proto-Pama-Nyungan could well have fully argued cases for their subgroup status on the basis been a high-status language, and its usage a sign of the of shared innovations, at least some of which (e.g. par- status of performers of the new ceremony. Thus, adigmatic irregularities) are more likely to be reten- Pama-Nyungan could have spread out with relatively tions from an earlier proto-language than resulting minor (though still discernible) demographic changes. from diffusion. Examples of such shared innovations Both models are of course speculative. Nevertheless peculiar to the Arandic languages are phonological they are noteworthy in that they attempt to forge links changes (including complete loss of all initial conso- between linguistic, anthropological, and archeological nants, centralization of vowels in unstressed syllables, evidence. They also make potentially testable pre- and stress shift to the second syllable of words), and dictions concerning the distribution of bio-genetic 806  PÀNINI " markers and substrate influences on modern Pama- Evans, Nicholas. 1988. Arguments for Pama-Nyungan as a Nyungan languages. genetic subgroup, with particular reference to initial lami- nalization. Aboriginal Linguistics 1. 91–110. ———. 2003 Introduction: comparative non-Pama-Nyungan and Australian historical linguistics. Comparative studies in References non-Pama-Nyungan, ed. by Nicholas Evans, 3–25. Alpher, Barry. 1990. Some proto-Pama-Nyungan paradigms: a Canberra: Pacific Linguistics. verb in the hand is worth two in the phylum. Studies in com- Evans, Nicholas and Rhys Jones. 1997. The cradle of the Pama- parative Pama-Nyungan, ed. by Geoffrey N. O’Grady and Nyungans: archaeological and linguistic speculations. Darrell T. Tryon, 155–71. Canberra: Pacific Linguistics. Linguistics and archaeology: aboriginal Australia in global Blake, Barry. 1988. Redefining Pama-Nyungan: towards the perspective, ed. by Patrick McConvell and Nicholas Evans, prehistory of Australian languages. Aboriginal Linguistics 1. 385–417. Melbourne: Oxford University Press. 1–90. Koch, Harold. The case for Pama-Nyungan: evidence from Bowern, Claire L. 2001. Karnic classification revisited. Forty Remains forthcoming inflectional morphology. Paper pre- years on: Ken Hale and Australian languages, ed. by Jane sented to 17th International Congress of Linguists, Prague, Simpson, David Nash, Mary Laughren, Peter Austin, and Czech Republic, July 24–29, 2003, to appear in Proceedings Barry Alpher, 245–61. Canberra: Pacific Linguistics. of the congress. Bowern, Claire, and Harold Koch (eds.) 2004. Australian lan- ———. The Arandic subgroup of Australian languages. In guages: Classification and the comparative method. Bowern and Koch, pp. 127–50. Amsterdam: John Benjamins. Miceli, Luisa. Pama-Nyungan as a genetic entity. In Bowern Dench, Alan. 2001. Descent and diffusion: the complexity of and Koch, pp. 61–86. the Pilbara situation. Areal diffusion and genetic inheri- O’Grady, Geoffrey N., Carl F. Voegelin, and Frances M. tance: problems in comparative linguistics, ed. by Alexandra Voegelin. 1966. Languages of the World: Indo-Pacific fasci- Y. Aikhenvald and Robert M.W. Dixon, 105–33. Oxford: cle 6. Anthropological Linguistics 8. 1–197. Oxford University Press. Wurm, Stephen A. 1972. Languages of Australia and Tasmania. Dixon, Robert M.W. 2002. Australian languages: their nature The Hague: Mouton. and development. Cambridge: Cambridge University Press. WILLIAM B. MCGREGOR PÁnini " Virtually nothing is known about the life of the great judging from the fact that PÁnini mentions the names Indian grammarian PÁnini, except for the fact that he of ten of his predecessors. " was a native of S´alÁtura," in northwestern India The language described by PÁnini is a variety of (GandhÁra); from his matronymic, we learn that his Sanskrit that does not correspond "exactly to the lan- mother’s name was DÁks¯i. None of the numerous guage of any known text. This fact has aroused suspi- commentators say anything" about the time in which he cion, especially among Western scholars in the lived, but, based on the variety of Sanskrit he nineteenth century, leading some to blame PÁnini of describes, he is now thought to have lived in the fourth " describing an invented variety. A further complication (or maybe fifth) century BCE. His grammatical trea- is given by the fact that even the earliest commentators tise, the AstÁ dhyÁ y¯i (‘Eight chapters’), is the most are at variance with PÁnini on some points: most like- "" important grammatical description of Sanskrit, and a ly, different varieties of "Sanskrit underlie their descrip- large part of the Indian grammatical tradition after tions. The philological tradition of PÁnini’s text is not PÁnini consists of commentaries on his work. " without problems either, as some minor" parts probably The style of the AstÁ dhyÁ y¯i appears quite striking if belong to later commentaries (notably to Patañjali’s considered from the" "point of view of Western gram- MahÁ bhÁ sya see below). matical tradition: it is constituted by 3,996 su¯tra, or " It is noteworthy that PÁnini occasionally mentions aphoristic rules, by which PÁnini not only describes differences between the variety " he describes and earli- grammatical rules but also defines " the way in which er varieties, notably Vedic Sanskrit. This has led some grammatical terms and constructions are used in his scholars to state that PÁnini had a historical view of metalanguage. PÁnini’s terminology constitutes his grammar: however, in view " of the existence of earlier own achievement only " in part: grammatical descrip- grammarians, who had mostly described Vedic, it tion already relied on a rather conspicuous tradition, seems more likely that PÁnini only wanted to mention " 807 PÀNINI " variants found in other treatises, without historical predecessor of both structuralism and generativism. It intentions. became apparent that PÁnini had a much more thor- Among the most important commentators of PÁnini " ough understanding of phonology than his Greek con- are KÁtyÁyana, who, according to tradition, lived" in temporaries, and his description of vocalic sandhi was the third century BCE and annotated about a third of even said to presuppose a theory similar to (early) PÁn ini’s work, and Patañjali, author of the generative phonology. Mah " Á bhÁ sya (‘Great commentary’, around 150 BCE). Perhaps the field in which PÁnini has enjoyed most Patañjali’s" commentary, which takes as a starting point admiration lately is case theory. In" the late 1960s, after KÁtyÁyana’s notes (which, in fact, we only know the outburst of Fillmorean Case Grammar, it was point- through the MahÁ bhÁ sya), and adds comments, partly ed out that PÁnini’s notion of kÁ raka corresponded to to the notes, and partly " to PÁnini’s text, also contains " what Fillmore called ‘deep cases’, or, to state it in more in the introduction some general " statements about the modern terminology, semantic roles. Thus, there is Vedas and about grammar. kÁ raka of the agent, which can be expressed in the PÁnini’s grammar became known to Western schol- nominative or instrumental, the use of the latter ending " end of the eighteenth century, with the begin- ars at the being demonstrated by means of passive constructions. ning of Indological studies; since then, its fortune in In fact, when describing the use of ‘affixes’ (meaning the West has undergone quite dramatic changes. both verbal and nominal affixes), PÁnini I mostly has a During the nineteenth century, the striking differences semantically based approach, rather " than an approach between the frameworks in which grammatical based on forms, as was usual within the Greco-Roman description was organized by PÁnini and by the Greco- tradition. Some later grammarians gave sematically Roman tradition, together with the " increasing convinc- finer definitions: for example, within the philosophical tion that insights into language could be gained only school of NyÁya the kÁ raka of the agent was conceived by means of historical investigation, progressively led as implying intentionality, so that only animate entities to a negative evaluation of PÁnini’s work. Perhaps the could be regarded as agents. On the contrary, PÁnini and worst opinions about PÁn ini " were held by the his followers (PÁnin¯iyas) called agents all subjects" of " American indologist Withney, who not only thought, " action verbs, including inanimate entities (in examples as many of his contemporaries, that the Indian gram- such as ‘the cart goes to the village’). marian described an invented language but also dis- missed his grammatical terminology as crude and References philosophically not well founded. In spite of this and other severe judgments, Indian Cardona, George. 1994. Indian linguistics. History of linguis- grammatical tradition influenced Indo-European linguis- tics, Vol. 1, ed. by G. Lepschy. Oxford: Oxford University Press. tics in its early development, partly providing a basis for Joshi, Venkateshshastri. 1980. Problems in Sanskrit grammar. the terminology, partly because the Indian use of listing Pune: 2 Dastane Ramchandra & Co. roots helped to develop the concept of morpheme. Rocher, Rosane. 1975. India. Current trends in linguistics, Vol. The twentieth century brought a complete reevalu- 13, ed. by T. Sebeok. The Hague, Paris: Mouton. ation of PÁnini’s methods, in some cases with enthusi- Vasu, S´r¯is´a a Chandra. 1962. The AstÁdhyÁy¯i of PÁnini "" Allahabad 1891, repr. Delhi: Motilal Banarsidass. " " astic appraisals. The systematic character of PÁnini’s " ¯tra SILVIA LURAGHI treatment of grammar, and the formulation of the su as grammatical rule appealed to scholars working See also India; Indian Traditional Grammar; San- within various frameworks; PÁnini was hailed as a skrit " Papiamento Papiamento (Papiamentu) is a language spoken north natively by about a quarter of a million people, most- of Venezuela on the Leeward Islands of the ly in the Caribbean, and by some 40,000 people living Netherlands, Antilles, Curaçao, and Bonaire, and on in the Netherlands. Although the official language on the island of Aruba, otherwise known as the ABC the ABC islands is Dutch, Papiamento pervades all islands. It is estimated that Papiamento is spoken levels of society, including the educational system and 808  PAPIAMENTO government. Papiamento also has a rich literary tradi- today, replete with elements of the Spanish language tion, an established presence in electronic and digital thought to be less prominent in the earlier stages of media, and several newspapers and magazines are also Papiamento. Thus, questions concerning the precise published in the language. The name is derived from origins of Papiamento continue and as it is not direct- the verb papia meaning ‘to speak’ and the suffix, ly related to any other Portuguese- or Spanish-based -mento. Thus, Papiamento can be literally interpreted creole, scholars can only continue to piece together as ‘(the act of) speaking’. linguistic and historical evidence to help shed light on Papiamento is a creole language, which emerged the genesis and subsequent evolution of the language. during the second half of the seventeenth century Papiamento has existed in tandem with Dutch on around the period of European colonization of the the islands for centuries; however, the Dutch influ- Caribbean islands. Like other creole languages, ence in the language is far less obvious than the Papiamento inherited structural characteristics from a Iberian influences, in both historical and contempo- mixture of West-African languages (the substrate lan- rary contexts. In spite of the fact that Dutch is the offi- guages). Approximately two thirds of its vocabulary cial language of the regions where Papiamento is comes from Iberian languages (the superstrate lan- spoken natively, it is quickly becoming far less com- guages), about a quarter from Dutch, and the rest from mon even in the areas of government administration, other languages including French and English. The education, and the judiciary system, realms of society language also has traces of Amerindian and West in which it was traditionally used, testifying to the African vocabulary as well. high level of social prestige affiliated with The history of the ABC islands provides some Papiamento—a fact that makes it somewhat unique insight into the social and linguistic context in which among creole languages. Papiamento’s social curren- Papiamento developed: originally populated by cy is also demonstrated by the fact that it has an offi- Amerindians speaking an Arawakan language, cial orthographic system, differing slightly on the Curaçao was settled by the Spanish in 1527. After a islands of Curaçao and Bonaire than on the island of century of Spanish occupation, the Dutch conquered Aruba. These spelling differences can be seen in the Curaçao in 1634 at which time most of the name of the language itself, while the former islands Amerindians and the Spanish were driven from the prefer Papiamentu as opposed to Papiamento, its island. Shortly thereafter, it became a depot for the Aruban counterpart. Additionally, after careful lan- slave-trading practices of the Dutch West-India guage planning efforts, in 1987, Papiamento was offi- Company. Not long after the fall of the Dutch cially integrated into the educational system of the Brazilian Empire in 1654, a substantial population of ABC islands. Portuguese Jews (and non-Jewish settlers as well) The phonemes (sounds) of Papiamento include arrived on the island. Despite the diverse linguistic almost all the consonants and vowels found in backgrounds of the island’s inhabitants, Papiamentu is American Spanish, as well as some Dutch vowels. A believed to have stabilized as a creole on Curaçao linguistic trait of Papiamento not found in either the around 1700 and spread to Bonaire and Aruba by the Iberian languages or Dutch, and which is rare in creole end of the century. languages in general, is the use of high (´) and low (`) Although this brief account of the islands’ history tones to distinguish between pairs of words (eg. párà demonstrates the importance of Dutch and Portuguese ‘bird’, pàrá ‘to stand, to stop’; wárdà ‘guard service’, in the incipience of the language, Papiamento clearly wàrdá ‘to wait’). This feature of the language is attrib- exhibits Spanish elements and in fact, a diary belong- uted to the West African, or substratal influences. West ing to a Jesuit missionary on Curaçao dating from African structural influences typical of creole lan- 1704, first referred to the language on the island as guages can be noticed in serial-verb constructions ‘broken Spanish’. This has caused some controversy such as Ela kore bai su kas or ‘She or He ran (go) surrounding the origins of the language with respect to home,’ or focus particles found in sentence-initial whether it is actually a Portuguese- or Spanish-based positions like Ta e buki m’a dunabu or ‘(Focus) I gave creole. Some scholars argue that the Portuguese con- you the book.’ Below are some Papiamento words that tribution to the language has been obscured by its con- illustrate the Iberian and Dutch contributions to the tinual contact with Spanish. Consequently, the earliest language: forms of Papiamento are thought to have exhibited the Portuguese: bai ‘to go’; preto ‘black’; bringa ‘to Portuguese contribution more so than the Spanish; fight’ (derived from brigar) however, the earliest written attestation of the lan- Spanish: Dios ‘God’; salida ‘departure’; ruman guage, a letter dating from 1776 written by a ‘brother’ (derived from hermano) Portuguese Jew to his wife, shows a remarkable Dutch: wikent ‘weekend’; spiel ‘mirror’ (derived resemblance to the language spoken on the islands from spiegel), and wak ‘to see’ (derived from waken). 809 PAPIAMENTO References Goilo, E.R. 1997. Papiamentu textbook. Aruba: VAD Printers. Kouwenberg, Sylvia, and Eric Murray. 1994. Papiamentu. Agard, Frederick B. 1985. Papiamentu grammars and the struc- Newcastle: Lincom Europa. ture of the language. For Gordon H. Fairbanks, ed. by Martinus, Frank. 1990. Papiamentu: the road to emancipation. Veneeta Z. Ascon and Richard L. Leed. Honolulu: Language reform: history and future, ed. by István Fodor University of Hawaii Press. and Claude Hagège. Hamburg: Helmut Buske Verlag. Anderson, Roger W. 1990. Papiamentu tense-aspect, with special Wood, Richard E. 1972. New light on the origins of attention to discourse. Pidgin and Creole Tense-Mood-Aspect Papiamentu: an eighteenth-century letter. Neophilologus 56. systems, ed. by Jon Singler. Amsterdam: John Benjamins. Arends, Jacques, Pieter Muysken, and Norval Smith (eds.) BETSY BARRY 1994. Pidgins and Creoles: an introduction. Amsterdam: John Benjamins. See also Caribbean; Pidgins and Creoles Paralanguage The term paralanguage refers to a wide variety of non- an individual, or a cultural, social, or linguistic group verbal behaviors relevant to communication. Vocal par- (see Laver (1980) for a thorough introduction to this alanguage includes voice quality, intonation, and voice subject; and Kent and Ball (2000) for an overview of dynamics such as loudness, pitch range, and rate of recent research). speech. Bodily paralanguage incorporates facial expres- Voice quality settings may be roughly divided into sion, hand gestures, posture, touch, and physical proxim- two categories: laryngeal and supralaryngeal. Laryngeal ity to others. These behaviors, all of which may co-occur settings (also known as ‘phonation types’) involve dif- with verbal communication, differ from language per se ferent patterns of vocal fold vibration. These settings in that they do not form part of a hierarchically struc- include modal voice, breathy voice, creaky voice, harsh tured, coded system in which a discrete symbol is arbi- voice, falsetto voice, whisper, and the various possible trarily associated with a particular meaning. combinations of these phonation types (e.g. harsh whis- While paralanguage does not convey meaning in pery voice). Modal voice involves quasiperiodic vibra- the same way as language, it plays important functions tion of the vocal folds, in that the vocal folds are open in communication. Paralinguistic behavior conveys the and closed for approximately equal periods of time, emotional state of speakers and listeners, expresses lending the voice a smooth, even quality. Other laryngeal semantic nuances that could not be inferred from lan- configurations may be thought of as departures from this guage alone, helps regulate the flow of interaction, and ‘neutral’ setting. For example, for a breathy voice, the provides cues regarding the sex, age, culture, and vocal folds are open longer than they are closed, pro- socioeconomic status of speakers. ducing audible friction. By contrast, for a harsh voice, The study of paralanguage has grown significantly the vocal folds are closed longer than they are open, and over the past 30 years. Technological advances such as the vibration is irregular, producing a rasping, uneven audio and videorecording, and more recently, the quality that is often enhanced by the vibration of the ven- widespread availability of computers and specialized tricular or ‘false folds’. For a whisper, the vocal folds do software analysis packages, have facilitated a more not vibrate at all: the vocal folds are closed along most systematic approach to the study of paralanguage. The of their length, and a triangular opening is left at the pos- sections below describe some key paralinguistic chan- terior part of the vocal folds, generating strong friction as nels: voice quality, intonation, and hand gestures (see air passes through. Some laryngeal settings are readily Siegman and Feldstein (1987) and Feldman and Rime identifiable by their pitch, as, for instance, the very high (1991) for an overview of these and other paralinguis- pitch of falsetto voice, involving the vibration of a small tic channels). portion of the vocal folds, and the very low pitch of creaky voice, in which the listener can hear each separate vibration of the vocal folds. Voice Quality Supralaryngeal settings involve different configura- Voice quality may be defined as a quasipermanent set- tions of the larynx, pharynx, soft palate, tongue, lips, ting of the vocal tract that characterizes the speech of and jaw, all of which affect the resonance of the voice. 810  PARALANGUAGE The larynx may be raised or lowered, the pharynx prominent, or ‘accented’ syllable in the tone group. widened or constricted, the soft palate raised or low- Within the British system of intonation analysis, there ered, the tongue and lips protruded or retracted, and are five major nuclear tones: falls, rises, levels, the jaw lowered or raised—all of which, separately or fall–rises, and rise–falls. Within the predominantly in combination, have distinct acoustic and auditory American autosegmental metrical system, (for exam- effects. For example, a speaker with a habitually low- ple, the ToBI, or ‘Tone and Break Index’ system), ered soft palate will have a nasal voice. A speaker with these and other contours are expressed as different pat- a lowered larynx will have a deeper sounding voice terns of high and low tones (‘H’ and ‘L’). (See than he or she would with a neutral or raised larynx, by Cruttenden (1997) for a thorough overview of intona- virtue of having produced a longer vocal tract with tion in the British and American traditions.) correspondingly lower formant frequencies. Intonation may have linguistic or paralinguistic The combination of various laryngeal and suprala- meaning. For instance, in English, an utterance pro- ryngeal settings contributes to the characteristic ‘tone duced with falling intonation is associated with a state- of voice’ or ‘accent’ that listeners come to associate ment or declarative sentence, while a rising intonation with different individuals and regional or social is often used for questions. Such uses of intonation are groups. For example, North American females are often considered linguistic. However, there are many more likely than their male counterparts to speak in a intonational variations that carry paralinguistic mean- habitually breathy voice—a habit that some women ing. For instance, some speakers use rising intonation may deliberately alter in order to sound more profes- for declarative sentences as a way of checking whether sional in a particular work or social setting. Certain or not the listener understands what they are saying, to regional accents of English may be distinguished by a suggest openness to an alternative point of view, or, particular voice quality setting, such as, for instance, possibly, to convey a lack of confidence in what they the nasal quality characteristic of speakers in the are saying. This particular usage is more common southern United States of America. Some foreign among young North American females than their male accents are stereotypically associated with a particular counterparts. The word ‘yes’, spoken slowly with a voice quality setting, such as the quasi-permanent rising then falling pitch, may suggest doubt or hesita- retraction of the tongue tip (or ‘retroflexion’) that tion on the speaker’s part, even though the word itself often characterizes an East Indian accent. conveys agreement. The same word, spoken slightly In addition to suggesting social, cultural, and lin- more quickly, with a falling then rising pitch, may sug- guistic characteristics of speakers, voice quality set- gest impatience or irritation on the part of the speaker. tings can be deliberately exploited by a single These are but a few of many possible examples of the individual in different communicative settings to con- paralinguistic use of intonation. vey paralinguistic meaning—whether to signal the Intonation, in combination with voice quality, and speaker’s actual emotional state or attitude toward other features of voice dynamics, such as variations in what he or she is saying. For example, in English- pitch range (wide vs. narrow), intensity (loud vs. soft), speaking countries, a whisper is often used to signal and rate of speech (fast vs. slow), provide the listener secrecy, while a breathy voice may be used to convey with significant cues regarding the speaker’s attitude intimacy or sexual attractiveness. A creaky voice may and/or emotional state. For example, a wide pitch range sometimes be used to suggest sadness or boredom, is often characteristic of a happy mood, and a more while lowered larynx voice may be used by speakers restricted pitch range is characteristic of a more who wish to sound serious or authoritative. depressed state. Anger is often associated in a loud, high-pitched voice, while fear may be expressed with a quieter, high-pitched voice, sometimes combined with Intonation a whispery quality. In a state of excitement, people Intonation refers to the variations in pitch or funda- often speak more quickly than when they are calm. The mental frequency that occur in speech, creating its qualities associated with a particular mood may be characteristic melodies. In intonational analysis, these characteristic of a speaker’s habitual use of the voice, melodies are broken down into ‘tone units’ or ‘intona- possibly suggesting more permanent personality traits. tion groups’. Tone units usually consist of five or six words, often constituting a grammatical constituent Gestures such as a sentence, clause, or phrase, but sometimes consisting of a single word, or even a single syllable. While speaking, people can often be seen moving their The major characteristic of a tone unit, independent of hands up and down, pointing, or drawing pictures in its length or grammatical status, is that it contains one the air, in ways that appear connected to what they are major pitch movement, realized primarily on the most saying. These movements, as opposed to other hand 811 PARALANGUAGE movements that they might happen to make while which the hands rise from a position of rest; a stroke speaking, such as smoothing their hair or removing phase, the most ‘energetic’ portion of the gesture, or lint from their pants (the latter are called “self-adap- the point at which the gesture appears to have reached tors”), are considered communicative gestures. its ‘destination’; and a recovery or offset phase, the Adam Kendon—a pioneer in the research on ges- point at which the hands return to a position of rest. tures—has characterized hand gestures as existing The stroke of a gesture almost always co-occurs with along a continuum, ranging from completely nonlin- the word or phrase with which it is meaningfully con- guistic gestures to the fully linguistic hand movements nected. Moreover, the stroke often co-occurs with the that comprise actual sign languages (see McNeill most prominent, or accented syllable in a tone unit, (1992:37) for a fuller description of Kendon’s contin- suggesting that the planning of gesture is closely uum). In the middle of this continuum are the semilin- linked with the planning of speech. guistic hand gestures known as ‘emblems’. Emblems In recent years, there has been a surge of interest in have a fixed form that is associated with a particular the study of gesture and other paralinguistic phenom- meaning, and may in this sense, be considered to be ena, such as facial expression, voice quality, and into- quasilexical. Emblems vary considerably from culture nation. Traditionally, these features have often been to culture. For example, the ‘thumbs-up’ gesture often considered to reflect the emotional state of speakers. used in English-speaking countries to signal a victory While this approach continues to form a significant or a generally positive attitude toward a particular sit- part of research on paralanguage, recent research has uation, has an obscene connotation in Iran. Some cul- also highlighted the important semantic information tures—Neopolitan Italian culture, for example—have conveyed through paralinguistic channels (see, for elaborate sets of emblems that are extensively used in example, McNeill 1992, 2000). Moreover, there is a day-to-day conversation. Because of their quasilexical growing interest in the complex ways in which lin- nature, emblematic gestures can often be used instead guistic and paralinguistic channels interact. This line of words in face-to-face interaction for communicative of research is leading to an enriched understanding of effect. communication as integrated, multimodal behavior— However, the majority of communicative hand ges- and perhaps challenging the traditional distinction tures are nonlinguistic, and co-occur with speech. between language and paralanguage. Nonlinguistic gestures have been categorized in a variety of ways over the past 50 years; but the cate- gories developed by McNeill (1992) are those most References commonly used in the current literature on gestures. Armstrong, David, William Stokoe, and Sherman Wilcox. Nonlinguistic gestures may be classified as iconic, 1995. Gesture and the nature of language. Cambridge, New metaphoric, deictic, or beats. Iconic gestures illustrate York, and Melbourne: Cambridge University Press. concrete entities in the world, while metaphoric ges- Birdwhistell, Ray. 1970. Kinesics and context. Philadelphia: University of Pennsylvania Press. tures depict abstract concepts. Iconic and metaphoric Cruttenden, Alan. 1986. Intonation. Cambridge, New York, and gestures have no fixed form, and are spontaneously Melbourne: Cambridge University Press. 2nd edition, 1997. invented by speakers to physically represent some Feldman, Robert, and Bernard Rime (eds.) 1991. Fundamentals aspect of what they are saying. For example, while of nonverbal behavior. Cambridge, New York, Port Chester, saying ‘and the guy just grabbed her purse’, a speaker Melbourne, and Sydney: Cambridge University Press, and Paris: Editions de la Maison des Sciences de l’Homme. might physically depict this action in an iconic ges- Kent, Raymond, and Martin Ball (eds.) 2000. Voice quality ture—thereby often conveying information not fully measurement. San Diego, California: Singular Publishing expressed in words, such as the precise movement Group. involved in the grabbing. Deictic gestures involve Laver, John. 1980. The phonetic description of voice quality. pointing, either to a location within the speaker’s envi- Cambridge: Cambridge University Press. McNeill, David. 1992. Hand and mind. Chicago, and London: ronment, or to an area in space that symbolizes a loca- University of Chicago Press. tion referred to in the accompanying speech. Beats are ———. (ed.) 2000. Language and gesture. Cambridge, New short, rhythmic movements of the hands, often physi- York, Melbourne, and Madrid: Cambridge University Press. cally oriented toward the listener (e.g. involving an Poyatos, Fernando. 1993. Paralanguage. Amsterdam, and open palm facing the listener) and do not represent Philadelphia: John Benjamins Publishing Company. Siegman, Aron, and Stanley Feldstein (eds.) 1987. Nonverbal anything in the accompanying speech. Rather, they are behavior and communication. Hillsdale: Lawrence Erlbaum used to emphasize the delivery of information, help Associates. structure discourse, and help regulate aspects of inter- ALLISON BENNER action, for instance, turn-taking routines. As noted above, nonlinguistic gestures are closely See also Communication Theory; Conversation synchronized with speech. Gestures may be divided Analysis; Emotion and Language; Signed Lan- into three phases: a preparatory or onset phase, in guages 812  PARSING Parsing A recognizer is an algorithm that takes a string as Context-Free Parsers input and either accepts or rejects it, depending on The term context free (CF) was introduced by whether or not the string is a sentence of a particular Chomsky in 1956. CF grammars consist of rules with grammar (Grosz et al. 1986). A parser is a recognizer a single symbol on the left-hand side. Such rules are that also shows all the ways in which the string can be always applicable, since they are in no way dependent derived, in the form of the invoked grammar rules on the nature of constituents elsewhere in the input arranged in trees or nested square brackets. text. Most of the CF parsing algorithms were devel- The most common way to represent a grammar is as oped in the 1960s for use with computer language a set of production rules, which are also called rewrite compilers, but pure CF grammars are not effective for rules. Phrase structure grammars are sets of production describing natural languages. Grammars that are not rules that specify the various ways in which a sentence CF are said to be context sensitive, where the rules can be decomposed into its constituent syntactic units, have been augmented with additional conditions, which themselves can be decomposed into smaller units, meaning that they only apply if certain things are and finally into individual words. The production rule found elsewhere in the input text. S → VP NP shows that a sentence can consist of a noun phrase followed by a verb phrase, and the rule ADJ → small|red|new shows that an adjective can consist of one Top-Down and Bottom-Up Parsing of the following individual words: small, red, or new. A top-down parsing strategy starts with the start sym- Symbols that are themselves further expanded by bol S and then searches through different ways to rules are called nonterminal symbols, which generally rewrite the symbols until either the input sentence is correspond to syntactic classes. Symbols that corre- reproduced or all different ways to rewrite the symbols spond to individual word strings that must be found in have been tried. A top-down parse using the simple the input sentence are called terminal symbols (Rich grammar above to derive the sentence ‘John ate the and Knight 1991). The nonterminal symbol S, which cat’ would proceed as follows: stands for the input sentence in its entirety, is called S the root of the grammar. An example of a simple gram- mar which derives the sentence ‘John ate the cat’ is
NP VP (rewriting S) given by Allen (1995:42).
NAME VP (rewriting NP)
John VP (rewriting NAME) (1) S → NP VP
John V NP (rewriting VP) (2) VP → V NP
John ate NP (rewriting V) (3) NP → NAME
John ate ART N (rewriting NP) (4) NP → ART N
John ate the N (rewriting ART) (5) NAME → John
John ate the cat (rewriting N). (6) V → ate (7) ART → the The bottom-up parsing strategy, on the other hand, (8) N → cat starts with the actual words of the sentence and uses the rewrite rules in reverse until either the S symbol The parse tree corresponding to this parse is shown alone is generated or all combinations of the rewrite in Figure 1. rules have been tried. A bottom-up parse of ‘John ate the cat’ might proceed as follows: S John ate the cat NP VP
NAME ate the cat (rewriting John)
NAME V the cat (rewriting ate) NAME V NP
NAME V ART cat (rewriting the)
NAME V ART N (rewriting cat) John ate ART N
NP V ART N (rewriting NAME) the cat
NP V NP (rewriting ART N)
NP VP (rewriting V NP) Figure 1. A tree representation of John ate the cat.
S (rewriting NP VP) 813 PARSING LR parsers read their input from left to right and Completed constituents are placed on a list called produce a rightmost derivation, i.e. expand the right- the agenda. The algorithm proceeds by selecting a most non-terminals of the production rules first. completed constituent from the agenda. For each matching rule in the grammar, an active arc is com- Chart Parsers bined with the completed constituent in the agenda. Whenever the agenda is empty, all possible syntactic In a standard parser, if an early segment can be ana- interpretations of the next word in the sentence are lyzed in more than one way, each step in the recogni- added. The parse ends when an active arc and a com- tion of subsequent phrases in the sentence will be pleted constituent combine to produce an entire sen- repeated once for each earlier analysis. Chart parsers tence S. If all possible parses of a sentence are (Winograd 1983) provide a means of eliminating such required, the process continues until the agenda is redundant computation. Kay described the chart as ‘a empty. The top-down chart parser is called the Earley kind of well formed substring table’. It allows the parser (Earley, 1970). parser to store the partial results of the matching it has done so far so that this work need not be duplicated. First, we will consider the bottom-up chart parser. Augmented Transition Networks Each word in the input sentence is considered in turn. An early model for natural language grammars was a The current input word is called the key. For each new transition graph called the finite state machine. This key, we look for rules that match a word sequence consists of a network of nodes (states) and directed involving the key—either a rule that begins with the arcs or transitions, marked by arrowheads pointing key, or a rule that has already been started by earlier away from the previous state and toward the next state. keys, which can either be extended or completed by Each arc has a symbol, representing terminal symbols the current key. Allen presents the following example (individual words such as ‘red’ or syntactic classes of a grammar: such as ‘adjective’) which may appear in an input sen- (1) S → NP VP tence. There may be more than one arc arriving at each (2) NP → ART ADJ N state or leaving each state. (3) NP → ART N There is one distinguished state called the start state (4) NP → ADJ N (with no inward arcs), and a number of distinguished (5) VP → AUX VP states called final states (with no outward arcs). (6) VP → V NP Starting at the start state, the words of the sentence to If we are parsing a sentence that begins with ART, be parsed are read off in turn. If the first word read off ART becomes the first key, and rules (2) and (3) are matches the symbol on one of the arcs leaving the start matched since their right-hand sides both begin with state, the next state is the one found by following that ART. The following notation is used to show that rules arc. The second word is read in, and if that word (2) and (3) can be continued from the point of the ART matches an arc leaving the second state, the arc is fol- (denoted by ‘o’) onwards. The partially matched rules lowed to the third state. This process continues, and if are stored in the chart, along with their locus (starting we have arrived at one of the final states as a result of position in the sentence) and position of the last word reading in the last word, the sentence has been accept- to match so far: ed by the augmented transition network (ATN) gram- mar. Sometimes an arc can return to the very same (2′) NP → ART o ADJ N (1,2) state that it left, to show that optional additional words (3′) NP → ART o N (1,2) of the type specified by the arc’s symbol are permitted If the next key is ADJ, then rule 4 can be started, at that point in the sentence. and rule 2′ can be extended: ATN are finite state machines augmented by allow- ing not only terminal symbols on the arcs, but nonter- (2′) NP → ART ADJ o N (1,3) minal symbols representing entire phrases (Woods (4′) NP → ADJ o N (2,3) 1970). These phrases are themselves represented by The chart keeps the record of all the rules found so far ATNs. For example, an ATN designed to recognize a which have been matched partially by the previous keys. prepositional phrase is shown in Figure 2. These partially matched rules are called active arcs. Active arcs are never removed from the chart; hence, their original and extended forms are simulta- PREP NP neously present. The chart also records the completely Start S2 Finish matched syntactic constituents (as defined by the grammar rules) found so far in the parse. Figure 2. A simple augmented transition network. 814  PARSING To make the transition from S2 to finish, there must for the same sentence are randomly generated. The be a noun phrase (NP) in the input sentence. In an likelihood of each rule being used in a particular ran- operation called a ‘push’, processing is suspended at dom parse depends on its probability. When the sam- the point the NP is requested, and a new ATN is ple is large enough, the most frequently occurring entered, which accepts subsequent words from the output parse tree over all the random tests is assumed input sentence. This is analogous to a subroutine call to be the most probable parse. in a computer program. When the final state of the new ATN is reached (showing that the sequence of input Tagging words since entering the new ATN was indeed an NP), an operation called a ‘pop’ takes place and processing Tagging is marking items in a text with additional of the input sentence resumes at the point in the origi- information, related to their linguistic properties. This nal ATN from where it left off. A called ATN may information is attached to the text items in the form of itself call other ATNs, and an ATN may call itself codes called tags. Here, we will mainly discuss the (recursion). The record of the sequence in which the automatic assignment of part of speech (POS) catego- ATNs have called each other is called a ‘stack’, analo- ry tags to words in a text; but tagging can also involve gous to a stack of plates, each plate representing an the assignment of semantic categories as a by-product ATN, which may be ‘pushed’ onto or ‘popped’ off the of word sense disambiguation. stack. The most recently called ATN is at the top of the The Brill (1995) tagger operates as follows. An ini- stack, but it cannot be ‘popped’ off until any ATNs it tial guess is made as to the POS of each word in the calls itself are completed. Advantages of ATNs are text by an initial state annotator. Simple methods of their generative power, efficiency of representation, doing this include (a) labeling all words with their and their ability to capture regularities in language. most frequent tag as found in the training corpus and (b) assuming all unknown words are nouns. The out- put of the initial state annotator is then compared with Probabilistic Parsing the manually annotated Penn Treebank, and a list of A deterministic parser consists of a set of rules, each transformations is learned that can be applied to the with equal a priori likelihood. It either generates just output of the initial state annotator, according to the one parse according to the ordering of the rules, or all nature of words in the vicinity of each tagged word. possible parses of a sentence, but does not mention The transformations have two components, a rewrite whether any one of these parses is any likelier than the rule (such as ‘change the tag from modal to singular others. A probabilistic parser, on the other hand, may noun’) and a triggering environment (such as ‘if the offer alternative rules for the decomposition of the preceding word is a determiner’). Using this transfor- constituents of a sentence, each with an associated mation, if the word ‘can’ has been given an incorrect probability value, e.g. VP → V NP 0.6, VP → V PP initial tag as in ‘the_DT can_MD rusted_VBD’, the 0.4. The overall probability of a parse is found by mul- modal (MD) tag is replaced by a singular noun (NN) tiplying together the probabilities of all the rules that tag, because the previous word has a determiner (DT) were called upon in the derivation of that parse. The tag. Other transformations are ‘change an infinitive likeliest parse among many possible parses is found by verb to a singular noun if one of the previous two tags the Viterbi algorithm. The strategy of exploring the is a determiner’, and ‘change a singular noun to a verb paths with high probability constituents first is called infinitive if the previous word is ‘to’. Best First Parsing. The CLAWS probabilistic tagger (Garside et al. The Penn Treebank contains over two million 1987) takes an input sentence such as ‘Henry likes words of American English parsed manually. It has stews’. and compares it word by word against the proved useful as a reference corpus for the comparison CLAWS lexicon, which shows that ‘Henry’ must be of different parsing algorithms, such as in the PARSE- tagged as a proper noun (NP); ‘likes’ and ‘stews’ could VAL evaluations (Black et al. 1991). In 1994, either be tagged as plural nouns or verbs (NNS or Magerman was able to derive rules automatically for a VBZ), and the full stop always has its own tag (.). For probabilistic parser through an analysis of the Penn each of the four possible tag sequences spanning this Treebank. The simplest approach is to count the num- region of ambiguity, a value is generated by calculat- ber of times each rule is used in the treebank and nor- ing the product of the frequencies per thousand for malize this to estimate the probability of each rule. successive tag adjacencies as found in the Brown cor- For certain grammars, the most probable parse can- pus, as shown below. The number 17 shows that in the not be found easily using the Viterbi algorithm. In such Brown corpus, the tag NP is followed by the tag NNS cases, a most probable parse may be found using 17 times per 1,000, while 983 times out of 1,000 it is Monte Carlo simulation techniques. Many parse trees followed by something else. 815 PARSING Value(NP-NNS-NNS-.)  17  5  135  11,475 useful than suffix removal, as the removal of a prefix Value(NP-NNS-VBZ-.)  17  1  37  629 may radically alter the meaning of a word: consider the Value(NP-VBZ-NNS-.)  7  28  135  26,460 group ‘bisect’, ‘dissect’, ‘insect’, ‘sect’. Value(NP-VBZ-VBZ-.)  7  0  37  0 Thus, the tag sequence ‘Henry_NP likes_VBZ Semantic Parsing stews_NNS ._.’ is found to be the likeliest. Since there A semantic grammar, like a syntactic grammar, con- are generally too many possible sequences to find a sists of production rules, but instead of specifying only value for them all, the Viterbi algorithm (Forney 1973) the syntactic constituents that may be present in a sen- is used, which means that only the most promising tence, may also specify domain-specific semantic cate- sequences are tested. The CLAWS tagger relies on gories. Allen (1995) gives an example in the domain of bigram (pairs of adjacent tags) frequencies, while in airline bookings, of a semantic grammar that should 1988, Church described a tagger based on trigram cater for such input phrases as ‘flight 457 to Chicago’. (sequences of three adjacent tags) frequencies. This would include such rules as ‘FLIGHT-NP
FLIGHT-N NUMB’ (A noun phrase referring to a flight consists of a noun referring to a flight and a num- Affix Stripping ber) and ‘FLIGHT-NP
FLIGHT-N to CITY-NP’ (A Affix stripping and morphological parsing are parsing noun phrase referring to a flight consists of a noun at the subword level. Instead of analyzing a sentence meaning ‘flight’, followed by ‘to’, then a noun phrase in terms of its constituent words, a word is analyzed in which is the name of a city). Different algorithms for terms of its constituent morphemes. Klavans and semantic parsing vary according to whether it is neces- Chodorow (1991) used an instructional morphological sary to produce a syntactic parse first, or whether one parser to teach morphological theory. Several sets of should proceed directly with the semantic parse. rules exist for the automatic removal and replacement of common suffixes (stemming), including those pro- References duced by Paice (1990), Lovins, and Porter. Recognizing suffixes automatically can help in POS Allen, James. 1995. Natural language understanding. Redwood identification, where, for example, any word ending in City, CA: Benjamin Cummings. Black, Ezra et al. 1991. A procedure for quantitatively compar- ‘ation’ must be a noun. The simplest form of stemming ing the syntactic coverage of English grammars. is the reduction of all regular nouns and verbs to their Proceedings of the fourth DARPA Workshop on Speech and inflectional root, which is the singular form for nouns Natural Language. San Mateo, CA: Morgan Kaufman. and the infinitive form for verbs. Brill, Eric. 1995. Transformation-based error driven learning The rule sets for the removal and replacement of and natural language processing: a case study in part-of- speech tagging. Computational Linguistics 21(4). 543–65. common suffixes are in the form of production rules, Earley, Jay. 1970. An efficient context-free parsing algorithm. in the form ‘old suffix → new suffix’, meaning that if Communications of the ACM 13(2). 94–102. a word ends in the old suffix, this must be removed Forney, G D Jr. 1973. The Viterbi Algorithm. Proceedings of the and replaced by the new suffix. The new suffix may be IEEE 61(3). 268. null, meaning that the old suffix is removed and is not Garside, Roger, Geoffrey Leech, and Geoffrey Sampson (eds.) 1987. The computational analysis of English, a corpus- replaced with anything. The rules must be presented in based approach. London: Longman. a fixed order. Four of the rules given by Paice are as Grosz, Barbara J, Karen Spärck Jones, and Bonnie Lynn follows: Webber (eds.) 1986. Readings in natural language process- ing. Los Altos, CA: Morgan Kaufmann. (1) ING → NULL Klavans, Judith, and Martin S Chodorow. 1991. Using a mor- (2) E → NULL phological analyser to teach theoretical morphology. (3) ION → NULL Computers and the Humanities 5. 281–87. Paice, Chris D. 1990. Method for evaluating stemming algo- (4) OLV → OLUT rithms based on error counting. Journal of the American Imagine that the word ‘resolving’ is given to each Society for Information Science (JASIS) 47(8). 632–49. Rich, Elaine, and Kevin Knight. 1991. Artificial intelligence. rule in turn. Rule (1) removes the suffix ‘ing’ and New York: McGraw-Hill. replaces it with nothing; hence, we are left with ‘resolv’. Winograd, Terry. 1983. Language as a cognitive process: syn- This does not end in ‘e’ or ‘ion’, hence, rules (2) and (3) tax. Reading, MA: Addison-Wesley. have no effect. However, rule (4) states that the current Woods, W A. 1970. Transition network grammars for nnatural suffix ‘olv’ must be removed and replaced by ‘olut’, language analysis. Communications of the ACM 13(10): 591–606. leaving the string ‘resolut’. The words ‘resolve’ and MICHAEL P. OAKES ‘resolution’ will also be reduced to ‘resolut’, rendering all three words equivalent. Prefix removal may be less See also Phrase Structure 816  PAUL, HERMANN Paul, Hermann Hermann Paul was one of the most influential linguists Other areas of linguistics Paul contributed to include of the late nineteenth and early twentieth centuries. lexicography, syntax, word formation, and language Paul was a neogrammarian (Germ. Junggrammatiker), pedagogy. He compiled a German dictionary, the a school of linguistics that flourished in his time and Deutsches Wörterbuch (German dictionary), which had an immense influence on linguistics thereafter, as remains in print today. Paul was also interested in the discussed further below. His interest in languages and theoretical aspects of lexicography, publishing various linguistics dates back to his schooldays, when he essays on lexicographic topics (although the titles of developed a taste for medieval German literature and these essays indicate that his main concern was with his Middle High German (superseding an earlier interest German dictionary). As for syntax, Paul recognized its in mathematics). importance early on; his Middle High German grammar In a brief autobiographical essay published just was one of the first such handbooks to include a section after his death, Paul suggested that his work could be on syntax (although it must be admitted that this section divided into two general categories: literary interpreta- was not in the first edition of the book; it was added to tions and textual criticism, with a focus on Middle the 2nd edition, published in 1884). His work in word High German, and the phonology and morphology of formation was generally an offshoot of his lexicograph- the Germanic languages. While Paul was certainly an ical work; he published an important article on word important contributor to these fields, this self-classifi- formation, ‘Zur Wortbildungslehre’ (‘On the doctrine of cation is something of an oversimplification, as he word formation’), in 1896. Finally, a volume on lan- made significant contributions to a number of other guage pedagogy, Über Sprachunterricht (On language fields of linguistics and German studies. teaching), was published posthumously in 1922. Paul was certainly an influential figure in literary Probably Paul’s most significant contribution to lin- studies and textual criticism. His dissertation and guistics was his work Prinzipien der Sprachgeschichte Habilitionsschrift both dealt with questions of Middle (Principles of the history of language), originally pub- High German literature. He authored a variety of arti- lished in 1880. This volume rapidly became a standard cles on literary topics, ranging from the life of handbook (five editions of the text were published dur- Hartman von Aue, an important Middle High German ing Paul’s lifetime, along with an English translation poet, to the question of whether a Middle High of the second German edition), and proved to have a German literary language ever really existed. As to lasting influence on the field. It is notable not only for textual criticism, he was the founding editor of the its incisive statements of many Neogrammarian ideas Altdeutsche Textbibliothek [Early German text but also for its innovativeness, and is rewarding read- library], and edited three of the volumes of the series, ing even today. all Middle High German texts. His Mittelhochdeutsche Finally, Paul’s activities as an editor cannot be Grammatik [Middle High German grammar], first underestimated. Together with his friend and former published in 1881 and republished many times (11 fellow student Wilhelm Braune, Paul founded a jour- editions appeared in Paul’s lifetime), remains a stan- nal, the Beiträge zur Geschichte der deutschen dard handbook in the field. Sprache und Literatur [Contributions on the history of Paul’s contributions to the study of Germanic the German language and [its] literature], which is phonology and morphology were also immense. He still generally referred to as the ‘PBB’, for Paul- published a monograph-length article entitled ‘Zur Braune Beiträge. Paul served as coeditor of this jour- geschichte des germanischen vocalismus’ (‘On the nal for nearly 20 years, from 1874–1891, and was a history of germanic vocalism’), as well as articles on consulting editor for 30 years after that. Paul was also the vowels of the inflectional and derivational sylla- the general editor of the Grundriss der germanischen bles in the oldest Germanic dialects and the ablative Philologie (Outline of Germanic philology). case in Germanic, among others. His most important Furthermore, as noted above, he was the founding edi- work in this area is the five-volume Deutsche tor of the Altdeutsche Textbibliothek (Early German Grammatik (German grammar), described in his obit- text library), and edited three of the volumes of the uary as ‘die einzige vollendete wissenschaftliche series (listed above). Grammatik der deutschen Sprache’ (‘the only com- Despite his many contributions to the field, Paul is pleted scientific grammar of the German language’). too often dismissed as ‘merely’ a historical linguist. 817 PAUL, HERMANN This viewpoint is entirely natural, since most of his abroad, ranging from Leonard Bloomfield, the dean of work was on historical linguistics, and because of American linguistics in the 1930s and 1940s, to William statements like the following, found in his Prinzipien: Labov, a prominent sociolinguist of today, were deeply ‘Es ist eingewendet, dass es noch eine wis- influenced by Neogrammarian thought. Given the senschaftliche Betrachtung der Sprache gäbe, als die importance of the Neogrammarians, and Paul’s impor- geschichtliche. Ich muss das in Abrede stellen’. [‘It has tance within the group, his place in the history of lin- been claimed that there is another way to view lan- guistics seems assured. guage than the historical. I must object to this.’] On the other hand, it has also been claimed that Paul was much Biography more than ‘merely’ a historical linguist. It has been argued that Paul was a forerunner of Structuralism; Paul Hermann was born in Salbkempierc bei more specifically, that certain concepts found in the Magdeburg, Germany on, August 7, 1846. He attended work of Ferdinand de Saussure, the well-known Swiss the University of Berlin 1866–1867. He received a linguist who is generally viewed as the intellectual Doctorate (1870) for his dissertation on Freidank, father of Structuralism (especially the distinctions Habilition (1872) for work on Gottfried’s Tristan, between synchrony and diachrony and between langue University of Leipzig. He was Privatdozent in Leipzig in and parole), were inspired by Paul’s work (e.g. by his 1872–1874. He was also Professor of German, differentiation between individual linguistic acts and University of Freiburg im Breisgau, 1874–1893; more general language use). The first of these view- Professor of German, University of Munich, 1893–1916; points must be rejected; a closer reading of Paul’s and Rektor, University of Munich, 1909–1910. He was a Prinzipien readily indicates that he was also interested corresponding member of the Wissenschaftliche in synchronic topics. The second of these viewpoints is Akademie in Vienna in 1919–1921. Hermann died in probably also an overstatement. It is clear that de Munich, Germany, on December 29, 1921. Saussure codified a number of ideas that were current linguistic theory at his time, but to trace the critical dis- References tinction between synchrony and diachrony to Paul, as Antal, Lázló. 1985. Some comments on the relationship some have, for example, is probably an exaggeration. between Paul and Saussure. Cahiers Ferdinand de Saussure Regardless of these ideas, it is clear that Paul was an 39. immensely important figure in linguistics. Paul played Henne, Helmut, and Jörg Kilian (eds.) 1998. Hermann Paul: an extremely important role in the development of the Sprachtheorie, Sprachgeschichte, Philologie. Tübingen: Neogrammarian school of linguistics; he has been Niemeyer. Jankowsky, Kurt R. 1972. The neogrammarians. A re-evalua- described by one historian of linguistics of the period as tion of their place in the development of linguistic science. the Neogrammarians’ ‘most fertile and formidable theo- The Hague: Mouton. retician’, and his Prinzipien codified many aspects of Koerner, E.F.K. 1972. Hermann Paul and synchronic linguis- Neogrammarian thought. More specifically, this same tics. Lingua 29. historian of linguistics (Kurt Jankowsky; the relevant Labov, William. 1994. Principles of linguistic change. Internal factors. Cambridge, MA: Blackwell. work is cited under ‘Further Reading’ below) has argued Paul, Herman. 1879. Zur geschichte des germanischen that Paul’s influence on his contemporaries can be traced vocalismus. Beiträge zur Geschichte der deutschen Sprache by examining five major aspects of his work. First, Paul und Literatur 6. rejected the equation of linguistics to the natural sci- ———. 1880. Prinzipien der Sprachgeschichte; transl. into ences, suggesting that the term ‘sound law’ should not be English by H.A. Strong as Principles of the history of lan- guage, 1889; reprint of 8th edition 1970. understood as a ‘law’ is understood in physics and chem- ———. 1881. Mittelhochdeutsche Grammatik. istry. Second, Paul distinguished between historical and ———. (ed.) 1894. Grundriss der germanischen Philologie. comparative linguistics. Third, Paul was interested not ———. 1896. Zur Wortbildungslehre. Sitzungsberichte der only in how languages change, but why they do so. philos.-philol. Klasse der bayerischen Akademie der Fourth, Paul emphasized individual psychology over Wissenschaften. ———. 1897. Deutsches Wörterbuch. group psychology in his psychological approach to lan- ———. 1916–1920. Deutsche Grammatik. guage. Fifth, Paul was deeply interested in linguistic ———. 1922. Über Sprachunterricht. geography, recognizing early on that a standard lan- ———. 1992. Mein Leben. Beiträge zur Geschichte der guage is a convenient abstract fiction, made up of a vari- deutschen Sprache und Literatur 46. ety of dialects and idiolects. A discussion of the place of Reis, Marga. 1978. Hermann Paul. Beiträge zur Geschichte der deutschen Sprache und Literatur 100. the Neogrammarians in the history of linguistics is clear- MARC PIERCE ly beyond the scope of this entry, but their influence on linguistics, especially historical linguistics, was consid- See also Historical Linguistics; Leskien, August; erable. Numerous prominent linguists, in America and Saussure, Ferdinand de 818  PEIRCE, CHARLES SANDERS Peirce, Charles Sanders Charles Sanders Peirce was a most original and high- of what anyone believed about it. As for his episte- ly versatile American philosopher, logician, and sci- mology, Peirce rejected the Cartesian attempt to use entist—by all standards, a veritable polymath. Son of self-knowledge as the doorway to knowledge about Benjamin Peirce, a distinguished Harvard mathe- the world, arguing instead that all knowledge is falli- matician, Charles was born in 1839 and grew up in a ble, although continuous inquiry makes it self-correc- family milieu that was most conducive to a promis- tive. He also contended that the ultimate reality is ing career in academia. Yet, except for a brief period mental, matter being nothing but ‘effete mind’, and as a lecturer in logic at the Johns Hopkins University that science provides us with the best method of arriv- (1879–1884), Peirce never did any systematic teach- ing at it. ing. He was a difficult person to get along with and Along with the Swiss linguist Ferdinand de the university refrained from renewing his contract Saussure, Peirce is also widely looked upon as a on charges of personal misconduct. His most sus- founding father of semiotics (or ‘semeiotic’, as he tained employment was with the U.S. Coast Survey, himself preferred to spell it). Unlike Saussure who where he worked for 31 years and made a successful claimed the sign relation to be dyadic—involving a career as a geodesist, until the termination of his sign (signifier) and a referent (signified) Peirce insist- appointment in 1891. Thereafter, he set up a private ed that it was irreducibly triadic and involved, in addi- practice as a chemical engineer and, thanks to the tion to the object and the sign, an interpretant (the intervention of his friend and admirer William James, interpreting thought or meaning). The sign, in was invited to give occasional series of lectures at Peirce’s conception, mediates between its object and Harvard. His last years were spent in penury and fail- its interpretant, which are therefore its correlatives. ing health. Central to Peirce’s thinking is the idea that language Peirce is widely known as the founding father of does not simply and unproblematically refer to a real- Pragmatism, a distinctively American philosophical ity external to it; it does so by representing it. This movement whose central tenet is that belief guides means, he claimed, all of cognition is necessarily action. Our ordinary beliefs as well as more sophisti- mediated by the process of sign interpretation or cated doctrines are nothing but dispositions to act in semiosis. Peirce also viewed semiotics as fundamen- certain ways and are to be judged by taking into tally a classificatory science, like chemistry and biol- account their practical consequences. Peirce therefore ogy. Furthermore, signs are not restricted to the rejected much of metaphysics—i.e. claims about the activity of the human mind but are of the order of ultimate reality of what there is—since no practical nature itself and, in this too, he differed from consequences necessarily follow from alternative Saussure, for whom ‘semiology’ was part of social metaphysical conceptions. He was averse to the sort of psychology. Unlike Saussure too, Peirce saw semi- intuitive rationalism as advocated by Descartes and otics as teleological, tending toward the truth of foundationalism as encapsulated in the Cartesian quest things as they really are. for first principles, as well as his celebrated ‘Method Peirce was a meticulous system-builder and his of Doubt’. As he famously put it in an essay published mind worked typically by positing trichotomies. in 1878 titled ‘How to make our ideas clear,’ truth is Building on his own triadic characterization of sign simply that opinion which is destined to be agreed to relation, Peirce initially distinguished three divisions by everyone who is disposed to finding it and so-called of signs: based on the nature of the sign itself, the reality is made up of the objects represented in that relation between the sign and its object, and the role opinion. of the interpretant in representing the object. In their The idea of truth as resulting from a convergence of turn, these three divisions harked back to three onto- opinion was interpreted by many of his followers to logical categories: quality (firstness), relation (sec- mean that there is no more to it than a matter of mere ondness), and representation (thirdness). The first consensus. William James for one regarded truth as division, the one based on the nature of the sign itself, ‘whatever proves to be good in the way of belief’. But yielded a three-way distinction among ‘qualisign’ it is clear from Peirce’s later writings that he sub- (appearance), ‘sinsign’ (token), and ‘legisign’ (type). scribed to an objectivist conception of reality—the Under the second division, the one obtained by taking idea that there is a reality out there that is independent into account the relation of the sign to its object, 819 PEIRCE, CHARLES SANDERS Peirce distinguished icons (that bear a certain resem- Arts and Sciences on January 30, 1867. He was blance to the object), indices (that signify by direct Assistant at Harvard Observatory, October 1869– causal interaction with the object in question), and December 1872 and was elected to the National symbols (that are associated with their object via con- Academy of Sciences on April 20, 1877. He was vention). Finally, under the third division, the one appointed lecturer in Logic at Johns Hopkins in where the interpretant enters the picture, yet another 1879 and was dismissed on charges of misconduct triad was postulated: rhemes (predicative signs), in 1884. He was elected to the American Association prepositional signs, and arguments. The arguments for the Advancement of Science in August 1880. He were further divided into the three categories—abduc- purchased a farm near Milford, Pennsylvania on tion (hypothesis/diagnosis), deduction (tautological May 10, 1888; over the next two years, he recon- reasoning), and induction (generalization)—the first structed the existing farmhouse and bought more two leading to qualitative or conceptual understand- land, and named the estate ‘Arisbe’. The manuscript ing and the last providing important quantitative of ‘How to reason’ was rejected by both Macmillan details to it. and Ginn Co (1894) and that of ‘New elements of Toward the end of his life, Peirce grew increas- mathematics’ was rejected by Open Court (1895). In ingly impatient with the way the word ‘pragmatism’ 1902, a grant application for ‘Proposed memoirs on was being used (and, in his view, abused) by other minute logic’ was rejected by Carnegie Institution. writers including William James and John Dewey He gave lectures on ‘Pragmatism’ at Harvard March and, in 1905, in a paper called ‘What pragmatism is’ 26–May 17, 1903 and Lowell lectures on ‘Some top- published The Monist, pleaded that the word ‘prag- ics of logic’ November 23–December, 17, 1903. He maticism’ be used instead to refer to his own doc- was invited to deliver the Harvard Philosophy Club trine, and distinguished it from that of the others, lectures on ‘Logical methodeutic’ April 8–13, 1907. adding that the neologism was ‘ugly enough to be Peirce died at Arisbe on April 19, 1914, of cancer. safe from kidnappers’. There followed two other papers (forming a trilogy): ‘Issues of pragmaticism’ (1905) and ‘Prologomena to an apology for prag- References maticism’ (1906). Apel, Karl-Otto. 1981. Charles S. Peirce: from pragmatism to Peirce died in 1914, an almost forgotten man. pragmaticism. English translation by John Michael Krois. Early commentators often described him as ‘a wasp New Jersey: Humanities Press. Auspitz, Josiah Lee. 1994. The wasp leaves the bottle: Charles in the bottle’, someone who constantly came up Sanders Peirce. The American Scholar 63. against formidable odds and did everything but avail Brent, Joseph. 1993. Charles Sanders Peirce: a life. himself of the only way out of his troubles. Starting Bloomington, Indiana: Indiana University Press. with the last quarter of the twentieth century or so, Brock, J. E. 1974. Peirce’s conception of semiotic. Semiotica 14. however, there has been a resurgence of interest in Freeman, Eugene (ed.) 1983. The relevance of Charles Peirce. La Salle, IL: The Hegeler Institute. Peirce’s work and a widespread recognition of the Hookway, Christopher. 1985. Peirce. London: Routledge & importance of his views to such areas as epistemolo- Kegan Paul. gy, linguistics, anthropology, psychology, and cogni- Ketner, Keneth Laine (editor) 1995. Peirce and contemporary tive science. thought. New York: Fordham Univerity Press. Moore, Edward C. 1993. Charles S. Peirce and the philosophy of science. Tuscaloosa, London: The University of Alabama Biography Press. Charles Sanders Peirce was born on September 10, Murphey, Murray G. 1993. The development of Peirce’s phi- losophy. Indianapolis, Indiana: Hackett Publishing 1839 in Cambridge, Massachusetts (USA). In 1855, Company Inc. he entered Harvard College, graduating in 1859 and Peirce, Charles Sanders. 1931–1935. Collected papers of continuing as a ‘resident graduate’ for one year. The Charles Sanders Peirce, Vols. 1–6, ed. by Charles same year, he began his career as a field aid at the Hartshorne and Paul Weiss. Cambridge, MA: The Belknap Coast Survey, specializing in geodesy and working Press of Harvard University Press; Vols. 7–8, ed. by Arthur Burkes. Cambridge, MA: The Belknap Press of Harvard his way up until being appointed an Assistant. He University Press, 1958. entered Lawrence Scientific School at Harvard in ———. 1982. Writings of Charles S. Peirce: a chronological 1861 and two years later obtained a graduate degree edition, 5 vols. Bloomington, IN: Indiana University in Chemistry. He delivered Harvard lectures on ‘The Press. Logic of Science,’ in the spring of 1865 and Lowell Rosenthal, Sandra. 1994. Charles Peirce’s pragmatic pluralism. New York: State University of New York Press. Institute lectures on ‘The Logic of Science; or KANAVILLIL RAJAGOPALAN Induction and Hypothesis,’ October 24,–December 1, 1866. He was elected to the American Academy of See also Philosophy of Language; Sign Relationship 820  PERSONALITY AND LANGUAGE Personality and Language Personality and language are two terms that suggest between language and personality, and significant some promising interconnections, such as the likely advances in this area may be seen in the near future. effect of personality pathologies on speech, the effect So far, the most intensive attempts to explore the com- of personality traits on the use of certain linguistic— mon ground between the two areas of research have especially discursive—patterns, or the influence of been conducted by social psychologists and—to a individual learner differences on the acquisition of a lesser extent—second-language researchers. Research second language. The effect of language on the con- has been primarily focused on two issues. figuration of personality is apparent if we think of the The first issue is the effects of linguistic differences lexical terms used in a given culture to refer to per- on the attribution of personality features by listeners, sonality phenomena. The ways in which people whose evaluation is based solely on speech. These describe their own or others’ personalities and the fun- studies started with the seminal work of Wallace damental influence of a person’s speech style on oth- Lambert and his associates. They found that listeners ers’ perception of his or her personality are examples in Montreal rated the personality of a given speaker of the interrelationship of the two. Additionally, per- differently when he spoke in two distinct languages sonality may be seen as a key factor in the construction (i.e. French and English) and also that French-speak- of language and of the speech style of each individual. ing listeners tended to assign personality traits to the Notwithstanding these areas of inquiry, we are still speaker quite differently from English-speaking listen- far from claiming any conclusive direct effect of one ers. Other speech evaluation studies of evaluations over the other. Not surprisingly, it is speech (i.e. the assigned to Spanish and English speakers in the oral performance of language by a given speaker) that United States followed. Howard Giles and his associ- has been the primary focus of researchers interested ates explored the effect of different regional and social in the links between language and personality. These speech varieties on personality ratings. Their research researchers have generally avoided claiming strong uncovered stereotypes linked to speech differences, connections between personality and language as a and it was very valuable precisely in demonstrating system and have instead claimed a relationship that laypeople’s connections between language and between speech style and personality features, either personality are the result of previously existing stereo- self-assigned by the speaker or attributed by listeners. types. In all cases, two factors were identified: a com- Psychological research on personality has not been petence factor, which had to do with efficiency and extremely consistent in providing a clear model high status in society; and a solidarity factor, which against which to test any possible mutual influences of involved a high affective empathy with the speaker. A language and personality. Existing models range from common result of most studies conducted so far has viewing the structure of personality as composed of been that language varieties that enjoy high prestige one or two dimensions, to the three or five factors that and can be regarded as the varieties of the most pow- are currently the most widely accepted models. In the erful people tend to be not so well valued in terms of 1990s, the so-called Five Factor Model gained popu- personal empathy, and vice versa. More recently, some larity within the research community. This model was attempts have also been made to connect personality first proposed in the 1960s, but it started receiving ratings and nonnative proficiency, with provisional wide attention and empirical support three decades outcomes that also point at the important role that lan- later. However, it still has to be confirmed in studies guage variation may play in listeners’ perceptions of a involving non-Western societies, where a different cat- speaker’s personality, with all the major subsequent egorization of personality traits may be more accurate. consequences for interpersonal relations. This method In Western societies, the model seems to provide a of study is less subjective than self-reports or self- good account of personality structure with the follow- accounts of personality traits, which tend to be heavi- ing five factors: extraversion, agreeableness, conscien- ly mediated by the idea of what one would like to be tiousness, emotional stability, and openness (also like rather than what one really is or seems to be. referred to as intellect). Another positive aspect of attributions of personality The existence of a predominant model may eventu- by listeners is the fact that they are closely related to ally clarify research questions on the interplay social stereotyping. Thus, the study of the influence of 821 PERSONALITY AND LANGUAGE speech style on personality evaluation leads to the References uncovering of social stereotypes. Ellis, Rod. 1994. The study of second language acquisition. The second area of research that links language Oxford: Oxford University Press. and personality is second language acquisition, par- Furnham, Adrian. 1990. Language and personality. Handbook ticularly the role of different personality traits in the of language and social psychology, ed. by Howard Giles and ultimate achievement of second language proficiency. W. Peter Robinson. Chichester: John Wiley & sons. Giles, Howard, and Peter F. Powesland. 1975. Speech style and Surveys conducted among second language teachers social evaluation. London: Academic Press. and students show that personality is considered to be Goldberg, Lewis R. 1981. Language and individual differences: a primary factor in second language achievement. the search for universals in personality lexicons. Review of Empirical studies have mainly focused on a single personality and social psychology, Vol. 2, ed. by L. W. personality factor: extraversion. Extroverted people Wheeler. Beverly Hills, CA: Sage. Lambert, Wallace E., R.C. Hodgson, R.C. Gardner, and S. seem to be more successful in acquiring basic com- Fillenbaum. 1960. Evaluational reactions to spoken languages. municative abilities because of their willingness to Journal of Abnormal and Social Psychology 60. 44–51. interact with other people and search for communica- Llurda, Enric. 2000. Effects of intelligibility and speaking rate tion opportunities. Introverts are less engaged in com- on judgements of non-native speakers’ personalities. IRAL municative interactions but may benefit from an — International Review of Applied Linguistics 38. 289–99. Ryan, Ellen B., and Miguel A. Carranza. 1975. Evaluative reac- increased attention to the formal aspects of language. tions of adolescents toward speakers of standard English and On the whole, research has not found one type of Mexican American accented English. Journal of Personality learner to be inherently better than another, but per- and Social Psychology 31. 855–63. sonality profiles seem to affect the preferred path to Saucier, Gerard, and Lewis R. Goldberg. 1996.The language of learning and, therefore, the final outcome of the personality: lexical perspectives on the five-factor model. The five-factor model of personality; Theoretical perspec- process. tives, ed. by Jerry S. Wiggins. New York: Guilford. Personality and language should be viewed as two Skehan, Peter. 1989. Individual differences in second-language uniquely human characteristics that are bound to inter- learning. London: Edward Arnold. act; however, at this point, there is a need for much Taylor, Harold C. 1934. Social agreement on personality traits more research to investigate the relationship between as judged from speech. Journal of Social Psychology 5. the two. ENRIC LLURDA See also Discourse Analysis Philippine Spanish Creoles The Spanish-based Creoles in the Philippines devel- of Luzon and are spoken by relatively small commu- oped after the Spanish conquest in the sixteenth centu- nities in the towns of Cavite and Ternate, but the ry and are commonly known as Chabacano. Despite its dialect of Ermita, a district of Manila, can now be despicable original meaning (‘tasteless’, ‘vulgar’), this regarded as extinct. Southern Mindanao Creole is spo- name is used by the speech-community as a self-des- ken on the island Mindanao in the south of the archi- ignation. Alongside Papiamento and Palenquero in the pelago. The highest degree of vitality of Southern Americas, Chabacano is one of the three Spanish- Mindanao Creole, and of Chabacano generally, can be based Creoles worldwide. It shares interesting features observed in the city of Zamboanga and the surround- with other European-based Creoles in Asia, a fact dif- ing area in the extreme western part of Mindanao ferentiating it from the Atlantic creoles. Further, unlike (approximately 300,000 speakers). Creole is also used the latter, Chabacano has been acquiring a mixed char- here as a lingua franca by diverse ethnolinguistic acter; especially from the twentieth century onward. groups. It is codified for public and private purposes The total number of Chabacano speakers can be with a hispanic orthography; there are bible transla- estimated at no more than 500,000. Geolinguistically, tions, literary writings, and until recently press. Public Chabacano can be divided into two main varieties: oral use has recently been increasing, there is broad- Manila Bay Creole and Southern Mindanao Creole. casting in the language, and it is used in church and Dialects of the former are found on the northern island partially in school. A subvariety of Zamboangueño is 822  PHILIPPINE SPANISH CREOLES also spoken in the town of Cotabato; however, little is of the source languages, but in many cases they have known about the vitality of the subvariety of Southern different functions. The word-formation system seems Mindanao Creole found in Davao. to be unusually complex for a Creole. The plural of The genesis and history of Chabacano, especially nouns is marked with the particle maga (el maga Zamboangueño, is somewhat complex due to different péhro ‘the dogs’). In Zamboangueño, first-person plu- contact situations in time and space. Keith Whinnom ral pronouns have a Philippine form encoding inclu- (1956) suggested that Chabacano grew out of the sion or exclusion of the addressee. The basic word Portuguese creole of the Indonesian island of Ternate, order of V(erb)-S(ubject)-O(bject) stems from which had been transported to the Philippines in the Austronesian contact, e.g. man-J ángge yo ‘I’m going seventeenth century. He considered Zamboangueño a shopping’. This is exceptional compared to other layer of Manila Bay Creole. This theory was contest- European-based creoles that are typically SVO. As in ed principally by John Lipski (1992), who affirmed a Philippine languages, human proper names in subject typical creolization of a Spanish-based pidgin with function are marked with si (ta-kantá si Maria ‘Maria Tagalog influences for Manila Bay Creole. For the is singing’). Human and definite objects are obligato- special case of Zamboangueño, he proposes a devel- rily marked with kon, perhaps a merger of an opment of different stages from the eighteenth century Austronesian element and Spanish con ‘with’, onward: it arose in the garrison, absorbing elements (konosé-le kon ese muhér ‘he knows that women’). from Manila Bay Creole later on. During the nine- There are many discourse particles of Philippine ori- teenth century, Hiligaynon elements were introduced gin (e.g. daw for reported information). However, through contact from ships coming from Manila hav- Chabacano also has features typical of Atlantic cre- ing made a stop in Central Philippines. Afterward, it oles, e.g. serial verb constructions (pwéde asé salé ‘to was re-hispanicized by native speakers from Europe be able to make to go out’), no syntactic passive con- and Mexico who came to Zamboanga. It was also struction, lack of an equative copula (soltéro el hénte influenced by Cebuano, like Hiligaynon, a Visayan ‘the man is a bachelor’), identity of ‘have’ and ‘there language, which was brought by large-scale immigra- is’ (both tyéne), and preverbal markers for tense, tion. From the 1930s onward, the influence of English mood, and aspect (e.g. ya-andá yo na Samboanga ‘I and, after World War II, that of the national language went to Zamboanga’, ta-kusiná le ‘(s)he is cooking’ Pilipino have been increasing, especially affecting the and ay-kantá silá ‘they will sing’). The lexicon princi- lexicon. Consequently, Zamboangueño does not fit pally consists of Spanish vocabulary (83%), with some into the classification of plantation, fort, and maritime Philippine (c. 15%) and English (2.5%) words. It can Creoles; nor can the dichotomy of exogene and endo- be assumed that, more recently, the influence of gene Creoles adequately describe the situation of English has increased and that of Spanish has dimin- Zamboangueño. In contrast to the Atlantic creoles, ished. Very few words are of Portuguese origin (e.g. slavery played a marginal role. Whatever the different gumitá  port. gomitar ‘to vomit’). The importance of theories on the origin of Chabacano, first documented Mexican input is shown by words like J angge ‘market’ in Hugo Schuchardt’s Kreolische Studien (1883;  tiangue (a Mexican word of Nahuatl origin). Creole studies), in the beginning there was a Despite the close geographical contact with pidginized form of Spanish used for intercommunica- Zamboanga, the local languages, Tausug, Subanon, tion by different ethnic groups, namely, European and Yakan have nearly no influence on the lexicon. Spaniards, Mexicans, West-Austronesian speaking groups, and Chinese immigrants. A pidgin created by the latter is still spoken by shopkeepers in Davao References (‘Bamboo Spanish’). Fernández, Mauro (ed.) 2001. Shedding light on the Chabacano The current structure of the language, especially language [Special Issue of Estudios de Sociolingüística. concerning the vocabulary and word structure, is heav- Linguas, sociedades e culturas 2, 27–56]. Vigo (Spain): ily influenced by Philippine languages. The sound sys- University Press. tem shows Philippine features like the merging of Forman, Michael L. 1972. Zamboangueño texts with grammat- ical analysis: a study of Philippine Creole Spanish. Ph.D. [o]/[u] and [e]/[i] in nonaccented syllables and the Dissertation. Cornell University. occurrence of the glottal stop. Some of the productive Frake, Charles O. 1971. Lexical origins and semantic structure morphemes for word formation in Zamboangueño in Philippine Creole Spanish. Pidginization and creolization originally come from Spanish, and many are from of languages, ed. by Dell Hymes, 223–42. Cambridge: Hiligaynon or Cebuano (e.g. the verbalizing prefix Cambridge University Press. Lipski, John M. 1992. New thoughts on the origins of man-: kwénto ‘story’ → man-kwénto ‘to tell’ or the Zamboangueño (PCS). Language Sciences 14. 197–231. adjectivizing prefix ma-: pwérte ‘strength’ → ma- Papia. Revista de Crioulos de Base Ibérica 12 (2002) [Special pwérte ‘strong’). Their form is often identical to those Issue dedicated to Chabacano]. 823 PHILIPPINE SPANISH CREOLES Riego de Dios, María. 1989. A composite dictionary of Phonology and morphology of Creole languages, ed. by Philippine Creole Spanish (PCS). Studies in Philippine Ingo Plag, 253–63. Tübingen: Niemeyer. Linguistics 7. 1–210. Whinnom, Keith. 1956. Spanish contact vernaculars in the Schuchardt, Hugo. 1883. Kreolische Studien IV: über das Philippine Islands. Hong Kong: Hong Kong University Press. Malaiospanische der Philippinen [Creole Studies IV: on the PATRICK O. STEINKRÜGER Malayo-Spanish of the Philippines]. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften zu Wien 105. See also Austronesian, Papiamento, Pidgins and 111–50. Steinkrüger, Patrick O. 2003. Morphological processes of word Creoles, Spanish and Iberoromance Languages, formation in Philippine Spanish Creole (Zamboangueño). Tagalog and Philippine Languages Philology Etymologically, philology is the love of words (or, per- Etruscan, they are funerary commemorations, from haps, the study of love). In the broadest sense, it covers which we learn only of family relationships. everything having to do with the study of language— The task of the epigrapher is to record inscriptions grammar, texts, history, civilization. Indeed, until a cen- accurately (and photography can rarely substitute for tury or so ago, what is now called linguistics was simply the sharp eye and steady hand of the draftsman), to part of philology, and even today, the term ‘comparative publish them first promptly and then systematically, philology’ can still be found as a synonym for ‘histori- and to provide such tools as charts of letter forms that cal linguistics’, especially within Indo-European stud- will assist in the interpretation of newly discovered ies. But here we should look at those branches of inscriptions. A separate branch of epigraphy is numis- philology that have not been co-opted by (descriptive, matics, the study of coins and their inscriptions. historical, or theoretical) linguistics. Thus narrowed, philology can be considered to be Paleography the study of texts, with all that implies: first, determin- ing exactly what the author of a text actually wrote; Paleography is the study of manuscripts, written with then, determining what the author said; then, deter- pen or brush and ink on flexible material, namely mining what the author meant. Correspondingly, a typ- papyrus, skin, or paper; textiles, wood, or leaves. ical edition—whether of a book of the Bible, a play by These are perishable materials, so only rarely have Shakespeare, a novel by Joyce, or an inscription from they survived from more than a few centuries ago. The an Indian temple—will contain the text itself, notes on exceptions are due to extremely dry climates, as near every detail, and an interpretative commentary. the Nile and Jordan Rivers, areas where thousands of Traditionally, the interpretative work of philology Egyptian papyri and the Dead Sea Scrolls have been relies on the prior accomplishments of four subdisci- found. In extremely rare cases, careful curatorial atten- plines: epigraphy, paleography, diplomatics, and textu- tion has preserved precious documents for a thousand al criticism. years or more, but our knowledge of ancient texts results far more from a continual tradition of copying than from ancient attestations. (The tradition was fair- Epigraphy ly quickly superseded by printing beginning c. 1450.) Epigraphy is the study of inscriptions on hard materi- The paleographer’s job is primarily to study the al, such as stone and metal. People have been carving development of handwriting over generations of copy- important texts into walls for as long as they have been ists, using dated manuscripts to anchor changes in sty- writing—in Egypt, Mesopotamia, China, Anatolia, listic trends and exercising judgment in assigning Greece, Arabia, Rome, Iran, India, Inner Asia, and undated ones to this or that range of dates—although it Mesoamerica, in rough chronological order—and must be recalled that any individual scribe’s hand will often such inscriptions are the only surviving written not change much over their professional career, so a evidence of past civilizations. Sometimes they are offi- date based strictly on paleographical grounds cannot be cial decrees, and by reading between the lines we can more precise than within a generation or so. There are learn much about the societies that prompted them. as many fields of paleography as there are scripts with Sometimes, as is almost exclusively the case with manuscript traditions—cuneiform, Egyptian, Chinese, 824  PHILOLOGY Greek, Latin, Indian, Southeast Asian, etc. (there are no the unfamiliar inscription is accompanied by an general surveys of world paleography outside histories inscription in a familiar script and language, and it can of writing like Diringer (1968) and Jensen (1969); Urry be assumed that the texts are at least roughly equiva- (1974) considers only Greek and Latin). lent. This state of affairs is known as a BILINGUAL, and both the first decipherment, of the Aramaic language Diplomatics Palmyrene by Jean-Jacques Barthélemy in 1754, and the best known, of Egyptian hieroglyphics by Jean- Diplomatics is the study of the form of documents. Its François Champollion in 1822, involved bilinguals, principal concern is detecting forgeries, so it deals although these two accomplishments were vastly dif- principally with the physical properties of manuscripts ferent in difficulties. Pope (1999) includes detailed and is the aspect of philology farthest removed from descriptions. linguistics. When no bilingual is available, the ingenuity of the philologist is required to discover a VIRTUAL BILIN- Textual Criticism GUAL. The most challenging, and for its impact on our Textual criticism is the study of the history of the con- understanding of human history the most important, tent of texts. Whenever a scribe makes a copy of a decipherment was of CUNEIFORM. Late-eighteenth-cen- manuscript by hand, however carefully, mistakes are tury travelers brought back to Europe drawings of introduced. If a work was popular, many copies would inscriptions they found at the ruins of the capital of the be produced, and copies of copies, and so on, every Persian Empire, Persepolis, which was destroyed by one of them differing very slightly from all the others. Alexander in 330 BCE. Each inscription was in three By noting the correspondences among the differences different scripts, which were assumed to represent the in different copies, the chain of transmission can be same content in three languages (hence, they are called TRILINGUALS), but none of the scripts was familiar. The reconstructed, and even if the author’s original manu- script has not survived, it may be possible to recon- characters in all of them were composed of incised struct the author’s original text with considerable wedges (cuneus in Latin), whence the name. certainty, with successive changes being removed. The first success in interpreting cuneiform came in Different scribal traditions handle the problem of 1802. A German high school teacher, Georg Friedrich scribal error in different ways. The rabbis who codified Grotefend, surmised that because the simplest of the the text of the Hebrew Bible, known as Masoretes, did three scripts always appeared in the most prominent things like counting the number of words and even let- position among the three, it probably wrote the ancient ters in each book. Scribes in the Islamic tradition includ- Persian language of the Empire. A few years earlier, ed elaborate COLOPHONS in their manuscripts, naming the Frenchman Antoine Isaac Sylvestre de Sacy had not only the scribe and the date and circumstances of deciphered royal inscriptions of a successor Iranian copying (as is often found elsewhere) but also a full list empire, the Sassanian, finding that they usually began of all intervening scholar/copyists back to the original with a genealogy in the form ‘X, great king, son of Y, author. And Chinese scribes sometimes bypassed the great king ….’ Grotefend knew the names of the problem by making copies of major works by printing kings—Darius, Xerxes, and so on—from the Greek them off woodblocks or even stone inscriptions of the Histories of Herodotus and sought similar formulas at texts, thus coming full circle in the fields of philology. Persepolis. He soon found the expected recurring pat- An interesting offshoot of textual criticism is that it is terns in the inscriptions, and although he himself was often possible to determine that a given text has been not an Iranian philologist, his breakthrough discovery translated from another language, and that other lan- made it possible for specialists like the Danes guage identified with considerable certainty. Thus, R.H. Christian Lassen and Rasmus Rask to determine the Charles was able to say of the Book of Enoch, which basics of Old Persian. Edward Hincks, a Church of was known in full only in Ge‘ez, with fragments surviv- Ireland (Anglican) clergyman, showed that each of the ing in Greek and Latin, that portions—chapters 1–5 and 36 characters of Old Persian script represented a 37–104—were originally written in Hebrew, and chap- Consonant–Vowel (CV) syllable. ters 6–36 were originally written in Aramaic The second most complicated of the three (1912:lvii–lxx). Decades later, fragments of the Aramaic cuneiform scripts had about 100 different characters original were identified among the Dead Sea Scrolls. (none of them the same as in Old Persian), so it was not surprising when, by comparing what seemed to be the same names in corresponding Persepolitan inscrip- Decipherment tions, it too was shown to record syllables—both CV The philological task most relevant to linguistics is the and VC. Its language, which is today called Elamite, is decipherment of disused scripts. Usually, these are still not well understood. Hincks was one of the schol- brought to light by archeological excavations. Rarely, ars who did that work; but more importantly, he went 825 PHILOLOGY on to decipher the third, most complicated script, historical linguistics would be limited to deductions which used several hundred characters. By the time that can be made by comparing attested languages. Hincks got involved, in the mid-1840s, numerous But philology has applications that are not so obvious: cuneiform inscriptions had been brought from the investigating the use and distribution of languages heartland, Mesopotamia, and they too were in the third themselves reveals patterns in the human career that script, so Hincks had a great deal of material to com- do not emerge when only the contents of historical pare. This time, the second script did constitute a sub- documents are considered. On at least one occasion a set of the third script, but the most useful aspect of the philologist has used his professional skills in creating inscriptions was their repetitiveness: what seemed to a fantasy universe; Tom Shippey (2003), besides intro- be the same formulas appeared in different texts, or ducing the work of philology to the committed reader, even within one text, but they contained small diver- shows how J.R.R. Tolkien’s legion of imitators gener- gences. (Some of the best examples came from an ally fail for lack of the kind of profound background extensive inscription in what we now call Urartian, Tolkien created through his many invented lan- which was published in 1840.) A sequence like A B C guages—and with example after example reveals that D E F G might substitute for A B H E F G, and Hincks Tolkien’s deep learning in Germanic philology, as well saw that this could mean that C D stood for as Celtic and Finnic, underlies feature after feature of C1V1–V1C2, while H stood for C1V1C2. He also dis- Middle Earth. covered patterns like iprus–iparras, a sure sign that the language was Semitic; we now call it Akkadian References (and the words were verbs). He found, rather incredi- bly, that characters could have more than one pronun- Charles, R. H. 1912. The book of Enoch …. Oxford: Clarendon. ciation—and he suggested that this was because the Daniels, Peter T. 1994. Edward Hincks’s decipherment of Mesopotamian cuneiform. The Edward Hincks bicentenary pronunciations found their origin in more than one lectures, ed. by Kevin J. Cathcart. Dublin: University language, what we now call Sumerian alongside College, Department of Near Eastern Languages. Akkadian. Most importantly, he realized that some of Daniels, Peter T. 1995. The decipherment of ancient Near the characters were being used not for the sounds they Eastern scripts. Civilizations of the ancient Near East, ed. by represented, but to stand for meanings. These charac- Jack M. Sasson et al. New York: Scribner’s. Daniels, Peter T. 1996. Methods of decipherment. The World’s ters—which he initially referred to as ‘ethnological writing systems, ed. by Peter T. Daniels and William Bright. boulders,’ since they were like alien rocks standing in New York: Oxford University Press. a plowed field—were remnants of Sumerian routinely Diringer, David. 1968. The alphabet: a key to the history of used in Akkadian texts. mankind, 3rd edition, 2 vols. New York: Funk & Wagnall’s. Thus, between 1846 and 1852, Edward Hincks pro- Herde, Peter. 1974. Diplomatics. Encyclopædia Britannica, 15th edition, Macropædia; since 1985 included under vided the key to the Akkadian, Urartian, and Sumerian ‘History, The study of.’ languages. One often hears of H.C. Rawlinson in this Jensen, Hans. 1969. Sign, symbol and script, 3rd edition, trans- connection, but he was never afterward able to explain lated by George Unwin. London: George Allen & Unwin. how he deciphered cuneiform; it is now known that, in Kenney, E. J. 1974. Textual criticism. Encyclopedia Britannica, his post in Baghdad, he was kept informed of Hincks’s 15th edition, Macropædia; since 1985 included under ‘History, The study of.’ work and used it in his analysis of the great inscription Larsen, Mogens Trolle. 1996. The conquest of Assyria: excava- on a cliff at Behistun, Iran, which he copied with great tions in an antique land 1840–1860. London: Routledge. difficulty—but which became available too late to be Ostler, Nicholas. 2005. Empire of the word: a language history useful in the decipherment. of the World. London: HarperCollins. Another celebrated decipherment was accom- Pope, Maurice. 1999. The story of decipherment: from Egyptian hieroglyphs to Maya script, 2nd edition. New plished by Michael Ventris in 1952, when he realized York: Thames & Hudson. that the Linear B tablets from Crete and Mycaenae on Puhvel, Jaan. 1974. Epigraphy. Encyclopædia Britannica, 15th the Greek mainland were actually in an archaic form edition, Macropædia; since 1985 included under ‘History, of Greek, not some unknown predecessor language. In The study of.’ this and several other cases, the virtual bilinguals were Robinson, Andrew. 2002. Lost languages:the enigma of the World’s undeciphered scripts. New York: McGraw-Hill. known place names that seemed to occur in texts Shippey, Tom. 2003. The road to Middle Earth: how J. R. R. found in those places; another is the decipherment of Tolkien created a new mythology. Boston: Houghton the Maya glyphs. Mifflin; 1st edition, 1982. Urry, William G. 1974. Paleography. Encyclopædia Britannica, 15th edition, Macropædia; since 1985 included under The Uses of Philology ‘History, The study of. PETER T. DANIELS It is self-evident that without philology to make texts available for study, history could not be written and See also Aramaic; Semitic Languages; Sumerian 826  PHILOSOPHY OF LANGUAGE Philosophy of Language Rather like the scholastic philosophers of the late ‘ontological argument for the existence of God’, Middle Ages, twentieth-century philosophers have which mostly simply runs, God is that being who pos- seen philosophy as linguistic analysis, as the attempt sesses all possible perfections; existence is a perfec- to discern the logical structure of reality through dis- tion; therefore, God exists. This argument confounds cerning the formal structures—the superficial or deep the hypothetical predications ‘If something is Divine, grammar—of the language in which we report or think then it has various Perfections’ and the existential reality. Philosophers are particularly interested in cer- ‘Something exists that is Divine’. tainty, in necessary truths as opposed to mere chance Russell was particularly praised for his ‘paradig- events. Nineteenth-century formalist mathematicians, matic’ analysis of ‘definite descriptions’, phrases of similarly, honed a sense of mathematics as syntactical- the form The so and so. Consider ‘The present King of ly grounded in formal language or in a consistent set France is bald.’ If you think this statement has a sub- of linguistic conventions. Since philosophers con- ject/predicate logical form, and believe that statements ceived of their enterprise as a search for helpful logi- are either true or false, then you seem to have to say cal truths and not as dependent on any experiential that either ‘The present King of France is bald’ is true truths (philosophers do not do experiments), twentieth or ‘The present King of France is not bald’ is true. The century philosophy became logicolinguistic analysis, problem, of course, is that there is no present King of and hence philosophy of language became, in short France (when Russell wrote and subsequently). It is and for much of the past century, philosophy. no solution to say ‘The present King of France’ means Linguistic philosophy was not just seeing the fea- nothing, because then ‘The present King of the United tures of language that revealed the world’s categorical States’ and ‘The round square’ and countless other structure but also seeing beyond language’s perhaps phrases would presumably also mean, or stand for, misleading surface features. To give a standard exam- nothing; but it is evident that all such phrases differ in ple, the verb to be in English plays three vastly differ- meaning. If a revolution occurred in the United States ent logical roles. Is can mean identity, as in 2 is (the in 2084, there might then be a King of the United same as) 2, Dubya is (the same as) President George States, but that would not mean there would be a round W. Bush, or heat is (the same as) the average motion square and a French King as well (or that nothing had of molecular particles. Is, however, can also mean changed to something). Russell insists that ‘The pres- predication, as in The sky is (has the property of ent King of France’ does not mean anything all by being) blue, Roses are (have the property of being) itself. Rather, ‘The present King of France is bald’ red, or The earth is (has the property of being) spher- means ‘There exists an x that bears the predicate of ical. Finally, Is can mean existence, as in There is being kingly of France, and if any y also bears that (exists) a Santa Claus or God is (exists) as opposed to predicate, y  x, and x is bald.’ God is (exists) not. Bertrand Russell (1919), while Following Russell’s lead, in the 1920s, the logical imprisoned as an antiwar demonstrator in World War positivists of the Vienna Circle insisted that logico- I, wrote in Introduction to mathematical philosophy mathematical truths provide the formal structure with- that it was important to keep these three senses whol- in which the observational truths of experience array ly separate in linguistic analysis (so important that themselves. Theoretical terms, such as ‘vital spirit’, Russell added that he would declaim so even if he ‘electron’, or ‘virus’, are only acceptable if they can be were ‘dead from the waist down and not merely in specified completely in observational terms. The prison’). In the mathematical logical notation that Vienna Circle philosopher W.V.O. Quine (1960) of Gottlob Frege and Russell created in the late nine- Harvard University claims that by casting science into teenth century, these notions were indeed wholly sep- Russell’s austere predicate logic, you can most clearly arate and specifically represented. With a notation determine what science says has to be real. Indeed, that puts predicates in capitals and writes names and such a translation into predicate logic will strip off the variables for individuals in lower case, ‘a  b’ means possibly misleading superficial features of actual lan- two names stand for the same individual, ‘Pa’ means guages. Philosophical linguistic analysis, conceived in predicating P of individual a, and ‘∃x[Dx]’ means, if this fashion, cares nothing for the phonology of lan- D predicates divine, that there exists a God. Russell guage, for literal physical sound streams and their stressed that this notation commendably dissolved the transformation into the sharper, leaner, and deeper 827 PHILOSOPHY OF LANGUAGE structures that fluent speakers hear. Moreover, the bittern—I got a clear view and, growing up in the exclusive concern is with the truth or falsity of sen- Fens, I’d know that booming anywhere’. Philosophical tences that describe the world, collectively feeding concern with semantic, performative, and pragmatic what scientific generalizations we can muster about aspects of language has meant some fruitful interac- the world. As Quine put it, startlingly, all of science is tion with linguistic science, particularly given the con- held up as one sentence to nature—but that is just to cern with syntax and logical form stressed by Noam insist that we collectively assert and understand this Chomsky and other generative linguists since the scientific sentential consensus, not that anyone actual- 1960s. ly says this one sentence. Philosophy of language, as But still other philosophers have come to feel that so understood, has little concern with how natural lan- philosophy is well rid of an exclusive emphasis on phi- guage is used in the acts and interactions of everyday losophy as linguistic analysis. In the middle decades life. Quine makes this clear when, at the beginning of of the twentieth century, philosophers concerned with Word and Object, he approvingly cites Otto Neurath’s values emphasized metaethics or the ‘logic of the lan- metaphorical remark that we are like seamen who guage of morals’; more recently, philosophers have must make repairs in our boat while at sea. Here, ‘we’ addressed specific normative issues. Similarly, many means all scientific-minded humans, and ‘the boat’ recent philosophers have constructed rational choice, means whatever natural and artificial symbol systems and social contract theories. Thinkers as diverse as the we use in our collective scientific description of the logician Saul Kripke and the linguist Noam Chomsky world, with Quine highlighting the austere notation of have argued that necessary truths are more central to mathematical logic. science than the real but often trivial analytic truths of By the mid-twentieth century, however, philoso- language. Heat is the average motion of molecular phers began to shift from a concern with the true/false particles is not, Kripke argues, an analytic or lexical logical relationship between sentences and the world truth, but rather a scientifically discovered necessary to a more expansive, but also more narrowly linguistic, physical truth. In the same vein, water is H2O asserts concern with sentences as speech actions of speakers, a physical necessity, given the basic combinatory who carry individual responsibility to us for what they properties that form the fabric of atomic physics. say, indeed for what they do in saying so. In the 1950s, Similarly, perhaps, the linguist’s claim that natural lan- the Oxford University philosopher J.L. Austin, among guages are generative and transformational follows others, drew attention to performatives, sentences that, from the structural character of the human linguistic when uttered by the appropriate person in the appro- faculty as a matter of biological necessity (it certainly priate circumstances, do something. If I say, ‘I prom- does not follow from a lexical entry attributed to nor- ise to return your $20 tomorrow’, I am not describing, mal speakers, most of whom do not have ‘generative truly or falsely, some peculiar mental state; rather, by and transformational’ in their everyday vocabulary— my speaking, I make a promise. If I say, ‘By the power or ‘average motion of molecular particles’, for that vested in me by the State of New Jersey, I appoint you matter). In any case, few philosophers today would be Port Commissioner,’ I make an appointment. ‘I prom- wholly content to characterize philosophy as linguistic ise X’ and ‘I appoint Z’ are explicit performatives, analysis. where the main verb of the sentence explicitly indi- cates what action is performed. However, ‘I certainly will return your $20 tomorrow’ will, given appropriate References circumstances, constitute a promise. Similarly, if an Austin, J. L. 1961. Philosophical papers. Oxford: Clarendon officer says to his subordinate, ‘You will move your Press. men to that bridge,’ he has given an order. Indeed, Ayer, A. J. 1936. Language, truth, and logic. London: Gollancz. Austin maintained that every use of language, or every Chomsky, Noam. 1986. Knowledge of language. New York: Praeger. speech act, has a performative aspect. For example, if Kripke, Saul. 1972. Naming and necessity, in Semantics of nat- I say ‘There’s a bittern in your garden,’ I should know ural language, ed. by Gilbert Harman and Donald Davidson, a little something about bitterns and have had some Dordrecht, Holland: D. Reidel. opportunity to identify the bird. If I have no idea of Quine, W. V. O. 1960. Word and object. Cambridge, MA: MIT what bitterns look or sound like, and indeed, have not Press. Russell, Bertrand. 1919. Introduction to mathematical philoso- been in or near your garden, I have no right to say phy. London: Allen & Unwin. what I did. Given that I meet those minimal require- JUSTIN LEIBER ments, I may demur if you ask, ‘How do you know it’s a bittern?’ pleading perhaps ‘Well, I don’t know it’s a See also Austin, John Langshaw; Chomsky, bittern but it is a large, white-feathered marsh bird’, or Noam; Quine, Willard van Orman; Russell, I may take a further plunge and say, ‘Oh, I know it’s a Bertrand 828  PHONEME Phoneme The phoneme [from Greek phonema—a sound] is rec- Phonetic Context ognized as the smallest abstract sound unit of a lan- Every uttered word is a particular combination of guage, for example, map consists of three phonemes sounds and is called the phonetic context of a sound. /m/, /æ/, and /p/. Each word in a given language must The English phonetician Daniel Jones gives the fol- consist of a permitted sequence of phonemes. lowing definition of the phonetic context: Phonemes may be defined as the class of sounds of a language that distinguish words from one another. …phonetic context of a sound is to be understood to The phoneme is a fundamental unit of phonology— mean the sounds next to it or near it in the sequence of the study of the sound systems of languages and the which it is a part, together with its duration (length), relationships between those sounds. Different lan- stress and (if voiced) voice-pitch. guages have various numbers of phonemes; for Minimal pairs provide the same phonetic context instance, one Brazilian language—Piraha—is said to with only one different phoneme, as in bat and cat; have only ten phonemes, while one of the African lan- where the sounds [b] and [k] occur in the same pho- guages—!Xu—is known to have 141 phonemes. netic context, they differentiate the meaning of the The term phoneme was introduced in the late 1870s, words, and, therefore, are two different phonemes. and the phoneme theory was first developed by Jan Baudouin de Courtenay and his pupils as a response to the need to systematize the sound patterns of the lan- Allophones guage. Since then, the phoneme and its features have A phoneme may have several variants or allophones. As been defined in various ways by different linguists. As opposed to phonemes, allophones do not distinguish Baudouin understood it, the term originally meant a words, and they never occur in the same phonetic con- ‘mental image’ of a real physical sound, and speech was text. For example, cat and king possess a common the act of producing sounds as close to that mental sound [k]. However, the sound [k] is pronounced slight- image as possible. However, during the twentieth cen- ly differently: in the first instance, it is articulated—or tury, linguists came to a different view. Lev Scerba, who pronounced—further back in the throat. In the second had initially followed Baudouin, shifted to the function- instance, the tongue—in an attempt to combine the al aspect of the phoneme, that is differentiating words. sounds [k] and [i]—moves upward and forward produc- Daniel Jones also described a phoneme as a physical ing a palatalized [k]. These two sounds vary only in the concept. If Nikolai Trubetskoy defined the phoneme as way they are articulated (pronounced); they cannot the smallest distinct unit and incapable of further subdi- occur in the same phonetic environment, i.e. the palatal- vision, Roman Jakobson saw it as the sum of features ized [k] cannot occur before [æ] as in cat for the simple that distinguish the phonemes from each other. reason that our tongues are physically unable to make these two sounds together. The two sounds vary slight- Contrastive Function ly in articulation; however, the difference is not signifi- cant, and may not even be picked up by a native English It is traditionally accepted that the phoneme functions as speaker. Thus, these two variants of /k/ are not distinc- the minimal contrastive unit of speech. In other words, tive in English; they are not different phonemes, but the phoneme differentiates the meaning of words. The rather allophones of the same phoneme /k/. native English speaker easily discerns the difference in Daniel Jones gives the following examples of the meaning between the words in minimal pairs, i.e. word allophones of the English phoneme /t/: pairs that differ in only one sound element: man and pan, The principal [t] is used before vowels in the where [m] and [p] differentiate between a human and a strongly stressed position as in table or ten. kitchen utensil, thus indicating that they possess differ- ent phonemes; bag and big, where [æ] and [I] make pos- A dental [t] is used before [th] as in eighth. sible the distinction between an object and a size; den A retracted [t] is used in [tr] as in train. and then, where [d] and [ð] clearly indicate that one An unaspirated or barely aspirated [t] is used in word means a ‘lair’ and the other one means ‘at that weakly stressed positions as in letter. time’; or man and men, where [æ] and [e] identify one A laterally exploded [t] is used in [tl] as in kettle. man as opposed to two or more of them. A nasally exploded [t] is used in [tn] as in mutton. 829 PHONEME For a native language speaker of a given language, opposed to each other on features that do not create a it is not really important how carefully allophones are phonemic opposition in English. pronounced to understand what is being said. Bilateral opposition represents only pairs of However, it is more important that phonemes are pro- phonemes, such as voiced–voiceless pairs /z/ - /s/, where nounced properly in order for the listener to get the the third element is not possible. Multilateral opposition, correct message; instead of My brother bought you a on the other hand, may have three or more elements, big present, the speaker must not relay My brother such as /p/, /t/, and /k/, which are all stop consonants, or bought you a pig present. /f/, /ð/, /θ/, /s/, which are all fricative consonants. Languages differ with respect to which sounds are Privative opposition can be illustrated by perceived as distinctive (phonemic) and which sounds voiced–voiceless pairs, in which one member of each are mere variants or allophones of phonemes. English pair lacks a voiced feature (/d/-/t/, /b/-/p/). Equipollent distinguishes between [l] and [r] as in lay and ray, which opposition is where voiceless /p/, /t/, /k/ each have a make them different English phonemes. Korean does not different place of articulation. Gradual opposition see any distinction between these two sounds, i.e. they involves gradation of a feature in the phonemes, for are perceived by Koreans as allophones of one phoneme. example, in vowels /i/, /e/, /a/, the difference is in aper- Moreover, although English distinguishes between [d] ture, or close–open continuum of tongue height. and [ð] as in than, Malay does not distinguish them as In many languages, the phonological opposition two phonemes but as allophones of one phoneme. Of can be neutralized. For example, in English, when /s/ course, many languages make other distinctions that an occurs in the word-final position, it is pronounced as English native speaker may not be able to hear. [s] after a voiceless plosive: lips, rats, sacks, and as [z] after the voiced plosive: dogs, bags, ribs. In the Russian language, a voiced consonant can Distribution never occur at the end of the word. When the voiced Each language imposes limitations on the distribution consonant is found at the word-final position, it must of phonemes, i.e. each particular phoneme can occur be pronounced as voiceless. Neutralization in Russian only in certain phonetic contexts in that language. In can be illustrated by the following pair: pok spelled as English, there are no words that can start with the [rok] (type of music), and pog spelled as [rog] (a nasal [N] sound. There are no such words as ngull, horn); both are pronounced as [rok]. In the latter ngreed, ngoose. In Chinese or Vietnamese, however, instance, the voiced [g] sound is neutralized and is similar sound constellations are quite possible (e.g. pronounced as the voiceless [k]. names such as Ng or Nguyen). The distribution pattern of the phonemes in a lan- Major Early Works on the Phoneme: guage can be identified by analyzing which phonemes Baudouin de Courtenay, Lev Scerba, may occur in the following positions: Ferdinand de Saussure, Daniel Jones syllable-initial syllable-medial syllable-final This section is a brief excursion into the major contri- word-initial word-medial word-final butions of world linguists in relation to the notion of The English phoneme /p/ can be found in all of the the phoneme only, and not phonology, which is above-mentioned positions, for example: beyond the scope of this paper. The term phoneme was introduced in 1875 to indi- post apt map cate the basic unit of human speech. Although it is said packet upper ketchup that the term was first used by French linguist Other English phonemes can occur only in certain Dufriche-Desgenettes in 1873, it was Jan Baudouin de positions, for example, /œ/ as in rat, stamp, sachet, ant Courtenay and Mikolaj Kruszewsky of the University can only be found in syllable- or word-initial positions of Kazan (Russia) who, between 1875 and 1895, con- and syllable- or word-medial positions. ducted an extensive study of sounds, and established the phoneme theory. The Kazan School focused its research on the phoneme as ‘the psychological equiv- Phonological Opposition alent of the speech sound’. The phoneme, according to The phonological opposition, developed by the Prague Baudouin, was impressed in the speaker’s mind. In the School, illustrates the network of connections and speech process, the speaker was aiming at uttering a relations between the phonemes. For example, propor- sound similar to the psychological impression or the tional opposition features two or more pairs of mental image; however, the pronunciation was always phonemes, such as /p/ vs. /b/, /k/ vs. /g/, which are determined by the real phonetic context. opposed to each other on the basis of voiced features, The work of Baudouin influenced the Russian lin- while /r/ and /l/ are in isolated opposition and are guist Lev Scerba (1880–1944), who attended Baudouin’s 830  PHONETIC TRANSCRIPTION lectures in St. Petersburg. Scerba accepted the psycho- the phoneme was based on the structuralist ideas of de logical definition of the phoneme, but later concentrat- Saussure. For example, a sound can be a vowel (a, i, o, ed on the functional aspect of the phoneme, i.e. the e, u) or a consonant (b, d, g, z). If the sound is a con- differentiation of words. Scerba also introduced such sonant, it can be voiced (b, g, d) or voiceless (p, k, t), phonological notions as complimentary distribution, nasal (m, n) or nonnasal (all others). Vowel, voiced, which is a characteristic of a sound that never occurs and nasal are all examples of distinctive features. in the same phonetic context. This can be illustrated by American structuralism was developed by allophones of a phoneme, all of which occur in differ- American linguists Edward Sapir and Leonard ent phonetic contexts but are variants of the same Bloomfield. Sapir adhered to the idea that a native phoneme. Since the allophones are different realiza- speaker knows the phonemic system of their language tions of the same phoneme, it can be said that they are through intuition. For Sapir, the phonemic system was in complementary distribution to each other. a mental reality that existed independently of the act of Ferdinand de Saussure (1857–1913), a Swiss lin- speech. Bloomfield saw the phoneme as reality, a bun- guist, is considered the founder of modern linguistics. dle of distinctive features, but not as the intuition of a His main conception assumed that language is a sys- speaker. He believed that phonemic analysis was the tem of mutually defining entities. He believed that lan- process of isolation of its distinctive features through guage was a structured system. This view had a major minimal pairs. According to Bloomfield, finding a dis- impact on the later development of structuralism in tribution of redundant, nondistinctive features was linguistics, which studies the language as a system and also part of the phonemic analysis. not the individual elements of it. Structuralism was further developed in Europe by Trubetskoy and Jakobson, and in America by Sapir and Bloomfield. References Daniel Jones (1881–1967), who was a lecturer of Abercrombie, David. 1965. What is a letter? Studies in phonet- phonetics at University College, London, favored the ics and linguistics. London: Oxford University Press. idea of the phoneme as a practical tool for teaching ———. 1967. Elements of general phonetics. Edinburg: University Press. purposes. His contribution to linguistics was in devel- Bloomfield, Leonard. 1933. Language. New York: Holt. oping a more practical approach to phonology. He Hawkins, Peter. 1984. Introducing phonology. London: found phonology very useful for acquiring good pro- Routledge. nunciation by people learning a foreign language, as a Jakobson, Roman. 1962. Selected writings, Vol.1. The Hague: basis of nonphonetic branches of linguistics such as Mouton. Jones, Daniel. 1962. The phoneme: its nature and use, 30. morphology and grammar, and as an important tool for Cambridge: W. Heffer & Sons Ltd. creating alphabets as simply as possible. Kramsky, J. 1974. The phoneme. Munich: Wilhelm Fink Verlag. The Prague Linguistic Circle, which later came to Trubetzkoy, Nikolai Sergeevich. 1958. Grundzuge der be known as the Prague School, began its meetings in Phonogie, Gottingen: Vandenhoeck & Ruprecht; third edi- 1926. It consisted mainly of Russian and Czech lin- tion, 1962, as Principles of phonology, translated by C.A.M. Baltaxe, Berkeley: University of California Press, 1969. guists. Among them were the Russian linguists Yallop, Colin. 1995. English phonology. Sydney: Macquarie Nikolai Trubetskoy and Roman Jakobson. The contri- University. bution of Trubetskoy lies in the developing of the ALFIA ABAZOVA notion of phonological opposition, and relations between the members of an opposition. Jakobson’s See also Baudouin de Courtenay, Jan Ignacy view of the phoneme as a sum of distinctive features NiecisBaw; Bloomfield, Leonard; Feature Theory; led to the creation of a new approach to phonological Sapir, Edward; Saussure, Ferdinand de; Struc- description. This approach of distinctive features of turalism; Trubetzkoy, Nikolai Sergeyevich Phonetic Transcription In speech, different letter(s) may represent the same language known in the world in an unambiguous way, sound (e.g. see, grief, key) or the same letter may the International Phonetic Association developed a set represent different sounds (e.g. bad, many, want). of symbols known as The International Phonetic In order to be able to express speech sounds of each Alphabet (IPA). The use of sequences of phonetic 831 PHONETIC TRANSCRIPTION symbols to represent speech is known as phonetic can be distinguished with regard to voicing, manner, transcription. The concept of a phonetic alphabet is to and place of articulation. Voicing refers to the pres- have one symbol for one sound. Over the years, dif- ence or absence of vocal fold vibration (/b/ versus ferent alphabets with modified IPA symbols have been /p/, /d/ vs. /t/). Manner of articulation refers to the developed for specific aims (as many symbols are not way in which speech sounds are produced (stop, used), but the IPA is considered the standard. fricative, nasal, trills, etc.), and place of articulation indicates the place of constriction in the vocal tract (lips, alveolar ridge, velum, pharynx, etc.). Some History languages in the world contain nonpulmonic sounds, In 1886, the International Phonetic Association was sounds that are not produced through air from the founded in Paris by language teachers who wanted lungs: clicks and voiced implosives are indicated by phonetic notation to be used in schools, as a method special symbols, while ejectives are indicated by a of acquiring a realistic pronunciation of foreign lan- plosive and an apostrophe. From an articulatory guages. In 1897, it was named L’Association point of view, vowels can be classified according to Phonétique des Professeurs de Langues Vivantes or the features ‘open’, ‘close’, ‘front’, and ‘back’. A the International Phonetic Association in English. vowel is considered ‘open’ when the space between The IPA is used, among other things, to indicate pro- the tongue and the roof of the mouth is large and nunciation in a dictionary, to record a language, and ‘closed’ when the tongue is near the roof of the to transcribe speech for automatic speech recogni- mouth. In front vowels, the tongue is fronted and tion. The IPA is based on the Roman alphabet, but raised toward the alveolar ridge; in back vowels, the other symbols have been included to cover the wide tongue is near the back of the mouth (palate). variety of sounds found in the languages of the Although the vowel space is continuous, most vow- world. Phonetic transcription is placed between els are placed in relation to one of the eight cardinal square brackets [ ] and there is no capitalization or (reference) vowels: /∑i, e,
, a, , , o, u/. punctuation. An overview of the different characters Several symbols are foreseen to designate supraseg- and symbols of the IPA is given in various charts mental information, i.e. intonation groups (), duration elsewhere. The symbols will be discussed in more (), stress (), and to distinguish between words in tone detail after highlighting some of the principles under- languages (e.g. Chinese, Thai). Some of these symbols lying the phonetic alphabet. are iconic in the sense that the shape of the line indi- Phonetic transcription does not only vary from lan- cates the height and possible movement of pitch. For guage to language but also from speaker to speaker. example, [maÜ] with falling intonation means ‘to Depending on dialect, fluency, etc., the same word can scold’ and [ma!] with fixed high intonation means be transcribed differently. Speech segments can be ‘mother’ in standard Chinese. However, it is also pos- divided into two major categories: consonants and sible to transcribe pitch height by adding diacritics to vowels. They are described with reference to how they existing segments, e.g. [ó bà] meaning ‘it perched’ vs. are produced and their auditory characteristics. It [ó ba] ‘he or she hid’ in the West African language requires considerable skill to transcribe an unknown Yoruba. In this transcription, the symbols are not icon- language. During speech production, the vocal tract ic, i.e. the accent does not indicate a rising or falling varies continuously and the division between subse- pitch, it means ‘high’ ‘mid’, or ‘low’. quent speech sounds may not be clear. Another set of marks, called diacritics, has been Only linguistically relevant speech sounds are tran- included to further refine the transcription of phona- scribed, not personal voice quality. A separate sign tion. This has been done to restrict the total number of exists for each distinctive sound (e.g. /k/ for cat, track, characters. Diacritics are used to indicate aspiration, kite, quick, monarch). The IPA also provides symbols e.g. [p ha] ‘pie’, to indicate vowel centralization [ë] for suprasegmental information (word boundaries, relative to the cardinal vowels, breathy and/or creaky stress, intonation) and for refining the pronunciation of voices, nasalization [fε˜] ‘fin’, meaning ‘end’ in an utterance (e.g. aspiration). French, and many other features. IPA charts Broad and Narrow Phonetic Transcription The IPA charts give an overview of the different Speech sounds can be transcribed ‘broadly’ or ‘nar- vowels, consonants, clicks, and other sounds occur- rowly’. A broad transcription is a phonemic tran- ring in the languages of the world. The charts are scription, while a narrow transcription captures more more than a list of symbols; they show how the dif- phonetic details of the speech sounds. For instance, ferent types of sounds can be classified. Consonants the word ‘toast’ can be transcribed phonemically as 832  PHONETICS [tst] or phonetically as [th st]. Both types References of transcription require some knowledge of the Albright, R.W. 1958. The International Phonetic Alphabet: its phonology of the language to solve ambiguities background and development. International Journal of related to the segmentation of utterances. And even American Linguistics 24(1). 3. when the likeliness of occurrence of certain (combi- Handbook of the International Phonetic Association. 1999. A guide to the use of the International Phonetic Alphabet, nations of) phonemes is known, the alignment 204pp. Cambridge: Cambridge University Press. between the physical speech sounds and the phonet- Ladefoged, P. 1982. A course in phonetics, 2nd edition. New ic transcription can still be problematic. Phonetic York: Harcourt Brace and Jovanovich. transcription has proved to be a very useful skill, but Ladefoged, P., and Maddieson, I. 1996. The sounds of the it cannot be done automatically. A proper analysis of world’s languages, 426pp. Oxford: Blackwell Publishers. Roach, P. 1991. English phonetics and phonology, 262pp. the speech signals requires considerable knowledge Cambridge: Cambridge University Press. about the segmental and suprasegmental aspects of a language. ASTRID VAN WIERINGEN Phonetics Phonetics is the scientific study of speech. It is con- Phonetics overlaps most directly with Phonology, cerned with all aspects of the production, acoustics, Psycholinguistics, Sociolinguistics, and Dialectology and perception of speech in the languages and dialects and also with Syntax, Semantics, and Pragmatics in of the world. The starting point for almost any phonet- modeling the prosodic aspects of speech. ic investigation is the identification of certain land- Phonetics is relevant to disorders of speech com- marks from a cross-sectional view of the vocal tract munication and to speech pathology. In the last quar- from which both consonants and vowels of the lan- ter of the twentieth century, phonetics has contributed guages of the world can be classified (see the section in various ways to the field of speech technology, on Basic Classificatory Principles). The classificatory which is concerned with the development of computer system can be used to describe both the various types systems for the automatic generation (synthesis) and of speech sounds that occur in different languages and recognition of speech. their dialects, as well as those aspects of pronunciation that are used contrastively, that is, to signal differences Basic Classificatory Principles of meaning (see the section on Phoneme and Phonetic Variation). Phoneticians also analyze how languages The range of possible sounds that can occur in the make use of pitch, duration, and loudness to commu- world’s languages and from which a language makes a nicate meaning distinctions at both the level of the selection can be described in terms of various interac- word and the utterance (see the section on Prosody). tions between the vocal organs (Figure 1). The great Phonetics is concerned with the way in which spo- majority of speech sounds are produced when air is ken communication is accomplished between speakers expelled from the lungs, passing through the larynx and hearers. This includes both the relationship between two shelves of muscular tissue known as the between speech sounds and the neurological, physio- vocal folds (Figure 2). When the vocal folds are apart, logical, and neuromuscular aspects of speech (see the as they are in quiet breathing, they do not obstruct the section on Speech Motor Control and Physiology), as passage of air from the lungs. Many speech sounds are well as the resulting acoustic signal that is decoded by produced with open vocal folds and they are known as the listener (see the section on Speech Acoustics and voiceless. In many other speech sounds, including all Speech Perception). vowels, the vocal folds can be drawn together and Phonetics is a highly interdisciplinary field and is made to vibrate very rapidly, that is, they repeatedly informed by theory and methodology from alternate between being open and closed—such Linguistics, Cognitive Science, Computer Science, sounds are voiced. There is a direct relationship Electronics, Signal Processing, Acoustics, Neurology, between the rate at which the vocal folds vibrate and Anatomy, and Physiology. Within Linguistics, the sound’s pitch. In most languages, there are pairs of 833 PHONETICS A 5 6 4 B c 7 1 d 3 b a 2 C e D E Figure 1. A cross-sectional view of the vocal tract of an adult male taken from an X-ray image (adapted from Laver, J, Figure 3. The configuration of the vocal organs for the pro- 1994). Upper lip (1), lower lip (2), upper front teeth (3), alve- duction of [n]. The direction of the air from the lungs is shown olar ridge (4), hard palate (5), velum or soft-palate (6), uvula by the lines and arrows. The velum is lowered allowing air to (7), tongue tip (a), tongue blade (b), tongue front (c), tongue enter the nasal cavity and pass out through the nose. back (d), tongue root (e), nasal cavity (A), oral cavity (B), pharyngeal cavity (C), larynx (D), and trachea (E). The places of articulation are obtained from the following combinations: actions of the velum or soft palate (Figure 3). When bilabial (12), labiodental (23), dental (a3), alveolar the velum is raised, the air passes out of the mouth or (a4), apical postalveolar (a5), laminal postalveolar (b5), oral cavity and sounds produced in this way are said to palatal (c5), velar (d6), uvular (d7) and pharyngeal (eC). be oral. When the velum is lowered, the air enters and passes through the nasal cavity and exits through the nose: such sounds are nasal. Many languages have pairs of sounds that are distinguished only by whether or not the velum is raised. For example, in English [b] in ‘bad’ and [m] in ‘mad’ are produced in the same way, except that [b] is oral and produced with a raised vocal fold vocal fold velum while [m] is nasal and produced with a lowered velum. Within the oral cavity, there are three other impor- tant landmarks: the upper front teeth, and behind them the alveolar ridge, which extends in an arch into the hard-palate (Figure 1). Figure 2. A photograph of the vocal folds taken from above The tongue is a highly flexible and mobile vocal (from Ladefoged, 2001b). The space between the vocal folds organ that has a very dense concentration of different is called the glottis. Here the vocal folds are apart allowing the muscles. The tongue is subdivided into the tip, blade, air to pass from the lungs through the trachea into the rest of the vocal tract. front, and back; in the rest position, these four land- marks lie roughly below the alveolar ridge, the back of the alveolar ridge, the hard palate, and velum, respec- sounds that differ only in whether or not the vocal tively. The tongue is attached to the epiglottis just folds vibrate, such as [f] (in ‘fan’, voiceless) and [v] above the vocal folds and includes a near-vertical sec- (in ‘van’, voiced). Voiced sounds are usually accom- tion in the pharynx, the root. panied by vibrations that can be felt by placing a hand The upper and lower lips are involved in various at the level of the throat and then switching voicing on ways in speech production: they may be protruded as and off as in a repeated production of [fffvvvfffvvvf- in the vowel of ‘soon’ or completely closed as in [b] fvvv]. and [m] in ‘bad’ and ‘mad’. The air from the lungs can be made to pass out of In producing any consonant, there is a point in the the vocal tract either through the mouth (the oral cav- vocal tract at which there is the greatest degree of nar- ity) or the nose (the nasal cavity) depending on the rowing and therefore the greatest obstruction to the 834  PHONETICS flow of air: its location is the consonant’s place of College of London. The Height  Backness space is articulation and it can vary anywhere between the lips used to define the phonetic differences between vowels and vocal folds (Figure 1). Three commonly occurring of different languages and dialects. There is independ- places of articulation are: bilabial (e.g. [p], [b], [m] ent experimental evidence to show that the extent to in ‘pan’, ‘ban’, ‘man’), alveolar (e.g. [t], [d], [n], [l] which listeners judge two vowels to be similar depends in ‘tie’, ‘die’, ‘nine’, ‘lie), and velar ([k], ["], [ŋ] as in to a large extent on the distance between them in this ‘cot’, got’, ‘sing’). space. Various acoustic analyses in the last 40 years For any place of articulation, there can be variations have shown that height and backness are correlated, in stricture, which defines the extent to which the flow respectively, with the first two resonances or formant of air out of the mouth is obstructed. When the frequencies of the vocal tract. obstruction is complete, as in [b d " p t k m n ŋ] in the Vowels can also vary in whether they are rounded above examples, the consonant is phonetically a stop. (produced with protruded lips) or unrounded (the lips A fricative is produced when the air is forced through are unprotruded). There is a preference across lan- a very narrow opening at high speed, often producing guages for front vowels to be unrounded (e.g. ‘heed’, a hissing sound due to the air becoming turbulent: ‘head’, ‘had’ have unrounded front vowels) and back examples are [s] (‘sip’) and [ʃ] (‘ship’). If the obstruc- vowels to be rounded (e.g. ‘hoard’, ‘who’ have round- tion is further reduced, the flow of air is no longer tur- ed back vowels). But there are many languages with bulent but laminar. A consonant produced in this way front rounded (French, German) and back unrounded is an approximant; examples in English include [w] vowels (Japanese, Vietnamese) and in some lan- (‘we’), [j] (‘you’), [l] (‘led’) and in some English guages, the presence or absence of lip-rounding may dialects [ɹ] (‘red’). [l] in ‘led’ is also a lateral, in which be the only feature that distinguishes between vowels the air flows over the side of the tongue, and is distin- (e.g. in French: high, front, unrounded [vi] (‘vie’, guished from central sounds in which the air flows ‘life’) vs. high, front, rounded [vy] (‘vu’, ‘seen’)). over the center of the tongue. The International Phonetic Alphabet provides a set The two principal features for describing vowel of symbols for transcribing the sounds of the world’s production are height and backness. Height is analo- languages based on classificatory labels or features of gous to stricture and defines the degree of opening in the kind discussed above. Each symbol therefore pro- the mouth. The smallest opening, which is never less vides information both about the sound that is pro- than for approximants, occurs in high or close vowels duced and its relationship with other sounds. For as in ‘heed’ while the greatest opening is in low or example, [p] is a voiceless bilabial oral stop, which open vowels like ‘had’. Backness is analogous to the means that like [f] and [t], it has open vocal folds place of articulation defined earlier: vowels can vary (voiceless), like [m] and [b] it has a constriction at the between front, in which the narrowing occurs at the lips (bilabial), like [b], [f], and [t] it has a raised velum hard palate, and back in which the greatest point of (oral), and like [b], [m], [t], and [d] the passage of air narrowing is often in the pharynx. through the mouth is completely obstructed at some Height and backness together form a two-dimen- stage during its production (stop). sional space that defines the ranges within which vowel The set of features and their relationships with each production (Figure 4) occurs. The edges of the vowel other have been considerably informed by a detailed space are marked by the cardinal vowels, defined by instrumental analysis of the sounds of the world’s lan- the phonetician Daniel Jones in 1918 at the University guages in the last 40 years (e.g. Ladefoged and Maddieson 1996). Other research seeks to understand why certain feature combinations tend to be preferred by languages, whether babies are born with detectors for identifying features, and the extent to which listen- ers make use of features in perceiving speech. Phoneme and Phonetic Variation The consonants at the beginning of the words ‘pin’ and ‘bin’ are different from each other not only phoneti- cally (the first is voiceless, the second voiced) but also in a more abstract phonological sense: they communi- Figure 4. The vowels and their phonetic symbols of the inter- national phonetic alphabet. The vertical dimension is Height, cate a difference of meaning and are allophones of the horizonatal dimension is Backness. When the symbols separate phonemes. The two ‘p’ sounds in ‘pin’ and appear in pairs, the one to the right represents a rounded vowel. ‘spin’ are also phonetically very different from each 835 PHONETICS other in most accents of English: the one in ‘pin’, tran- Scots English there is no such phonemic opposition scribed as [ph], is aspirated and produced with a puff and so these words are indistinguishable. And while of air that can be felt by saying the word and holding most English dialects have two very different allo- a hand near the lips, whereas the one in ‘spin’, tran- phones of /l/ depending on its position in the syllable, scribed as [p], is unaspirated and is produced without Southern Irish English makes no such distinction. any puff of air. But while [ph] and [p] are phonetically Therefore, /l/ in ‘leaf’ and ‘feel’ are phonetically very very different, they do not communicate a difference similar in Southern Irish English but quite different in in meaning and are allophones of the same phoneme most other dialects. /p/. The fact that these sounds do not distinguish Phonetic variation occurs at the level of the individ- between meanings in English can be shown by saying ual speaker. This variation is not simply due to the the word ‘spin’ but with an aspirated [ph]—the pro- anatomical differences between speakers (which is nunciation might be unusual, but native speakers of responsible for considerable acoustic differences in English would nevertheless still identify the word the speech of men, women, and children) but also ‘spin’. Allophones of the same phoneme occur pre- because speakers may have developed an idiosyncrat- dictably in different contexts: whereas [p] occurs in ic speaking style: a given speaker might have exces- the context after /s/ (‘spin’, ‘spoke’, ‘spring’, etc.), [ph] sive nasalization or a tendency to protrude the lips occurs in a different context at the beginning of during speech production. Developing a model that stressed syllables (‘pin’, ‘pan’, ‘put’, etc.). can represent speaker-specific aspects of pronuncia- A type of transcription that makes use of only the tion is part of voice quality research and it has impor- phonemes of a dialect or language is known as a broad tant applications both in disorders of voice and speech transcription: in such a transcription, all the phonetic as well as in the forensic analysis of speech that is characteristics that are predictable from context, and, increasingly used in criminal investigations. therefore, redundant for distinguishing between mean- ing, are filtered out. A narrow transcription is one that Prosody includes some phonetic detail, usually at least all the allophones that are common to the speakers of a par- The same utterance can be said in many different ways ticular dialect. to communicate paralinguistic effects of emotion such Every language has a finite number of phonemes as happiness, sadness, surprise, anger, and so on. out of which it assembles its words to construct differ- These paralinguistic effects are usually brought about ences of meaning. An important goal in phonetics and by manipulating the prosody or the timing, pitch, and phonology is to establish both the phoneme inventory loudness of an utterance. But these manipulations can of a language and the relationships between the be used linguistically to distinguish either between phonemes and their allophones. All of these differ words (word-level prosody) or between utterances markedly between languages and dialects. For exam- (sentence- or utterance-level prosody). Prosodic ple, while [p] and [ph] are allophones of the same effects usually extend over, and modify, more than a phoneme in English, in Korean they are allophones of single consonant or vowel; this is why the units of separate phonemes: as a result, exchanging them can prosody are sometimes called suprasegmentals. produce a meaning difference in Korean but not Word-level prosody includes differences due to English. The same phonemic opposition is neverthe- quantity that are largely communicated by timing dif- less likely to give rise to phonetic differences between ferences (these are marked phonetically by a colon: two languages. For example, both English and French thus, [t:] is a long version of [t]). Quantity differences create differences of meaning by distinguishing can occur in the consonant (e.g. Italian: [fato], ‘fact’ between two phonemes /b/ and /p/ (‘bin’ vs. ‘pin’ in vs. [fat:o], ‘done’) or in the vowel (German: [lam] English; ‘beau’ (‘beautiful’) vs. ‘peau’ (‘skin’) in (‘lamb’) vs.[la:m] (‘lame’)) or in both (Finnish: French). But in French the corresponding phonetic [mut:a] (‘but’) vs. [mu:ta] (‘other’) vs. [mu:t:a] (‘to distinction is between a fully voiced [b] (the vocal change’)). Word-level prosody also includes tone dif- folds vibrate while the lips are closed) and an unaspi- ferences in which word distinctions are based on pitch: rated [p], whereas in English the distinction is between for example, in Thai, the syllable [na] has different a devoiced [b#] (the vocal folds only start vibrating word meanings depending on whether it is produced after the lips come apart) and an aspirated [ph]. with a high-falling pitch (when it means ‘face’), a low- Not only languages but dialects of the same lan- falling pitch (when it is a name), or a high-rising pitch guage can differ in their phoneme inventory and how (when it means ‘aunt’)—see Ladefoged (2001a) for these phonemes are produced, or realized phonetical- examples and sound files. ly. Whereas most English dialects have a phonemic The syllable and word-stress are part of word-level difference between /u:/ in ‘pool’ and /υ/ in ‘pull’, in prosody. Almost all languages organize their phonemes 836  PHONETICS into syllable units and there is psycholinguistic evi- In the same way that word-stress concerns the rela- dence to show that children can identify syllables from tive prominence of syllables in a word, accentuation a very early age. Much of the phonetic variation (such has to do with the relative prominence of words in an as the two types of /l/ discussed above) can be attrib- utterance. In many, but by no means all, languages, uted to the structural position of a sound in the syllable. shifting the accentuation can evoke a different mean- In many languages, consonants at the end of a syllable ing that depends on the discourse structure of a dia- are phonetically much weaker than at the beginning of logue. Consider that when said in isolation, the most a syllable. This is the source of many sound changes in prominent or accented word of the sentence ‘I don’t which syllable-final consonants weaken or delete com- like classical music’ is usually ‘music’. But if the same pletely (e.g. Latin ‘septem’  French [set] (‘sept’, sentence were said in reply to: ‘Have you ever listened ‘seven’), processes that also occur synchronically (e.g. to Beethoven?’ then the accent in the same reply the /t/ of ‘fast’ is perhaps never pronounced in a context would shift to ‘like’ to indicate that ‘classical music’ is like ‘fast speech’, so that ‘fast’ rhymes with ‘pass’). part of the background or old information in the dia- Word-stress has to do with the relative prominence logue that is shared between the speaker and listener. or salience and, to a certain extent, the relative loud- Accented words tend to be produced with greater ness of syllables: for example, in the word ‘abra- clarity precisely because they often carry new infor- cadabra’, the first and fourth syllables are more salient mation and are therefore much more difficult to pre- than the others and are said to be stressed, while the dict from the context in which they occur. For a related others are unstressed. In English, the position in the reason, function words, which rarely carry new infor- word of the syllable with the strongest word-stress, mation and which are therefore much less likely to be known as the primary stressed syllable, can vary (e.g. accented, are less clearly produced and may be entire- ‘pattern’ vs. ‘admit’), whereas in other languages like ly deleted: consider that ‘the man in the moon’ can be Icelandic and Polish it has a delimitative function and understood as such simply by lengthening the [n] of occurs at a fixed position in the word. In some lan- ‘man’ and deleting the vowel in ‘in’ entirely. guages, there are a handful of words that differ only in Speaking style can have a considerable influence on their word-stress pattern: German ‘übersetzen’ means the utterance’s prosody. At faster rates of speech, or in ‘to transport’ (e.g. by ferry) when it has primary stress a more casual conversational setting, many of the on the first syllable but ‘to translate’ when it has pri- types of weakening and deletion discussed in this sec- mary stress on the third syllable. In English, tion are likely to be in greater evidence. unstressed syllables (and the sounds within them) can be considerably shortened and often completely delet- Speech Motor Control and Physiology ed (e.g. the second unstressed syllable in ‘Tower Bridge’ can be deleted resulting in a pronunciation Speech motor control is concerned with the neurolog- close to ‘Tar Bridge’). This very clear phonetic differ- ical and physiological bases of speech production. ence between stressed and unstressed syllables that Investigations in this area are directly relevant to artic- affects vowels and consonants in English is much less ulatory synthesis in which the aim is to be able to syn- is evidence in languages such as French and this is one thesize speech using a model of the speech production of the attributes that contributes to the very differently mechanism. sounding rhythm of these two languages. Studies of speech motor control make extensive use Utterance-level prosody includes intonation and of instrumentation for the analysis of the vocal organs accentuation, both of which can be used linguistically including: laryngography and direct fiberoptic laryn- to provide a range of different meanings. Intonation goscopy for measuring and viewing the activity of the depends phonetically on the rise and fall in pitch and vocal folds; aerodynamic techniques for measuring can mark syntactic differences in many languages such airflow and air pressure at various points in the vocal as distinguishing a statement from a question. tract; and electropalatography, which records the pat- Languages, and indeed dialects of a language, have tern of contact between the tongue and the roof of the quite different associations between meaning and into- mouth. Midsaggital electromagnetic articulometry nation. For example, whereas in most accents of (EMA) has been in use for about 15 years for record- English, the intonation falls toward the end of a state- ing the movement and velocity of the jaw, the lips and ment, in Belfast-English it usually reaches a plateau the tongue; this technique has to a certain extent and may rise; and whereas English, in common with replaced electromyography, which measures the elec- many languages, has a rising intonation in questions trical activity associated with muscle contraction. requiring a ‘yes—no’ answer (‘Did Marianna make the Considerable progress has recently been made with marmalade?’), the intonation of such ‘yes–no’ ques- magnetic resonance imaging (MRI), which provides tions in Greek, Hungarian, and Romanian usually falls. exceptionally clear cross-sectional images of the vocal 837 PHONETICS tract, and with ultrasonic techniques, which have been suggests that feedback may be used to develop an used for recording the movement of the vocal folds internalized model of feedback: this would allow and the tongue based on measuring reflections of high- speech production control to be a good deal more frequency sound waves between tissue and air. rapid because, instead of having to wait for actual A major issue in speech motor control is under- feedback to determine whether speech has been pro- standing a type of phonetic variation known as coar- duced as intended, speakers could calculate the feed- ticulation. This comes about because, in contrast to the back they would be expected to get and make any way in which letters are written on a page, sounds in necessary corrections before speech was actually pro- sequence overlap with, and therefore influence, each duced. other. Consider for example the production of ‘pan’: the last two phonemes /an/ are oral–nasal meaning that Speech Acoustics and Speech Perception the velum is raised for the /a/ vowel and lowered for /n/. But since the velum can not jump instantaneously This branch of phonetics is concerned with the analy- between these two states, it already starts to lower dur- sis of acoustic speech signals. Major progress in this ing the /a/ making the vowel nasalized well before the area has been possible as a result of at least three tongue tip touches the alveolar ridge for /n/. The pro- major technological developments: the invention of duction of /a/ is therefore influenced by the following the spectrograph in 1948, allowing an acoustic signal /n/: it is because of this type of coarticulation that the to be represented in terms of its time, frequency, and /a/ in ‘pan’ is phonetically very different from the /a/ amplitude components; progress in speech synthesis in other contexts such as ‘pad’. Coarticulation is ubiq- technology in the 1950s and 1960s, in which utter- uitous and, since speech sounds can be communicated ances could be generated synthetically by machine; at a faster rate if they overlap with each other in time, and advances in digital speech processing in the proponents of the motor theory of speech perception 1960s, in which computer algorithms were developed (see next section) have argued that coarticulation is for the rapid calculation of the frequency content of a necessary to ensure that speech is produced sufficient- signal and for representing acoustic speech signals in ly rapidly. Coarticulation may be anticipatory as in the terms of a small number of parameters. ‘pan’ example, in which a sound is influenced by a fol- When a speech sound is produced, an acoustic sig- lowing sound, or perseverative in which the influence nal is created whose characteristics are entirely is from a preceding sound. Various experiments have dependent on the actions of the vocal tract that give shown that it is very difficult to tie coarticulation rise to it. In articulatory-to-acoustic relationships, the definitively to any category boundary: coarticulatory aim is to break down the acoustic signal into a number influences can spread across phoneme, syllable, and of components that can be related to individual vocal word boundaries; hence, it would be quite possible for tract actions out of which the speech sounds are the velum to start lowering during the vowel of ‘saw’ formed: we would like to know, for example, how the in anticipation of the following /n/ in ‘saw another’. acoustic signal is changed by opening and closing the Explaining how listeners manage to recover the velum, protruding the lips for a rounded vowel like phonemic and linguistic content of an utterance [u], opening the mouth for [a], and so on. A major intended by the listener in the light of these ubiquitous breakthrough in our understanding of such relation- coarticulatory effects is a major research undertaking ships was in the development in the 1950s and 1960s in speech motor control and its relationship with of the acoustic theory of speech production. With such speech perception. a model, it became possible to reduce the complex Another major issue is the extent to which speech shape of the vocal tract to a relatively small number of motor control is regulated by feedback—that is, infor- parameters (for example, vowels can be modeled by mation that is relayed back to the higher centers of the the cross-sectional areas and lengths of four intercon- brain from hearing oneself speaking (auditory feed- necting cylinders) and then to use these parameters to back) and detecting the movement of, and contact predict the likely acoustic output. This type of model, between the vocal organs (tactile and proprioceptive which was developed primarily by Gunnar Fant in the feedback). Research in this area has been informed 1950s and 1960s has been central to a range of both by studying the way in which fine motor control research areas in phonetics, including synthesizing in adults who have become deaf in later life deterio- speech and predicting the types and distribution of rates, and from so-called immediate compensation vowels and consonants that are most likely to occur in experiments, in which investigators seek to determine the world’s languages. the extent to which a vocal organ like the jaw can When an acoustic signal reaches the listener, it is compensate for the lack of movement in the lips, if transformed in various ways as it passes through the these are artificially immobilized. Recent research ear and is represented by electrical impulses in the 838  PHONETICS auditory nerve. Auditory phonetics seeks to model use of prosody (in particular, syllable-structure and these types of transformations in order to obtain a word-stress) to divide the stream of speech into sepa- more accurate representation of how listeners actually rate words. Developing a model of how the available perceive an acoustic signal. Some researchers hold the information from the acoustic signal interacts with the view that the auditory transformations reduce the con- listener’s knowledge of the language is a major chal- siderable acoustic differences between male and lenge both to understanding how speech is perceived female speech. by humans and to developing computers to accom- Speech perception is concerned with how listeners plish an analogous task. retrieve phonemes and linguistic units from the acoustic signal. The application of speech synthesis to Conclusions speech perception—allowing the individual contribu- tion of acoustic cues to the perception of speech to be Phoneticians have made considerable progress in investigated—was pioneered at the Haskins understanding how abstract phonemic and prosodic Laboratories in the United States of America in the units are related to the physical characteristics of early 1950s. Their experiments led to the influential speech sounds in the production and perception of motor theory of speech perception, in which it is pro- speech in the languages of the world. The acoustic the- posed that listeners extract or decode phonemes from ory of speech production and the synthesis of intelligi- the acoustic signal by first reconstructing the speech ble speech from text are two examples of significant production strategies that could have given rise to scientific discoveries in the last 50 years that have very them: that is, listeners hear not the acoustic signal but clear practical applications in technology. the movements of the vocal tract. In Kenneth Stevens’ The dramatic improvement in the storage and lexical-access from features model by contrast, which analysis of speech data by computer in the last decade is partly based on the well-known quantal theory of has provided the tools for analyzing a far greater speech perception, it is proposed that there are land- range of speaking styles of naturally occurring spon- marks of acoustic stability that allow listeners to taneous speech. However, most of the phonetic access the mental lexicon directly without recourse to knowledge is still derived from stylized laboratory speech production. In Björn Lindblom’s hyper and speech of English and there is a great lack of compre- hypo (H&H) model of speech, the notion that listeners hensively annotated databases from other languages make use of either invariant acoustic or articulatory and in particular from endangered languages. New landmarks is rejected. Instead, speech is said to vary experimental approaches to measuring the vocal along a continuum from hypo- to hyperarticulation, organs have advanced our understanding of speech depending on the speaker’s desire to produce speech production control. There is now recorded material with a minimum of effort (hypoarticulation) and the spanning several decades, which can be used to pro- need to produce speech clearly (hyperarticulation) at vide an experimental basis to sound change. Careful e.g. points of accent and information focus (see analyses of all these new types of data will lead to above). more sophisticated models and new discoveries in Speech perception is further complicated by the phonetics in the future. way in which decoding linguistic information from the speech signal interacts with ‘top-down’ processing, References that is, with the knowledge of the language that the lis- tener brings to bear in recognizing and understanding Clark, J., and C. Yallop. 1995. An introduction to phonetics and speech. It is quite clear that listeners do not recognize phonology, 2nd edition. Oxford: Blackwell. Fant, G. 1973. Speech sounds and features. Cambridge, MA: speech by first decoding the speech signal into MIT Press. phonemes and subsequently transforming these into Johnson, K. 1997. Acoustic and auditory phonetics. Cambridge, words: a listener instead often recognizes a word well MA: Blackwells. before the acoustic signal for that word has occurred in Ladefoged, P. 2001a. A course in phonetics, 4th edition. New its entirety. Listeners can identify words that may be York: Harcourt Brace College Publishers. Ladefoged, P. 2001b. Vowels and consonants: an introduction to present only in a substantially impoverished form in the sounds of language. Oxford: Blackwell. the acoustic signal, or perhaps not even physically Ladefoged, P. 2002. Phonetic data analysis: an introduction to present at all (see the examples above). A further com- fieldwork and instrumental techniques. Oxford: Blackwell. plicating factor is that, in contrast to the presence of Ladefoged, P., and I. Maddieson. 1996. The sounds of the white spaces between printed words on the page, there World’s languages. Oxford: Blackwell. Ladd, D. 1996. Intonational phonology. Cambridge: is no direct information in the acoustic signal about Cambridge University Press. where one word ends and the next begins. Various Laver, J. 1994. Principles of phonetics. Cambridge: Cambridge experiments have suggested that listeners may make University Press. 839 PHONETICS Laver, J., and W.J. Hardcastle. 1997. The handbook of Pickett, J. 1998. The acoustics of speech communication. phonetic sciences. Oxford: Blackwell. Boston: Allyn and Bacon. Lehiste, I. 1967. Readings in acoustic phonetics. Cambridge, JONATHAN HARRINGTON Ma: MIT Press. ———. 1970. Suprasegmentals. Cambridge, MA: MIT Press. See also Spectral Analysis Phonology The aim of phonology is to examine the way sounds substitution of one allophone for another one only are organized in languages and to explain the varia- results in a different pronunciation of the same word. tions that occur. While it is physically possible to pro- Phonemes are the ‘contrastive’ sounds of a language: /c/ duce a wide range of sounds, only a relatively small and /s/ are different phonemes, because ‘cat’ and ‘sat’ number of these are used in a language. For example, contrast in meaning. Allophones are the phonetic varia- a nasal preceding a plosive, such as in mbeke does not tions of distinctively used sounds. Sounds that are not occur in English at the beginning of a word, but it does distinctive are also known as ‘redundant’. In English, in several African languages. This is not because aspiration is a redundant feature. However, in Thai, English speakers cannot produce the specific sequence aspiration is distinctive, as the use of aspiration affects of sounds, but because of the way speech sounds in a the meaning of the utterance. By substituting one sound particular language are organized. for the other (the ‘commutation test’), it is possible to In phonology, the most characteristic properties of determine the phonemes of a language. Pairs of words different sounds, known as features, are compared to that differ by just one sound, such as ‘cat’ and ‘sat’ or develop rules underlying the use of sounds in groups of ‘robe’ and ‘roam’, are called minimal pairs. languages. Systematic surveys of a representative num- Other criteria also help to determine whether ber of languages are necessary to be able to generalize speech sounds belong to the same phoneme or not. about sound systems, and to relate the findings to other One is complementary distribution, which refers to the areas of language (e.g. syntax and morphology). The situation where two sounds should not occur in the UPSID database (UCLA Phonological Segment same environment. For instance, in English, aspirated Inventory Database), developed by Ian Maddieson and plosives occur at the beginning of words, but not in colleagues in the 1980s, contains inventories of over consonant clusters. It can be predicted which allo- 450 languages of the world. This database, as well as phone is pronounced in which context. Where we find the Stanford Phonology Archive, has become a popular the one, we do not find the other: they are mutually tool in teaching and in phonological and phonetic exclusive, never occurring in the same phonetic envi- research. A detailed description of a language is not ronment. Another criterion is free variation: sounds only of interest for the sake of having a survey of dif- that occur in the same place in a word can belong to ferent languages of the world but is also essential to be the same phoneme only if they do not change the able to teach foreign learners a language, or to treat meaning of the word. For instance, substitution of children who are not acquiring the sound system of glottal stop [] for [t] as in ‘butter’ does not change the their language properly, or people who have lost the meaning of the word. Therefore, /buer/ and /butter/ capacity to speak through injury or illness. In addition, are not minimal pairs. Also, sounds ought to display a speech technology communication systems require reasonable amount of physical similarity to belong to knowledge of the structure of a language. the same phoneme. In summary, when two sounds are in complementa- ry distribution or free variation, they are allophones of Phonemes and Allophones the same phoneme. A classical approach in phonology is to begin by establishing the phonemic system of a language, i.e. Distinctive Features to determine which sounds are phonemes and which are allophones. Substitution of one phoneme for A phoneme can be described by several articulatory another will result in a word with a different meaning; and acoustical features. One of the main aims of 840  PHONOLOGY phonology is to identify the set of distinctive features and this should be reflected in the phonological rep- required to describe the sounds of a language. resentation. This is possible if it is assumed that each Distinctive features serve to distinguish one phoneme distinctive feature is free to act independent of the from another. It is important to describe these detailed other features it may be associated with. This aspects of speech sounds, in order to understand how approach consisting of ‘interconnecting levels’ sets of sounds are related. (tiers) is known as autosegmental phonology: each Tables 1 and 2 list some distinctive features (not feature can influence a neighboring sound segment, exhaustive) of some English vowels and consonants. If irrespective of whether other associated features do a feature is present, it is marked with a () sign; if it so. In particular, feature geometry describes the is not present, it is marked with a minus () sign. (nonlinear) action of one feature on another by rep- The features ‘front’, ‘back’, ‘high’, ‘low’, and resenting their relationship as hierarchical (tree) ‘round’ refer to the position of the tongue in the structures. vocal tract. For front sounds, the body of the tongue is fronted re its neutral position (for the /ə /, a Rules schwa), for back sounds it is retracted re its neutral position. For high sounds, the body of the tongue is Phonological rules make statements about which allo- raised re the / ə /, while for low sounds it is lowered phones of a phoneme will occur in a specific context. re the / ə /. Rounded sounds are produced with pro- The change of one item to another item in a certain truding lips. environment can be stated by a rule. The validity of the Voicing refers to the presence or absence of vocal rule can then be tested against other examples to deter- fold vibration. The feature ‘continuant’ distinguishes mine exceptions between sounds within and across between stops (nasal stops included) and other languages. Rules are important for discovering the sounds. Anterior sounds are produced with the tongue universal principles governing the use of sounds in tip at or before the alveolar ridge in the vocal tract, languages. It is necessary to specify the item(s) affect- and coronal sounds are produced with the tongue tip ed, the change(s) that takes place (indicated by ‘→’), or blade raised (it includes some palatal consonants). and the environment in which the change occurs (indi- Strident sounds are fricatives or affricates (stop  cated by ‘/’). In linear rule writing, the original state of fricative) with high-frequency noise (a hissing affairs is given on the left of the arrow and the struc- sound). Nasal sounds are produced with the velum tural change and environment on the right of the arrow. lowered, resulting in airflow through the oral and For example, in English, vowels are nasalized before a nasal cavities. nasal stop. This can be stated as follows: The notion of distinctive features is not only rele- ~ ]/ ___ /n/ e.g. /æ/→ [æ vant for analysis purposes but also for the descrip- tion of phonological processes. Some phonological It is also possible to use brace notation to describe one processes, such as for instance assimilation, affect process in two different environments. The following certain combinations of features more than others, example states that /t/ can be pronounced as a glottal TABLE 1 Possible Distinctive Features of Some English Vowels (Not Exhaustive) u ɔ o ɑ  % e ε i i ə Front            Back            High            Low            Round            TABLE 2 Possible Distinctive Features of Some English Consonants (Not Exhaustive) p t k b d g m n f s ɵ ʃ v z ð h l r w j Voice                     Continuant                    Strident                     Anterior                     Coronal                     Nasal                    841 PHONOLOGY stop before a consonant or the end of a word (such as Optimality Theory in ‘button’). Until now, phonological processes have been /t / → []/ —– C # addressed by means of rules and derivations (deriva- tional Generative Phonology). Application of one rule often affects the subsequent application of some other By using parentheses and braces it is possible to for- rule. However, during the last decade, Optimality the- mulate rules of greater complexity. Moreover, the ory has obtained a dominant role in phonology. This Greek alphabet, or alpha notation, can be used to approach has introduced an alternative way of model- match features in different places in the rule. ing the relationship between words and sounds. While some rules state how features are affected by Rather than trying to define how a language′s the context of other features (feature-changing rules, words may be derived by combining given phonemes such as nasalization, glottalization, flapping), other according to certain rules, Optimality theory rules can affect the entire segment. For example, in approaches the problem from the opposite direction: English a schwa is inserted between a final liquid and in principle, any combination of features/sounds is a nasal, as in [filəm] (instead of /film/). This example assumed to be possible, but each language imposes of insertion can be described as: certain constraints in a language-specific way on the  endless possibilities.  cons  cons nas  ∅ → ə/ son The constraints are assumed to be universal while —– # -nas they may be ranked differently, which allows for lan- guage-specific grammars. While phonology attempts to model knowledge of a where ‘cons’ refers to consonantal sounds, ‘son’ to language, it should also reflect the fact that languages sonorant sounds, i.e. sounds that contain considerable are learnable. Proponents of Optimality theory argue energy (such as /m,l,j/), and ‘nas’ to nasal sounds. that apart from dealing with phonological phenomena, Given the presence of certain features, one can pre- it also includes a better learning theory than tradition- dict the value of other features in that segment. For al (derivational) approaches. In addition, Optimality example, there are no rounded front vowels in English. theory has already proven to be a productive new tool If a vowel is specified as [-back] it is also [-round]. It in other domains of linguistic analysis, such as mor- is therefore not necessary to specify [-round]. A blank phology, syntax, and even semantics. indicates that the feature is predictable by a phonolog- ical rule of the language. Similarly, if a phoneme is [nasal] it is also [voiced] in English. Sometimes, References nasal phonemes are ‘devoiced’ after an initial /s/ Anderson, Stephen. 1974. The organization of phonology. New [snu' p]. However, this does not affect the meaning of York: Academic Press. the word, as is the case in Burmese where /ma/ means Chomsky, Noam, and Morris Halle. 1968. The sound pattern of ‘health’ and /ma' / means ‘order’ (therefore, they are English. New York: Harper & Row; Paperback edition 1991, Cambridge, MA: MIT Press. different phonemes). Clements, George N. 1992. Phonological primes: features or gestures?. Phonetica 49. 181–93. Durand, Jacques, and Francis Katamba. 1995. Frontier of Suprasegmentals phonology. London: Longman. Until now, we have considered the individual phonemes Goldsmith, John. 1990. Autosegmental and metrical phonology. Oxford: Blackwell. of a sound system. However, several phonological Goldsmith, John. 1995. The handbook of phonological theory. processes affect units that are larger than phonemes, Oxford: Blackwell. such as syllables, words, phrases, and sentences. In Jakobson, Roman. 1990. On language. Cambridge, MA: some languages, words contain (with some exceptions) Harvard University Press. only front or back vowels (e.g. Turkish). The analysis of Kager, R. 1999. Optimality theory. Cambridge: Cambridge University Press. phonological features in terms of units larger than a Ladefoged, Peter, and Ian Maddieson. 1996. The sounds of the phoneme is dealt with by prosodic phonology, and World’s languages. Oxford: Blackwell. autosegmental phonology. Another branch of phonolo- Maddieson, Ian. 1984. Patterns of sounds. Cambridge: gy, suprasegmental phonology, deals with aspects of Cambridge University Press. pitch, loudness, tempo, rhythm, and tone. A phonologi- Roca, Iggy. 1994. Generative phonology. London: Routledge. Spencer, Andrew. 1996. Phonology: theory and description. cal theory concerned with organizing phonemes into Oxford: Blackwell. groups of relative prominence that emphasizes the rela- Vihman, Marilyn. 1996. Phonological developments: the tionship between phonemes and rhythm, and intona- origins of language in the child. Oxford: Blackwell. tional stress is known as metrical phonology. ASTRID VAN WIERINGEN 842  PHRASE STRUCTURE Phrase Structure Sentences are not just linear strings of words. There is in (2a). Again, this is not possible for likes the films, as ample evidence that words in a sentence can form a shown by (2b). Replacement thus gives the same indi- unit that excludes other words in the same sentence. cation of constituency as does redistribution. Such units are referred to as constituents. Constituents (2a) What does John like? (Answer: The films by are hierarchically organized, forming larger phrases. P.P. Pasolini). For example, a sentence will often consist of a subject (2b) *What does John by Pierre Paolo Pasolini? and a predicate. The constituent that forms the predi- (Answer: Like the films). cate can in turn consist of several smaller constituents, such as a verb, a direct object, an adverbial, and so on. The smallest constituents of a sentence are words, These constituents, too, may be complex and contain which are further organized into phrases, which in turn even smaller constituents. The internal structure of are further grouped into larger phrases, and so on. sentences and smaller constituents is generally Phrasal constituents are built around a word that referred to as ‘phrase structure’. functions as the head of the phrase. The head of a Evidence for the assumption that the internal organ- phrase is that word in the phrase which in a sense is its ization of a sentence consists of hierarchically ordered most important part: it cannot be omitted, and it deter- constituents is manifold. One piece of evidence con- mines most of the syntactic properties of the phrase as cerns the fact that the words in a sentence can be redis- a whole. For example, subjects of sentences are very tributed in certain ways such that the result is again a often phrases that are built around a noun. The phrase grammatical sentence. In such a redistribution process, can consist of just this noun, as in (3a), but more mate- certain words stay together. This indicates that they rial can be added, as in (3b), where the head music is form a unit—a constituent. For instance, in the sen- modified by an adjective preceding it, and (3c) where tence in (1a), the phrase the films by Pierre Paolo a relative clause following the noun is added. In this Pasolini functions as the direct object of the verb likes: way, arbitrarily large phrases can be formed. The head it is the thing that is liked. It thus forms a semantic cannot be left out, however, as illustrated in (3d). unit. That it also forms a syntactic unit is indicated by (3a) [Music]NP is his favorite pastime. (1b): the phrase can be placed at the beginning of the (3b) [Classical music]NP is his favorite pastime. sentence as a whole (a process technically known as (3c) [Classical music written in the twentieth ‘topicalization’) and the result is still grammatical. century]NP is his favorite pastime. This cannot be done with just any string of words that (3d) *[Classical written in the twentieth century]NP happen to stand next to each other. This is shown by is his favorite pastime. (1c), where the attempt to topicalize likes the films, to the exclusion of the rest of the direct object, results in If a phrase is built around a noun, the resulting ungrammaticality. This shows that this string does not structure is a noun phrase, abbreviated as NP. Any form a syntactic unit on which syntactic rules such as phrase headed by a noun has a syntactic distribution topicalization can operate. Thus, likes the films is not a that is determined by its head. For instance, phrases constituent of the sentence in (1a). (Ungrammatical headed by a noun may appear in subject, direct object, examples are marked by an asterisk, following the indirect object positions, etc. In contrast, phrases common convention in the literature). headed by an adjective (APs) cannot normally func- tion as such. (1a) John likes the films by Pierre Paolo Pasolini. (1b) The films by Pierre Paolo Pasolini, John likes. (4a) [The girl]NP gave [the boy]NP [a new cd]NP. (1c) *Likes the films, John by Pierre Paolo (4b) *[Pretty]AP gave the man a new cd. Pasolini. Adjectival phrases, on the other hand, can act as Another indication that phrases are structured and modifiers to nouns (e.g. the AP classical modifies the consist of constituents is that certain strings of words noun music within the NP classical music in (3b), can be replaced by one word, whereas others cannot. whereas NPs do not ordinarily appear in this position In (1a), the films by Pierre Paolo Pasolini can be (bar exceptional cases like a London bus). The distri- replaced by the question word what (which in English bution of phrases headed by a verb (VPs such as cook has to be placed at the beginning of the sentence), as pasta or like music), phrases headed by prepositions 843 PHRASE STRUCTURE (PPs such as in the cupboard and under the stairs), and rule in (8c), which combines a preposition and its phrases headed by an adverb (AdvPs such as very complement. quickly or yesterday) is yet different. (8a) NP → Det (Adj) N (PP) Depending on the properties of the head itself, other (8b) VP → V NP (Adv) material within a phrase besides the head can also be (8c) PP → P NP obligatory. Certain lexical items require additional ele- ments to be present in a phrase that they are the head It is worth noting that the NP mentioned in (8c) can of. This can be due, at least in part, to their meaning. be rewritten in accordance with (8a), meaning that it For instance, the verb like has to combine with an can contain a PP. This PP can be rewritten again in object (5a). Omission of this so-called complement to accordance with (8c), so that it contains an NP, which the verbal head leads to ungrammaticality (5b) in this can be rewritten yet again in accordance with (8a), and case. However, not all complements are obligatory. so on, ad infinitum. This results in sequences like in The verb eat can combine with an object (5c), but need (9). In principle, there is no limit to such sequences, not necessarily do so (5d). In the latter case, an object although for practical purposes they will stop at some is implicitly understood. point. This shows a pervasive property of the phrase structure of natural languages: it is recursive. (5a) John likes music. (5b) *John likes. (9) I saw the mouse in the hole in the wall of the (5c) John is eating plums. house next to the river beside the meadow in (5d) John is eating. the county next to … Apart from heads and complements, phrases may Sentences are formed by combinations of several contain additional, purely optional, material. Such phrases, in the simplest of cases an NP that functions optional elements are referred to as adjuncts. In VPs, as the subject of the sentence and a verb phrase that they are most often adverbials such as time, place, and functions as the predicate (as for instance in [[John]NP manner adverbials; in NPs, modifying relative clauses, [likes Mary]VP]S). This is expressed by the rewrite rule for example, can act as adjuncts. Adjuncts differ from in (10). heads and complements in not being unique in a (10) S → NP VP phrase. Whereas a phrase can only have one head, and (barring some problematic cases) also only one com- Rewrite rules were replaced in favor of a theory of plement, adjuncts can be freely stacked. This is illus- phrase structure, which proved to be very influential. trated in (6). This theory is known as X’-theory (pronounced as X bar theory). It was first proposed in Chomsky (1970), (6) John often walks his dog in the park, in the and further elaborated in such works as Jackendoff spring, on a nice Sunday afternoon, while (1977), Fukui (1986), and Speas (1990). It originated whistling The blue Danube. from the idea that all phrases, regardless of the catego- An early formalism of expressing the possible con- rial status of their head (i.e. whether they are NPs, VPs, stituent structures of sentences made use of so-called PPs, AdjPs, or AdvPs), are built along the same struc- rewrite rules (also known as phrase structure rules; see tural schema, given by the tree diagram in Figure 1. Chomsky 1957, 1965). They have the general form of X is a variable standing for a head of any category (7), where XP stands for a phrase of any category, and (N, V, P, A, Adv), YP is the complement of the head, Y, Z, and W for its constituent parts. The rule in (7) while ZP is called its specifier. X’ and X” are the pro- states that the phrase XP consists of Y, Z, and W (in jections of the head X. When the head X combines technical parlance, XP is said to be rewritten as Y Z with the complement YP, it projects up to the X’ (pro- W). Y Z W can be words, but can also be phrases nounced X bar) level. The X’ constituent further com- themselves. bines with a specifier leading to the projection of the X” (pronounced X double bar) level. In the classical (7) XP → Y Z W Simplifying somewhat, the structure of NPs can be X" (=XP) expressed by the rewrite rule (8a). Rule (8a) states that NPs consist of a determiner, optionally an adjective, the head noun, and optionally a PP. The structure of a ZP X' VP can be given by the rule in (8b), which states that a VP consists of a verb, an NP (functioning as an X YP object), and (optional) adverbs. The structure of a PP can be represented as given by the phrase structure Figure 1 844  PHRASE STRUCTURE version of the theory, the X” level is considered to be daughter of X” and sister of X’, and the complement the highest level up to which the head can project. It is can be defined as the immediate daughter of X’ and therefore also referred to as the XP level, designating sister of X. a full phrase. The schema in Figure 1 cannot accommodate all Examples of how phrases of various types fit into the possible constituent parts of phrases. In particular, the general scheme in Figure 1 are given in Figure 2. it cannot accommodate the adjuncts in a phrase, of Like the complement and the head, the specifier is which there can be arbitrarily many (see above). To unique. For example, while a noun can take a posses- accommodate adjuncts, it is assumed that bar levels sive phrase as specifier, and can take a determiner as can be reiterated. When an adjunct is added to a specifier, it cannot take both: phrase, the head does not project to a higher bar level; instead, the bar level is just repeated. In this way, an (11a) Mary’s copy of Dracula adjunct that is added at the double-bar level can be for- (11b) a copy of Dracula mally distinguished from the (unique) specifier of the (11c) *a Mary’s / Mary’s a copy of Dracula head: whereas both the adjunct and the specifier are There are several structural relations that can be daughters of an X”-node, only the specifier is the sis- defined over the tree diagram in Figure 1. Any two ter of an X’-node; the adjunct is the sister of an X”- constituents that are immediately dominated by the node (Figure 3). same node are called sisters. X and YP are sisters, Because the specifier can also be distinguished being immediately dominated by the X’ node, as are from adjuncts by the very fact that adjuncts can be X’ and ZP, the latter two being immediately dominat- stacked, whereas the specifier is unique, it is not clear ed by the XP node. The XP node is said to be the that it is essential to distinguish between the two for- mother of X’ and ZP, while the X’ node is similarly the mally as well (see below). mother of the X and YP nodes. The specifier of the An assumption built into most modern phrase struc- head X can therefore be defined as the immediate ture theories is that all nodes are maximally binary branching. This means that no node can have more VP than two daughters. If this assumption is adhered to, ‘flat’ structures of the type in Figure 4, for a VP that NP V′ contains both a direct object and an indirect object, are impossible. His brother V NP The consequence is that there must be a more hier- archical structure inside VP. Evidence for this is said (a) loves ice-skating to come from examples like (12a, b). These show that PP the indirect object can act as antecedent for a reflexive element (for instance, himself) inside the direct object, AdvP P′ X" right P NP (b) across the bridge WP X" NP YP X' AP N′ beautiful N PP X ZP (c) pictures of Amsterdam Figure 3 AP VP AdvP A′ V NP NP very A PP (d) fond of music give mary a book Figure 2 Figure 4 845 PHRASE STRUCTURE but not the other way around. Evidence from other DP types of sentences shows that the antecedent for the reflexive has to be higher in the structure than the D′ reflexive itself. The data in (12a, b) therefore indicate that the indirect object and the direct object are not on the same level in the structure. Instead, the indirect D NP object must be higher than the direct object. (12a) Mary showed Bill a double of himself. the house (12b) *Mary showed a double of himself Bill. Figure 5 The scheme in Figure 1 is not intended to express a linear order of constituents. The theory of phrase IP structure in the first instance is about the hierarchical relationship between phrases. Different principles DP I' must then determine whether a complement precedes or follows the head, as well as whether the specifier John I VP precedes or follows the head. Syntactic theory distinguishes between two classes is V' of words: lexical and function words. The position of function words in phrase structure theory has always V DP been rather unclear. Recently, such elements have come to be regarded by some as heads of phrases, on par with lexical categories. In line with this develop- singing a song ment, items like determiners (such as the and a in Figure 6 English), complementizers (such as that, if, whether), auxiliary verbs and modals (such as to be, can, will), and even inflectional affixes for tense and agreement CP (such as the English –ed suffix in past tense forms) are C' regarded as the heads of phrases. This view has vari- C IP ous consequences for how the internal structure of if DP I' phrases is supposed to be built up. For instance, a phrase like the house is no longer seen as an NP, but john I VP rather as a determiner phrase (DP), headed by the is V' determiner D, which takes the NP house as its com- plement, as represented in Figure 5. V DP Similarly, sentences are now regarded as projec- playing the cello ... tions of the inflection we see on the verb, and hence as inflectional phrases or IPs (Figure 6). Subjects are in Figure 7 the specifier position of IP. In the same vein, sentences introduced by a complementizer are regarded as com- Classical X’-theory not only needs the possibility of plementizer phrases or CPs. The complementizer head repeating bar levels to accommodate all the possible in the CP takes an IP as its complement, which in turn constituents of a phrase, but it suffers from the opposite takes a VP as its complement (Figure 7). The specifier problem as well: it sometimes forces more syntactic position of CP can also host elements in some sen- positions to be present in the structure than are actually tences, such as question words. necessary. For instance, a head can take a specifier while The proliferation of functional categories in the the- not taking a complement. Whereas (13b) is as well oretical inventory also has consequences for the notion formed as (13a) is, in classical X’-theory there neverthe- of adjunction. With every additional functional projec- less is a complement position present in (13b) as well. tion, an additional specifier position becomes avail- (13a) John’s collection of mushrooms able. This means that it is at least possible to discard (13b) John’s collection the assumption that bar levels have to be repeated in order to accommodate adjuncts. Instead, every adjunct It is even possible that a phrase contains just its may be regarded as a specifier of a particular func- head, without a specifier or a complement, as for tional projection, as extensively argued in Cinque instance in (14b) (vs. (14a), with specifier and com- (1999). plement present). If phrases always have the form of 846  PIDGINS AND CREOLES Figure 1, then in such cases there is both a specifier Cinque, Guglielmo. 1999. Adverbs and functional categories: a position and a complement position, neither of them cross-linguistic perspective. New York, Oxford: Oxford University Press. containing any lexical material. Chomsky, Noam. 1957. Syntactic structures. The Hague: (14a) (I expect) [Mary to win the race] Mouton de Gruyter. ———. 1965. Aspects of the theory of syntax. Cambridge, (14b) (I expect) [to win] MA: MIT Press. Partly because of this inelegant aspect of X’-theory, ———. 1970. Remarks on nominalization. Readings in English transformational grammar, ed. by Roderick A. a more flexible view of phrase structure has been adopt- Jacobs and Peter S. Rosenbaum. Waltham, MA: Ginn. ed (see Speas 1986; Chomsky 1994), usually referred to ———. 1994. Bare phrase structure. Cambridge, MA: as Bare Phrase Structure theory. Its basic assumption is Distributed by MIT Working Papers in Linguistics. that a head projects as often as it combines with anoth- Fukui, Naoki. 1986. Theory of projection in syntax. Stanford, er phrase, no more and no less. One consequence of this CA: CSLI Publications, Center for the Study of Language and Information, and Tokyo: Kurosio Publishers. is that it is no longer possible to distinguish between Grimshaw, Jane. 2003. Words and structure. Chicago: specifiers and adjuncts in structural terms (see also the University of Chicago Press. previous point on functional structure). Instead, the Jackendoff, Ray. 1977. X’ Syntax: a study of phrase structure. specifier can be distinguished relationally, since it Cambridge, MA: MIT Press. establishes a unique relation with a head. Kayne, Richard S. 1984. Connectedness and binary branching. Dordrecht: Foris. Pollock, Jean-Yves. 1989. Verb movement, universal grammar, and the structure of IP. Linguistic Inquiry 20. 365–424. References Speas, Margaret. 1986. Adjunctions and projections in syntax. Abney, Stephen. 1987. The English noun phrase in its sentential Ph.D. Dissertation. Cambridge, MA: MIT. aspect. Ph.D. Dissertation. Cambridge, MA: MIT. AMELA CAMDZIC AND PETER ACKEMA Pidgins and Creoles Although until recently creolists were more or less between a ‘true’ creole and its related ‘lexical donor agreed on the definition of ‘pidgin’ and ‘creole’, this is language’ (also called ‘lexifier language’, ‘superstrate no longer so. Also, the adage that ‘a creole is a nativized language’, or simply ‘superstrate’). The total number pidgin’ (Hall 1966) is no longer universally accepted. It of speakers, including those who only use a creole as a has been observed, for example, that at least some of the second language (a pidgin is a second language by def- French-lexicon creoles do not have a pidgin ancestry inition), may be estimated at around 100 million. By (Chaudenson 2001). Further, the more we learn about far, the largest group, especially among the creoles, is the transition from pidgin to creole, the less certain we formed by those that have derived the bulk of their lex- are about where to draw the line between the two (Baker icon from an Indo-European language such as 1995). At the same time, creolists are becoming aware Portuguese, French, English, Spanish, and Dutch. It is that it may not be justified to group pidgins and creoles no coincidence, of course, that these are the languages together, even though in the past they have often been spoken by the European nations that played the leading treated as if they belong to one category. In spite of all role in the European expansion. It was the contact these difficulties, some definitions will be presented between the European languages of the explorers and here, if only to give the reader an idea of the distinc- colonizers, on the one hand, and the non-European lan- tions. A pidgin is a nonprimary language that is the guages of the people with whom they came into con- result of language contact. A creole is a primary lan- tact, on the other (referred to as ‘substrate language(s)’, guage that is the result of language contact. or simply ‘substrate’), that gave rise to the emergence Before explaining these definitions, a brief survey of of these pidgins and creoles. Other Indo-European lan- the pidgins and creoles of the world is useful. guages, apart from the ones mentioned above, that have According to the most extensive survey that has been played a role in the emergence of pidgins and creoles made until now (Smith 1995), the total number of pid- include German, Russian, Italian, and Hindustani. gins and creoles is around 350, including both the ones Although these European-lexicon pidgins and cre- that are extinct as well as varieties that are intermediate oles have attracted the lion’s share of the attention of 847 PIDGINS AND CREOLES linguists, there is a significant and linguistically just as pidgin or a creole comes into being). This means, first, important group of pidgins and creoles whose lexicons that there will be no pidginization or creolization with- are based on non-Indo-European languages. The lexifi- out contact, and, second, that contact does not automati- er languages in this group include Arabic, Japanese, cally lead to pidginization or creolization. When we look Chinese, and a number of Bantu, Austronesian, at pidginization and creolization as language contact Australian, Papuan, and Amerindian languages. An phenomena, the following questions present themselves: important difference with the group of European-lexi- (1) What other types of language contact are there, con pidgins and creoles is that in many cases no Indo- apart from pidginization and creolization? European language was involved in their formation. (2) In what ways do pidginization and creoliza- Because of this, the study of non-European pidgins and tion differ from other types of language con- creoles may serve to correct any biases that may have tact? resulted from the focus on European pidgins and cre- (3) What are the factors that decide whether lan- oles that has characterized the field of pidgin and cre- guage contact will lead to pidginization or ole linguistics for a long time (see Thomason 1997, creolization? 2001). An important question in this regard is whether the features of pidgins and creoles that are sometimes These three questions will guide the remainder of regarded as universal might not be an artifact, resulting the discussion. from the fact that all the major lexifier languages are What other types of language contact are there, structurally similar, belonging as they do to only two apart from pidginization and creolization? branches (Germanic, Romance) within one and the The linguistic effects of language contact are same language family, Indo-European. To get an idea dependent on a number of factors, the most important of the variety as well as the geographical distribution of of which is whether or not imperfect language learn- the world’s pidgins and creoles, the maps in Holm ing plays a role in the contact situation. On the basis of (2000) are a good starting point. As far as the Pacific is this, two types of language contact can be distin- concerned, much more detailed information can be guished (Thomason 2001): gleaned from the splendid three-volume language atlas (a) Shift-induced interference. When people give edited by Wurm et al. (1996). Unfortunately, a similar up their native language while adopting anoth- work for the Atlantic region is not (yet) available. er one, the new language often displays fea- As mentioned earlier, the definitions given above tures from their native language. In this case, require some comments. First of all, what is meant by the changes in the affected language are a ‘primary’ and ‘nonprimary’ language? If you are mul- result of imperfect learning. tilingual, your primary language is the one with which (b) Borrowing. When people maintain their native you feel most at home, the one you use most often, language, it may still be influenced by anoth- even though it may not be your native language. When er language. This happens when native speak- we say that a pidgin is a nonprimary language, it means ers incorporate features from another that pidgin-speakers have at least one other language in language into their native language. In this their repertoire—their primary language— and that the case, the changes in the affected language are pidgin is used only as an auxiliary language, when not a result of imperfect learning. communicating with speakers with whom they have no other language in common. Creoles, on the other hand, The crucial difference between the two is that in the are primary languages by definition, although, again, case of shift-induced interference, speakers introduce this does not mean that they are necessarily native lan- changes in a language that is not their native language, guages. Second, if pidgins are nonprimary languages, while in the case of borrowing they introduce changes is it justified to view them as ‘true’ languages in the in a language that is their native language. Although first place? Although this is an interesting question in the end-results of these two types of language contact itself (when is something a language?), it is too com- are sometimes hard to distinguish, the processes by plex to be dealt with here. For the purpose of this essay, which these results come about are very different, for however, pidgins are referred to as languages. As for example, in the order in which the different compo- creoles, there is no question about it: they are lan- nents (lexicon, phonology, etc) of the language system guages just like any other. are affected. While borrowing always starts with The one thing that pidgins and creoles have in com- words and only involves the grammatical system later, mon, according to the definitions given above, is that it is the other way around in shift-induced interference. they are a product of language contact. Language con- The results of the two processes may range from tact, therefore, is the key concept in these definitions: it minor changes in the lexicon, such as adoption of a is a necessary, although not sufficient, condition for loanword, to major structural changes, such as pidginization and creolization (the process by which a changes in basic word order. 848  PIDGINS AND CREOLES In what ways do pidginization and creolization dif- it seems fair to say that they all share the concept of fer from other language contact phenomena? autonomicity as a basic feature of pidgins and creoles. Pidginization and creolization differ from other lan- What are the factors that decide whether language guage contact phenomena (except, perhaps, so-called contact will lead to pidginization or creolization? ‘mixed languages’; see relevant lemma) in that neither This is one of the fundamental questions regarding pidgins nor creoles can be justifiably viewed as changed pidginization and creolization. Since little attention has versions of the languages in contact, while this is the been devoted to it, especially from a historical point of case with borrowing and shift-induced interference. In view, the discussion here is necessarily tentative. It other words, pidgins and creoles are autonomous lan- should also be noted that what follows is restricted to guages. This entails that they have a grammar of their European-lexicon pidgins and creoles (by far the largest own, one that is not derivative of or dependent on a pre- group), and that it is not necessarily valid for pidgins existing language. The autonomy of the grammar is the and creoles that are lexically based on other languages. result of a process whereby grammatical ‘components’ It is hoped, however, that insights into the conditions for (rules, patterns, processes) that were initially related to the genesis of the European-lexicon pidgins and creoles one or more of the contributing languages, have become will also shed some light on the other cases. independent. The transition from dependency to inde- The European pidgins and creoles both arose in the pendency occurs when these components start interact- historical context of the European expansion. This par- ing with one another to produce a new grammar, rather ticular phase in history began around 1430 when than forming an incoherent collection of elements Prince Henry (‘the Navigator’) started sending out derived from other languages. Portuguese ships to explore the northwestern coast of The idea of pidgins and, especially, creoles as Africa. Although the exact date when the contacts autonomous languages is shared by most creolists, between Europeans and Africans for the first time led including those who view creolization in terms of to the emergence of a pidgin or creole language in not either borrowing or shift-induced interference. The known, it is clear that the genesis of the European-lex- view of creolization as a case of (extreme) borrowing icon pidgins and creoles took place in the context of is held by Lefebvre (1998), who sees Haitian Creole as the European expansion, whether in Africa, America, the result of a process of relexification, whereby words Asia, or the Pacific. The age of European expansion from the slaves’ native language(s) were replaced by was a very special, perhaps even unique, phase in the similar words from French, while the grammatical history of the world, which may be characterized by structure of the native language(s) was maintained. A the following features: very different view is represented by Chaudenson (a) The introduction of trade and production on a (2001), who believes that—especially the French-lex- truly global scale. icon—creoles are the result of a process of language (b) The presence of contact situations character- shift (from the slaves’ native languages to French), ized by unusually wide social, cultural, and whereby some features may have been introduced as a psychological gaps between the parties result of interference. No matter how far apart, both involved (Europeans on the one hand and views agree that creoles are autonomous languages, Africans, Asians, Native Americans, and not modified versions of one or more of the contribut- Polynesians on the other). ing languages. The idea of creoles as autonomous lan- (c) The use of forced labor and forced displace- guages is also inherent in Thomason’s (2001) view of ment of large numbers of people (slave trade, pidginization and creolization as cases of ‘abnormal’ indentured labor). language transmission. While in ‘normal’ transmission (d) The creation of a new type of society known as both the lexicon and the grammatical apparatus are ‘the plantation complex’ (Curtin 1998). transmitted from one and the same ‘parent language’ to one or more ‘daughter languages,’ in cases of With regard to the last feature, it should be added ‘abnormal’ transmission, there are at least two ‘par- that, although plantations in the strict sense were not ents’, one for the lexicon and one for the grammatical present in every single situation where pidgins and system. As a result of this, a pidgin or a creole cannot creoles developed, the concept of ‘plantation complex’ be a variety of either one of the parent languages; may be interpreted in a sufficiently wide sense to therefore, it is an autonomous system. Finally, in gen- encompass the latter situations as well. erative approaches to pidginization and creolization Although the exact relationship between the pres- (e.g. DeGraff 1999), the question of autonomy hardly ence of particular external conditions on the one hand arises, as every language—whether pidgin, creole, or and the emergence of pidgins and creoles on the other is something else—is seen as an instantiation of not very clear, a few things may still be said about it. For Universal Grammar. Although this brief discussion example, there is the issue of access to the target lan- does not exhaust the list of theories of creole genesis, guage. With the increasing disproportion between the 849 PIDGINS AND CREOLES European and non-European segments of the popula- ———. 1995. Some developmental inferences from historical tion and the concomitant change from the small-scale studies of pidgins and creoles. The early stages of creoliza- tion, ed. by Jacques Arend, 1–24. Amsterdam, Philadelphia, ‘homestead society’ to the large-scale ‘plantation socie- PA: Benjamins. ty’ (Chaudenson 2001), access of the non-Europeans to Chaudenson, Robert. 2001. Creolization of language and cul- the European language decreased. However, this cannot ture. Revised in collaboration with Salikoko Mufwene be the whole story because there are a number of cases, [translation of Des îles, des hommes, des langues. Paris: especially in Spanish colonies, where such dispropor- L’Harmattan. 1992]. London and New York: Routledge. Curtin, Philip. 1998. The rise and fall of the plantation complex, tion did not lead to pidginization or creolization 2nd edition. Cambridge: Cambridge University Press. (McWhorter 2000). Other factors, such as lack of moti- DeGraff, Michel, (ed.) 1999. Language creation and language vation on the part of the enslaved to learn the language change: Creolization, diachrony, and development. of their masters (Baker 1990), may have been involved Cambridge MA: MIT Press. as well. However, there is another, more elusive, factor Hall, Robert. 1966. Pidgin and creole languages. Ithaca, NY: Cornell University Press. that has received little attention until now. If we look at Holm, John. 2000. An introduction to pidgins and creoles. the period of the European expansion from a contem- Cambridge: Cambridge University Press. poraneous rather than from a modern perspective, we Lefebvre, Claire. 1998. Creole genesis and the acquisition of cannot help but recognize that this unique phase in his- grammar: the case of Haitian Creole. Cambridge: tory confronted all those who were involved in it with Cambridge University Press. McWhorter, John. 2000. The missing Spanish creoles: an entirely ‘new world’. Never before had so many dif- Recovering the birth of plantation contact languages. ferent parts of the world been in contact on such a large Berkeley CA: University of California Press. and intensive scale. Although the fact that this new con- Smith, Norval. 1995. An annotated list of creoles, pidgins, and tact situation was accompanied by a true ‘explosion’ of mixed languages. ed. by Jacques Arends, Pieter Muysken, new languages does not necessarily mean that the one and Norval Smith, Pidgins and creoles: an introduction, 331–74. Amsterdam, and Philadelphia, PA: Benjamins. was caused by the other, the synchronicity of the two Thomason 1997. Thomason, Sarah, (ed.). Contact languages: ‘events’ is striking. It is hoped that future research, A wider perspective. Amsterdam/Philadelphia PA: Benjamins. especially into the historical dimension of pidginization Thomason, Sarah. 2001. Language contact: an introduction. and creolization, will shed more light on this issue. Philadelphia: Georgetown University Press. Wurm, Stephen, Darrell Tryon, and Peter Mühlhäusler (eds.) 1996. Atlas of languages of intercultural communication in References the Pacific, Asia and the Americas, 3 vols. Berlin: Mouton Arends, Jacques, Pieter Muysken, and Norval Smith (eds.) de Gruyter. 1995. Pidgins and creoles: an introduction. Amsterdam, JACQUES ARENDS Philadelphia, PA: Benjamins. Baker, Philip. 1990. “Off target?” Journal of Pidgin and Creole See also Language: Contact-overview; Lexical Bor- Languages 5. 107–19 rowing Pike, Kenneth Lee Kenneth Pike, like his predecessors, based his theo- the highlands of Oaxaca, Mexico. The following year retical contribution firmly on linguistic field work. (1936), he taught the phonetics component of the He followed their lead in attempting to explain the Townsend’s course, using his personal field experi- ways in which the languages of North America struc- ence to explain phonetic concepts. His introduction tured words, phrases, and clauses. Unlike his teach- to the initial writings of Edward Sapir (1921) and ers, he added a needed emphasis on meaning and Bloomfield (1933) motivated him to attend the 1937 context to earlier analytical methods. Pike’s interest Linguistics Institute of the Linguistics Society of in linguistics began when he attended the second ses- America, where he also met the leading American sion of linguistics training led by William Cameron linguists of the day. He studied under Sapir from Townsend in 1935 at a remote Arkansas farm. After 1937 to 1942 at the University of Michigan, focusing this rudimentary training, Pike applied his new on phonetics and descriptive linguistics. Pike was knowledge to the study of the complex tone patterns, later hired by the university and taught linguistics morphology, and syntax of the Mixtec language in there from 1948 to 1979. 850  PIKE, KENNETH LEE Pike spent much of his early efforts (1935–1948) on with the elements that could ‘fill’ that position. In phonetics and phonology (phonemics) and published English, for example, the subject position (slot) of a textbooks in these areas that were used extensively for sentence can have (be filled by) a Noun Phrase, a pro- many years. His personal field studies and his experi- noun, or proper noun, but not by an adverb or a con- ence as a consultant to other researchers in Latin junction. Eventually, Pike would add semantic America led Pike to investigate phonetics and phonol- information to his tagmemic model. In the previous ogy (then called phonemics). His Phonetics (1943) example, Pike noted that the role or function of a was the most complete book of its kind. Pike’s pho- Subject in English can be either the person who does the netic studies provided insights into what he called con- action (an Agent) or the person who is affected by the trast, variation, and distribution. Contrast refers to the action (a Patient or Undergoer), as in a passive clause. ways in which sounds are different from one another Using principles of contrast, variation, and distribu- in a particular language. Variation describes the fact tion from phonemics (phonology) and phonetics, Pike that individual sounds are not always pronounced the coined the terms ‘emic’ and ‘etic’ in 1954 to discuss same way, although native speakers perceive them as word and clause structures. Pike’s seminal work the same sound. Distribution deals with the locations Language in relation to a unified theory of the structure in which a sound appears. These three themes contin- of human behavior presents his philosophy of language ued throughout Pike’s writing and formed the basis of and culture that formed the basis of the emic/etic dis- his later work on morphology and syntax as well. tinction later used in anthropology and other fields. Pike pioneered the study of tone languages and With his wife and coteacher, Evelyn, Pike developed a helped many other linguists analyze the diverse tone unified approach for studies of morphology, syntax, systems of the world. One important contribution was and discourse at a time when other linguists limited using a constant tone pattern or frame to analyze sim- syntax to the clause or sentence level. Their ilar tones. By keeping the frame (such as ‘This is a…’) Grammatical analysis (1982) has been widely used as constant, linguists could better hear the contrastive a textbook in descriptive linguistic methodology. tones of the words being studied. Pike’s early students at the University of Michigan Pike’s studies in intonation were likewise an essen- include Alton Becker, Ruth Brend, Charles Fillmore, tial tool for many scholars and people involved in John Gumperz, Ilse Lehiste, and Velma Pickett. His pio- teaching English to speakers of other languages. Pike neering work in tagmemics has also been developed by applied intonation studies to poetry as well and pub- Walter Cook, Soedarjanto, Linda K. Jones, and Robert lished examples of poems (his own included) marked Longacre, among others. Other linguistic theories over- for intonation patterns reflecting how the poem was shadowed Pike’s work at times, but his teaching and read by different speakers. writing influenced hundreds of field linguists and helped In 1936, Pike developed a ‘monolingual demonstra- them to document indigenous languages around the tion’ in which he showed students how to learn a lan- world. After his retirement from the University of guage without relying on English translation. Using a Michigan in 1979, Pike served as an adjunct professor of variety of sticks, leaves, and other common objects, linguistics at the University of Texas–Arlington. Pike would elicit language data from a speaker of a Pike’s contribution to linguistics was overshadowed foreign language without using an intermediary lan- when tranformational approaches to syntax became guage. Within an hour, Pike would be able to explain popular. Although tagmemics handled a larger portion major features of the sound system, morphology, and of the overall linguistic analysis, it did not handle such syntax of the language. The lecture was very popular, areas as active vs. passive transformations or the ways both on campus and in nonacademic settings. It pro- in which certain patterns can be derived from other, vided a clear illustration of the role of a hypothesis and more basic, patterns. experimentation in linguistic studies. In 1977, the Pike was one of the first people involved in the University of Michigan distributed a video version of Summer Institute of Linguistics (SIL), a Christian one such demonstration of Pike working with a organization that sponsors linguistic and cultural Javanese speaker. research as well as literacy and translation programs in Pike is best known for his work in tagmemics, a the- indigenous language communities. When Pike was not oretical approach to morphology and syntax that inte- teaching at the University of Michigan, he split his time grates language forms and grammatical functions. between his long-term research among the Mixtec peo- Based on the concepts of sounds (phonetics) and sound ple and helping his SIL colleagues. In 1951, Pike and patterns (phonemics), Pike proposed that there were his coworkers published a translation of the New similar features in morphology (morphemes) and syn- Testament in San Miguel Mixtec, the first such transla- tax (syntagmemes). He defined the tagmeme as a com- tion completed by SIL. From 1936 through 2000, he bination of particular grammatical ‘slots’ or positions was regularly involved with SIL schools, first in 851 PIKE, KENNETH LEE Townsend’s training program in Arkansas, then at the International Phonetic Association, 1945; member, University of Oklahoma (1941–1987), and later at other Linguistics Society of America; and President, 1961. locations. As President of the Summer Institute of Pike was a member of the Linguistics Association of Linguistics from its incorporation in 1942 until 1978, Canada and the United States and its President in 1977. Pike oversaw the development of SIL’s worldwide lin- He was also a member, American Academy of Arts and guistic research and training in over fifty countries. Pike Sciences, 1973 and member of the National Academy personally supervised the beginnings of the SIL schools of Science, 1985. He was nominated for the Nobel at the University of Oklahoma (1941), in Australia Peace Prize in 1982–1996. He received Fulbright lec- (1950), and in Great Britain (1951). He was also active tureship in the USSR in 1988. Kenneth Lee Pike died in in the Linguistic Society of America and the Linguistic Dallas, Texas on December 31, 2000. Association of Canada and the United States. Pike was the author of more than 20 books and 200 References articles, including articles on the interaction of his per- Brend, Ruth M. (ed.) 1972. Selected writings to commemorate sonal spiritual beliefs and his academic work. He was the 60th birthday of Kenneth Lee Pike. The Hague: Mouton. also a prolific poet and frequently incorporated his Brend, Ruth M. (compiler) 1987. Kenneth Lee Pike bibliography. poems in his linguistic writing and teaching. He was Bloomington, Indiana: Eurasian Linguistic Associa-tion. an outstanding consultant who assisted other linguists Brend, Ruth M., and Kenneth L. Pike (eds.) 1977. The Summer to see the patterns in the languages they studied. He Institute of Linguistics: its works and contributions. The Hague: Mouton. frequently coauthored articles with younger col- Headland, Paul, Kenneth L. Pike, and Marvin Harris (eds.) leagues and was coeditor of linguistic volumes in such 1990. Emics and etics: the insider/outsider debate. Newbury diverse settings as Indonesia, Nepal, and Papua New Park, CA: Sage. Guinea. Pike lectured in 43 countries covering five Pike, Kenneth Lee. 1943. Phonetics: a critical analysis of pho- continents and received eight honorary degrees and netic theory. Ann Arbor, MI: University of Michigan Press. ———. 1945. The intonation of American English. Ann Arbor, numerous other honors including the Presidential MI: University of Michigan Press. Merit Medal from the Philippines (1974). His work ———. 1947. Phonemics: a technique for reducing languages with SIL led to his being nominated for the Nobel to writing. Ann Arbor, MI: University of Michigan Press. Peace Prize each year from 1982 to 1996, based on the ———. 1948. Tone languages. Ann Arbor, MI: University of positive impact of the linguistic, literacy, and transla- Michigan Press. ———. 1967. Language in relation to a unified theory of the tion efforts of SIL in over 1,000 languages. structure of human behavior. The Hague: Mouton. ———. 1982. Linguistic concepts. Lincoln, NB: University of Nebraska Press. Biography Pike, K.L., and Evelyn Pike. 1982. Grammatical analysis. Kenneth Lee Pike was born in Woodstock, Connecticut Dallas, TX: Summer Institute of Linguistics. ———. 1993. Talk, thought, and thing: the epic road to conscious on June 9, 1912. He received his Th.B. (1933) and knowledge. Dallas, TX: Summer Institute of Linguistics. Ph.D. (1942) for studies in phonetics, mentored by ———. 1977. Seasons of life: a complete collection of Kenneth Edward Sapir, University of Michigan. He was a mem- L. Pike’s poetry, compiled and ed. by Sharon Heimbach. ber of the Summer Institute of Linguistics, 1935–2000; Huntington Beach, CA: Summer Institute of Linguistics. President, 1942–1979; and President Emeritus, 1979– Pike, Eunice V. 1981. Ken Pike: Scholar and Christian. Dallas, TX: Summer Institute of Linguistics. 2000. He was also Associate Professor of Linguistics, The Summer Institute of Linguistics provides additional infor- University of Michigan, 1948–1953; Professor of mation about Kenneth Pike at http://www.sil.org/klp/. Linguistics, 1954–1979; Chair of Linguistics Department, 1974–1979; and Professor Emeritus PETER J. SILZER 1979–2000. He was Permanent Council Member, See also Mexico; Tone Languages Pintupi and Pama-Nyungan Languages The Pintupi language is an Australian indigenous lan- northwest of Alice Springs in the Northern Territory. It guage spoken in the Western Desert, eastern Central is mainly spoken in and around the Papunya Indigenous Australia, which is situated approximately 400 km Community Settlement. Pintupi belongs to a group of 852  PINTUPI AND PAMA-NYUNGAN LANGUAGES languages which includes Ngaanyatjarra, Pitjantjatjara/ While these words seem very similar in structure, Yankunytjatjara, Luritja/Pintupi, with the last two lan- just the addition of a longer ‘r’ sound gives a com- guage groups being very similar in formation (Simpson pletely different meaning to the word. The sound ‘ng’ 1993:136, 142). These languages form part of the is also often used at the beginning, or in the middle, of Pama-Nyungan grammar group of Australian languages words, making the language difficult to pronounce for and exclude other indigenous languages from the north English speakers because we are used to using this of Western Australia and the north and northwest of the sound at the end of words, such as in the word ‘sing’ Northern Territory (Yallop, C., p. 16). According to an (Walsh and Yallop 1993:xi). Inquiry by the Australian Government in 1992, 90 of the Hansen also found that it was very important to use original 250 Australian indigenous languages are still the correct words in different social situations, and living languages and approximately 3,000 people still because of the strange combination of sounds, this was speak the eastern Western Desert group of languages. difficult for him to do. His goals were based on the use The Western Desert, western Central Australia group of of the Bible in traditional indigenous language and languages include Manjiljarra, Yulparija, Kukatja, culture, and he was obviously very committed to his Ngaanyatjara, and Ngaatjatjarra languages, with work. He had estimated that it would take him at least approximately 1,000 people speaking these lan- 15 years to work out how to write the language by just guages. This brings the total number of people speaking listening and transcribing the sounds into an alphabet, the Western Desert languages of Central Australia to a dictionary, and a grammar, let alone have enough approximately 5,000, making it the largest group of information about the culture to translate the Bible. speakers of traditional indigenous languages in their There is now a Pintupi/Luritja to English Dictionary natural environment in Australia. In 1995, Peter available that has been written by him. Trudgill wrote in his book Sociolinguistics: Another enormous problem in doing this type of In Australia, for instance, there used to be about 200 translation is that the Aboriginal people had never aboriginal languages. Of these, 50 are already dead, and heard of things such as a synagogue, and similar con- another 100 are very close to extinction. Perhaps as few cepts such as ‘meeting place’ had to be used instead. as 30 will make it to the year 2000. (p. 177) The indigenous word for ‘ceremonial offering’, ‘kunatinpa’, could possibly have been used or incor- It has therefore become an important task for lin- porated here. guists today to do as much as they can to work toward Some other examples of Hansen’s Pintupi/English the preservation of these endangered languages. translations are (with literal translations and/or expla- Australian indigenous people have been either dis- nations): couraged or forbidden to speak their own languages since colonization, and many Aboriginal languages synagogue tjuwuku tjaatji have been lost to the world. These languages are ‘oral’ the Jew’s church only and have no written form apart from drawings, king mayutju pulka such as is found in caves, in their natural settings, and big boss it is only in recent times that linguists have been able angel nganka ngurrara to do more extensive fieldwork to give written form to one who belongs in the sky the surviving few indigenous languages. Pintupi is one pharisee tjuwuku luwuku mikunytju of these languages, and translators, such as Aboriginal a lover of the Jew’s law Bible translator Ken Hansen, who lived in a Pintupi Holy Spirit katutjaka kurrunpa community for several years, have done significant God’s personal spirit work in this area. God katutja In his article on ‘Translating for the Pintupi’, Ken the one who pertains above Hansen says that when he first began to learn to speak Christ kirritja Pintupi his progress was very slow because the sounds Christ pronounced in the Pintupi way of the indigenous language were more complex than Christ katutjalu tjamatatjunkula wantirriyantja the sounds of English. It was therefore very difficult to the one whom God promised to send transcribe into a written form. long ago For example, Hansen says the Pintupi language has As can be seen from these translations, the effect of words such as: British colonization has been that many words that (1) ngarrinpa lying originally come from English origins would not have and ngarinpa standing been in existence in indigenous language prior to this (2) wangka talk event. They have since been added to the Australian and wanka alive indigenous language and culture. This is particularly 853 PINTUPI AND PAMA-NYUNGAN LANGUAGES the case with words such as ‘big boss’, because many The following are Pintupi concepts with their indigenous men became stockmen on cattle properties, English meanings relating to feelings: and the owner was called the ‘big boss’, ‘mayutja kunta shame pulka’, which was then transcribed into Traditional rama mad indigenous language. This is a concept that is impor- ngulu fear tant to modern social use of indigenous language and rarru anger has meant the creation of new Aboriginal word usage pukulpa happy (Simpson 1993:137–78). One of these words is the ngarru happy Western Desert word for ‘spirit’, ‘kardiya’, which is now also used to refer to ‘white people’. These concepts have been interpreted by As can be seen in modern word usage in indige- researchers, such as Meyers, as: nous language is the current, and extremely signifi- (1) ‘Mad, crazy persons are said to be “deaf”, cant, social event taking place in Australia today, the “rama”, because they take no notice of what is Reconciliation Process, and the use of the word being said to them. “Rama”, or “not hearing”, ‘sorry’. The Indigenous people are asking the therefore relates to people who do not take Australian Government and the Australian people as notice of advice and are therefore not thinking.’ a whole to say ‘sorry’ for the past injustices that (2) ‘The English and Pintupi concepts of “fear” or have been done to the Australian Indigenous people “ngulu” are the same and generally mean to be (Haebich 2000:569). These injustices include taking frightened or afraid of the consequences of the indigenous children from their natural parents one’s actions. Men’s sacred objects and ritual and communities and placing them with white fami- paraphenalia are often described as “ngulu”, or lies ‘for their own good’ (Haebich 2000:567–8). “frightening and dangerous”.’ Although the present Australian multicultural popu- (3) ‘“Anger” or “rarru” means to “get wild” or lation was not specifically involved in perpetrating “very angry” because he does not feel sorry for these injustices, it is seen as a token of respect his actions. Pintupi descriptions of “rarru” toward the indigenous people, and an acknowledge- means that their “ears are closed” and they are ment of the shame that the people who were respon- not thinking of the consequences of their sible for these injustices should feel, for committing actions.’ these acts. The associated feelings for this can be (4) ‘When a relative comes to visit the Pintupi say incorporated into words such as ‘kunta’, or ‘shame’, one becomes happy, or rejoices, “pukularrin- by saying ‘sorry’ to what is now left of the indige- pa”. Or men bringing back meat make people nous community. Part of the meaning of the Pintupi happy. It is the opposite of anger, fighting, and word ‘kunta’ relates to the English words and con- sorrow, and is tied to a major Pintupi image of cepts for ‘shame’ and ‘respect’. When used to mean sociality and cooperation, says Meyers. ‘respect’, it is usually symbolic of ‘shyness’ in not Ceremonies that include song and dance make wanting to show disrespect, such as refusing a per- people happy, and “pukulpa” was traditionally son to his face without excuses. Someone who is used to stop fighting with an approaching being respectful of others shows embarrassment by enemy.’ saying whatever they have is less significant than the other persons’ possessions, and this type of shame is Another way in which indigenous language use is called ‘kuntarrinpana tjanampa’. Children are being preserved in Australia today is in bilingual edu- socialized by shaming, and a child is said to be deaf cation programs in schools. Pintupi/Luritja is being or unheeding, ‘patjarru’, when young, but needs to taught bilingually in Walungurra School, Northern be respectful when older and to be aware of ‘kunta’ Territory, since 1983, with Luritja language being as they learn of the need to show respect within the taught at Papunya (Black 1983:207–20). Another community (Mawr, B:126–157). These emotional place of bilingual education is of the Diwurruwurru- concepts have a great significance in the Pintupi jarru Community, where the Walpiri language, (also a understanding of social behavior and it could be fur- Pama-Ngungan language) is being taught at ther said that this concept is of importance to the Lajamanu, Yuendumu, Nyirripi, and Willowra schools, indigenous people as a whole. It can then be an inte- near Katherine in the north of the Northern Territory gral part of the Reconciliation Process of saying (Bavin 1993:85). Kriol is being taught at Barunga ‘sorry’, therefore incorporating an important indige- School, also in this area (Rhydwen 1993:155–8). In nous concept of being sorry for the bad things that some of these communities, linguist/teachers work someone has done within the indigenous concept of alongside indigenous women while the children are their own culture. taught traditional ways of food gathering. The names 854  PITCAIRNESE of ‘bush tucker’ (food), including plant and animal Dixon, R.M.W. 1980. The languages of Australia. Cambridge: names, and names for ‘body parts, are taught in Kriol Cambridge University Press. Eades, Diana. 1993. Language and the law: White Australia v and Walpiri Language by the Indigenous women in Nancy. In Walsh and Yallop. communities near Katherine, and the linguist/teacher Haebich, Anna. 2000. Broken circles: fragmenting indigenous teaches the same names in English. families 1800–2000. Fremantle, Western Australia: In other bilingual teaching programs, everyday Fremantle Arts Centre Press. classes are taught in both traditional indigenous lan- Hansen, Kenneth. 1977. Pintupi/Luritja dictionary, 2nd edition. Alice Springs, Northern Territory: Summer Institute of guage, and English. A significant problem that arises Linguistics, Australian Aboriginals Branch. because indigenous language is an oral not written lan- Hansen, Kenneth C. 1980. Translating for the Pintupi. guage, is that new words have to be invented to explain Nungalinya Occasional Bulletin No 22. Darwin, Northern how to write things such as commas and full stops, and Territory: Nungalinya Aboriginal and Torres Strait Islander also in the teaching of mathematics. This is how and Church Training and Research Centre, University of the Northern Territory. where linguists and teachers are working together with House of Representatives Standing Committee on Aboriginal indigenous people to preserve Australia’s original lan- and Torres Strait Islander Affairs. June 1992. Language and guages. It is especially important that this work be culture — a matter of survival. Report of the Inquiry into done as quickly as possible so that indigenous lan- Aboriginal and Torres Strait Islander Language guages are not only lost to their original users, but to Maintenance. Australian Government Publishing Service. Meyers, Fred R. 1976. Cultural concepts of the emotions. the world as a whole as globalization and the common From: to have and to hold: A study of persistence and use of English as the predominant language take over. change in Pintupi social life. Ph.D. thesis. Bryn Mawr. Myers, Fred R., 1986. Pintupi country, Pintupi self: sentiment, place and politics among Western Desert Aborigines. References Washington, DC: Smithsonian. Amery, Rob. 1986. Languages in contact: the case of Kintore Rhydwen, Mari. 1993. Kriol:the creation of a written language and Papunya (with particular reference to the Health and a tool of colonization. In Walsh and Yallop. Domain). Language in Aboriginal Australia 1. Schmidt, Annette. 1990. The loss of Australia’s Aboriginal lan- Bavin, Edith. 1993. Language and culture: socialisation in a guage heritage. Canberra: Aboriginal Studies Press. Walpiri community. In Walsh and Yallop. Simpson, Jane. 1993. Making dictionaries. In Walsh and Yallop. Black, Paul. 1983. Aboriginal languages of the Northern Tindale, Norman. 1974. Aboriginal tribes of Australia. Los Territory. Batchelor, NT: School of Australian Linguistics, Angeles: University of California Press. Darwin Community College. Trudgill, Peter. 1995. Sociolinguistics : an introduction to lan- Blake, Barry. 1991. The handbook of Australian languages. guage and society. England: Penguin Books. Cambridge: Oxford University Press. Walsh, Michael, and Yallop, Colin, (eds.) 1993. Language and Devlin, Bruce. 1990. Some issues related to vernacular lan- culture in Aboriginal Australia. Canberra, Australia: guage maintenance: a Northern Territory view. ed. by, C. Aboriginal Studies Press. Walton and W. Eggington. Language maintenance, power Wierzbicka, Anna. 1997. Understanding cultures through their and education in Australian Aboriginal contexts, Darwin: key words. New York: Oxford University Press. Northern Territory University Press. JOANNE MARIE BAUMGARTNER Pitcairnese Pitcairnese is the name of one of two closely related might die out because none of the children were varieties of a language initially formed on Pitcairn speaking it spontaneously, declared it an official lan- island and subsequently transplanted to Norfolk island guage of the island. by descendants of the nine English-speaking sailors Although there is some disagreement about who mutinied on HMS Bounty in 1789 and their whether PN is a creole rather than a variety of English, Polynesian companions from Tahiti and Tubuai, one of the conditions under which it developed were clearly the Austral Islands of the Tahiti group. Linguists use exceptional because we know the place and time of its the name ‘Pitcairn-Norfolk’ (PN) to refer to both, but formation, as well as details of the origins of the pri- speakers call the variety spoken on Pitcairn ‘Pitkern’ mary individuals involved. PN is one of the few cases and that on Norfolk Island ‘Norfolk.’ In 1996, Pitcairn in which it is possible to directly trace the survival of Island Council, prompted by the concern that Pitkern some British dialectal features to some of the sailors. 855 PITCAIRNESE The Bounty, sent to Tahiti to transport breadfruit Young family sailed the 3,700 miles back to Pitcairn. trees to the West Indies as a source of food for slaves, Five years later, a second party returned. Today, fewer contained no pressganged men and was probably the than 50 speakers remain on Pitcairn Island. Present- first British ship to sail with an all-volunteer, hand- day Norfolk is spoken by about 700 people; the major- picked crew of 45. The cliché ‘motley crew’ well ity of the island’s population is not descended from the describes the mutineers, for they covered not only a original Pitcairners. Some 700 other descendants of wide social spectrum ranging from their leader, the original population live in Australia, New Zealand, Fletcher Christian, master’s mate, born of an aristocrat- and elsewhere. ic family originally from the Isle of Man, to common Because the mutineers were outnumbered by two to seaman John Adams, a Cockney and orphaned son of a one, the odds were that Pitcairn would become pre- Thames waterman, brought up in a poor house, but they dominantly Polynesian in language and culture, partic- also covered a wide range of regional varieties of ularly because all the women were Polynesian (most of English. Four of the nine mutineers were English, two them from Tahiti) and would have been the primary Scottish, one American (from Philadelphia), one part- caregivers to the children. There are also hints in both West Indian from St. Kitts in the Caribbean, a well- boatswain’s mate James Morrison’s and midshipman educated son of a navy captain (Edward ‘Ned’Young), Peter Heywood’s accounts of the mutiny that the and one, John Williams, assistant armorer, was a native sailors had begun to intersperse Tahitian words in their of Guernsey and spoke French, as Captain William speech to one another. Midshipman Roger Byam Bligh noted in his list of identifying traits of his men. recalls that by the time the Bounty sailed with the by Of Peter Heywood, he remarked that he spoke with the now heavily tattooed sailors after a stay of more than ‘Manx, or Isle of Man’ accent. five months in Tahiti, all the men knew some Tahitian In their search for a place to hide from the British words. Some were said to be quite fluent and could be navy, the mutineers initially sailed back to Tahiti and heard carrying on conversations with one another in to Tubuai, where they picked up six men and 12 which barely a word of English was spoken. Yet, the women, one with a daughter, before settling in 1790 language of power on Pitcairn island was clearly on a remote Pitcairn island more than 2,000 kilometers English, and many linguists argue that PN was not a from Tahiti. This small, two- by one-mile, harborless fully developed pidgin or creole because there was no island with its steep cliffs and lack of reefs or anchor- point in its history in which English was not the pri- age was at the time uninhabited and incorrectly chart- mary language of the community. Stable pidgins rarely ed on contemporary maps. Pitcairners’ first contact develop where contact involves only two languages. with the outside world occurred 18 years after settle- Moreover, the descendants seem to identify more ment, in 1808, when Captain Folger of the Topaz from strongly with their English rather than Tahitian lineage, Boston was greeted in English by Thursday October and the predominant influence on the vocabulary, pro- Christian, Fletcher Christian’s son, and was asked nunciation, and sentence structure of PN is English. whether he knew Captain Bligh. The main influence of Tahitian survives in the form of Apart from this brief encounter, the settlers lived a large element of Tahitian vocabulary, particularly for there for the first 33 years in almost complete isola- flora, fauna, and foodstuffs, e.g. buhe, ‘sea eel’ (Tahitian tion, thus permitting no outside influences during the puhi), and miti, ‘coconut cream sauce’ (Tahitian miti). formative period of the language. It was not, however, Tahitian influence can also be seen in certain semantic a tranquil time by any means, with nearly 90% of the (meaning) shifts that have affected English words, such founding male population murdered by 1800. Disputes as hand, meaning both ‘hand’ and ‘arm’ in PN, parallel- over land and the women led to the violent deaths of ing Tahitian rima. Otherwise, most of the vocabulary is all the men except four Polynesians and four muti- ultimately of English origin, including a number of neers, one Englishman (John Adams), a Scot (William dialects. Words can sometimes be traced to the dialects McCoy), one Cornishman (Matthew Quintal), and one of individual mutineers, e.g. the negative imperative (‘do West Indian from St. Kitts (Edward Young). Young, not’) dune from the Scots probably spoken by John McCoy, and Quintal died by 1800, leaving only John Mills, gunner’s mate from Aberdeen, or William McCoy Adams, then 33 years old, who remained with nine of Rosshire, or moge, ‘thin’ (and possibly also the nega- surviving women and the 23 children born on the tive form kaa/kannt, ‘cannot’), from Edward Young’s island. Outsiders did not enter the colony until the West Indian English. Other Scots forms include bole, ‘to 1820s. The first separation in a community that had make a small hole in anything’, devil’s needle, ‘dragon- lived together as a family for more than 60 years fly’, gaggle, ‘to cackle’, and possibly tayte, ‘potato’. occurred in 1856, when the 194 Pitcairners were reset- There are also archaisms such as dub, ‘to square and tled on Norfolk Island, originally a penal colony. smooth (timber)’, tardy, ‘late’, paunch, ‘stomach’, and a Eighteen months after their arrival, 17 members of the few words of American origin, such as corn, ‘maize’, 856  PITCAIRNESE and candy, ‘sweets’, possibly because of the presence of translation of Tahitian taua, composed of English Isaac Martin, a mutineer from Philadelphia, or because ‘thou’  ‘me’, with the initial sound of ‘thee’ of the influence of American whalers who visited the changed to /h/. This pronoun contrasts with uklun island. Other dialect forms include dunnekin, ‘lavatory’, (‘we/us plural’), whose etymology is unclear. PN, and grub, ‘food’. like Tahitian, has dual marking in forms such as yu A variety of terms with original nautical reference, tu, ‘you two’, contrasting with yoli, ‘you plural’, i.e. such as all hands, ‘everyone’, deck, ‘floor’, and flog, three or more. This may also reflect archaic English ‘spank’, have extended their meanings in PN. ‘all ye’ or ‘you all’. Similarly, heave ho, a cry of sailors in raising the The sound system of PN also shows some influence anchor up, survives in PN heway me (heave me away) from Tahitian. Lack of a significant distinction ‘to lift up/ throw/take away/ remove’, as does heway between /r/ and /l/, a feature common to many (‘heave away’), now obsolete in English. However, Polynesian languages, including Tahitian, is found in this once nautical usage with the meaning ‘to fling, PN forms such as stolly, ‘lie’ (English ‘story’), morla, throw or cast, to haul up or raise by means of a rope’ ‘tomorrow’, and tomolla, ‘day after tomorrow’ (both does exist in dialectal usage in Scots, where heave from English ‘(to)morrow’). This feature (along with away means ‘to throw away’, as well as in some variation between /v/ and /w/, also similar to Tahitian) English dialects in which heave means ‘to lift/raise’. is a feature typical of St. Kitts too, which may suggest Although PN was formed independently of Pacific additional influence from the speech of Edward Jargon English, an English-based foreigner talk used Young. Unlike many pidgins and creoles, PN has pre- in encounters between Europeans and Pacific islanders served consonant clusters under some circumstances, and found on almost all islands in Polynesia and whereas Tahitian has none. However, final t/d are often Micronesia by the 1830s because of the whaling absent in PN, a common feature of colloquial and industry, it shares some features in common with other dialectal English. PN pronunciations such as /flaid/ for Pacific pidgins and creoles descended from it, e.g. ‘Floyd’ and /kloi/ for ‘cry’ may reflect a sound shift bimorphemic (i.e. two-part) question words such as that probably originated in London and was well under whatawe, ‘how’ (English ‘what way’), verbs generally way by the mid-nineteenth century. It is possible that unmarked for tense (past, present) and aspect (whether John Adams is the origin of this feature in PN, the action is completed or not), use of dem (English although it could also have arisen later by independent ‘them’) as a plural marker, e.g. dem ai, ‘the eyes’, and innovation or contact with other varieties, such as the absence of the copula ‘be’, e.g. whatawe hem, ‘how is one spoken by Young. The vowels of words such as he?’ The sailors probably knew some rudiments of price and choice still rhyme in St. Kitts today in many Pacific Jargon English with adaptations made during people’s usage. The /ie/ diphthong (double vowel their stay at Tahiti. These features could also have sound) of gate could reflect West Indian or English been acquired through contact with visiting whalers. dialectal influence. Despite the presence of two Scots Between 1813 and 1852, 400 visits of ships are among the founding English speakers, PN has no trace recorded, three quarters of which were American, of /r/ after vowels in words such as cart, barn, etc. mostly whaling ships. For a time, there was a steady market for the islanders’ vegetables. Norfolk islanders References also had contact with the whaling trade in the late nineteenth and early twentieth centuries. Many Buffett, Alice, and Donald C. Laycock. 1988. Speak Norfolk worked as crewmen on whaling ships. They were also today. Norfolk Island: Himii. Dening, Greg. 1992. Mr. Bligh’s bad language. Passion, power in contact with the varieties of Melanesian pidgin spo- and theatre on the bounty. Cambridge: Cambridge ken by Solomon Islanders and New Hebrideans. In University Press. 1867, the Melanesian Mission established a training Harrison, Shirley. 1972. The language of Norfolk Island. B.A. school for Solomon Islanders on Norfolk. From the Thesis. Sydney: Macquarie University. 1890s until World War II, Norfolk Islanders worked as Harrison, Shirley. 1985. The social setting of Norfolk speech. English World-Wide 6(1). 131–53. plantation managers and cadets in the Solomons. Laycock, Donald C. 1990. The status of Pitcairn-Norfolk: cre- Otherwise, however, PN has far more in common with ole, dialect, or cant. Status and function of languages and the Caribbean creoles, no doubt attributable to the varieties, ed. by Ulrich Ammon. Berlin: de Gruyter. influence of Edward Young from St. Kitts. The pos- Mackaness, George. 1938. A book of the ‘Bounty’ and selec- sessive construction makes use of English ‘for’ as in tions from Bligh’s writings. London: J.M. Dent & Sons, Ltd. Ross, A.S.C., and Moverley, A.W. 1964. The Pitcairnese lan- dem ai fo yoen, ‘your eyes’. guage. New York: Oxford University Press. A few grammatical features, particularly the per- SUZANNE ROMAINE sonal pronouns, reflect Tahitian influence, e.g. the first-person inclusive dual hemi (‘you and I’), a literal See also Pidgins and Creoles 857 PLACE OF ARTICULATION Place of Articulation Consonants are formed by creating a constriction in the lower articulator is the apex of the tongue and that the vocal tract; in producing a consonant, the place of the upper articulator is the upper teeth. Occasionally, articulation is the part of the vocal tract having the when the lower articulator is obvious or unimportant, greatest constriction. The symbols used for the only the upper articulator is named: e.g. velar, used description of sounds are taken from the IPA chart, alone, is interpreted as meaning dorso-velar. The which lists the International Phonetic Alphabet. major places of articulation are shown in Figure 1. The major constriction, which may be a complete Places of articulation can be divided into four gen- or partial closure, is made by the articulators coming eral categories: labial, coronal, dorsal, and guttural. At together, generally by a lower articulator moving each place of articulation, sounds with different man- toward an upper articulator. The lower articulators are ner of articulation can be produced: stops have com- elements of the lower jaw—the lower lip, the lower plete closure at the indicated place, whereas fricatives teeth, and the tongue. The upper articulators are the are pronounced with partial closure, which produces a upper lip, the upper teeth, the palate, the velum, the hissing sound. uvula, and the rear wall of the pharynx. Places of artic- Labial sounds are made with one or both lips and ulation usually have a compound name giving the include bilabials, labio-dentals, and linguo-labials. lower and upper articulators, with the name of the The lower lip articulates with the upper lip to form a lower articulator first. Thus, apico-dental indicates that bilabial consonant. The term bilabial is used rather Palatal Velar Postalveolar Alveolar Uvular Bilabial Dental Labio-dental Pharyngeal Figure 1 858  PLACE OF ARTICULATION than labio-labial. Bilabial stops [p, b] and the nasal the back of the lower teeth and produces a slightly dif- [m] are extremely common in languages; indeed, a ferent sound. language without them (e.g. some Oto-Manuean lan- Two fricatives are given in the International guages) is noteworthy. The bilabial fricatives [ϕ, β], Phonetic Alphabet as alveolo-palatal; these are differ- however, are rather uncommon; languages tend to ent from the postalveolars. They are produced with the have bilabial stops and labio-dental fricatives. tip of the tongue near the lower teeth and with the The lower lip articulates against the upper teeth to blade quite close to the back of upper teeth. The sym- form labio-dental consonants. Labio-dental fricatives bols are [-] for the voiceless fricative and [.] for the [f, v] are very common. Labio-dental stops do not occur voiced fricative. These sounds occur in Mandarin as distinctive sounds in any language. The labio-dental Chinese and in Polish. nasal [(] occurs as a variation of [m] in English words For retroflex consonants, the tip of the tongue curls ) such as symphony [ si(fəni], in which it takes the place back, and the underside of the tip articulates with the of articulation of the following labio-dental fricative. area at the back of the alveolar ridge. The body of the Linguo-labial sounds are very rare sounds made tongue is quite concave. Retroflex consonants are with the tip or blade of the tongue articulating with the symbolized as [/01 2 34]; they use the alveolar sym- upper lip. The IPA uses a diacritic [ ] underneath a bols modified by a lower hook. These sounds are com- labial symbol for this place of articulation: . mon in languages of India. Coronal sounds are made with the apex, or lamina, The English r sound is usually called retroflex; how- of the tongue. They include dentals, alveolars, alveo- ever, in practice, the place of articulation may vary lo-palatals, postalveolars, and retroflex sounds. from alveolar [ɹ] to retroflex [5]. Further, many speak- Dental sounds can be made with either the tip of the ers of English use a bunched r, in which the tip of the tongue (apico-dental) or with the blade (lamino-dental). tongue is near the lower teeth and the body of the The dental fricatives are voiceless [θ] and voiced [ð]; tongue is pulled up and back toward the palate. these sounds both occur in English as in the words thin Although the articulations of the retroflex and bunched and then, respectively. Other dental sounds are shown by r are very different, the acoustic effect is quite similar. using the alveolar symbol with the diacritic [' ]; e.g. [t' d' Dorsal sounds include those made with the front or n' 'l r' ]. Lamino-dental fricatives [s' z' ] can be made with back of the tongue: palatal, velar, and uvular. the blade of the tongue near the back of the upper teeth, With palatal sounds, the front of the tongue articu- in addition to the apico-dental fricatives [θ ð]. Danish lates with the palate. The tip of the tongue points has a dental approximant [ð +]. Interdentals are made by down, often touching the lower teeth. Italian has a thrusting the tongue slightly forward so that the tip pro- palatal lateral [ʎ], as in figlio, and a palatal nasal [1], trudes between the teeth; they can be symbolized as as in signore; German has a voiceless palatal fricative [θ, ð,]. In English, apico-dentals are normal, although [ç], as in ich. The palatal stops [c 7] and the voiced some people occasionally use interdentals in emphatic or fricative [ 8] are not common. exceptionally careful speech. Dorso-velar sounds are made with the dorsum of the Like the dentals, alveolar sounds may be made with tongue articulating with the velum. The dorso-velar either the tip or blade of the tongue, known according- stops are [k "] as in English could, good; the dorso- ly as apico-alveolar or lamino-alveolar. The symbols velar nasal is [ŋ] as in English sing. The fricatives are for either are [t d n s z]; all of these sounds occur in [x], as in German Bach, and [γ], the voiced equivalent. English. Alveolar and dental sounds are both extreme- Frequently, these sounds are simply called dorsals. ly common in languages; it is rare, however, for a lan- The uvular consonants are made with the dorsum of guage to have both, except for fricatives. the tongue articulating with the uvula. They are like Postalveolar sounds are made with the blade of the velar sounds made very far back in the mouth. The sym- tongue articulating with the area at the back of the bols are [q] and [g] for the stops, [n] for the nasal, and alveolar ridge. The fricatives [ʃ ] and affricates [χ] and [ʁ] for the fricatives. Uvular sounds are found [tʃ d] are quite common in the languages of the in Arabic and Inuktitut. Some French accents have an world. English has these sounds in shin [ʃ], leisure [], approximant [ʁ +] or fricative [ʁ] for the r sound. chin [tʃ], and gin [d]. The guttural sounds include the pharyngeal, The alveolar fricatives [s z] and the postalveolar epiglottal, and laryngeal places of articulation. fricatives [ʃ ] involve some difficulty in describing Pharyngeal consonants are made by moving the their point of articulation accurately. Different speak- root of the tongue backward toward the pharyngeal ers hold their tongues in somewhat different positions. wall. The root of the tongue is the vertical part, form- With the alveolars [s z], the airstream hits the back of ing the forward wall of the pharyngeal cavity. Many the upper teeth. This causes a turbulence that gives people cannot make a complete pharyngeal stop. A them a distinctive sound. For [ʃ ], the air stream hits pharyngeal nasal stop is physiologically impossible 859 PLACE OF ARTICULATION because, if no air flows through the pharynx, then no [t]) labialization [t=], palatalization [t>], velarization air can flow through the nasal passage. For the frica- [t?], and pharyngealization . The IPA has alternative tives, the symbols are [©] and ; these sounds are symbols for either velarization or pharyngealization . found in Arabic and other Semitic languages. Labialized sounds involve lip rounding; they are Epiglottal sounds are extremely rare, but they occur quite common. Palatalized consonants have the front in some Semitic and Caucasian languages. For them, of the tongue more toward the palatal region than the epiglottis moves backward to articulate with the usual. Palatalized consonants are common in some pharyngeal wall. The symbols are for the stop and Slavic languages. Velarized consonants have the for the voiceless and voiced fricatives. tongue toward a position that is more velar than usual, Laryngeal sounds involve the vocal folds; they i.e. a high back unrounded vocalic position. The include the glottal stop and the fricatives [h] and English dark [@], occurring at the ends of words such [ ]. From a phonological point of view, these sounds as ball, is an example of a velarized lateral; here, the can all be thought of as consonants with a laryngeal dorsum of the tongue is raised slightly toward the place of articulation, with the glottal stop as a laryn- velum. Pharyngealized sounds are found in Arabic; geal stop and the sounds [h] and [ ] as laryngeal frica- they involve a retraction of the root, thus effecting a tives. From a purely phonetic point of view, however, narrowing of the pharynx. each of these could be categorized differently: as a state of the glottis involving complete closure of the References vocal folds, [h] as a voiceless vowel, and [ ] as a breathy voiced vowel. Glottal stop and [h] are very Catford, J.C. 1977. Fundamental problems in phonetics. Bloomington, IN: Indiana University Press. common; [ ] is somewhat rare, although it occurs Catford, J.C. 1988. A practical introduction to phonetics. between vowels in English, as in ahead. Oxford: Oxford University Press. In general, the major categories of place of articu- Gimson, A.C. 1980. An introduction to the pronunciation of lation (labial, coronal, dorsal, and guttural) are inde- English. London: Edward Arnold. pendent of each other. Thus, it is possible to make two Hardcastle, William J., and John Laver (eds.) 1997. The hand- book of phonetic sciences. Oxford: Blackwell. stops at the same time, e.g. a dorsal [k] and a labial [p]. International Phonetic Association. 1999. Handbook of the Such a sound is said to have a double articulation; for International Phonetic Association. Cambridge: Cambridge example, with [kp], the place of articulation is labial- University Press. velar. It is symbolized as [kp], with the tie-bar show- Ladefoged, Peter. 1993. A course in phonetics. Ft. Worth, TX: ing that the [k] and the [p] are made simultaneously. Harcourt Brace and Jovanovich. Ladefoged, Peter. 2000. Vowels and consonants: an introduc- English [w] is a doubly articulated labial-velar approx- tion to the sounds of languages. Oxford: Blackwell. imant: the lips are rounded and the dorsum of the Ladefoged, Peter, and Ian Maddieson. 1996. Sounds of the tongue is partially raised toward the velum. Many world’s languages. Oxford: Blackwell. English speakers pronounce a double articulation in Laver, John. 1994. Principles of phonetics. Cambridge: kitten . Swedish has a doubly articulated voice- Cambridge University Press. Minifie, Fred, Thomas J. Hixon, and Frederick Hixon. 1973. less postalveolar-velar fricative [<]; this is the same as Normal aspects of speech, hearing, and language. a simultaneous [ʃ] and [x]. Englewood Cliffs, NJ: Prentice-Hall. Secondary articulations involve the addition of a Pullum, G.K., and W.A. Ladusaw. 1996. Phonetic symbol lesser, or secondary, articulation to a greater, or primary, guide. Chicago: University of Chicago Press. articulation. Thus, by adding lip-rounding to a [t], a Rogers, Henry. 2000. The sounds of language: an introduction to phonetics. Harlow, England: Pearson. labialized [t=] is produced. The primary place of articu- HENRY ROGERS lation is alveolar; the secondary articulation is labial. The common secondary articulations are (shown with See also Anatomy of the Articulatory System Plurality The majority of languages have singular and plural guages, some languages of Southern Asia, Baltic lan- forms of words to indicate a single object or a set of guages) have a dual form of words to indicate a set of more than one object. Some languages (Philippine lan- two objects. Dual nouns are used in archaic dialects of 860  PLURALITY Lithuanian, and some relics are still known in Russian, during the storm. The tree has fallen during the storm. Ukrainian, and some other Slavic languages. For Some trees have fallen during the storm. They can example, in Russian the word eye has the form: gπaA have both singular and plural forms. Most of the count [glaz] (singular), gπaAa [glaza] (plural). The plural nouns form their plurals by adding -(e)s. However, form is actually a relic of the dual. The origin of the there are a number of irregular words, and words bor- dual in these languages can be explained by the fact rowed from other languages, from Greek and Latin in that some things exist only in a pair: like ears, eyes, particular, that create their plurals in specific ways (see knees, legs, and sides. examples below). There are also some English words In most of the modern languages, the dual form has that can only be either singular or plural. disappeared. Only two categories, i.e. the singular and The plural form of count nouns is formed by adding plural (more than one object) are used; for example, -(e)s to the noun: the singular word leg in English takes the plural form flower flowers of legs; in Russian it would be HOg a [noga]—HOg u lady ladies [nogi]; and in Turkish the same word would be bacak balloon balloons in singular and bacaklar in plural. lunch lunches There are also languages that do not have a specif- bush bushes ic form to indicate plural objects. This can be illustrat- fairy fairies ed by examples from Malay, where a noun mata (eye) potato potatoes would normally imply a plural meaning (there are usu- ally two eyes on a face). The same principle would A number of irregular nouns in English form their apply to kaki (legs). When necessary, the plural form plurals differently, i.e. by changing their vowel or by in Malay is created by repeating the noun—kamus- adding -en. kamus (dictionaries), kanak-kanak (children). man men The plurality of nouns and pronouns as objects in woman women the sentence affects subject–verb agreement. For exam- foot feet ple, the singular English pronoun I requires a singular goose geese verb—I have a beautiful vase for these flowers. While mouse mice the plural form of I–we requires a plural verb—We louse lice have a beautiful vase for these flowers. The singular child children noun sister requires a singular verb—My sister goes to ox oxen the movies every Sunday. The plural sisters requires a plural verb—My sisters go to the movies every Sunday. Those count nouns, which end with -f, form plurals The following English nouns are grouped accord- by changing -f to -ves, for example, ing to objects they refer to: wolf wolves wife wives (1) Sister, house, table, book, mother, minister, scarf scarves teacher, doctor; (2) Milk, coal, chocolate, mud, water, gold, infor- Words borrowed from other languages, Latin and mation, sugar, cream; Greek in particular, create their plurals according to (3) Love, hatred, friendship, hope, charity, justice; the foreign (original) rule, for example: (4) Family, group, committee, class, parliament. datum data stratum strata The first group refers to concrete, tangible objects, or larva larvae persons. They are count nouns. The second and third criterion criteria groups of nouns indicate substance (2) or abstract (3) nouns that are not discrete, and therefore cannot be Some English nouns are used only in their singular counted. They are mass nouns. The last group repre- form, such as music, chemistry, physics, ceramics, sents entities or bodies that consist of a number of mathematics, etc. At the same time, there are nouns members. These nouns can be counted. They are col- that are only used in the plural: jeans, trousers, scis- lective nouns. sors, spectacles, measles, etc. However, the language allows for the following phrases if there is a need to indicate the number of some of the above-mentioned Count Nouns words. For example—Tanya knew she desperately Count nouns can be preceded by the word some or the needed another pair of jeans. The office assistant has articles a, an, and the. For example: A tree has fallen ordered two pairs of scissors for the department. 861 PLURALITY There are cases of nouns that do not change for plu- used with the definite articles—The love John and ral, e.g. sheep, deer, fish, salmon, etc. Also, there are Matt shared for their mother finally ended the war words that have two different plural forms that convey between the brothers. The chocolate made in this different meanings, such as: town is delicious. Since the mass nouns refer to abstract, not discrete craft: craft (boats) objects, they cannot be counted. Therefore, they usu- crafts (handicrafts) ally do not allow for plural morphology. However, index: indexes (alphabetical lists they allow for other means of measurement: of names, subjects, etc.) indices (algebraic A bottle of milk; exponents) Three cups of tea; medium: mediums (at spiritual Bars of chocolate; séances) Patches of mud; media (newspapers, A useful piece of information. television) The appearance of such words as sugars or lights is possible only if they refer to concrete, discrete entities; Collective Nouns for example: Switch off the lights please. Can I have Collective nouns are the nouns that indicate a group, a two sugars in my coffee please? body, or an organization that consists of a number of members, for example family, class, parliament, com- mittee. They form their plurals in the same way the References count nouns do by adding (e)s: families, classes, par- Dummet, Michael. 1993. Grammar and style for examination liaments, committees. The fact that a collective noun candidates and others. London: Duckworth. refers to a number of people does not affect the sub- Haegeman, Liliane, and Jacqueline Gueron. 1999. English grammar and generative perspective. Oxford, UK: ject–verb agreement in a sentence. Such words in their Blackwell Publishers. singular form are used with singular verb, e.g.—My Horan, Ronald S. 1996. Exploring English: a complete analysis family enjoys skiing. The Social Club Committee usu- of modern grammar. Leichhaardt, Australia: Honeysett ally plans the recreational activities of the employees. Press. Hudson, Richard. 1984. Word grammar. Oxford, UK: Blackwell. Mass Nouns Johansson, Stig. 1980. Plural attributive nouns in present-day English. Lund: CWK Gleerup. Mass nouns are not usually used with the definite arti- Bach, Emmon, et al. (ed.) 1995. Quantification in natural lan- cle; for instance—Loyalty is his most valuable char- guage, Vol. II. Dordrecht: Kluwer Academic Publishing. acteristic. Sugar adds sweetness to any drink. Reid, Wallis. 1991. Verb and noun number in English: a func- However, in the following examples the mass tional explanation. London: Longman. nouns are modified by further information and can be ALFIA ABAZOVA Polish and West Slavic Languages The present-day West Slavic language area is made up Slavic used by these settlers, and, by the end of the primarily of the national territories of Poland, the tenth century, features distinctive of the three West Czech Republic and Slovakia, as well as a small part Slavic subdivisions of Lechithic, Czecho-Slovak, and of south-eastern Germany. By the end of the sixth Sorbian had emerged. Apart from one or two isolated century CE, migrations of Slavic-speaking people had remaining pockets of Slavic, the more westerly settled in a wide area of northwestern Europe, and Slavic-speaking areas were gradually resettled by these communities extended as far west as the River German speakers in the middle ages, resulting in Elbe. By the ninth century CE, distinctly West Slavic approximately the West Slavic language area as it elements had become noticeable within the Common exists today. 862  POLISH AND WEST SLAVIC LANGUAGES The modern West Slavic languages are Polish, million people in that country. It has been estimated Czech, Slovak, and Upper and Lower Sorbian. To that there are several million more speakers outside these is also often added a further member, Cassubian, Poland, including immigrant communities in the whose status is disputed between its being a separate United States of America and elsewhere. language or a dialect of Polish. Until the mid-eigh- The earliest evidence that we have for Polish dates teenth century, another West Slavic language, from the twelfth century and consists solely of proper Polabian, also existed. Polish, Cassubian, and names within an otherwise Latin document. The earli- Polabian make up the Lechithic subgroup of West est surviving complete sentence in Polish dates from Slavic, Czech, and Slovak make up the Czecho-Slovak the thirteenth century and the earliest complete text subgroup, and Upper and Lower Sorbian make up the dates from the fourteenth century. Sorbian group. Other divisions have been suggested in Within present-day Polish, five main dialect areas the past, involving the merging of Sorbian with either are usually recognized: Wielkopolska is an area Lechithic or Czecho-Slovak, but this three-group toward the northwest of Poland, centered around model is the most widely accepted. Poznan, Mazowsze is in the northeast and includes West Slavic is distinguished from South and East Warsaw, MaAopolska is in the southeast, S`la sk is a Slavic primarily on the basis of phonetic differences, small area toward the southwest, and Kaszuby is such as the retention of /tl/ and /dl/ clusters, which have another small area in the north around the port of been reduced to /l/ in South and East Slavic languages. Gdansk. In addition to these, there are also mixed West Slavic languages also have fixed word stress, while dialect areas, both in the extreme northeast and in the most other Slavic languages have moveable stress. west along the border with Germany. These latter All the West Slavic languages use a Roman-based areas were part of Germany until the end of World alphabet, with each language having a set of accented War II and were then settled by Polish speakers of and/or special characters. Polish has the largest set of various dialects. special characters, including ‘hooked’ vowel charac- An increasingly standard form of literary Polish, ters that represent nasalized vowels. drawing mostly on the dialects associated with the Two features of the West Slavic languages have main cultural centers of Poznan and Gniezno (in the been of particular general interest to linguists. Both are Wielkopolska area) and Cracow (in the MaAopolska shared with other Slavic languages, but one is especial- area) can be identified from the sixteenth century, when ly closely associated with work on West Slavic. there was a flourishing ‘golden age’ of Polish literature. The first feature is the category of aspect, which This literary language was also influenced to a certain stands alongside tense to indicate whether an action or extent by Czech, as well as by borrowings from Latin. event is completed or ongoing, as opposed to whether Following a period of decline for written Polish in the its occurrence is in the past, present, or future. This is early eighteenth century, a revivalist movement evident in the existence of pairs of related verbs, such emerged in the late eighteenth and early nineteenth as the Lower Sorbian cytas´ and docytas´, where cytas´ centuries, which aimed to codify the standard lan- focuses on the general act of reading, while docytas´ is guage. Although members of this movement took note used for a completed reading session (meaning rough- of the earlier literary language of the sixteenth and sev- ly ‘read through to the end’). enteenth centuries, it was not primarily an archaizing The other feature is word order within the sentence. movement (in contrast to the codification of Czech, All the West Slavic languages mark grammatical func- outlined below) and consequently the differences tions by the use of special word endings, rather than by between the codified written and contemporary spoken word order as is the case in English. They thus permit languages were not all that great. Throughout this peri- considerable flexibility in word order, and they use od, and also thereafter, there was a strong opposition to this to give prominence to particular pieces of infor- the adoption of distinct regionalisms into standard mation within the sentence. This phenomenon has Polish. A version of Polish based on the standard writ- sometimes gone by the name of ‘functional sentence ten language has become adopted as the standard spo- perspective’ and its theoretical status within general ken variety. Today, most Polish speakers who speak a linguistics originated with work on Czech by the dialect are also able to speak this standard language, Prague School in the 1930s. and they use it in more formal and public settings. Distinctive features of Polish include the nasalized vowels mentioned earlier, the marking of gender as Polish and Cassubian well as number and person in certain forms of the Polish has the largest number of speakers of the West verb, and the use of third-person (‘he’ or ‘she’ or ‘it’) Slavic languages. It is the official language of the verb forms instead of second-person (‘you’) forms in Republic of Poland and is thus spoken by around 38.5 respectful address. 863 POLISH AND WEST SLAVIC LANGUAGES The Kaszuby dialect (Cassubian in English) is, as six sets of noun endings from earlier Slavic are better stated earlier, often recognized as a separate language preserved in Czech than in any of the other West Slavic within the West Slavic group, although within Poland languages. Jungmann was more concerned with widen- there is a greater tendency to see it as a dialect of ing the use of Czech, although his attitude was also Polish. It certainly possesses greater differences from essentially purist and he was responsible for introduc- standard Polish than do the other dialects, and Poles ing a number of distinctively Czech new words in pref- from outside the Cassubian region have difficulty in erence to borrowing from other languages. understanding it. Its status as a separate language is Besides standard spoken Common Czech, Czech strengthened by a strong Cassubian ethnic identity. It divides into three main dialect areas: Bohemian, also has its own literary tradition and is used regularly Hanák (or Central Moravian), and Lach (or Silesian). in the local media. Despite these outward attributes of Additionally, there are two transitional dialect areas: a a relatively cohesive language, Cassubian has a high southeast Moravian area, whose dialects are closely degree of internal dialect variation. There have been related to Slovak, and a mixed Czech-Polish dialect various attempts to codify the language, but, to date, area. Standard Common Czech is linked most closely there is no formally recognized standard. Arguments to the central Bohemian dialects, and particularly to against the status of Cassubian as a language are its the dialect of the main urban center, Prague. There is, close relation to Polish and the fact that the Cassubian however, a degree of regional variation even within area is politically a part of the Republic of Poland. Common Czech, and some now suggest that these are References will sometimes be found to a further of sufficient magnitude to argue that two distinct West Slavic language, Slovincian, which is now regional versions of Common Czech exist—Bohemian extinct. This was located geographically adjacent to Common Czech and Moravian Common Czech. As Cassubian, to the northwest of the main Cassubian- with Polish, there are also mixed dialect regions in speaking area. Although the residents of that area areas along the border with Germany, which had been referred to their language as Slovincian, it is almost German-speaking prior to the end of World War II. universally recognized by linguists as having been a subdialect of Cassubian. A combination of earlier Slovak Germanization and postwar population movements led to the extinction of Slovincian by the 1950s. Slovak is the official language of Slovakia, spoken by around 4.6 million people there. A further 300,000 or so speakers are resident in the neighboring Czech Republic. Czech Slovak is closely related to Czech and the two lan- Czech is the official language of the Czech Republic, guages are, broadly speaking, mutually intelligible. This spoken by between 9 and 10 million inhabitants, as is particularly the case with the southeast Moravian well as by around 60,000 inhabitants of neighboring dialects of Czech. However, there are also some differ- Slovakia and by immigrant communities elsewhere. ences between the two languages in both sound and The earliest surviving example of Czech dates from word structure. This is especially true of the written lan- 1057 CE and consists of a set of glosses on a Latin guages, owing to the different language varieties that manuscript. From the thirteenth century onward, a were drawn on during the period of their codification. prolific Czech literary tradition exists, which flowered As with Czech, the earliest surviving evidence for especially in the sixteenth century. Slovak comes from a set of eleventh-century glosses. Czech is unusual in that there is a larger than aver- However, in contrast, no substantial literary tradition age difference between the standard written and spoken existed for Slovak prior to around 1780: instead, languages. The standard spoken language—known as Czech took on the role of the written language, Common Czech—represents the natural evolution of although Slovak writers did introduce some distinc- the Czech language over time. In contrast, standard tively Slovak features into their written Czech. A fur- written Czech stems from essentially artificial codifica- ther similarity with Czech is in the date that, and the tion by linguists in the early to mid-nineteenth century, process by which, the eventual standard written lan- the most influential of whom were Josef Dobrovský guage did develop. Written Slovak was not codified and Josef Jungmann. Dobrovský, who was mainly until the 1840s, when a standardized form was pro- responsible for codifying the grammar of Czech, based duced by L’udevít Štúr in association with a group of his grammar on substantially earlier literary usage from other Slovak intellectuals. However, unlike Czech, the golden age of the sixteenth and seventeenth cen- where much earlier written usage exerted a strong turies. Thus, in comparison with Common Czech and influence on the newly codified language, the Slovak other West Slavic languages, standard written Czech written language was based primarily on a synthesis of shows a number of archaic features. For instance, the the contemporary spoken Central Slovak dialects. This 864  POLISH AND WEST SLAVIC LANGUAGES must at least in part have been a necessity, owing to the around Cottbus in Brandenburg), spoken by approxi- paucity of a written Slovak tradition, but it also mirrors mately 20,000 speakers. Upper and Lower Sorbian the codification of Polish, where contemporary usage have a number of differences from one another, and had more influence than it did with Czech. these occur at all language levels, including sound Slovak has three main dialect areas—West, Central, structure (e.g. Lower Sorbian has góra in place of and East—with the West Slovak dialects being closest Upper Sorbian hora), vocabulary (e.g. Lower Sorbian to their Czech neighbours. has jo for ‘yes’, whereas Upper Sorbian has haj), and grammar (e.g. Upper Sorbian makes more use of pos- sessive adjectives than Lower Sorbian). Sorbian (Upper and Lower) In contrast to the two standardized varieties, the fea- The Sorbs are an indigenous Slavic minority who live tures that distinguish the several Sorbian dialects show in the region of eastern Germany known as Lusatia. a rather complex distribution, given the small size of Until 1990, Lusatia was part of the former German the language area. Nevertheless, broadly speaking, it is Democratic Republic. It now belongs to the federal possible to make a primary distinction between Upper states of Saxony and Brandenburg. In earlier times, and Lower Sorbian dialect areas, with a transitional Sorbian dialects extended eastward beyond the River area, showing features of both, located in between. Neisse into what is present-day Poland, as well as Having coexisted for such a long time alongside westward and northward into further regions of German, Sorbian has been affected by a number of Germany. However, Sorbian is now wholly contained Germanic influences, several of which are not shared by within the Lusatian region. the other West Slavic languages. These influences have Today, there have been estimated to be around occurred not only at the vocabulary level but have also 60,000–70,000 speakers of Sorbian, although all affected the grammar of Sorbian. For example, most Sorbian speakers are now bilingual in Sorbian and Slavic languages do not have articles (such as a, an, and German, and a number of these speakers use the the in English). However, under German influence, German language for most everyday purposes. Sorbian Sorbian has made use of its demonstrative pronouns and has survived most strongly in small, closely knit rural the number ‘one’ as, respectively, definite (the) and communities, and has been noted as being especially indefinite (a/an) articles, although this practice is dis- strong in a group of villages to the north of Bautzen in couraged in the standard written languages, which have Saxony. However, although its strongest base is among been subject to Slavic purist movements. Lower Sorbian the rural community, Sorbian is not a low-prestige rural probably retains more German influences than Upper dialect and has a lively literary and scientific culture. Sorbian, which was subject to a more intensive push Apart from proper names, the earliest surviving toward purism. German influence on Lower Sorbian was evidence for Sorbian vocabulary consists of glosses in formerly also evident at the orthographic level, with Latin manuscripts dating from the twelfth century. The much Lower Sorbian publishing using the German earliest continuous Sorbian texts date from the six- Fraktur type until around the time of World War II. teenth century. These earlier Sorbian texts were writ- As a small minority language within a German- ten in local dialects with distinct differences from one speaking context, the number of speakers of Sorbian another. In the eighteenth century, three standard writ- has declined substantially over the past century. ten varieties emerged: Lower Sorbian, Catholic Upper However, there are lively movements to preserve its use Sorbian, and Protestant Upper Sorbian, with Upper and the future is not necessarily bleak. The rights of the Sorbian thus being split on denominational lines. Sorbs have been legally guaranteed within a bilingual (Since the Lower Sorbians were Protestants, the issue framework since the end of World War II, firstly by the did not arise for that variety.) At that time, only reli- former German Democratic Republic and subsequent- gious texts were published in Sorbian. From the mid- ly by the federal states of Brandenburg and Saxony. To nineteenth century onward, alongside an increase in promote and maintain the language, a number of initia- nonreligious writing, the denominational split in tives exist, including the provision of bilingual school- Upper Sorbian was gradually diluted, although it was ing from an early age (the WITAJ initiative). not until after World War II that the present twofold division of Sorbian was firmly established. Polabian There are two present-day standard varieties of Sorbian. Upper Sorbian is the language of the Sorbs in Polabian, sometimes known as Dravaeno-Polabian, is an Upper Lusatia (centered around Bautzen in Saxony) extinct West Slavic language. The last native speaker is and is the larger of the two varieties with roughly reported to have died in 1756, although there are records 40,000–50,000 speakers. Lower Sorbian is the of someone able to recite the Lord’s Prayer in Polabian language of the Sorbs in Lower Lusatia (centered as late as 1798. Polabian was formerly spoken in a small 865 POLISH AND WEST SLAVIC LANGUAGES part of the Lüneburg area, which is situated in north References Germany between Hamburg and Hannover. It was thus Auty, Robert. 1960–1961. Dialect, Koιv` and tradition in the surrounded by German-speaking territory from an early formation of literary Slovak. Slavonic and East European point in its history and was geographically isolated from Review 39. the other Slavic languages. (Contrast Sorbian, which has Auty, Robert. 1976. Problems in the formation and development retained a border with both Polish and Czech.) Not sur- of the Czech literary language. Slavic linguistics and language teaching, ed. by Thomas F. Magner. Cambridge, MA: Slavica. prisingly, therefore, it shows evidence of a very strong Comrie, Bernard, and Greville G Corbett (eds.) 1993. The German influence, especially from the regional Low Slavonic languages. New York: Routledge. German varieties, both in vocabulary and grammar. De Bray, R.G.A. 1980. Guide to the West Slavonic languages. Very little Polabian material has survived, and Columbus, OH: Slavica. much of what has is in the form of isolated words and Olesch, Reinhold. 1989. Gesammelte Aufsätze. 1. Dravaeno polabica. Cologne and Vienna: Böhlau. phrases compiled into dictionaries in the last century Polan´ski, Kazimierz. 1993. Polabian. In Comrie and Corbett. of the language’s existence. Only a small number of Rothstein, Robert A. 1993. Polish. In Comrie and Corbett. complete texts, such as the Lord’s Prayer and a folk Short, David. 1993. Czech. In Comrie and Corbett. song, have been preserved. A figure of around 6,000 Short, David. 1993. Slovak. In Comrie and Corbett. recorded vocabulary items has been suggested. This Stone, Gerald. 1972. The smallest Slavonic Nation: the Sorbs of Lusatia. London: The Athlone Press. has nevertheless provided enough evidence to suggest Stone, Gerald. 1993 Cassubian. In Comrie and Corbett. that at least three dialects existed. Stone, Gerald. 1993. Sorbian. In Comrie and Corbett. Despite this lack of material, and in view of its Townsend, Charles E. 1990. A description of spoken Prague unusual isolation from related languages, Polabian has Czech. Columbus, OH: Slavica. received continued and substantial attention from lin- ANDREW WILSON guists, including notable figures such as Trubetzkoy. See also Indo-European 5: Slavic Politeness Politeness refers to how a speaker chooses to phrase an interaction thus presents a dilemma for interactants; utterance in a particular context, a choice that can they want to maintain each other’s face, but they are reflect the speaker’s view of the interpersonal context. also motivated to perform actions that are face-threat- There are a number of different approaches to polite- ening. This conflict is solved by engaging in face- ness. In a social normative view, to be polite is to speak work, or, more precisely, by being polite. formally and to behave in accordance with situation- According to Brown and Levinson (1987), polite- specific rules regarding appropriate linguistic behavior. ness is conveyed by deviating from maximally efficient A variant of this approach has been to regard politeness communication; politeness is thus roughly equated as one component of pragmatic competence (Leech with indirectness. There are many ways this can be 1983). In this view, there are conversational maxims accomplished, and Brown and Levinson organized (e.g. tact and generosity) that stipulate which linguistic politeness into five superstrategies. They provided evi- form is preferred over another in any particular context. dence for these strategies from three different lan- Currently, the most popular approach to politeness, guages and argued that the linguistic strategies for based on the work of Brown and Levinson (1987), conveying politeness are universal. The least polite views politeness as the linguistic means by which peo- strategy is to perform an act without any politeness, or ple manage each other’s ‘face’. The concept of face to perform the act directly. The most polite strategy is (Goffman 1967) refers to the identity that people pres- to perform the act with an off-record form. Off-record ent during the course of their interactions with others. forms are deliberately ambiguous and hence perform a Face is assumed to be universal and of two types: neg- face-threatening act indirectly (e.g. It’s warm in here as ative face, or the desire to have autonomy of action, a request to have the window opened). Falling between and positive face, or the desire for closeness with oth- these two extremes are on-record (i.e. relatively unam- ers. One’s face is fragile and subject to continued biguous) acts with redress emphasizing either positive threat during social interaction. For example, criti- face or negative face. The former, termed ‘positive cisms threaten the recipient’s positive face, and politeness’, functions via an exaggerated emphasis on requests threaten the recipient’s negative face. Social closeness or solidarity with the hearer. Examples 866  POSSESSIVES include the use of slang and familiar address terms, One of the most important issues surrounding polite- jokes, and presumptuous optimism (You’ll loan me ness is its status as a cultural universal. Although few your car, won’t you?). The latter, termed ‘negative would argue the claim that politeness exists in all cul- politeness’, functions via attention to the recipient’s tures, the claim that positive and negative faces are uni- autonomy. This is accomplished by symbolically less- versal desires motivating the form that politeness takes ening any imposition on the hearer. A common method has been questioned. For example, it has been argued is to use conventionally indirect forms (e.g. Could you that negative face is relevant only in Western cultures, or shut the door?). Research has provided partial support cultures in which there is an emphasis on individual for this typology. It appears, however, that politeness autonomy. Although Brown and Levinson (1987) mar- strategies are very diverse and may not be completely shaled impressive evidence for the cross-cultural validity captured with this scheme. Also, it is unclear whether of their politeness strategies, empirical support for their politeness should be equated with indirectness; very claim that they are universal has been rare. Clearly, there indirect strategies may sometimes backfire and may be is great cultural variability in politeness. The crucial perceived as manipulative. question is whether this variability is a result of differing One of the most important features of a face-man- cultural conceptions of face or whether these differences agement view of politeness is the attempt to specify the can be explained at a lower level of abstraction. Face relationship between politeness and the interpersonal wants may be universal, but cultures will vary in terms of context. Brown and Levinson (1987) proposed that what threatens face, who has power over whom, how politeness will vary as a function of the weightiness (or much distance is typically assumed, and so on. degree of face-threat) of the act to be performed. Recently, there have been some new developments Weightiness is contextually determined and assumed to in politeness theory. Some researchers have used it as be an additive weighting of the following three vari- a framework for explaining how people interpret indi- ables: the intrinsic (and culturally bound) degree of rect utterances. Others have begun to examine nonver- imposition of the act itself (e.g. asking for a loan of $5 bal politeness. Finally, many researchers are now is less imposing than asking for a loan of $500), the using politeness theory in applied contexts as a means power of the hearer relative to the speaker, and the of examining and explaining linguistic behavior in degree of social distance between the interlocutors. specific contexts, such as courtroom trials, classroom Thus, increased weightiness (hence, in general, greater interaction, and political campaigns. politeness) occurs as a function of increasing imposi- tion, hearer power, and relationship distance. Note that References in this way the politeness of an utterance cannot be Brown, P., and Levinson, S. 1987. Politeness: some universals in determined in isolation, but rather only in a specific language usage. Cambridge, UK: Cambridge University Press. interpersonal context. For example, an utterance per- Fraser, B. 1990. Perspectives on politeness. Journal of Pragmatics ceived as polite when used with an underling may not 14. 219–36. be polite when directed toward a superior. Goffman, E. 1967. Interaction ritual: essays on face to face Research has provided partial support for these behavior. Garden City, NY: Anchor. Holtgraves, T. M. 2001. Politeness. Handbook of language and ideas. Many studies have demonstrated that greater social psychology, 2nd edition, ed. by H. Giles and W.P speaker power is associated with decreased politeness. Robinson. London: Wiley. Fairly consistent support has also been found for the Lakoff, R. 1973. The logic of politeness: or, minding your p’s imposition variable, with increasing imposition asso- and q’s. Papers from the ninth regional meeting of the ciated with increasing politeness. The effects of rela- Chicago Linguistic Society, 292–305. Leech, G. 1983. Principle of pragmatics. London: Longman. tionship distance have been the most problematic; Watts, R. J. 2003. Politeness. Cambridge, UK: Cambridge whether greater politeness is associated with increas- University Press. ing relationship distance is not clear. THOMAS M. HOLTGRAVES Possessives Possessive marking is used to indicate that two nouns modifier, the possessor, and the other noun (phrase) is or noun phrases stand in a modifying relationship, the modified, the possessed. According to the position namely that of possession. One noun (phrase) is the of the modifier, two kinds of possessives are available 867 POSSESSIVES in English: one is marked by the genitive case (my The only difference between noun phrases and verb father’s books) and the other by a preposition (the phrases (or sentences, for that matter) lies in the ele- books of my father). ments that are used to mark the exact relationships According to traditional grammars, the genitive between the individual items. In the case of noun possessive refers to a specific entity or individual. It is phrases, the possessive is the relevant grammatical more frequent with modifiers referring to animate marker for the subject. The markers ’s and of thus were entities, i.e. living things, but certain inanimate modi- considered to be ‘semantically null’, i.e. they do not fiers can also take the genitive, such as geographical contribute any particular meaning to the sentence. names (Canada’s farms), temporal nouns (yesterday’s Chomsky (1986) assumed that both are simply overt paper), and the names of institutions (the university’s realizations of the genitive (case). council). The use of inanimate modifiers also depends In fact, possessive markers have no semantic content on the modified noun and is generally possible with in cases of so-called ‘inalienable’ possession. For modified nouns referring to places (the earth’s sur- example, the expression John’s picture may have two face, the water’s edge, at the road’s end) and meas- ‘inalienable’ readings: (1) ‘John has created the pic- ures (arm’s length). The preposition of is used in ture’ vs. (2) ‘John is in the picture’. Both are cases of partitive constructions (a piece of wood, a kind of let- inalienable possession, because nobody can take away ter) and expressions denoting a place of origin (Jesus the fact that ‘John created the picture’, or nobody can of Nazareth), quality (a man of affairs), and material take John out of the picture (without destroying the pic- (a sword of steel). ture). In other words, the entities John and picture are The genitive modifier can function as a determiner inseparable and presuppose each other (i.e. an inherent and thus excludes the use of other determiners (such as relationship). If John is construed as a creator of the the or a). For example, it is not possible to say the my picture, this noun is inserted in the subject position father’s book. To use a determiner, the so-called ‘dou- [John picture], which automatically leads to the geni- ble genitive’ construction must be used—that is, a book tive expression John’s picture. If, on the other hand, of my father’s. The double genitive can be used with the John is the object of picture, the base structure would indefinite article a and with demonstrative pronouns be [picture John], and the default phrasing would be such as that or this (a/that/this book of my father’s). The the picture of John. However, if no creator is specified, definite article can generally not be used in double gen- i.e. if the subject position is left empty in the base struc- itive constructions: the book of my father’s is not a pos- ture, English allows ‘promotion’ of the object, which sible English phrase. However, if there is a modifying again yields the genitive John’s picture. Hence the relative clause, use of the definite article becomes pos- potential ambiguity of English genitive expressions. sible: the book of my father’s that is on the table. However, Mona Anderson (1983) argued that the Genitive and prepositional markings can reflect the genitive ’s does contribute independent meaning in subject/object distinction in constructions like the cases of ‘alienable’ or ‘true’ possession. Thus, the enemy’s destruction of the city. Here, the nouns enemy expression John’s picture can also be interpreted as and city receive the same interpretation as in the sentence ‘The picture that John owns’. This means that the rela- the enemy destroys the city. In both cases, the enemy has tion between John and picture is not inherent but is a subject-like function, while the city is the object. This established by means of the ‘word’ or ‘lexical unit’ ’s. parallelism was an important piece of evidence in a heat- Note that ’s can be labeled a word, since it fulfills a ed linguistic debate concerning nominalization, the function similar to ‘full’ words, namely possess or process of deriving a noun such as destruction from the own. In other words, in this case the phrase John’s pic- corresponding verb, here destroy. In grammatical theo- ture is projected not from the noun picture but rather ries of the 1960s, it was assumed that a noun phrase like from the possessive ’s, and the base structure would be John’s presentation of the project is derived from the [John ‘s picture], similar to [John possess picture] or synonymous sentence John has presented the project. [John own picture]. This view changed with Noam Chomsky’s widely The idea of lexical ’s has been commonly accepted quoted article ‘Remarks on nominalization’ (1970). since the second half of the 1980s, and it has been Chomsky argued—convincingly to many linguists— extended to other structures. For instance, his book has that any noun can ‘project’ structures containing a sub- been analyzed as deriving from the base structure [he’s ject and an object. Thus, for phrases like the enemy’s book], where he is the subject and book is the object of destruction of the city, the base structure would be the word ’s. [enemy destruction city]. For comparison, the base In a crosslinguistic perspective, the distinction structure for The enemy destroyed the city would be between genitive and prepositional possessives is not [enemy destroy city]. Nouns and verbs thus ‘generate’ universal. In French, for example, a genitive marker or ‘project’ very similar structures at an abstract level. analogous to the English ’s does not exist. Thus, the 868  PRAGMATICS English expression Robert’s friends is translated in This change was caused by the increasing loss of case French as les amis de Robert (literally the friends of inflections and the resulting increased importance of Robert). Russian, on the other hand, has no preposi- function words. tion corresponding to the English of. For instance, two In conclusion, the possessive markings ’s and of are constructions are available to translate the English usually realizations of genitive case in English and expression Robert’s friends: (1) Robertovy druzja, establish a grammatical relation between two nouns, where -ov- is a possessive marker, and (2) druzja but English ’s also has a lexical use. The manifestation Roberta, where -a is a morphological manifestation of of genitive case varies across languages and historical- genitive case. ly also within a given language. Historically, the marker ’s originates from the gen- itive case singular marker -es used in Old English. For example, in the sentence Him sceamode þæs mannes References ‘He felt ashamed for the man’, this ending is observed Allen, Cynthia L. 1997. The origins of the ‘Group Genitive’ in on the noun stem mann ‘man’ preceded by the genitive English. Transactions of the Philological Society 95.1. form of the demonstrative se ‘the, that’, i.e. the literal 111–31. Anderson, Mona. 1983. Prenominal Genitive NPs. The translation would be something like ‘To-him there- Linguistic Review 3. 1–24. was-shame the-of man-of.’ Similarly, þurh Chomsky, Noam. 1970. Remarks on nominalization. Readings geswicenysse yfeles ‘by cessation from evil’ is literal- in English transformational grammar, ed. by Roderick A. ly ‘through cessation evil-of’, and cyninges botl ‘a Jacobs and Peter S. Rosenbaum. Waltman, MA: Ginn and king’s palace’ is ‘king-of palace.’ Since the genitive Company. Chomsky, Noam. 1986. Knowledge of language: its nature, ori- ending already sufficiently performed the function of gin, and use. New York: Praeger. expressing possession, the preposition of did not mark McCawley, James D. 1988. The structure of noun phrases. The possessive phrases in Old English. Prepositional mark- syntactic phenomena of English, 2 vols. Chicago and ing becomes much more frequent in Middle English. London: The University of Chicago Press. The latter is characterized by a common use of the so- Taylor, John R. 2001. Possessives in English: an exploration in cognitive grammar. New York: Oxford University Press. called ‘group genitive’ constructions like the king of EGOR TSEDRYK France’s son corresponding to þæs cyninges sune Frances ‘the-of king-of son France-of’ in Old English. See also Case Pragmatics Pragmatics is the study of language use. Whereas the is derived from Greek pragma, which means ‘action’. study of grammar focuses on the language system, Action is defined as intentional behavior. Pragmatics pragmatics offers a complementary perspective on lan- studies verbal communication as a complex form of guage, providing an insight into the linguistic choices intentional behavior, which is interactive, i.e. partner that users make in social situations. Pragmatics is, for oriented, and symbolic, i.e. conventionalized and cul- instance, interested in how people pay compliments, ture specific. In the English-speaking world, pragmat- engage in small talk, or write e-mails. The commu- ics was originally considered part of sociolinguistics. nicative functions of utterances or texts, and the speak- Pragmatics is a relatively young field of study. So ers’ or writers’ intentions behind them, are of far, there is no coherent pragmatic theory. Pragmatics particular interest. Historically, the emphasis was on developed from linguistic, philosophical, and socio- spoken language. logical approaches to language use. The so-called The term ‘pragmatics’ was first used in semiotics, pragmatic turn in linguistics was brought about by the the general theory of signs. In this theory, pragmatics writings of a group of philosophers known as speech pertains to the relationship between signs and their act theorists in the late 1960s. Until then, mainstream users. In linguistics, pragmatics deals with verbal linguistics had focused on linguistic forms and struc- signs (words, utterances, texts) and how they are used tures, neglecting meaning and ignoring communica- by humans in communication. The term ‘pragmatics’ tive functions and language users. 869 PRAGMATICS Initially, pragmatics was dominated by the speech act A distinction can be made between pragmalinguis- theory. The basic insight of this tradition is that speech is tics and sociopragmatics. ‘Pragmalinguistics’ is con- action. Whenever we say something, we do not only pro- cerned with the verbal resources available for realizing duce sound waves, utter words, and produce sentences any given speech act. By contrast, ‘sociopragmatics’ but we also perform an action. If a mayor says, I declare focuses on the polite norms governing the selection of this bridge open, then this bridge is open. In this sense, resources relative to social situations. For example, the mayor has performed an action. pragmalinguistics identifies the word choices, mean- Speech act theory aims at establishing how many ing patterns, and sentence constructions that are used fundamentally different ways there are of doing things to pay a compliment, whereas sociopragmatics deter- with words. Bridge-opening belongs to ‘declarations’, mines who may compliment whom on what in which a class of speech acts that all require an institutional situations. Pragmalinguistics is language specific, and context. Four further classes have been identified, sociopragmatics culture specific. which are more likely to occur in everyday communi- In empirical pragmatics, five types of inquiry can cation. These are ‘directives’, ‘commissives’, ‘expres- be distinguished. Studies may focus on speech act sives’, and ‘assertives’. Typical examples for these realization in one language alone, or they may com- classes are requests, promises, apologies, and state- pare speech acts across languages. The second type is ments, respectively. known as contrastive pragmatics. It aims at answering In linguistics, the study of language use focuses on questions such as these: how do speakers of English the individual speech acts that belong to these funda- and speakers of Japanese formulate complaints? By mental classes, e.g. requests, promises, and apologies, contrast, cross-cultural pragmatics examines differ- and on the linguistic resources used to perform them. ences and similarities in social norms and cultural val- Empirical pragmatics aims at systematically determin- ues governing polite choices. A typical question is ing the different strategies that people actually use to this: how do members of US American culture and realize a particular speech act. A classic case is members of British culture refuse offers? A fourth requesting. For instance, if you want a person to take branch of empirical pragmatics is interlanguage prag- out the garbage, you may say: Take out the garbage!, matics, which is the study of language use in a second You ought to take out the garbage, How about taking or foreign language. The fundamental question here is out the garbage?, Can you take out the garbage?, the following: how do language learners perform and Garbage day tomorrow, etc. These strategies differ in interpret speech acts in the target language? Finally, their degree of directness. The most direct strategy is applied pragmatics focuses on problems of communi- to use an imperative construction, and the least direct cation in practical contexts, such as courtrooms, med- strategy is to only hint at the desired action. Directness ical interviews, or international business encounters. must not be confused with politeness. The degree of Both natural and elicited data are analyzed in directness is a feature of any strategy independent of empirical pragmatics. Natural data, which occur in context, whereas the politeness value of an utterance everyday situations, are recorded by using field notes, depends entirely on the social situation. audiotape, or video. Further sources are large corpora Theories of verbal politeness have also contributed of spoken language, such as the Santa Barbara Corpus to the development of pragmatics. According to one of Spoken American English. Alternatively (or addi- particular politeness theory, any speech act may, at tionally), a wide range of research instruments is used least potentially, interfere with the wants and needs of to elicit data. Tools include role plays, questionnaires, hearers. This is obvious if a speaker wants a hearer to and interviews. Fictional material is also used, e.g. do something. It is less obvious in the case of social dialogue in narrative prose or drama. niceties, but a compliment, for example, may also be All approaches to language use discussed so far focus interpreted as harassment. Given the threatening poten- on speakers and their intentions, rather than on hearers tial of all speech acts, speakers are thought to calculate and their interpretations. How hearers understand utter- the imposition involved in the respective situation and ances is dealt with in a theory of conversational maxims, then select from a range of direct and indirect strategies which also developed in speech act philosophy. available for realizing the act in question. Also, any of According to this theory, hearers assume that speakers these strategies can be modified in a number of ways. behave cooperatively at all times (cooperative principle). For instance, speakers can say, Could you perhaps take More specifically, hearers expect speakers to observe a out the garbage, darling?, where perhaps and darling set of conversational maxims, involving ‘Don’t lie’ serve to modify the basic strategy, Could you take out (maxim of quality), ‘Say as much as is necessary, no the garbage? Modifications are motivated by polite more and no less’ (maxim of quantity), ‘Be relevant’ considerations and aim at reducing the social risks (maxim of relation), and ‘Be brief’ (maxim of manner). involved in performing the act in question. If an utterance does not, for example, seem to relate to 870  PRAGMATICS what was said before, or if maybe more or less is said portive moves, head exchanges and supportive than is necessary, then the hearer tries to work out an exchanges, and so on. For example, the utterance Can interpretation that makes sense in the given context. This you lend me ten dollars? I’ve left my purse at home process is called ‘inferencing’. A pragmatic theory of consists of a head move, which is the request, and a communication and cognition that focuses on the supportive move, which is an explanation. process of understanding utterances is relevance theory. The type of macropragmatics sketched here is It is based on the conversational maxim of relation. based on British discourse analysis, which was devel- In everyday communication, people do not always oped in linguistics. An alternative approach is conver- observe the maxims of conversation. For example, sation analysis, which originates in American people lie, talk too much, or make obscure statements. sociology. In conversation analysis, the term ‘adjacen- Such behavior can be explained by assuming that a cy pair’ is used for the exchange unit. Another unit politeness principle may conflict with the cooperative identified in conversation analysis is the ‘turn’. principle. In a particular situation, conversationalists Conversationalists take turns at talk. A turn is defined may rate politeness maxims higher than the conversa- as a speaker’s contribution at a particular point in a tional maxims. For instance, a ‘white lie’, which vio- conversation. It is surrounded by speaker switches. lates the maxim of quality, can sometimes be more Turns may differ considerably in length. Proto- considerate than telling the truth. Generally speaking, typically, a turn consists of a second pair-part to maxims of politeness minimize unfavorable effects for match the previous speaker’s first pair-part and a first the hearer or increase favorable effects. For example, pair-part that must be matched by the next speaker. In positive evaluations expressed in compliments are conversation analysis, the mechanisms of turn-taking often maximized, whereas impositions involved in and turn-allocation for encounters involving two or requests are usually reduced. more speakers have been identified, and simultaneous Most theories of language use and most empirical speech and interruptions have also been investigated. approaches are speech act-based, i.e. they focus on Conversation analysis differs from discourse analysis utterances in isolation. Speech act-based pragmatics is in using only naturally occurring conversation, where- referred to as micropragmatics. Its shortcomings are as discourse analysis also works with other data types, remedied in macropragmatics, which provides a fuller including fictional dialogue. picture of language use by accounting for the dialogi- Most research in pragmatics deals with present-day cal and interactive nature of human communication. language. Recently, however, an interest has also Macropragmatic approaches do not focus on the developed in earlier periods. Two branches have speaker or hearer alone, but rather on participants in evolved, namely, historical pragmatics and diachronic communication who alternately adopt the role of pragmatics. Historical pragmatics deals with previous speaker or hearer. Also, they challenge the view that stages of language, e.g. Old English or Middle the speech act is the central communicative unit and English. By contrast, diachronic pragmatics studies claim that the ‘exchange’ is more important. An development across time. In both types, two alterna- exchange minimally consists of two speech acts. The tive perspectives can be adopted. On the one hand, the first act functions as an initiating ‘move’, and the sec- way in which a particular communicative function, ond act as a responding ‘move’. Moves indicate the e.g. a speech act, was realized at different times may interactive function of an act relative to the immediate be examined. On the other the researcher may start verbal context. Typical exchanges are question– with a particular expression, e.g. a routine formula, answer, offer–refusal, and greeting–greeting. Between and describe its uses in certain periods or in the course initiation and response, countering moves can occur, of history. It is not always possible to keep these two as in Can you take out the garbage? (initiating move) perspectives apart. Typical research topics in historical Why don’t you take it out yourself? (countering move) pragmatics include greetings in the fifteenth century Okay, okay. Calm down, I will (responding move). An and politeness strategies in Shakespeare’s plays. exchange is closed when an outcome has been Topics for diachronic pragmatics include, for instance, reached. insults in Old, Middle, and Modern English or changes In macropragmatics, complete communicative in apologizing through the ages. Although all histori- events such as conversations, interviews, debates, etc. cal sources are in the written mode, some include a are analyzed. Communicative events are considered considerable amount of information on spoken lan- hierarchical structures. The hierarchy, starting with the guage. For example, court reports may incorporate smallest unit, involves acts, moves, exchanges, detailed accounts of what defendants or witnesses sequences, and phases. At each of these levels, there actually said. Finally, recent and ongoing pragmatic are obligatory units (heads) and optional units (sup- change may be studied in e-mails and chat-room con- portives). Thus, we distinguish head moves and sup- versations, because netiquette is only just emerging. 871 PRAGMATICS Another recent development in pragmatics is the politeness phenomena, and discourse strategies on the study of synchronic variation in contemporary lan- basis of natural, elicited, and also fictional data. The com- guage use. Sex differences have been examined to parison of language use across languages and cultures is some extent, but regional, social, and ethnic variation also an important empirical research component. Applied has not received much attention in pragmatics. pragmatics focuses on problems of communication in Sociolinguistic research suggests that we can expect practical contexts. Most pragmatic research investigates speech act realizations, politeness markers, discourse present-day language, but recent developments also strategies, and turn-taking, etc., to vary across regions, include historical and diachronic perspectives. urban and rural communities, social classes, and eth- nic groups. This is an area for future investigation. References In short, pragmatics is the study of language use. Pragmatics is not a component of the language system, Blakemore, Diane. 1992. Understanding utterances: an intro- but it offers a different perspective on verbal phenomena. duction to pragmatics. Oxford, Cambridge, MA: Blackwell. Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper, It examines how linguistic resources are used in commu- (eds.) 1989. Cross-cultural pragmatics. Norwood, NJ: nication and investigates a speaker’s (or writer’s) inten- Ablex. tions and a hearer’s (or reader’s) interpretations. Jucker, Andreas, (ed.) 1995. Historical pragmatics. Amsterdam: Language use is considered a complex form of social Benjamins. action. The general question addressed in pragmatics is Leech, Geoffrey. 1983. Principles of pragmatics. London and New York: Longman. how language functions in the lives of human beings. The Levinson, Stephen. 1983. Pragmatics. Cambridge: Cambridge focus can be on utterances (micropragmatics) or on University Press. longer stretches of discourse (macropragmatics). Mey, Jacob (ed.) 1998. Concise encyclopedia of pragmatics. Pragmalinguistics is concerned with the structural Oxford: Elsevier. resources that a language provides for conveying particu- Mey, Jacob. 2001. Pragmatics: an introduction, 2nd edition. Oxford and Cambridge, MA: Blackwell. lar intentions, whereas sociopragmatics considers lan- Trosborg, Anna. 1995. Interlanguage pragmatics. Berlin and guage use relative to social situations. There is no New York: Mouton de Gruyter. coherent pragmatic theory to date. Theoretical pragmat- Verschueren, Jef. 1999. Understanding pragmatics. London and ics draws on a number of disciplines, including philoso- New York: Arnold. phy, sociology, and anthropology. Speech act theory, Wierzbicka, Anna. 1991. Cross-cultural pragmatics. Berlin and New York: Mouton de Gruyter. politeness theory, discourse analysis, and conversation KLAUS P. SCHNEIDER analysis provide major contributions. Empirical pragmat- ics involves the investigation of speech act realizations, See also Speech Acts Predication Although predication is extremely important in expressed in the clause—and its arguments—the contemporary linguistic research, it is a very old concept entities involved in the event—the semantic analysis of tracing back to Aristotelian logic, where it emerged to predication has been mainly interested in how this link- indicate the semantic relationship existing between the ing takes place. The most common answer to this ques- event described in a proposition and the individuals par- tion is found in Theta Theory—that is, the lexico- ticipating in such an event. Due to its origin in the semantic module of the transformational theoretical Western philosophic tradition, the linguistic analysis of framework known as Government and Binding. Since it predication has been approached for a long time from a is chiefly concerned with the predicate’s assignment of semantic perspective. From the 1980s onward, however, theta roles to its different arguments, it specifies the its study took a different course due to the advent of sev- exact number of arguments a particular predicate eral transformational theories mostly devoted to exam- requires and their semantic nature; that is, if they take ine the syntactic behavior of predication. part in the predication as agents, patients, experiencers, Being traditionally accounted for in semantic terms themes, etc. It must be noted, however, that, being as the connection between a predicate—the event unlimited in number, there is no agreement as regards 872  PREDICATION the name and catalogue of these detailed semantic traits Although the previous syntactico-semantic alloted by a predicate to its arguments, generally called characterization applies to predication in general, some theta roles. Regardless, the main principles underlying refinements must be made because not all instances of Theta Theory are formulated in the well-known Theta- predication exhibit the same grammatical behavior. Criterion, which postulates the following restrictions on Predication cannot be considered, therefore, a the assignment of theta roles: (a) every argument in a homogeneous linguistic phenomenon, but rather a predication must receive one theta role from its predi- heterogenous one involving two distinct classes of pred- cate; (b) the same theta role cannot be assigned to more ication: primary and secondary predication. The former than one participant in the predication. According to corresponds to the prototypical notion of predication Theta Theory, then, and as illustrated in Mary gave because, being syntactic and semantically independent, some money to her little brother, a predicate like give it embodies the basic and matrix predicative relation- needs three arguments bearing the following theta roles ship occurring at the level of the clause: John painted to be completed: an agent (Mary), a theme (some the house, John eats carrots. The latter lacks, in turn, money), and a recipient (her little brother). such syntactico-semantic autonomy and must always From the syntactic viewpoint, predication is defined appear, consequently, complementing an instance of via saturation: a predicate is a maximal projection— primary predication: John painted the house red, John that is, a full phrasal category—and, consequently, an eats the carrots raw. The aforementioned contrast is, open function, which must be syntactically closed by furthermore, extremely important because it determines the appropriate arguments to form a grammatically the status of the subject of each type of predication: acceptable chain. Despite being the maximal projection whereas the subject of a primary predication is system- with the strongest predicative capacity, the verbal atically, like John in the previous examples, the predi- phrase is not the only one able to function as predicate: cate’s theta-marked external argument, the subject of a The girl read her cousin a book. The adjectival (Bill is secondary predication displays a dual function: it is the fond of Jan), the nominal (This ratty piece of leather is secondary predicate’s theta-marked external argument a wallet), the prepositional (Those sisters are on at the same time as a theta-marked argument of the pri- drugs), and the adverbial phrases (She feels a bit down mary predication; notice, for instance, that in the exam- today) can also perform, as illustrated between brack- ples above, the house and the carrots are, on the one ets, the predicative function. Although different predi- hand, the subjects of the secondary predicates red and cates demand a different number of arguments, as the raw, and on the other, the direct objects of the primary previous examples show, they all share the property of predicates painted and eats. having, at least, one compulsory argument: their sub- Besides these contrasts, two semantic types of ject. This argument (The girl, Bill, This ratty piece of secondary predication must be distinguished: resultative leather, Those sisters and She in the previous examples) and depictive secondary predication. The former, as differs from the rest of possible arguments that a pred- illustrated by the house red in the previous example, icate may have both semantically and syntactically: signals a potential result of the action denoted by the semantically, because it constitutes the starting point verb and can only be predicated by the direct object for the predication; that is, it is the participant the pred- argument of the primary predication; the latter, in turn, ication is about; and syntactically, because it is the does not maintain any semantic relation with the verbal predicate’s only external argument. Although the sub- predicate it complements, because, as shown by raw in ject and predicate constituents are unanimously said to the aforementioned example, depictive secondary pred- be in a relation of mutual c-command—that is, they are icates denote intrinsic and temporary properties of the syntactic sisters dominated by exactly the same maxi- subject or direct object arguments of the primary predi- mal projections—the two main transformational cation, which must exist at the time of the verbal action. approaches to the study of predication explain the sub- ject–predicate syntactic relationship in different ways: References the Predication Theory, for instance, considers that subject and predicate are independent structural com- Aarts, Bas. 1995. Secondary predicates in English. The verb in contemporary English. Theory and practice, ed. by Bas ponents, whose syntactic relation has to be marked by Aarts and Charles F. Meyer. Cambridge: University Press. indexing at surface structure, and the Clausal Theory Dik, Simon C. 1981. Predication and expression: the problem defends, in turn, that they both form a single con- and the theoretical framework. Predication and expression stituent having clausal status either with or without in functional grammar, ed. by A. Machtelt Bolkestein, Henk verb. If this is the case, the clause containing the pred- A. Combé, Simon C. Dik, Casper de Groot, Jadranka Gvozdanovi_, Albert Rijksbaron, and Co Vet. London: ication relationship is catalogued as small. An example Academic Press. of a small clause is, thus, represented by Jill a nice per- Frawley, William. 1992. Linguistic semantics. New Jersey: son in Rachel considers Jill a nice person. Lawrence Erlbaum Associates. 873 PREDICATION Nakajima, Heizo. 1990. Secondary predication. The Linguistic Stowell, Tim. 1983. Subject across categories. The Linguistic Review 7. 225–309. Review 2. 285–312. Napoli, Donna Jo. 1989. Predication theory. A case study Williams, Edwin S. 1980. Predication. Linguistic Inquiry 11. 1. for indexing theory. Cambridge: Cambridge University 203–38. Press. Williams, Edwin S. 1983. Against small clauses. Linguistic Rapoport, Tova R. 1990. Secondary predication and the lexical Inquiry 14. 2. 237–308. representation of verbs. Machine Translation 5. 31–55. BEATRIZ RODRÍGUEZ-ARRIZABALAGA Rothstein, Susan D. 1985. The syntactic forms of predication. Bloomington: Indiana University Linguistics Club. See also Thematic Structure Professions for Linguists Linguists study aspects of language, languages, and The sounds of language are the focus of phoneticians language use. With this diversity of focus, linguists in the fields of acoustic phonetics (e.g. the study of the may be found in many different professions. Linguists physical properties of speech sounds using instrumental are used in both private and public sectors, in jobs techniques of investigation to provide an objective related directly to linguistics, as well as in professions account of speech patterns, which can be related to the where a knowledge of language or language use is way sounds are produced and heard), auditory phonet- highly relevant, although not directly required. ics (e.g. the study of the way people perceive sound, as The nature of language has interested scholars for mediated by the ear, auditory nerve, and brain), or artic- over 2,000 years. The earliest works of a linguistic ulatory phonetics (e.g. the study of the use of the vocal nature date to some of the first writings known to organs (articulators) to produce the sounds of speech). humankind. Such early linguistic writings include dis- The sound systems of language are of interest to course on the nature of language and humanity (e.g. phonologists, who study the sound systems of lan- Socrates), the reconstruction of the spoken language guages and the general or universal properties dis- of the Vedas (e.g. Panini in India), and the translation played by these systems. of texts and codification of languages (e.g. the trilin- Phonetics and phonology are also relevant to the gual Rosetta Stone, Varro’s grammar of Latin). The teaching of pronunciation, speech pathology and description of child language acquisition has also fas- speech therapy, accent reduction, voice and drama cinated scholars through the ages, including scientists coaching, elocution and public speaking, forensic lin- (e.g. Charles Darwin), psychologists (e.g. Piaget), guistics (e.g. voice recognition, voice printing), sound philosophers (e.g. Rousseau), and others. In fact, the engineering, and the growing area of speech synthesis disciplines within which linguists have worked have and voice activation in the computer industry. included philosopher (e.g. Plato, Locke, Hume), gram- Syntactians study the structure (or grammar) of lan- marian (e.g. Dionysius Thrax, Cicero), philologist guages, including the way words are combined to form (e.g. Saussure), anthropologist and field linguist (e.g. sentences, relationships between the elements of sen- Boas, Sapir), psychologist (e.g. Skinner, Piaget), sci- tence structure, and the rules governing the arrange- entist (e.g. Bloomfield), and ‘linguist’ within an inde- ment of sentences in sequences. Meaning in language, pendent field of research (e.g. Chomsky). in terms of relations such as synonymy and antonymy, In their professions, linguists may focus on the is the domain of semanticians. Issues such as to what sounds of the language (e.g. phonetics, phonology), the extent semantic concepts are universal is a current area words of the language (e.g. semantics), how words of focus. combine (e.g. syntax, discourse), how speakers acquire In addition to academic research, syntacticians and and/or use language (e.g. sociolinguistics, psycholin- semanticists work on dictionaries and grammars of guistics, applied linguistics, dialectology), how the different languages, develop and study computer lan- human brain processes language (e.g. psycholinguistics, guages and computer translation, preserve endangered neurolinguistics), how to program machines to carry out and dying languages through field research, provide voice recognition/voice synthesis (e.g. computational forensic analysis for the legal system, develop lan- linguistics), or even how animal communication sys- guage tests, develop language policy and other educa- tems, or languages, work (e.g. zoosemiotics). tional curriculum, work in language reconstruction 874  PROFICIENCY TESTING (with archeologists), etc. Linguists also work for intel- ods, and observations); pragmatics (the study of the ligence agencies as spies, translators, interpreters, conditions on language use deriving from the social code-breakers, and computer programmers. situation); semiology (the study of signs and their use, Professions for linguists related to the acquisition of focusing on the mechanisms and patterns of human first/child language include developmental linguistics communication and on the nature and acquisition of and developmental psychology, psycholinguistics, knowledge); semiotics (the study of language as one early childhood education, deaf education, teaching, type of sign system, along with bodily gestures, cloth- speech therapy, speech pathology, textbook and cur- ing, and the arts); sociolinguistics (the study of how riculum development, and software development. language is integrated with human society with refer- Similarly, professions for linguists related to the acqui- ence to race, ethnicity, class, sex, and social institu- sition of second or additional languages include tions); and zoosemiotics (the study of animal applied linguistic research, ESL/EFL teaching, foreign communication systems). and modern language teaching, bilingual/bicultural In the new millennium, with globalization and the education, teacher training, translation, proofreading, explosion of new communication technologies, the marketing, textbook writing and publishing, editing, knowledge and expertise of linguists are increasingly speech therapy, speech pathology, software develop- in demand in the fields of education, business, technol- ment, website design and marketing, on-line teaching, ogy, health, law, and communications. The field that recruiting students for language schools (e.g. agents), began with philosophical explorations of language and owning language schools and other related businesses, humanity is now recognized as being in the forefront of developing and implementing multicultural policies, human/machine interfaces in the twenty-first century. facilitating minority language rights, language policy, language planning in business and government servic- References es, and as freelance consultants (e.g. providing pro- gram development/analysis, giving seminars on gender Brand, Amy. A short road from academic research to academic publishing: one linguist’s trip. Linguistic Enterprises issues in the workplace, cross-cultural training, etc.). Advice. http://web.gc.cuny.edu/linguistics/enter/elgin.htm. Some additional areas within linguistics in which Crystal, David. 1987. The Cambridge encyclopedia of lan- linguists are employed include anthropological lin- guage. Cambridge, UK: Cambridge University Press. guistics (the study of languages in relation to social Haden Elgin, Suzette. Another item in the freelance linguist’s and cultural patterns and beliefs); areal linguistics (the survival kit: Seminars, Linguistic Enterprises Advice, http:// web.gc.cuny.edu/linguistics/enter/elgin.htm study of geographical areas that are characterized by Levi, Judith, Questions and answers about language consulting shared linguistic properties); audiology (the study of with lawyers, Linguistic Enterprises Advice, http://web.gc. hearing and hearing disorders) biolinguistics (the cuny.edu/linguistics/enter/elgin.htm study of the biological preconditions for language O’Grady, William, and Michael Dobrovolsky. 1996. development and use); dialectology (the study of Contemporary linguistic analysis. Toronto, CA: Copp Clark Ltd. dialects, dialect geography, or linguistic geography); Wylie, Dovie. But Mom, who is going to pay you to know these historical linguistics (the study of language variation things? Linguistic Enterprises Advice. http://web.gc.cuny. and/or change); neurolinguistics (the study of the basis edu/linguistics/enter/elgin.htm in the human nervous system for language develop- Other Work-Related Websites ment and use); philology (the study of language histo- Dave’s ESL Café: http://www.eslcafe.com/search/jobs/ Linguist List: http://www.emich.edu/
linguist/jobsindex.html ry, including historical study of literary texts); American Association of Applied Linguistics: http://www. philosophical linguistics (the study of the role of aaal.org language in relation to philosophical concepts, as well Linguistic Enterprises: http://web.gc.cuny.edu/linguistics/enter/ as the philosophical status of linguistic theories, meth- KAREN WOODMAN Proficiency Testing The term proficiency testing embodies two main view- ability to use a language for some extralinguistic points: an earlier, traditional one referring to tests of purpose, and the more recent ‘Proficiency Movement’ 875 PROFICIENCY TESTING in foreign language assessment associated with the English Test (OET) was developed in Australia to ascer- American Council on the Teaching of Foreign tain whether migrant candidates for licensure in the Languages (ACTFL). health professions are able to use English with sufficient Traditionally, proficiency testing has been contrast- ability to work with English-speaking patients and ed mainly with achievement testing in which the test clients; the Japanese Language Test for Tour Guides was content and method are based on a language teaching also developed in Australia; the Test of English for syllabus. The purpose of an achievement test is to find International Communication (TOEIC) is a US test out how much students have learned of what has been aimed primarily at international business people; and the taught. By contrast, proficiency tests are not linked to Proficiency Test in English for Air Traffic Controllers is any syllabus but rather attempt to measure how well a a European test for trainee air traffic control officers. learner can use the language regardless of the circum- These and other specific-purpose language tests were stances in which it was acquired. Lado (1961) was the developed without reference to any particular course of first to make this contrast explicit: instruction but rather on the basis of an analysis of a par- [Achievement] tests...attempt to measure how much of a ticular target language use situation (Douglas 2000). language a student knows. When the...tests are thought The distinction between achievement and proficien- of independently of the learning experience, they may cy testing is not entirely rigid, either. For example, then be referred to as proficiency tests. Proficiency tests Bachman (1990) argues as follows: measure how much of a foreign language a person (not necessarily a student) knows. (p. 369) Whether or not the specific abilities measured by a given proficiency test actually differ from those meas- Note that Lado speaks of the purpose of the profi- ured by a given achievement test will depend, of course, ciency test as measuring how much of the language a on the extent to which the theory upon which the profi- person knows, reflecting his ‘discrete point’ approach ciency test is based differs from that upon which the syl- to test development wherein specific grammatical labus is based (p. 71). points are tested more or less in isolation. At about the Note that Bachman defines proficiency tests as the- same time, other scholars were calling for proficiency ory-based, and indeed, it is generally the case that pro- tests that required language performance in contexts of ficiency tests are related to some theory of what it use. For example, Carroll (1961 [1972]) argued that means to know and use a language. Since Hymes’s an ideal English language proficiency test should make (1972) formulation of communicative competence, it possible to differentiate, to the greatest possible proficiency test developers have used theories about extent, levels of performance in those dimensions of the components of language knowledge to guide them performance which are relevant to the kinds of situa- in deciding what to test. The current, most well-known tions in which the examinees will find themselves after framework is that proposed by Bachman (1990) and being selected on the basis of the test. (p. 319) Bachman and Palmer (1996) of communicative lan- Thus, Carroll explicitly links proficiency testing to guage ability, consisting of grammatical, textual, func- performances relevant to future situations, and this has tional, and sociolinguistic language knowledge, plus become the standard view of proficiency: ability to use strategic competence, which directs the use of one’s a language for some extralinguistic purpose (Davies et language knowledge in actual communication. al. 1999). The so-called Proficiency Movement, although The extralinguistic purpose that proficiency testing more recent than proficiency testing itself, has its ori- is most often associated with is admission to academ- gins in the early days of the assessment of the oral lan- ic programs. For example, the two best-known profi- guage skills of US Government employees, most ciency tests, the Test of English as a Foreign Language notably the Foreign Service Institute (FSI), the Peace (TOEFL) and the International English Language Corps, the Central Intelligence Agency (CIA), and the Testing System (IELTS), are both used primarily for military (Lowe 1988). The FSI (now the Interagency admission to university programs in the United States, Language Roundtable [ILR]) Oral Proficiency the United Kingdom, and Australia. These tests were Interview became the standard for the testing of spo- developed to answer the question of whether a candi- ken language ability, and the American Council on the date for admission to an English-speaking academic Teaching of Foreign Languages referred to it in institution possesses sufficient English ability to cope responding to a mandate from the 1979 President’s with academic work. Commission on Foreign Languages and International However, proficiency tests need not be limited to aca- Studies to establish a series of nationally recognized demic situations only. The focus on future uses of the descriptors that would facilitate assessment of the pro- language naturally leads to tests of languages for specif- ficiency of both students and teachers in foreign lan- ic purposes (LSP). For example, the Occupational guage programs. The ACTFL Proficiency Guidelines 876  PRO-FORMS (1986, 1999) thus ‘represent the combined efforts of Bachman, Lyle, and Adrian Palmer. 1996. Language testing in groups of educators to provide an operational defini- practice. Oxford: Oxford University Press. Carroll, John B. 1972. Fundamental considerations in testing tion for [proficiency], to represent it as phenomena for English language proficiency of foreign students. observable and evolving’ (Galloway 1987: 26). The Teaching English as a second language: a book of readings, Guidelines are directed primarily at assessing the for- ed. by H.B. Allen and R.N. Campbell. New York: McGraw- eign language proficiency of college and university Hill. (Originally published in Testing, Washington, DC: students in the United States, and consist of descrip- Center for Applied Linguistics, 1961.) Davies, Alan, Annie Brown, Cathie Elder, Kathryn Hill, Tom tors of performance in speaking, listening, reading, Lumley, and Tim McNamara. 1999. Dictionary of language and writing at Superior, Advanced, Intermediate, and testing. Studies in language testing, Vol. 7. Cambridge: Novice levels, with the latter three levels divided into Cambridge University Press. high, mid, and low categories. Douglas, Dan. 2000. Assessing languages for specific purposes. The focus of the Proficiency Movement, as embod- Cambridge: Cambridge University Press. Galloway, Vickie. 1987. From defining to developing proficien- ied in the ACTFL Guidelines, is the assessment of lan- cy: a look at the decisions. Defining and developing profi- guage production or performance rather than ciency: guidelines, implications, and concepts, ed. by Heidi internalized language knowledge or ability for use Byrnes and Michael Canale. Lincolnwood, IL: National (Galloway 1987; Lowe 1988), thus distinguishing the Textbook Company. Movement from the more traditional understanding of Hymes, Dell. 1972. On communicative competence. Sociolinguistics, ed. by J.B. Pride and J. Holmes. the concept of proficiency described above. Harmondsworth, UK: Penguin. Lado, Robert. 1961. Language testing: the construction and use of foreign language tests. London: Longman, Green, and References Company. American Council on the Teaching of Foreign Languages. Lowe, Pardee. 1988. The unassimilated history. Second lan- 1986. ACTFL proficiency guidelines. Hastings-on-Hudson, guage proficiency assessment: current issues, ed. by Pardee NY: American Council on the Teaching of Foreign Lowe and Charles Stansfield. Englewood Cliffs, NJ: Languages. Prentice-Hall Regents. American Council on the Teaching of Foreign Languages. 1999. McNamara, Tim. 1996. Measuring second language perform- ACTFL proficiency guidelines—Speaking. Hastings-on- ance. London, New York: Longman. Hudson, NY: American Council on the Teaching of Foreign DAN DOUGLAS Languages. Bachman, Lyle. 1990. Fundamental considerations in language See also Second Language: Learning; Second Lan- testing. Oxford: Oxford University Press. guage: Teaching Pro-forms In linguistics, a pro-form is generally taken as an ele- and one/ones, and adjectives by such and verb phrases ment used in place of other linguistic element(s). by do. Pronouns are the most common pro-forms used to sub- The term ‘pro-form’ was probably first used by stitute for a noun or a noun phrase. ‘She’, for example, Jerrold Katz and Paul Postal (1964) as a mechanism to is a third-person singular pronoun used in place of a explain both syntactic and semantic aspects of the sub- singular human female animate noun such as a woman stitutions in the above examples. Syntactically, the as in, ‘A woman is coming to see you. She has called pro-constituent guarantees the recoverability of a sub- you earlier.’ stitution or deletion. Semantically, the pro-form calls Although pronouns have often been used as exam- for interpretation by retrieving its equivalents. ples of pro-forms, there are other linguistic elements The term pro-form, since its introduction, has often that have comparable properties but do not substitute been used alternately with pronoun, and now it seems for a noun or a noun phrase. ‘So’, as in ‘He thinks it to replace pronoun. However, some linguistic ele- will rain tonight but I don’t think so,’ substitutes for a ments seem to have comparable properties to pro- whole clause, i.e. ‘it will rain’. nouns but they are not substitutes for nouns. In fact, Here and there are pro-forms for locative expres- there are many other classes of words than nouns that sions; bare noun phrases may be substituted with what get a different form in the following mention in a text. 877 PRO-FORMS Thus, the term pro-form seems appropriate to be used substitutions. However, some recent work in natural collectively for any kind of substitution. language processing (e.g. by Daniel Hardt 1993) In the literature of generative grammars, a pro-form includes pro-form as one category of elliptical forms. is often found as an element that assumes the process In terms of semantics, a pro-form has no meaning of substitution. In these theories, pro-forms can be in itself, rather it requires a retrieval of meaning from used as one of the key tests for constituency in syntax: a previously mentioned element, or antecedent, i.e. the if a string of words can be replaced by a pro-form, element for which it substitutes. In other words, pro- these words form a constituent, i.e. a structural unit forms are semantically bound by other elements. within the sentence. To illustrate, in the sentence ‘The student read a book in the library,’ the pronoun she can References substitute for the noun phrase ‘the student’, a there can Hardt, Daniel. 1993. Verb phrase ellipsis: form, meaning and stand in for the prepositional phrase ‘in the library’, processing. Ph.D. dissertation. University of Pensylvania. and a did can substitute for the verb phrase ‘read a Katz, Jerrold, and Paul Postal. 1964. An integrated theory of book in the library’. Consequently, we may conclude linguistic descriptions. Cambridge: Cambridge University that these three types of phrases are constituents. Press. Moreover, there are some other terms that are Radford, Andrew 1997. Syntactic theory and the structure of English: a minimalist approach. Cambridge: Cambridge loosely used in place of pro-forms. One of these is University Press. ellipsis. Andrew Radford (1997) considers ellipsis a Van Valin, Robert D. Jr. 2001. Introduction to syntax. process by which redundant information in a sentence Cambridge: Cambridge University Press. is omitted. Pro-forms, however, are not omissions but PRANEE KULLAVANIJAYA Proper Nouns Proper nouns (usually known as ‘proper names’ by by Victoria. It does not indicate any attributes as philosophers) are basically names of specific people, belonging to these individuals. Sometimes, we do not places, organizations, months, festivals, and so forth. even know whether it is a woman’s name or a place Typical examples are John, London, Oxfam, January, name. With reference to the semantic information and Christmas. The investigation of the semantic about the entities, pure proper nouns are lexically meaning of proper nouns goes back to the Greeks. In opaque and the entities they designate are normally 200 BCE, the Greek grammarian Dionysius Thrax unpredictable. On the other hand, a common noun is gave one of the most often used definitions: a proper used to subsume similar entities under a generic noun signifies an individual being, whereas a common concept. For example, various books, irrespective of noun signifies a general substance. As time has gone material, size, or purpose, are subsumed under the by, philosophers (e.g. John Stuart Mill, Gottlob Frege, generic concept book. Thus, the use of a common Bertrand Russell, and Saul Kripke) and linguists (e.g. noun for an entity is completely determined by the Otto Jespersen, Alan Gardiner, and John Algeo) have meaning of the noun. Common nouns such as chair put forward a number of criteria for the semantic are semantically transparent because they have lexical description of proper nouns with reference to the tra- meaning and carry information. ditional assumption of a proper–common dichotomy. It has been argued that while proper nouns have no One of the earliest and most famous criteria was meaning in isolation, they carry descriptive meaning introduced by the English philosopher John Stuart when applied in a specific context to a particular per- Mill in A system of logic (1843). He suggests that a son or place. For example, the proper noun John is used pure proper noun is a denotative sign without connota- in English-speaking countries as a name for a male. tion. In contrast to a common noun that both denotes However, this culturally specific information should and implies attributes, a proper noun is used to denote not be identified with the meaning carried by a com- a specific or unique individual entity regardless of mon noun. The information associated with a proper what properties it has. For example, a proper noun like noun is usually changeable, whereas the meaning of a Victoria simply denotes the individuals who are called common noun is basically constant. 878  PROPER NOUNS Within a given context of utterance, proper nouns While numerous philosophers and linguists have together with pronouns are two distinct kinds of sin- considered proper nouns and common nouns as two gular definite referring expressions that enable the discrete categories and the boundary between them as hearer to denote the actual referent from the class of unambiguous, Jespersen (1924) holds that no sharp potential referents. Unlike pronouns that have some line can be drawn between proper and common nouns, descriptive meaning, proper nouns have reference but the difference being one of degree rather than of kind. not sense. They identify their referents by making use Later, Quirk et al. (1985) explicitly stated that the class of the unique and arbitrary association that holds of proper nouns has unclear boundaries and the degree between a proper noun and its referent. Moreover, of membership involves the notion of gradience. There unlike pronouns that are used by different speakers to are some circumstances in which proper nouns (e.g. denote different entities in different contexts, proper the Avon, the Crimea, the Himalayas) behave like nouns are always claimed in any particular context to common nouns by taking the definite article and the refer to one and only one entity, although a proper plural form. On the other hand, a number of common noun may name many individuals. It follows that a nouns with unique denotation are sometimes capital- proper noun merely serves as an identification mark by ized and thus enjoy a similar semantic function as singling out an entity from among similar entities. Its proper nouns e.g. Fortune, Wealth, Fate. function is precisely that of labeling. In the 1970s, the American philosopher Saul Kripke References (1972, 1980) espoused the view that proper nouns are ‘rigid designators’ that designate the same individual Algeo, John. 1973. On defining the proper name. Gainesville, in all possible states of the world whatever their char- FL: University of Florida Press. Frege, Gottlob. 1892. Über Sinn und Bedeutung. Zeitschrift für acteristics happen to be. According to Kripke, the Philosophie und philosophische Kritik, Vol. 100; as On sense proper name John is a rigid designator because it picks and reference, translated by Max Black, in Translations from out particular individuals who happen to be John the Philosophical Writings of Gottlob Frege, ed. by Peter without any defining characteristics in whatever state Geach and Max Black. Oxford: Basil Blackwell, 1960. of affairs we entertain. However, the common noun Gardiner, Alan. 1940. The theory of proper names: a controver- sy essay. London: Oxford University Press; 2nd edition, chair refers to the same thing in the world that happens 1954. to fit the defining characteristics. Jespersen, Otto. 1924. The philosophy of grammar, London: In some languages such as English, it has long been Allen and Unwin; New York: Norton, 1965; Chicago and held that the definition for typical proper nouns should London: The University of Chicago Press, 1992. also be laid down on the basis of three additional cri- Kripke, Saul A. 1972. Naming and necessity. Semantics of nat- ural language, ed. by G. Harman and D. Davidson. teria: orthographic, morphological, and syntactic. Dordrecht, Netherlands and Boston, USA: D. Reidel Orthographically, it is a graphic convention that Publishing Co.; as Naming and necessity. Oxford: Basil English proper nouns are marked with initial capital- Blackwell, revised edition, 1980. ization. By the morphological criterion (i.e. the rela- Lyons, John. 1977. Semantics, 2 vols. Cambridge: Cambridge tionships between the parts that make up the proper University Press. Marmaridou, Sophia A.S. 1991. What’s so proper about names? noun itself when it is more than a single morpheme), a A study in categorisation and cognitive semantics. Athens: typical English proper noun has no plural suffix in the University of Athens. strictest sense e.g. Britain—*Britains. By the syntactic Mill, John Stuart. 1843. A system of logic, ratiocinative and criteria (i.e. the relationships of the proper noun with inductive, being a connected view of the principles of evi- other items in the same grammatical construction), dence, and the methods of scientific investigation. London: John W. Parker; in Collected works of John Stuart Mill, Vol. typical English proper nouns are not preceded by the VII, ed. by J.M. Robson. Toronto and Buffalo: University of article or other determiners since the references that Toronto Press, and London: Routledge and Kegan Paul, 1973. proper nouns have is by their very nature self-deter- Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan mining or inherently definite. Unlike common nouns, Svartvik. 1985. A comprehensive grammar of the English proper nouns normally lack the contrastive definite- language. London and New York: Longman. Russell, Bertrand. 1950. An inquiry into meaning and truth. ness (e.g. *an America, *some Americas; an orange, London: George Allen and Unwin; London and New York: some oranges). Another major syntactic criterion is Routledge, revised edition, 1995. that proper nouns lack modification. When they have Ullmann, Stephen. 1962. Semantics: an introduction to the sci- the normal unique denotation, they can only accept ence of meaning. Oxford: Basil Blackwell. nonrestrictive modifiers (e.g. Mary, who is my sister, is Zabeeh, Farhang. 1968. What is in a name? An inquiry into the semantics and pragmatics of proper names. The Hague: studying abroad—*Mary who is my sister is studying Marinus Nijhoff. abroad.). Proper nouns require the when they are mod- GRACE YUEN WAH TSE ified by a restrictive clause (e.g. The Mary I know is my friend’s wife), but not when they are unrestricted. See also Naming; Reference 879 PROSODY Prosody Prosody refers to the (variations in) rhythm, intona- subglottal pressure and an increase in the tension of tion, and stress patterns in speech. Its role is to empha- the laryngeal muscles (thereby increasing the stiffness size words, to segment a sentence into smaller units, to of the folds and reducing their local mass) result in an change the meaning of an utterance (not of a speech increase in the fundamental frequency. Prosody can be sound!), and to resolve syntactic disambiguity (e.g. used grammatically, by producing a question with a ‘They fed her dog biscuits’ Who ate the biscuits? She rising pitch, and a statement with a falling/flat pitch. or the dog?). Apart from these language-dependent Intonation also helps to group words, especially in aspects, prosody is also used in a more universal man- syntactically ambiguous phrases (e.g. a light house ner, to provide cues to the state of the speaker and to keeper). However, there is quite a complex interaction convey the speaker’s mood (happiness, impatience, between stress, intonation, and duration in the case of boredom, etc.). Prosodic patterns are very important syntactic ambiguities. At some word boundaries, such during speech development. They appear in the as between ‘gray tie’ vs. ‘great eye’, duration is the infant’s speech long before the first word: babies often primary cue, although the ambiguity can also be speak nonsense with the correct intonation! Moreover, resolved through differentiation in intonation. as boundaries between words are not clearly marked in speaking, infants use prosodic information to ‘seg- Stress ment out’ words in the continuous stream of speech sounds. This process is termed ‘prosodic boot-strap- Stress usually refers to relative prominence and ping’. And the importance of prosody also becomes increased intensity due to increased physical effort. clear when listening to foreign speakers: even if they Note, however, that stress is not perceived as loud- articulate well, they are not easily understood if their ness, and that intensity is not the only acoustic corre- use of stress and timing is incorrect. The same can be late of greater effort. An increase in stress often co- said of deaf speakers: their speech is often unintelligi- occurs with an increase in fundamental frequency. ble because they cannot master the prosody of their After all, both stress and intonation are brought about language well. In noisy environments prosody helps by an increase in respiratory effort and subglottal the listener to understand the speaker. Moreover, spec- pressure. The minimum size of the unit of stress trally distorted speech with the original prosody is eas- placement is the syllable. At the word level, stress on ier to understand than the original speech signal with the first or second syllable is used to differentiate distorted prosody. Prosody does not only contribute meaning (cf: ‘DIgest’ and ‘to diGEST’ or RECord and significantly to the intelligibility of speech but also to to reCORD). This is termed lexial stress. At the sen- its naturalness. Although present-day speech synthesis tence level, stress is used to draw attention to certain is highly intelligible, it still sounds artificial. More words (emphatic stress) or to differentiate between research on speech prosody is needed to make it sound two words (contrastive stress). Languages are spoken natural or pleasant enough to listen to for long stretch- with a certain rhythm. English is a stressed-timed lan- es of time. guage due to the fact that stressed syllables occur at The most important prosodic features are intona- regular intervals. French and Japanese are referred to tion, stress, and quantity. The acoustic correlates of as syllable-timed languages because the duration of these features are fundamental frequency (perceived as their syllables remains relatively constant. Stressed pitch), intensity, and duration (although there is no syllables have a higher intensity and fundamental fre- one-to-one correspondence). quency than their unstressed counterparts. Apart from these voice source factors, the timing of the articula- tory movements of the vowels and consonants pro- Intonation duces differences in duration, depending on the Variations in tone (word level) and intonation (sen- position of the utterance in a word or phrase. In many tence level) are used to change the meaning of words, languages, stressed syllables are longer than to change the function of the utterance (declarative vs. unstressed ones. The acoustical correlates of stress interrogative), to signify attitude and emotion, etc. (especially duration) are, however, language depend- These changes in fundamental frequency are related to ent: in Czech, there is hardly an increase in duration the rate of vibration of the vocal folds. An increase in on stressed syllables. 880  PSYCHOLINGUISTICS Quantity However, significant changes in tempo may convey information about the mood of the speaker or the cir- Quantity is the linguistic term for ‘constrastive dura- cumstances under which the utterance is made. tion’. Each individual speech sound has an ‘intrinsic duration’ to specify its identity, but the duration of speech segments or larger units can yield contrastive References information, depending on context or the phonological Cooper, W., S. Eady, and P. Mueller. 1985. Acoustic aspects of system of the language. For example, the duration of contrastive stress in question–answer context. Journal of the vowels in Finnish can be constrastively short and long. Acoustical Society of America 77. 2124–56. In Norwegian, a long vowel is followed by a short con- Cutler, A., and D.R. Ladd. 1983. Prosody: models and meas- urement. New York: Springer-Verlag. sonant and vice versa. In English, the duration of a Cutler, A. 1991. Linguistic rhythm and speech segmentation. vowel preceding a voiced consonant is 1.5 times Music, language, speech, and brain, ed. by J. Sundberg, L. greater than that of the same vowel preceding a voice- Nord, and R. Carlson. Houndmills and London: MacMillan. less consonant (sight197 ms, side297 ms; Peterson Lehiste, I. 1970. Suprasegmentals. Cambridge, MA: MIT Press. and Lehiste 1960). Vowel duration, and consonant Lehiste, I. 1972. The timing of utterances and linguistic bound- aries. Journal of the Acoustical Society of America 51. duration to a lesser extent, is also affected by the num- 2018–24. ber of syllables in a word, by syllable stress (where it Lehiste, I. 1996. Suprasegmental features of speech. Principles interacts with fundamental frequency), and by the type of experimental phonetics. Mosby-Year Book, 226–44. St of word (function words such as ‘in’ or ‘on’ are rela- Louis, MO: Mosby. tively short). Very often, the last syllable of a word or Maeda, S. 1976. A characterization of American English into- nation. Cambridge, MA: MIT Press. the last word of an utterance is longer than the same Peterson, G.E., and I. Lehiste. (1960). Duration of syllable syllable or word in other positions. This temporal nuclei in English. Journal of the Acoustical Society of effect, called prepausal lengthening, results from a America 32. 693–703. slowing down in the speaking rate and it signals phrase Terken, J. 1991. Fundamental frequency and perceived promi- and sentence boundaries. At word level, duration can nence of accented syllables. Journal of the Acoustical Society of America 89. 1768–76. signal the syntactic structure of a sentence, e.g. ‘a gray ASTRID VAN WIERINGEN tie’ vs. ‘a great eye’. At a sentence level, relative dura- tion does not affect the meaning of individual words. See also Soundwave Analysis Psycholinguistics Psycholinguistics is the discipline of psychology that pheme, and the clause. Psycholinguistic experiments studies the mental mechanisms of language process- provide information about the psychological reality of ing—speaking, listening, reading, and writing—in linguistic units and the way linguistic information is both a native and a second tongue. Psycholinguistics represented and processed in the mind of the language also studies the processes underlying the acquisition of user. language, how language processes break down in lan- A crucial difference between linguistics and psy- guage pathologies such as dyslexia and aphasia, and cholinguistics is the latter’s focus on the mental how these processes relate to brain function. process. For instance, in natural speech we can easily Psycholinguistics borrows many of its theoretical produce two words per second. How do speakers find constructs from linguistics. Levels of processing, dis- these words so quickly in their (vast) mental lexicon, tinguished in theories of language comprehension or which contains at least 10,000 words? What happens language production, correspond to linguistic levels, to these processes when we are unable to say a word, such as semantics (meaning), the lexicon (vocabu- but it is on the tip of our tongue? As another example, lary), syntax (sentence structure), morphology (units when listening to sentences, we can sometimes be led that make up words), phonology, and phonetics up the ‘garden path’. A classic example is ‘the horse (sound systems). Furthermore, the processing units raced past the barn fell’. Here, one builds up a struc- that these theories assume correspond to linguistic tural representation of the sentence, but upon reading units, such as the phoneme, the syllable, the mor- the last word (‘fell’), it turns out to be wrong. How do 881 PSYCHOLINGUISTICS readers build up a structural representation? What Research in the 1960s and 1970s embraced went wrong with this process in the example? Chomsky’s theory of transformational grammar. This The scientific inquiry into language processing theory provided representations of sentences (called began in the second half of the nineteenth century. deep structure and surface structure), and procedures Rudolf Meringer (1859–1931) started to collect that transformed these representations. However, little speech errors. In contrast to his contemporary convincing evidence was found for a correspondence Sigmund Freud (1856–1939), he studied the linguistic, between linguistic transformations and psychological rather than psychological, properties of such incidents. processes, and this approach was largely abandoned in Wilhelm Wundt (1832–1920) proposed the first theory the early 1980s. Research in the 1980s and the 1990s of language production, and Gustav Aschaffenburg was characterized by three major developments. First, (1866–1944), by using the technique of word associa- Jerry Fodor published his The modularity of mind tion, investigated the representation of word meaning. (1983), in which he presented the theory that many The Dutch ophthalmologist Fransiscus Donders cognitive processes are performed by dedicated pro- (1818–1889) introduced the method of mental cessing systems (modules) that operate independent of chronometry (measuring the time mental processes other processing systems. A large number of studies in take). Reaction time studies are still the most common modern psycholinguistics have been and still are cen- methodology in psycholinguistics. tered on this theory. Second, psycholinguists designed Around the same time, the French neurologist Paul computer models of language processing. Seminal Broca (1824–1880) tested a patient who had lost his work in this area was done by James McClelland and ability to speak. Monsieur Tan, as the patient was David Rummelhart (1989), in particular, on the mod- known, could produce only a single syllable (‘tan’). eling of word perception. Third, technical advances After Mr. Tan’s death, Broca examined the brain of made it possible to measure the activity of the brain this patient and discovered massive damage to an area while it was engaged in verbal tasks. Although still in in the front part of the brain, which is now known as its infancy, the use of such techniques can be consid- Broca’s area. Broca’s discovery, and subsequent stud- ered very promising in studying the representation and ies by the German neurologist Carl Wernicke processing of linguistic information in the brain. (1848–1905), marked the beginning of the inquiry into What are the mental processes involved in using the relationship between language and brain. language? In the following paragraphs, a brief In the first half of the twentieth century, psychology overview will be presented of theoretical issues in piv- was dominated by behaviorism. This approach rejected otal language processes: comprehending, speaking, the notion of ‘mental representation’ and ‘mental and ‘control’ processes. process’ but focused on what is directly observable: The first challenge in comprehending is to identify behavior. An exponent of behaviorism was Burrhus F. words. This is no easy task: the speech signal has to be Skinner (1904–1990), who published his Verbal behav- distinguished from noise, which may distort it. The ior in 1957. In Skinner’s approach, the language-acquir- speech signal is also very variable. One way of dealing ing child begins without any linguistic knowledge. It with this variation is ‘categorical perception’. That is, learns language by experience only. Sentences would be tokens of a speech sound can be acoustically quite dif- produced as associative chains, where one word would ferent, but they tend to be perceived as the same sound be the stimulus that triggered the next word. (e.g. as the speech sounds /b/ or /p/, but never some- A few years later, a young linguist from the thing in between). Massachusetts Institute of Technology, Noam Another challenge for the listener is that in contin- Chomsky (1928–), published an extremely sharp criti- uous speech, there are often no pauses between words. cism of Skinner’s book. Chomsky argued that How then do listeners know where one word ends and Skinner’s theory could not work—for instance, a child the next one begins? Psycholinguistic research shows is exposed to far too few examples of sentences to that this depends on the language: different languages learn the language by experience only. Instead, provide different cues to word onset (e.g. the pattern of Chomsky argued that the child is born with a tacit stress), and listeners use these cues. knowledge of ‘universal’ grammar, and learning Visual word recognition—reading words—poses means tuning this grammar to its own language. its own problems. English is an alphabetic language Around the same time, psychologists were aban- (as opposed to, e.g. Chinese). However, the correspon- doning behaviorism and became interested in linguis- dence between printed letters or groups of letters and tics again. In 1951, a conference was organized at sounds is often irregular (e.g. in the word ‘yacht’). which leading psychologists and linguists discussed This has led psycholinguists to distinguish two routes the possibilities of ‘merging’ the fields. At this confer- for reading: a lexical route, in which letters or groups ence, the term ‘psycholinguistics’ was launched. of letters directly access a word, and a sublexical 882  PSYCHOLINGUISTICS route, in which letters are transformed into a phono- is whether these data also imply that there are two sep- logical code that guides word recognition. The lexical arate lexical representations. route would be used for reading irregular words; the Another task of formulation is the construction of a sublexical route would be used for reading ‘nonwords’ sentence structure. A key finding is the observation (e.g. ‘climp’). Support for this theory comes not only that sentence structure is ‘persistent’. Speakers have a from distinct patterns of reading disorder (dyslexia) tendency to reuse a structure that has been recently but also from experiments in which the speed of read- used before, even if the two sentences do not share any ing irregular and regular words is measured. meaning, words, or prosody. This finding strongly sug- Once words are identified, we construct a structural gests that structured representations of sentences are representation of the sentence, so that we can infer who psychologically real. It also provides a technique for did what to whom. This process is called parsing. Many determining how sentences are mentally represented. studies of parsing have analyzed structural ambiguities. A final task of formulation is to retrieve pronuncia- In a sentence such as ‘the man hit the girl with the gui- tions. A major problem here is that words in real tar’, the guitar can be in the possession of the man (and speech do not correspond in a one-to-one fashion to used for hitting the girl), or it can be in the possession words in their citation form: in connected speech, of the girl being hit. These interpretations correspond to sounds from one word are often ‘merged’ with those of different sentence structures. Studies with ambiguous another (e.g. ‘give it’ will often be spoken as ‘gi-vit’). sentences have revealed some of the processes involved Another issue is whether certain representations that in parsing. A very important debate centers on the ques- are distinguished in theoretical phonology are psycho- tion of what kind of information is available to parsing. logically real. In some views, only syntactic information is used. The earlier discussion focused mainly on content- According to other views, parsing is influenced by a directed processes in language use. Yet, there is variety of factors (e.g. plausibility, or the frequency with increasing interest in control processes, which deter- which a word occurs in a certain structure). mine when the content-directed processes should Sentence understanding does not stop with parsing. engage, and in mental resources, such as memory and We have to integrate the word meanings and the sen- attention, which are the ‘fuel’ on which the content- tence structure into a sentence meaning, and we have directed processes run. to relate this meaning to prior discourse. In one study, First, language is predominantly used in interaction brain responses were measured to sentences such as ‘I with others. There is a growing body of evidence that spread my bread with socks’, and it was discovered partners in a dialogue coordinate with each other: that that such stimuli evoke distinct patterns in the elec- is, they converge on using the same linguistic struc- troencephalogram signal. This finding has proved use- tures, such as words, syntax, and description schemes, ful for studying these integration processes. to denote the same things in the external world. How Speaking starts with the decision to engage in a and why we accomplish such coordination is an communicative act. This is based on the speaker’s important item on the agenda for study. intentions, background knowledge, and a mental Second, the language processes may contain errors. model of previous dialogue. On the basis of these Fortunately, there are self-monitoring processes dedi- sources of information, the first step in speaking is the cated to intercepting errors and ‘repairing’ them. How generation of a message. This is the interface between these monitoring processes operate—and how it is thought and language: the message is not yet lan- possible that they do the job as fast as they do—is, at guage, but it is specific to language. In most theories present, a matter of debate. of language production, the message contains con- Third, the language user is constrained by factors cepts for the things and actions the sentence is about, such as attention and memory. In recent years, there has perspective (is ‘in front of’ relative to you or me?), and been an increasing interest in how limitations in such focus (which information is given, and which is new?). ‘resources’ affect processing. Debate here centers on the The next step consists of formulation of this mes- nature of such resources: whether they are comparable sage: retrieval of the words and construction of a sen- to the same short-term memory in which one briefly tence. There is much evidence that word retrieval stores the digits of a phone number, or whether there are consists of two steps: a process that is based on mean- specialized resources for language processing. ing and a process that is based on form. This distinc- tion is based on research that analyzed collections of References speech errors and research that measured reaction Carrol, David. W. 1999. Psychology of language. Pacific Grove, times in word production. In these tasks, so-called dis- CA: Brooks/Cole. tracter words with either a form or a meaning relation Harley, Trevor A. 1995, 2001. The psychology of language: from to the target word have different effects. A major issue data to theory, 2nd edition. Hove, UK: Psychology Press. 883 PSYCHOLINGUISTICS Fodor, Jerry A. 1983. The modularity of mind. London, Dijkstra, Ton, and Koenraad De Smedt (eds.) 1996. England and Cambridge, MA: MIT Press. Computational psycholinguistics. London, UK: Taylor & Gernsbacher, Morton-Ann (ed.) 1994. Handbook of psycholin- Francis. guistics. San Diego, CA: Academic Press. McClelland, James, L, and David E. Rummelhart. 1989. Parallel Levelt, Willem J.M. 1989. Speaking: from intention to articula- distributed processing: explorations in the microstructure of tion. London, England and Cambridge, MA: MIT Press. cognition, Vol. 2: Psychological and biological models. Brown, Colin M., and Peter Hagoort (eds.) 1999. The neurocog- London, England and Cambridge, MA: MIT Press. nition of language. Oxford, UK: Oxford University Press. ROBERT J. HARTSUIKER 884  Q Quantification A sentence such as John is happy consists of a subject noun phrase (i.e. the determiner and the noun) (John) and a predicate (is happy). The predicate attrib- contribute independently to the meaning of a sentence: utes the property it describes (‘being happy’) to the the determiner quantifies over the whole proposition individual the subject refers to (‘John’). When the sub- of which the noun is a part. This runs counter to the ject noun phrase contains a determiner (a, the, some, common linguistic analysis of sentential expressions. every, etc.), this element plays a crucial role in the Another limitation is that the first-order logic analysis predication process. The contrast in meaning between is not designed to capture the content of the extensive the sentences A student is happy and Every student is variety of natural language determiners. For example, happy is caused by the different nature of the deter- it has been shown that the English determiner most is miner that expresses the number or quantity of indi- not expressible in first-order logic. The sentence Most viduals to which the relevant property applies. We say, students are happy does not mean that most individu- then, that these sentences differ in quantificational als under consideration are students and they are force. The first sentence states that there is (at least) happy (first-order logical meaning). Rather, it means one student who is happy in the situation under con- that most of those individuals who are students are sideration. The second sentence states that every indi- also happy. vidual under consideration is happy. This semantic Generalized Quantifier Theory developed during contrast can be expressed as a contrast in logical form, the 1980s, following the initial contribution of the as defended by the philosophers Gottlob Frege and philosopher Richard Montague, as an attempt to Bertrand Russell. In first-order logic, the determiner improve the shortcomings of traditional logical analy- a/some is treated as an existential quantifier (meaning ses of quantification in natural language. The func- ‘there is at least one individual for which it is true tional nature of the subject—predicate relation that we that…’), and the determiner every as the universal described at the beginning is reversed. A predicate still quantifier (meaning ‘it is true for all individuals expresses a property, but it is not treated as a function that…’). Quantifiers are assumed to bind a variable (x) that applies directly to the subject (argument). Rather, in a logical statement or proposition. Going back to the a subject is viewed as a higher-order expression that above sentences, the paraphrases of their respective expresses a property of the predicate or, equivalently, logical forms would be the following: There is at least a set of properties. In general, noun phrases express one individual x such that x is a student and x is happy generalized quantifiers (sets of properties). We can and For every individual x if x is a student then x is represent the sentences in the previous paragraph com- happy. positionally as some student (happy) and every student The first-order logic analysis of quantification has (happy), where the generalized quantifiers associated been criticized for positing that the constituents of a with some student and every student are respectively 885 QUANTIFICATION interpreted as the set of properties that some student or Determiners are also characterized by the type of every student has, and what the sentence asserts is that inferences that they license, namely, set-to-subset happiness is one of those properties. The sentence will inferences or set-to-superset inferences. For example, be true or false depending on whether this is the case the sentence No students smoked entails the sentence or not. No students smoked cigars. The property denoted by The meaning of a noun phrase such as some stu- smoked cigars is a subset of the property denoted by dent can also be determined in a compositional fash- smoked, so the determiner no licenses set-to-subset ion. The noun student expresses or describes, more inferences for its second argument. The same is true technically ‘denotes’, a set of individuals and the for its first argument: the sentence No students smoked determiner some relates it with another set of individ- entails the sentence No female students smoked, given uals, namely, that denoted by happy. This yields the that the property denoted by female students is a sub- following logical representation: a/some (student) set of the property denoted by students. The determin- (happy). Given the particular meaning of the deter- er function some has the opposite pattern and licenses miner some, this expression is true if the intersection set-to-superset inferences for its two arguments. Thus, of the sets denoted by students and happy is not we predict that the sentence Some female students empty, i.e. if there is at least one individual who is a smoke cigars entails both Some female students smoke student and who is happy. Other determiner meanings and Some students smoke cigars. can be characterized in a similar fashion: every (stu- Quantifiers can also be classified according to dent) (happy) is true if the set denoted by student is a whether they can occur in an existential construc- subset of the set denoted by happy; no (student) tion, i.e. a sentence that asserts the existence of (happy) is true if the intersection of the sets denoted something (There is…). Consider the structure There by students and happy is empty; more than three (stu- is/are Q student(s) in the garden. The determiners dents) (happy) is true if the number of individuals in some, three, no, fewer than five, and many can be the intersection of the sets denoted by students and substituted for Q and occur in this construction, happy is greater than three; most (students) (happy) is whereas the occurrence of determiners such as true if the number of individuals in the intersection of every, most, and all but three would make the sen- those two sets is greater than the number of the indi- tence ungrammatical. The determiners that can viduals in the intersection of the sets denoted by stu- occur in an existential construction are intersective, dents and not happy, i.e. the number of those students i.e. they express a relation of intersection of their who are happy is greater than the number of those stu- two arguments. On the other hand, those determiners dents who are not happy; etc. In sum, a natural lan- that cannot occur in an existential construction are guage determiner is a function with two arguments. characterized by expressing a relation of inclusion, The first argument corresponds to the noun denotation such as every, or proportionality, such as most. and the second argument to the content of the sen- Definite determiners, whether simple (the) or com- tence verb phrase. plex (the ten…), demonstrative (this) determiners, Natural language determiners also satisfy a series of and possessive determiners (my) do not occur in constraints that set them apart from their logical coun- existential sentences either. These determiners are terparts. For example, all natural language determiners all inherently context dependent and thus presup- are conservative or live on their first argument—moti- pose or do not assert existence. vating the intuition that the determiner has a close bond When more than one quantifier occur in a sentence, with the noun. The sentence Some sailors left is equiv- a form of semantic interaction called scope arises. alent to the sentence Some sailors are sailors who left. Scopal relations are determined by the different order If we substitute any other determiner for some in this of quantifiers in the semantic representation of a sentence, the equivalence still holds. The effect of this clause. For example, the sentence Every student read constraint is to make natural language quantification a book is ambiguous because it contains two general- inherently restricted to the first argument of the deter- ized quantifiers: every student and a book. Under one miner. Thus, in checking whether some (sailors) (left) interpretation, every student read a different book. is true, we do not have to consider those individuals Under the second, there is a unique book such that who are not sailors. In other words, in processing the every student read it. This ambiguity is a genuine sentence Some sailors left, we do not first check who scope ambiguity. In the first reading, the scope order left and then determine whether any of these individu- of the quantifiers is the one that respects the linear als are sailors; rather, we look at relevant sailors and order of the noun phrases; i.e. the subject noun phrase determine whether some of them left. is more dominant or has scope over the object: every 886  QUANTIFICATION student  a book. The universal generalized quantifi- farmer who has a donkey beats it, the indefinite a er every student takes scope over the existential gen- lacks its typical existential force. The sentence does eralized quantifier expressed by a book. The second not mean that every farmer beats one donkey or other reading is an inverse scope reading in that the scopal (at least one) that he has. Rather, the correct interpre- order of the quantifiers differs from the surface linear tation is that for every pair consisting of a farmer and order. In this reading, the existential quantifier takes a donkey owned by that farmer, it is also the case that scope over the universal quantifier: a book  every the farmer beats the donkey. The indefinite seems to student. have universal force here; so we may conclude that So far, we have considered only nominal quantifi- the universal quantifier headed by every behaves as a cation, i.e. quantification corresponding to the mean- binary quantifier here and associates both with the ing of noun phrases. However, other elements may restriction of the universal noun phrase and with the also contribute to the quantificational force of a sen- indefinite one. The same pattern can be observed in tence. Adverbs such as always, sometimes, and often the sentence Always, if a farmer owns a donkey, he are not merely temporal adverbs. In the majority of beats it, where the indefinite noun phrases a farmer contexts, they behave as adverbs of quantification. For and a donkey seem to inherit the universal force from example, the sentence Peter always drinks coffee does the adverb always. Some theories have concluded not mean that Peter drinks coffee at every moment, but that indefinites lack quantificational force and inher- rather that whenever Peter drinks, he drinks coffee. it their apparent force from other surrounding ele- Adverbs of quantification can thus be taken to express ments in the clause. In general, adverbs of quantification over situations. quantification are taken to transmit their quantifica- Interrogative generalized quantifiers are wh-phras- tional force to all the indefinites that appear in their es such as who, what, where, and which. Interrogative scope. This phenomenon is known as unselective determiners are also conservative. The interrogative quantification. Quantification thus covers a wide sentence Which students are rich? is equivalent to range of interesting phenomena and has triggered Which students are students who are rich? many debates among linguists and philosophers. Interrogative determiners are also uniformly intersec- tive, because the meaning of an interrogative deter- References miner depends on the intersection of its two arguments. For example, a complete answer to which Benthem, Johan van, and Alice ter Meulen (eds.) 1985. (students) (rich) specifies, in a situation, the intersec- Generalized quantifiers in natural language. Dordrecht: Foris. tion of the set of students and the set of rich individ- Gärdenfors, Peter (ed.) 1987. Generalized quantifiers: linguistic uals in that situation. Interrogative determiners and and logical approaches. Dordrecht: Reidel. existential determiners such as some have a majority Gutiérrez-Rexach, Javier. 1998. Semántica lógica y cuantifi- of their properties in common. From this point of cación nominal. Introducción a la teoría de cuantifi- view, it is not surprising that a significant number of cadores generalizados. New Orleans: University Press of the South. the world’s languages use the same lexical expres- ——— (ed.) 2003. Semantics: critical concepts. London: sion for interrogatives and existential determiners. Routledge. Sometimes, the quantificational force of a sentence Hawkins, John. 1978. Definiteness and indefiniteness. London: cannot be associated with any overt element in the Croom Helm. clause. The sentence Lions are fierce is usually inter- Heim, Irene. 1982. The semantics od definite and indefinite noun phrases. Ph.D. dissertation. University of preted as stating that most lions are fierce or that lions Massachusetts, Amherst. are normally fierce. To characterize the semantic prop- Kamp, Hans, and Uwe Reyle. 1993. From discourse to logic. erties of this sentence, the existence of hidden ele- Dordrecht: Kluwer. ments of diverse quantificational force has to be Keenan, Edward, and Leonard Faltz. 1985. Boolean semantics postulated. In the example under consideration, there of natural language. Dordrecht: Reidel. Lappin, Shalom. 1996. Handbook of contemporary semantic is a hidden element of generic force, Gen, so the logi- theory. Oxford: Basil Blackwell. cal representation of this sentence would be Gen May, Robert. 1987. Logical form. Cambridge, MA: MIT (lions) (fierce). Press. Nominal and adverbial quantification interact in Reuland, Eric, and Alice ter Meulen (eds.) 1987. The apparently unexpected ways. This interaction arises representatio of (in) definiteness. Cambridge, MA: MIT Press. mostly when an indefinite occurs in the scope of an Szabolcsi, Anna. 1997. Ways of scope taking. Dordrecht: adverb of quantification in relative constructions, Kluwer. conditional sentences, etc. In the sentence Every JAVIER GUTIÉRREZ-REXACH 887 QUESTIONS Questions The term ‘questions’ in everyday usage refers to utter- example, the use of the nonassertive form ‘anyone’ in ances inviting a response and that, when spoken, may Has anyone left? can be argued to render this question have specific intonation patterns. Questions can be neutral in that it leaves open whether the response will analyzed for their grammatical forms and also for their be affirmative or not. However, the use of the more functions in discourse. Discussion of questions gener- assertive form ‘someone’ in Has someone left? is ally involves both these aspects. The analysis of ques- argued to indicate speaker expectation of a positive tions can be used to relate linguistic findings to response, whereas the use of a negative form hasn’t in patterns of social behavior. Hasn’t anyone left? is argued to indicate a bias toward In English, questions can be identified and catego- a negative response. rized on the basis of syntactic and phonological forms. Generally, questions are understood to be based on Four types of questions in English can be distinguished: presuppositions, or underlying assumptions. For exam- ple, the question When does your brother finish col- (1) Wh-questions lege? is based on the presuppositions that (1) you have These are forms beginning with an interrogative a brother and (2) your brother is at college. element (who, which, what, why, when, how, and so Questions can be examined in relation to their prag- on): Where does Antony work? What can Antony do? matic functions, namely, their functions to request spe- The subject (Antony) and either an auxiliary (does) or cific types of linguistic responses. A discourse-based modal verb (can) are inverted in English. Wh- ques- analysis of questions involves a consideration of con- tions generally require more elaborate answers than textual factors, such as the situation, who knows what, those elicited by yes/no questions. and how the discourse unfolds. Identification of the functions is the primary interest, but identification of (2) Yes/no questions the forms that typically realize these functions may This broad category refers to questions to which the follow from this. hearer is expected to respond either with yes or no. From a pragmatic viewpoint, a question can be The following forms can be identified: identified as a question if it fulfills certain conditions: the act of questioning must be genuine, and the speak- a. Inversion of the subject and an auxiliary or er must believe in the presuppositions of the question modal verb and desire to know the response. For example, a Does the race start here? speaker (sitting in the driver’s seat in a car with some Can we finish this now? passengers at the back) produces the following utter- b. Statements with a question intonation ance: Are you ready? Without a response, the speaker then sets forth on the journey. This indicates that the In these questions, the speaker presents an utterance driver does not desire to know the response. Also, the in the form of a statement, but marks it as question by absence of any response from the passengers indi- a rising and falling tone. cates that the hearers do not interpret the driver’s You will be sure to bring it back soon? utterance as seeking a response. Thus, despite the (3) Alternative forms question-like form, the utterance would not be termed a question. In these forms, the speaker presents two or more In discourse-based analyses, questions are often alternative answers within the question. referred to as ‘elicits’ or ‘requests’, and they are cate- Will the new administrator work here or on the first gorized in relation to the specific functions they floor? require of the hearer. One such category is elicits of, or (4) Tag forms requests for, information. In the following example (data taken from Schegloff 1972:107), A requires These forms consist of a declarative utterance fol- information about the location of an address from B. lowed by a question. The computer is connected to the internet, isn’t it? A I don’t know where the—uh—this address is. Yes/no questions are said to either be neutral or B Well, where do—-which part of the town do you biased toward a positive or negative orientation. For live in? 888  QUINE, WILLARD VAN ORMAN Another category in a discourse-based analysis is to recall or recognize factual information, display that of elicits of, or requests for, confirmation. These questions functioning to check or test the student on function to invite the hearer to verify or disconfirm the information known to the teacher, and referential speaker’s idea or proposition, as the following exam- questions seeking information unknown to the teach- ple taken from Coulthard and Brazil (1981:84) illus- ers themselves. trates. Conversation analysts have analyzed questioning to identify rules underlying interaction. Questions are A So the meeting’s on Friday identified as the first part of a two-part sequence: ques- B Thanks tion (Q) and answer (A). The person who asks a ques- A No I’m asking you tion has the option of speaking again to ask another This example also illustrates that identification of question. Thus, there is a chaining rule providing for the exact function may be ambiguous and can only be the sequence Q-A-Q-A-Q-A…. This approach to the resolved as the discourse unfolds. A’s first utterance analysis of questions has been used to relate linguistic asks B to confirm that the meeting is on Friday. findings to power structures in society. For example, However, the function of this utterance is unclear, and the analysis of interaction may reveal one speaker ask- B understands it as information being provided by A. ing more questions, thus directing the discourse and B thus thanks A for the information. In utterance 3, A indicating asymmetrical power relationships. then makes explicit the function of the first utterance as an elicit of confirmation. Other discourse functions of questions have been References identified. Elicits of commitment require the hearer to Aijmer, K. 1996. Conversational routines in English: conven- promise further action (e.g. You will come on tion and creativity. New York: Longman. Wednesday?). Elicits of agreement require the hearer Coulthard, R.M., and D.C. Brazil. 1981. Exchange structure. Studies in discourse analysis, ed. by R.M. Coulthard, and to verify that the idea proffered by the speaker is self- M.M. Montgomery, 82–106. London: Routledge and Kegan evidently true. The following example taken from Tsui Paul. (1992:107) illustrates: Dillon, J.Y. 1990. The practice of questioning. London: (On a sunny day) Routledge. Greenbaum, S., G. Leech, and J. Svartik. 1972. A grammar of A Lovely day, isn’t it? contemporary English. London: Longman. B Yes, beautiful! Schegloff, E.A. 1972. Notes on conversational practice: formu- lating place. Studies in social interaction, ed. by D. Sudnow, The analysis of questions according to function has 75–119. New York: Free Press. been applied to the study of language use in specific Tsui, A.B.M. 1992. A functional description of questions. situations. For example, analysis of teacher talk has Advances in spoken discourse, ed. by M. Coulthard, London: Routledge. been shown to reveal various didactic functions. These ——— 1994. English conversation. Oxford: Oxford University didactic functions include higher cognitive level ques- Press. tions requiring the hearer to manipulate bits of previ- van Lier, L. 1988. The classroom and the language learner. ously learned knowledge to create a response, lower Harlow Essex: Longman. cognitive level questions requiring the hearer simply HELEN BASTURKMEN Quine, Willard van Orman Quine is one of the most important analytic philoso- time of his death in December 2000, Quine was the phers of the twentieth century, a fact eloquently attest- Edgar Pierce Professor Emeritus at Harvard, where he ed to by the fact that in 1987 a supplement to the had pursued an active career spanning over half a cen- Oxford English dictionary recognized his importance tury. His early work was heavily influenced by Rudolf as an eponym by including the entry ‘Quinean’ with Carnap and other logical positivists. But, as an empiri- the following definition: ‘Of, pertaining to, or charac- cist and (from the 1960s on) an advocate of a natura- teristic of Willard van. Quine or his theories’. At the listic epistemology, he distrusted all talk of meaning 889 QUINE, WILLARD VAN ORMAN equivalence, which was a mainstay of Carnap’s reduc- evidence, an entire network of beliefs and theoretical tionist program. predilections is involved every time a scientific A prolific writer, Quine is widely perceived as a hypothesis is put to test. This means, he went on to philosophical iconoclast. His early essay ‘Two dog- argue, that no scientific hypothesis can be tested in mas of empiricism’ (published in the 1953 collection) isolation (this claim is often referred to as had great impact on the philosophical community. In ‘Quine–Duhem thesis’), just as no single sentence can it, Quine challenged the distinction between ‘analyt- be said to lend itself to an analysis of its putative ic’ and ‘synthetic’ statements, one of the most funda- meaning. mental and well-entrenched distinctions of modern Skepticism concerning meaning and with it a whole philosophy—originally introduced by Kant but array of assorted concepts such as property, proposi- already foreshadowed in the work of Locke and tion, and necessity as well as the very concept of a Leibniz. According to Kant, an analytic statement concept led Quine to unveil one of his most influential was one whose predicate was contained in the con- and controversial doctrines, namely the thesis of the cept of its subject. In the case of a synthetic state- ‘indeterminacy of translation’. Quine argued that there ment, there was no such relation of inclusion can be no such thing as radical translation, i.e. trans- between the subject and the predicate. Post-Kantian lating, as it were, from scratch between two languages interpretation had resulted in an extension of the dis- totally foreign to each other. A radical translator or tinction to cover cases other than the simple sub- interpreter will only have access to certain external ject–predicate statements originally envisioned by facts, but these will be of no help when it comes to Kant and, with it, the key criterion for analyticity had deciding between alternative candidates for the trans- become the idea of a sentence being true by virtue of lation of an expression from one language to a radi- its meaning. The idea had been embraced by the log- cally different one. The impossibility of radical ical positivists of the Vienna Circle who had also translation stems from what Quine called the extended its scope to include, among other things, all ‘inscrutability of reference’ (later renamed ‘indetermi- propositions of mathematics. nacy of reference’). The native word ‘gavagai’ uttered Quine’s attack on the ‘analytic/synthetic’ distinc- by the informant as he points at a rabbit jumping past tion was based on his rejection of the notion of ‘syn- does not authorize one to hypothesize that the two onymy’, which he thought was a philosophical mare’s words ‘gavagai’ and ‘rabbit’ are translation equiva- nest. Synonymy was suspect because it rested on the lents; a number of alternative hypotheses, such as that notion of meaning that too was, Quine argued, inca- the informant might have been referring to specific pable of surviving critical scrutiny, notwithstanding rabbit parts or determinate rabbit movements, cannot the widespread use of the term in philosophical cir- be ruled out as equally plausible in light of the avail- cles. Whatever heuristic value there was to the distinc- able empirical evidence. tion was due to the fact that understanding a sentence At a broader level of application, Quine’s position was largely a matter of knowing what experiences meant that, when different theories all seem equally would make a given sentence true, a claim long made adequate in explaining a given set of behavioral pat- by empiricists—the so-called analytic statements sim- terns, there can be no choosing among them on ply being, under this interpretation, those that have grounds of greater psychological reality or whatever. been confirmed on all known occasions. In an important paper titled ‘Methodological reflec- Unlike the early Wittgenstein and the positivists, tions on current linguistic theory’, Quine challenged Quine saw philosophy as a natural ally of science. In Noam Chomsky’s contention that the best theory is his view, a scientific theory was an intricate web of the one that is not only observationally and descrip- ideas whose borders touched on concrete experience. tively adequate, but explanatorily so as well. Quine No part of the web was immune to revisions in the insisted that, given a native speaker who presumably light of new experience. Furthermore, revisions are has a consummate command of his or her language never strictly localized but rather affect several parts and a hypothetical language learner whose linguistic of the web at once. This meant that individual sen- performance is indistinguishable from that of the for- tences, considered in isolation, have no meaning, mer, there is no sense in which the former may be strictly speaking. In addition, Quine held that in talk- claimed to know the language in any deeper sense ing about the world one characteristically moves from than the latter. Quine’s point was that, insofar as they talking in certain terms to talking about them, that is, are meant to be explanations for certain verifiable from so-called material mode to formal mode—a behavioral phenomena, the ‘theory’ that the language process that he famously referred to as ‘semantic learner presumably brings to bear on the task of ascent’. On Quine’s holistic view of meaning and speaking the foreign language is just as good as the 890  QUINE, WILLARD VAN ORMAN one the native speaker is being claimed to have inter- Advanced Studies (Princeton, NJ, 1956–1957), nalized. Accordingly, Quine dismissed Chomsky’s American Philosophical Association (President 1957), celebrated notion of ‘linguistic competence’ as scien- American Philosophical Society, member (1957 –), tifically untenable. In the final analysis, Quine’s Centre for Advanced Studies in the Behavioral objection to such notions had to do with his firm con- Sciences (Palo Alto, CA, 1958–1959), British viction in the irreducibly public nature of language. Academy corresponding fellow (1959–), Instituto Toward the end of his career, he became more and Brasileiro de Filosofia corresponding member more convinced that one can only learn language (1963–), Centre for Advanced Studies (Wesleyan from experience with the world at large and that, University, Middletown, CT), Nicholas Murray Butler therefore, there is no way linguistics can avoid gold medal (1965), Columbia University (New York, embracing behaviorism. 1970), National Academy of Sciences fellow Quine will be remembered as one of the most chal- (Washington, DC, 1977), Institut de France (1978), lenging and dynamic philosophers of the second half Norwegian Academy of Sciences (1979), F. Polacky of the twentieth century. His hobbies included travel- gold medal (Prague, 1991), Charles University gold ing and collecting stamps. He could never bring him- medal (Prague, 1993), Rolf Schock Prize (Sweden, self to use a computer, contenting himself instead with 1993), and the Kyoto Prize (Japan, 1996). He retired a 1927 Remington typewriter whose keyboard had to as Edgar Pierce Professor Emeritus in 1978 from be modified to suit his special requirements of charac- a teaching career in which his pupils had included not ters and type faces. only influential philosophers like Donald Davidson but also the satirical songwriter Tom Lehrer and Theodore J. Kaczynski, the so-called ‘Unabomber’. Biography He died in Boston on December 25, 2000, aged 92. Quine was born in Akron, Ohio on June 25, 1908. He graduated from Oberlin College (Oberlin, Ohio) in 1930, and in 1932, barely two years later (an all-time References Harvard record), received his doctorate from Harvard, Barrett, R., and R. Gibson, (eds.) 1993. Perspectives on Quine with the famous English philosopher and mathemati- (Philosophers and their critics). London: Blackwell. cian Alfred North Whitehead as adviser. He traveled to Davidson, D., and J. Hintikka (eds.) 1969. Words and objection: Europe on a Sheldon Traveling Fellowship and spent essays on the work of W.V. Quine. Dordrecht, Holland: D. the next year in Vienna, Prague, and Warsaw, studying, Reidel. lecturing and meeting various members of the Vienna Gibson, R. 1982. The philosophy of W.V. Quine: an expository essay. Florida. Circle, among them Philip Frank, Moritz Schlick, Guttenplan, S. (ed.) 1975. Mind and language. Oxford: Oxford Alfred Tarski, A.J. Ayer, and Rudolf Carnap as well as University Press. Kurt Gödel. He started his teaching career at Harvard Hahn, L. and P. Schilpp, (eds.) 1986. The philosophy of W.V. in 1936 as an Instructor in Philosophy and taught there Quine. New York: Open Court. ever since, except for four years in the US Navy dur- Orenstein, A. 1977. Willard van Orman Quine. New York: Twayne. ing World War II, serving in Washington and working Quine, Willard van Orman. 1934. A system of logistic. with cryptoanalysts trying to break the German sub- Cambridge, MA: Harvard University Press. marine code. He was promoted to Associate Professor ———. 1940. Mathematical logic. New York: Norton. (1941), Professor (1948), and Edgar Pierce Professor ———. 1950. Methods of logic. New York: Holt. (1956). He paid two visits to Oxford: in 1953–1954 as ———. 1953. From a logical point of view. Cambridge, MA: Harvard University Press. Eastman Visiting Professor, and in 1973–1974 as ———. 1960. Word and object. Cambridge, MA: MIT Press. Savile Fellow of Merton College and Wolfson ———. 1963. Set theory and its logic. Cambridge, MA: Lecturer. He was awarded 18 honorary degrees by Harvard University Press, Belknap Press. international institutions, including University of ———. 1970. Philosophy of logic, Foundations of Philosophy Lille, Oxford University, Cambridge University, Series, Englewood Cliffs, NJ: Prentice-Hall. ———. 1972. Methodological reflections on current linguistic Uppsala University, University of Bern, and Harvard theory. Semantics of natural language, ed. by D. Davidson University. He received innumerable honorary fellow- and G. Harman. Doredrecht, Holland: D. Reidel Publishing ships and awards, including: Society of Fellows, Co. Harvard University (Junior Fellow, 1933–1936; Senior ———. 1974. The roots of reference. The Paul Carus Lectures, Fellow, 1949–1978), American Academy of Arts and Vol. 14. LaSalle, IL: Open Court. ———. 1976. The ways of paradox, and other essays. Sciences (fellow 1949 –), Harvard University Cambridge, MA: Harvard University Press. (Chairman, Philosophy, 1952–1953), Association for ———. 1977. Ontological relativity. New York: Columbia Symbolic Logic (President, 1953–1955), Institute for University Press. 891 QUINE, WILLARD VAN ORMAN ———. 1981. Theories and things. Cambridge, MA. and Romanos, G. 1983. Quine and analytic philosophy. Harvard: London: Harvard University Press. MIT Press. ———. 1989. Quiddities: an intermittently philosophical dic- Shahan, R. and C. Swoyer, (eds.) 1979. Essays on the philoso- tionary. Cambridge, MA: Harvard University Press. phy of W.V. Quine. Oklahoma: University of Oklahoma ———. 1992. Pursuit of truth. Cambridge, MA: Harvard Press. University Press. KANAVILLIL RAJAGOPALAN ———. 1995. From stimulus to science. Cambridge, MA: Harvard University Press. See also Philosophy of Language 892  R Reading Over the past several centuries, reading has gone from whether comprehension is necessary for something to being a relatively obscure intellectual endeavor limit- be considered reading. Studies of decoding printed let- ed to a small elite segment of the population to a rela- ter strings sometimes use nonwords rather than real tively common one. You are, in fact, reading at this words to eliminate the effects of individual differences very moment. Despite the fact that reading has become in vocabulary knowledge. If you can read, you will so common, and is something you probably do every have little trouble working through a list of nonwords day, picking a useful definition of reading turns out to such as HEZ, RAF, POTE, and HINKER, and the pro- be surprisingly difficult. cessing that is done overlaps with that done when reading real words. But without comprehension, is this reading? What Is Reading? A pragmatic answer to the question of what is Reading a newspaper article surely counts as reading. reading is to define reading as extracting meaning But does scanning the stock market results to see how from print, and to acknowledge that it can be useful to a favorite stock fared this week amount to reading? study parts of the reading process such as aspects of How about ‘reading’ a map? Is working through a decoding that are best viewed when people are asked mathematical proof, or a computer program, reading? to pronounce nonwords. In addition, it is important Is it reading when you decode and pronounce a non- to acknowledge that successful reading requires the word such as BOPE? coordinated execution of decoding processes that are Definitions of reading run the gamut from ‘decod- unique to processing print and comprehension ing printed symbols’ to ‘thinking with a book in front processes that may be common to listening as well as of you’. Differences in opinion about what constitutes reading. a useful definition of reading arise from different answers to two fundamental questions. First, for some- What Is Being Read? thing to be defined as reading, must it be unique to reading or can it be general to oral language com- Humans have used spoken language for at least prehension? Reading almost always is done for the 100,000 years and perhaps quite a bit longer. In con- purpose of comprehending meaning, but most com- trast, the earliest artifacts of writing, and hence read- prehension appears to be the same regardless of ing, are cave paintings in southern France and northern whether the message was initially processed through Spain, that are approximately 20,000 years old. The the eyes or the ears. Should a book on reading be paintings depict animals and occasional humans, and mostly about language comprehension or should it it is not clear whether they tell a story or are just pic- be mostly about what is unique to reading as opposed tures that serve some religious or magical purposes. A to listening? The second fundamental question is major advance in writing occurred around 5,000 years 893 READING ago in Sumeria. When goods were sent to market, they Reading involves the coordinated orchestration of just were accompanied by clay balls. Sealed inside the clay about every process ever studied by cognitive psy- balls were small pieces of clay that represented the chologists and linguists. At best, teachers put print in goods that were being transported. When the shipment front of children, and point out associations between reached the market, the clay ball would be broken print and sound or meaning, but the real work of open and the shipment would be inspected to make learning to read is done by cognitive and linguistic sure nothing had been diverted. Eventually, it became machinery that is not readily accessible to teachers or apparent that marks could be pressed into wet clay to learners. It is clear that learning to read an alphabetic represent the goods, saving the trouble of forming the system of writing is easier for children who have balls filled with small pieces. This, of course, marks awareness and access to the sound structure of their the birth of a real writing system. oral language (i.e. phonological awareness) that is Today, printed languages vary in terms of what they represented in writing. Thus, the spoken words ‘cat’, represent, with three major categories of writing. The ‘rat’, and ‘sat’, have different initial sounds and iden- Chinese writing system is called a logography. Each tical medial and final sounds. To a child who is aware character represents a morpheme or minimal unit of of these similarities and differences, the English sys- meaning. Beijing means ‘north capital”, and is repre- tem of writing will seem sensible in that these rela- sented in writing by two symbols, one for north and tions in sounds are represented in spelling. Cat, rat, one for capital. The Kana system of Japanese is a syl- and sat have different initial letters and identical labary. Each symbol represents a spoken syllable. medial and final letters. To a child lacking such an Finally, English, modern Korean, and many other writ- awareness, an alphabetic writing system will seem to ten languages are alphabets, in which a small set of be arbitrary and complex. characters are put together to make a much larger number of words. Dividing the world’s languages into References three categories is an oversimplification. Thus, some sounds as well as meaning are represented in Chinese Adams, Marilyn. 1994. Beginning to read: thinking and learn- ing about print. Cambridge, MA: MIT Press. writing, and English is best described as a morpho- Crowder, Robert, and Richard Wagner. 1991. The psychology phonemic representation system as English spellings of reading. New York: Oxford University Press. represent both sounds (phonemes) and (units of) Perfetti, Charles. 1985. Reading ability. New York: Oxford meaning (morphemes). Consequently, the printed University Press. word HEALTH has a spelling that makes apparent its Rayner, Keith, and Alexander Pollatsek. 1989. The psychology of reading. Englewood Cliffs, NJ: Prentice-Hall. relation to the morpheme ‘heal’ as opposed to a strict- Wagner, Richard, and Catherine McBride-Chang. 1996. The ly phonetic spelling such as HELTH. development of reading-related phonological processes. Annals of Child Development 12. 177–206. Wagner, Richard, and Joseph Torgesen. 1987. The nature of How Do Children Learn to Read? phonological processing and its causal relations with read- ing. Psychological Bulletin 101. 192–212. The obvious answer is that children are taught to read RICHARD WAGNER by teachers and parents. But a closer investigation of what is involved suggests it is immodest to assert that See also Handwriting; Literacy; Reading Impair- teachers or parents actually teach children to read. ment; Writing Systems Reading Impairment Reading impairment, reading disability, and dyslexia Nature of Reading Impairment all refer to a level of reading that is below expectations. Expectations for reading can be based on normative Dyslexia is perhaps the most commonly used term to data from age-matched peers, or based on an individual describe reading impairment, and surely the most mis- child’s oral language ability or cognitive ability. understood. The common view of dyslexia is that of a 894  READING IMPAIRMENT visual perceptual problem that results in seeing mirror backward eye movements or regressions. Conclusive images of words or letters. Thus, individuals with evidence was provided by careful studies in which dyslexia are reported to read WAS as SAW, or to con- normal readers were given material that was as diffi- fuse the letters ‘b’ and ‘d’. The origin of this view is cult for them to read as is grade-level reading material easily established. Children with reading problems for individuals with reading impairment, and individu- typically became obvious to teachers and parents at als with reading impairment were given very easy the second-grade level, and indeed they can be reading material that they could read as well as normal observed reading WAS for SAW and confusing the let- readers could read grade-level material. Under these ters ‘b’ and ‘d’. But it turns out that normal beginning conditions, the eye movements of normal readers dete- readers also make similar errors. The fact that words riorated to match the previously reported erratic eye in English are to be read from left to right as opposed movements of individuals with reading impairment, to right to left is arbitrary and must be learned. In and the eye movements of individuals with reading addition, the letters ‘b’ and ‘d’ are both visually con- impairment now looked normal. Additional confirma- fusable and similar in sound (i.e. both stop conso- tion came from the results of eye-movement training nants). Careful analysis of reading errors has shown studies. Although eye-movement training did result in that second-grade readers with dyslexia make no gains performance on eye-movement tasks outside the more reversal errors than do younger normal readers context of reading, reading performance did not who are at the same level of reading. Thus, second- improve. grade teachers observed only poor readers making For the vast majority of individuals with reading reversal errors, whereas kindergarten teachers would impairment, the problem appears to be based in lan- know that such errors are quite common. guage as opposed to the visual system, and is com- Another mistaken idea about reading impairment is monly compounded by ineffective instruction. that it results from erratic or inefficient eye move- Compared to reading-level matched controls, most ments. Reading requires highly sophisticated and individuals with reading impairment perform poorly coordinated eye movements characterized by ballistic on measures of phonological awareness and phono- movements called saccades and pauses called fixa- logical decoding, and have fewer words that can be tions. During the movements or saccades, little infor- decoded by sight. Phonological awareness refers to an mation is available to the eyes beyond a blur. Nearly individual’s awareness and access to the sound struc- all information is acquired during fixations. As you ture of an oral language. Phonological decoding refers read the words on this page, your perception is that to decoding words by sounding them out, as when one your eyes are moving smoothly across the page. This is asked to decode nonwords such as TANE. The is an example of a situation where perception does not underlying language problem for individuals with reflect reality. If you observe normal readers reading reading impairment is likely to be a subtle and not text by having them read directly across from you well-understood problem in forming accurate phono- while holding a book low enough so you can observe logical representations, which in turn leads to poor their eyes, you will indeed see that the eyes move phonological awareness and phonological decoding. across the page in a series of small, but observable Once beginning readers fall behind, they are exposed jerky movements. to reading instruction designed for more advanced Perform the same experiment on individuals with readers, which provides little assistance, until they are reading impairment and it will be apparent that their finally identified as having a reading problem and eyes move much more erratically, even moving in the more appropriate instruction is provided. wrong direction at times. Observations like these resulted in the belief that faulty eye movements were Distribution of Reading Impairment the origin of reading impairments, and also resulted in interventions based on eye-movement training. It turns Two important facts about the distribution of reading out that this view has it backwards. Indeed, the eye impairment are counterintuitive. First, despite the fact movements of individuals with reading impairments that boys outnumber girls by roughly four to one in are more erratic than those of normal readers, but the classes and clinics that serve individuals with reading erratic eye movements are the by-product, not the impairments, epidemiological studies in which all chil- cause, of the impaired reading. The eye movements of dren are tested reveal that reading impairment rates for individuals with reading impairments do not move boys and girls are roughly comparable. For every boy across the page as smoothly as do those of normal with a reading impairment, there is likely to be a girl readers because they are having trouble reading the with an equally severe reading impairment. The over- words. This also explains their greater frequency of representation of boys in classes and clinics appears to 895 READING IMPAIRMENT result from a referral bias that arises because boys are Foorman, Barbara (ed.) 2003. Preventing and remediating reading more likely to be disruptive than girls, and hence more difficulties: bringing science to scale. New York: York Press. Shankweiler, Donald, and Isabelle Liberman. 1989. Phonology likely to be referred for evaluation. and reading disability. Ann Arbor, MI: University of Second, reading impairment is not something that Michigan Press. one has or does not have. It is not an all or none phe- Spear-Swerling, Louise, and Robert Sternberg. 1996. Off track: nomenon. The distribution of reading performance is when poor readers become learning disabled. Boulder, CO: continuous, with no obvious breaks or bumps in the Westview Press. Stanovich, Keith, and Linda Siegel. 1994. Phenotypic perform- tails of the distribution. Where one draws the line is ance profile of children with reading disabilities: a regres- arbitrary. sion-based test of the phonological-core variable-difference model. Journal of Educational Psychology 86. 24–53. References Wagner, Richard, and Tamara Garon. 1999. Learning disabili- ties in perspective. Perspectives on learning disabilities, ed. Blackman, Benita (ed.) 1997. Foundations of reading acquisi- by Robert, Sternberg and Louise Spear-Swerling. Boulder, tion and dyslexia. Mahwah, NJ: Lawrence Erlbaum CO: Westview Press. Associates. RICHARD WAGNER Crowder, Robert, and Richard Wagner. 1991. The psychology of reading. New York: Oxford University Press. See also Reading Reference The term ‘reference’ is ambiguously used to designate It is also important to distinguish among different (a) the relation between a referring expression and that kinds of referring expressions. ‘That man over there’, to which it refers, (b) that to which a referring expres- ‘Smith’, ‘The President of the United States’ are sion refers (also known as the ‘referent’), and (c) the examples of singular definite expressions whereas act of referring to some extralinguistic entity or state ‘horse’, ‘gold’, etc. exemplify general expressions. of affairs, etc. by using a referring expression. Scholarly discussion about reference has tended to Sometimes the terms ‘denotation’ and ‘designation’ concentrate on singular definite expressions. Broadly are used interchangeably with reference in its sense speaking, theories of reference may be grouped under (a)—‘denotatum’ and ‘designatum’ being the corre- two conflicting approaches: descriptive vs. direct ref- sponding variants for the word in its sense (b). erence approaches. Sense (c) introduces, over and above the referring On the descriptive view, a referring expression does expression and the referent, a third element, namely its job by describing features or properties of the refer- the speaker, into the picture. This in turn introduces ent. The descriptive view hinges crucially on a distinc- the possibility that a speaker may succeed in referring tion between Sinn (sense) and Bedeutung (reference) to an object, although the referring expression she or proposed by the German logician and philosopher he uses is, strictly speaking, inaccurate as a description Gottlob Frege in order to explain the difference in cog- of it—as when someone says ‘That horse is agitated’, nitive value between identity statements of the form pointing at a certain restless quadruped that, on closer ‘x  y’ and ‘x  x’, where x and y stand for referring inspection, turns out to be a mule. A speaker may also expressions. Frege argued that two referring expres- successfully refer to an individual whose identity is sions may have different senses but nonetheless have otherwise unknown to everyone including himself or the same reference (i.e. referent), thus making the for- herself as when someone says ‘Smith’s murderer is mer equation more informative than the latter. He also insane’, meaning by the referring expression ‘Smith’s insisted that a referring expression can only have a ref- murderer’ something like ‘whosoever murdered erent provided it has a sense, although having a sense Smith’ rather than, say, ‘that man in the dock over was not a sure guarantee that the term will have a ref- there, about to be cross-examined’. These cases show erent (as in the case of ‘unicorn’ or ‘square circle’). that speaker reference and denotation may not always Bertrand Russell, who went on to spell out the coincide, raising the thorny issue as to which of them description theory more articulately, took a narrower should be regarded as basic. view of reference, maintaining that only ‘logically 896  REFERENCE proper names’ do the referring and they do it infallibly bearer. Not all direct theorists, however, subscribe to by naming objects known through ‘acquaintance’ that the thesis of causality. included sense-data, certain universals, and possibly Among modern defenders of the direct reference ourselves. As for descriptions, he argued—contra approach are Keith Donnellan, Saul Kripke, and Frege—that they are ‘incomplete symbols’, meaning- Hilary Putnam. Highlighting the pragmatic aspect of ful only in the sentential contexts in which they occur. reference, Donnellan argued that there are two distinct Ordinary proper names like ‘Bucephalus’ and ‘Mount uses of definite descriptions, namely ‘attributive’ and Everest’ are for Russell, as indeed they were for Frege ‘referential’ uses. In its referential use, a referring too, disguised descriptions. So a descriptive phrase expression such as ‘Smith’s murderer’ can succeed in such as ‘The present King of France’ in the sentence picking up the intended referent, namely, say, the man ‘The present King of France is bald’ does not qualify as in the dock awaiting the sentence, even though it may a genuinely referring expression, because a proper eventually turn out that the man in question is innocent analysis of its logical form will show that what the sen- of the charges. In other words, in its referential use a tence really says is that France currently has a (an only) referring expression can refer in spite of its sense, con- king and that he is bald. Thus paraphrased, the singular trary to what the descriptivist view predicts. In definite description no longer has a referring function Donnellan’s view, the Russell–Strawson debate was and its presence is explicated in terms of the existential on a nonissue, since the former was thinking exclu- quantifier and the variable within its scope as in sively of the attributive use of definite descriptions while the latter was interested only in the referential (∃x) (Kx & ( (∀y) (Ky → y  x) ) & Bx) use, but wrongly believing that the accuracy of the where ‘K’ and ‘B’ stand for the predicates ‘The present description was a necessary condition for the success King of France’ and ‘Bald’, respectively. of a referential use thereof. An interesting counter- Russell’s ‘theory of definite descriptions’ came argument to Donnellan’s thesis was offered by John under attack from Peter F. Strawson, who took up a Searle, who claimed that the two uses, far from being line of reasoning initiated by Frege and, departing mutually exclusive, can be explicated in terms of the from it in significant ways, insisted that anyone who number of ‘aspects’ under which a speaker is in a posi- utters the sentence in question in fact presupposes that tion to refer to the referent in question. there is presently a king of France, rather than assert- Kripke held that proper names are ‘rigid designa- ing it, as Russell’s analysis implied. For Strawson, tors’, meaning that they have the same referents in all whereas referring expressions indeed have their sens- possible worlds. So it is wrong to think that a proper es, it is the actual uses of those expressions that do the name such as ‘George W. Bush’ has the same sense as referring. What specific horse is being referred to by ‘The President of the United States of America’ the expression ‘that horse’ can only be resolved by because, although the two may indeed refer to the examining an actual use (utterance) of the expression same individual in the actual world, they may not do as part of, say, the making of a statement (and not so in some other possible world (say, the one in which merely the issuance of a sentence), along with the the Supreme Court decision had favored Al Gore attendant set of actual circumstances. In other words, instead). Strawson foregrounded the pragmatics of reference In the 1970s, Hilary Putnam developed the thesis of and claimed speaker reference to be paradigmatic, in ‘semantic externalism’ whose central tenet is that, in opposition to Russell who was primarily interested in many cases such as words referring to natural kinds sense (a) or the semantics of the term ‘reference’. like ‘gold’ and ‘water’, the reference of an expression In contrast to the descriptive approach, direct refer- cannot be thought of as a function of descriptions ence approach harks back to a view taken by J.S. Mill associated with it in the mind of the user. Thus, in in the nineteenth century. Unlike Frege and Russell, Putnam’s view, ‘water’ refers to the chemical com- who argued that proper names had senses and that pound H2O in all possible worlds, irrespective of these coincided with the sense of corresponding defi- whether or not one is aware of that scientific truism. nite descriptions, Mill had contended that they had no Direct reference theories run into difficulties when ‘connotations’ but only their ‘denotations’ (Mill’s called upon to explain the behavior of referring expres- terms for sense and reference, respectively). In other sions in what are known as intensional or referentially words, they simply referred, period, by functioning as opaque contexts as in ‘Juan believes that the Malvinas a label attached, as it were, directly to the object. Just rightfully belong to Argentina’, which has—on one how this direct contact with the referent is secured is interpretation—a different truth value than the sentence an open question. One suggested answer is that there ‘Juan believes that the Falklands rightfully belong to is a causal link between the use of a term and an orig- Argentina’ although ‘Falklands’ and ‘Malvinas’ desig- inal ‘baptismal’ episode of attaching the name to its nate the same group of islands (a fact that, as it happens, 897 REFERENCE Juan is not aware of). The usual answer that whatever Frege, Gottlob. 1949. Über Sinn und Bedeutung (1892). difference there may be between the two sentences is Readings in philosophical analysis, English translation by Herbert Feigl, ed. by Herbert Feigl and Wilfrid Sellars. New not of interest to semantics has not satisfied critics. York: Appleton-Century Crofts, Inc. Causal theorists are also hard put to it to explain Haack, Susan. 1978. Philosophy of logics. Cambridge: changes in the reference of a term over the years (as in Cambridge University Press. the case of ‘Yugoslavia’ whose referent has significant- Kripke, Saul. 1972. Naming and necessity. Semantics of natu- ly changed after successive wars) as well as natural kind ral language, ed. by Donald Davidson and Gilbert Harman. Dordrecht, Holland: D. Reidel Publishing Company. terms that are posited rather than effectively attested (as Lyons, John. 1977. Semantics, Vol. 1. Cambridge: Cambridge in the case of a missing link to account for the develop- University Press. ment of human beings from apes). Russell, Betrand. 1905. On denoting. Mind 14. Searle, John. 1979. Referential and attributive. Expression and meaning. Cambridge, UK: Cambridge University Press. References Strawson, Peter F. 1971. On referring. Logico-linguistic papers. Donnellan, Keith. 1966. Reference and definite descriptions. London: Methuen & Co. Philosophical Review 75. KANAVILLIL RAJAGOPALAN Register Broadly defined, ‘register’ refers to the way people use Linking register solely to levels of formality in language in different situations. The term is often used speech and writing, however, would be an oversimpli- as a full or partial synonym for style, speech variety or fication. Along with levels of formality, scholars also variation, field of discourse, and text type. It is the sub- study changes in register based on the speaker’s aware- ject of sociolinguistic, linguistic, and applied linguis- ness of a broader network of differences in situation, tic analysis. topic, addressee(s), or location. John Lyons emphasizes Most often, register is used to mean style. In that the complex sociolinguistic nature of register. He dis- sense, it refers to the ‘stylistic variety’, or ‘stylistic vari- cusses it in the framework of context, style, and culture, ation’ that occurs in a person’s speech or writing in dif- and more specifically as a feature of ‘communicative ferent social contexts. Typically, different registers vary competence’. He relates it to the appropriate choice of according to their degree of formality. For example, language with regard to ‘domain’, i.e. a cluster of expressions like ‘sure’, ‘no problem’, and ‘let’s have a social situations constrained by a common set of blast’ are associated with informal conversational styles behavioral rules. Within domain, Crystal and Davy or registers. Utterances like ‘in spite of our deliberate identify ‘province’, i.e. occupational or professional efforts to condemn fraud…’, on the other hand, are activity with no reference to the individuals engaged in used in formal written texts. In her Grammar book, D. it. For example, the language used in advertising, sci- Larsen-Freeman offers numerous examples of appropri- ence, law, and sports can be identified as being typical ate and inappropriate use of both formal and informal of the corresponding province, hence the use of ‘love’ registers. In the case of phrasal verbs, ‘put off’, ‘call in tennis and ‘sentence’ in law. off’, and ‘show up’ are common in informal registers, Fishman relates domain to ‘subject-matter’ on the whereas their one-word synonyms ‘postpone’, ‘cancel’, one hand, and to ‘locale’ and ‘role-relations’ on the and ‘arrive’, reflect higher levels of formality. The same other. Examples of these would be ‘family—home’ and applies to relative adverbs. For example, ‘1950 is the ‘religion—church’. Within these subject-locales, he year that I was born in’ would occur in more formal identifies certain role-relations, such as ‘mother-to- texts than ‘1950 is when I was born’. A connector, such son’ and ‘priest-to-parishioner’. Thus, a mother talks to as ‘moreover’, normally occurs in formal written texts. her son at a different level of formality if the conversa- It would be inappropriate in an informal conversational tion is held at home or not. Similarly, her choice of lan- setting like this, ‘Let’s go to the beach. *Moreover, let’s guage will be different if the topic was ‘her son’s drunk rent a boat’. It appears from these and other examples driving’ and ‘her son’s recent award’. She would also that there is an abundance of structures that reflect the choose a different register if she were to discuss these subtle nuances of formality, both within a separate lin- subjects with other people, at other places. Similar to guistic category and across different categories. Crystal and Davy’s use of ‘province’, Quirk defines 898  REICHENBACH, HANS register in relation to the ‘field of discourse’, or the to him, ellipsis and substitution, which are mostly activity in which a speaker is engaged. He categorizes found in dialogue, have a more local effect because of speech varieties as dependent and interdependent their shorter lexical reach. Typically, their reach according to ‘field of discourse’. Speakers who retain extends over no more than a few consecutive moves, regional features in their use of Standard English are ‘A: Would you like to come here? B: Could I bring a examples of the former. Interdependent speech vari- friend too? A: Sure. Which one? B: I’ll know tomor- eties, on the other hand, in a slightly broader use of row. A: Is it still raining outside?’ Such texts are relat- Fishman’s ‘locale’, are exemplified by topics that are ed not so much by ideational as by interpersonal associated with specific regions. Thus, discussions of meaning. Halliday also defines the features of oral and baseball are in American English. Similarly, coaching written texts or registers. Generally characterized by is handled by speaking rather than by writing. lexical simplicity, oral texts become complex by being Linguistically, register could also be examined on grammatically intricate. They use clauses to express the phonetic, morphological, syntactic, and semantic relationships and to achieve dynamic complexity. On levels. Yet, it can be most fully analyzed on the ‘text’ the other hand, written and technical texts, in particu- level, e.g. it is suprasentential or above and beyond the lar, are lexically dense and statically complex because sentence. In his An introduction to functional grammar, they use nominalization, or a lot of noun phrases, to M.A.K. Halliday discusses register in relation to ‘tex- express relationships. While they have simple clausal ture’, or ‘text structure’, which comprises theme and patterns, they pack ideas in nominal constructions, i.e. focus, lexical cohesion, reference, ellipsis and substitu- ‘advances in technology’ instead of ‘technology is get- tion, and conjunction, all of which have a semantic, not ting better’. Thus, nominalization, which helps to con- grammatical structure. He points out some of the dif- struct technical terms and develop step-by-step ferences in ‘theme and rheme’ in different text types, arguments, is often considered the most prominent such as narratives and instructions. In the following feature of texts or registers of expert knowledge, pres- example of a narrative, the participant (‘he’) remains tige, and power. the topical theme: ‘He went to many colleges but he didn’t complete any. He always fell into some difficul- References ty. He never hit upon the right thing’. In instructions Cheshire, Jenny. 1992. Register and style. International ency- that have stepwide structure, the theme, or given sub- clopedia of linguistics, ed. by W. Bright. New York: Oxford ject of one clause, becomes the rheme, or new object of University Press. the clause that follows it, e.g. ‘Turn the stove on. Put a Crystal, David, and D. Davy. 1968. Investigating English style. pot on the stove. Pour water into it. Add all ingredients London: Longman. Downes, W. 1994. Register in literature. The encyclopedia of to the water’. Thus, it would be inappropriate to use language and linguistics, ed. by R.E. Asher and J.M.Y. this kind of stepwide-structured register when writing Simpson. Oxford: Pergamon Press. an expository text, e.g. ‘I have twin sisters. My parents Fishman, J.A. 1972. The sociology of language. Rowley, MA: love them. They buy them many toys. One such toy is Newbury House. a huge stuffed cat. It now occupies the couch’. Halliday, M.A.K. 1994. Functional grammar. London, Melbourne and Auckland: Edward Arnold. Halliday also defines register by different types of Larsen-Freeman, Diane, and Marianne Celce-Murcia. 1999. The ‘lexical cohesion and reference’. He notes that they are grammar book, 2nd edition. New York: Heinle & Heinle. built through interlocking referential chain complexes Lyons, John. 1977. Semantics. London, New York and that vary from one register to another and produce a Melbourne: Cambridge University Press. certain global effect, e.g. ‘A boy called John. . . John . Quirk, Randolph, et al. 1985. A comprehensive grammar of the English language. London and New York: Longman. . . he . . . the lad . . . him . . .’ The global effect created Richards, Jack, et al. 1992. Longman dictionary of language by such overlapping referential chains is the source of teaching and applied linguistics. Harlow, Essex: Longman. the dynamic flow of discourse in narratives. According LILIA SAVOVA Reichenbach, Hans The philosopher Hans Reichenbach was a central figure and implications of relativity theory (Relativitätstheorie of logical empirism in the 1920s and 1930s. The starting und Erkenntnis a priori, 1920; relativity theory and cog- point of his research were the philosophical dimensions nition a priori), but also more general issues in the 899 REICHENBACH, HANS philosophy of space and time (Axiomatik der relativis- As for the simple tenses, Reichenbach proposes tischen Raum-Zeit-Lehre (1924; Axioms of relativistic the following tables, in which the direction of time is space–time), Philosophie der Raum-Zeit-Lehre (1928; represented as the direction of the line from left to Philosophy of space– time)). In 1930, he founded the right: journal Erkenntnis (Cognition, later called Journal of Unified Science), together with Rudolf Carnap from the Past Perfect Simple Past Present Perfect Vienna Circle. Furthermore, he developed a three- I had seen John I saw John I have seen John dimensional logic for the interpretation of quantum –X–X–X–> –X–X–X–> –X–X–X–> mechanics (Philosophic foundations of quantum ERS R,E S E S,R mechanics, 1944) and worked on probability. But linguists usually do not appreciate Reichenbach Present Simple Future Future Perfect as a philosopher. In linguistics, he became famous for I see John I will see John I shall have seen John not more than 12 pages of his complete works, name- –X–X–X–> –X–X–X–> –X–X–X–> ly for §51 ‘The tenses of verbs’ of his book Elements S,R,E S,R E SER of symbolic logic, 1947. According to Reichenbach, the tenses of verbs dis- If there is a progressive form, Reichenbach assumes play token reflexivity. The tenses determine time with that the event E covers a certain stretch of time. He reference to the time point of the act of speech, i.e. of calls these extended tenses. Here are two examples: the token uttered. The time point of the token is called the point of speech (S), and the time point at which the event being talked about took place is called the point Past Perfect, Extended Simple Past, Extended of the event (E). Using these two points means assum- I had been seeing John I was seeing John ing three tenses, namely E  S (the simple past in –XXXX–X––X–> –XXXX––X–> English), E  S (the simple present), and E  S (the ERS R,E S simple future). But obviously, there are more than these three tenses in English (and in other languages as As for adverbial modification, Reichenbach states well). This is the reason why Reichenbach introduced that in the presence of positional adverbs like now or a third parameter, namely the point of reference (R). yesterday, only modification of R is possible. This can What this parameter does becomes most easily clear in be illustrated with sentences like I had met him yester- narratives. Compare the following passage from W. day, where yesterday is the point of reference, but the Somerset Maugham’s Of human bondage: meeting (E) has occurred at some time before yester- day. Here, R is the carrier of the time position. But Philip ceased to think of her a moment after he had Therefore, Reichenbach speaks of the positional use of settled down in his carriage. He thought only of the the reference point here. future. He had written to Mrs. Otter, the massière to On the whole, there are 13 possibilities of ordering whom Hayward had given him an introduction, and had the three time points E, R, and S. But for English, in his pocket an invitation to tea on the following day. there are only six tenses in the traditional grammars. The events related in the simple past perfect The following table summarizes the possibilities, (had settled down..., had written to Mrs. Otter..., had Reichenbach’s terms for them, and the traditional given him an introduction...) are viewed as being terms: completed from the perspective of a specific point of reference, which is constituted by those events that Possibility Reichenbach’s Term Traditional Term are related in the simple past (ceased to think of her..., E_R_S Anterior past Past perfect thought only of the future..., had in his pocket an invi- E,R_S Simple past Simple past tation...). Thus, the point of reference is a kind of per- R_E_S Posterior past — spective point. R_S,E Posterior past — With some tenses, two of the three parameters E, R, R_S_E Posterior past — and S coincide. The simple past is such a case, with E E_S,R Anterior present Present perfect and R being simultaneous, and both E and R being S,R,E Simple present Present before S. With other tenses, the three parameters S,R_E Posterior present Simple future denote distinct times, e.g. in the case of the simple S_E_R Anterior future Future perfect past perfect: E and R are both before S, but E is also S,E_R Anterior future Future perfect before R. The different usages of the simple past and E_S_R Anterior future Future perfect the simple past perfect in the quotation above show S_R,E Simple future Simple future this quite clearly. S_R_E Posterior future — 900  RELEVANCE IN DISCOURSE With Reichenbach’s terms, the position of R relative appointment when the Nazis seized power and emigrat- to S is indicated by past, present, future, and the posi- ed to Turkey; he was Professor for Philosophy, tion of E relative to R is indicated by anterior, simple, University of Istanbul, 1933-1938. He moved to the posterior. United States, and was Professor for Theory of Science, As for the posterior future, it is used for instance in University of California at Los Angeles, 1938–1953. He I shall be going to see him. Sentences like I did not died in Los Angeles on April 9, 1953. expect that he would win the race express the posterior past—like posterior future this is commonly not classi- References fied as a tense. Reichenbach has been criticized for assuming the Comrie, B. 1985. Tense. Cambridge: Cambridge University Press. point of reference because this often coincides with Ehrich, V. 1992. Hier und jetzt. Studien zur temporalen und lokalen Deixis im Deutschen. Tübingen: Niemeyer. the point of the event or with the point of speech. Fabricius-Hansen, C. 1986. Tempus fugit: Über die Nevertheless, his work on tense is considered classic. Interpretation temporaler Strukturen im Deutschen. Within the modern theory of tense, there are many Düsseldorf: Schwann. reconstructions and adaptations of Reichenbach, e.g. Klein, W. 1994. Time in language. London and New York: Klein (1994), Ehrich (1992), Musan (2000), Comrie Routledge. Musan, R. 2000. The semantics of perfect constructions and (1985), Thieroff (1992), and Fabricius-Hansen (1986), temporal adverbials in German. Habilitationsschrift, to mention just a few. University of Berlin. Reichenbach, Hans. 1920. Relativitätstheorie und Erkenntnis a Biography priori. ———. 1924. Axiomatik der relativistischen Raum-Zeit- Hans Reichenbach was born in Hamburg, Germany, on Lehre. Braunschweig: Vieweg. September 26, 1891. He studied civil engineering, math- ———. 1928. Philosophie der Raum-Zeit-Lehre. Braunschweig: ematics, physics, and philosophy at Technische Vieweg. ———. 1935. Wahrscheinlichkeitslehre. Leyden: A.W. Stthoff. Hochschule Stuttgart and Universities of Berlin, ———. 1944. Philosophic foundations of quantum mechanics. Göttingen, and Munich; Ph.D. (1915, Erlangen) for Berkeley: University of California Press. mathematical work about probability; and Habilitation ———. 1947. Elements of symbolic logic. New York: (1920, Technische Hochschule Stuttgart) for philosophi- Macmillan. cal work about relativity theory. He was assistant ———. 1951. The rise of scientific philosophy. Berkeley: University of California Press. Professor, Technische Hochschule Stuttgart, 1920–1926; Thieroff, R. 1992. Das finite Verb im Deutschen. Modus— and Professor for Philosophy of Nature and Physics, Tempus—Distanz. Tübingen: Narr. University of Berlin, 1926–1933; he lost his teaching MONIKA RATHERT Relevance in Discourse Since discourse analysis is interested in stretches of more), quality (only say things that are true for which text longer than a sentence or an utterance, it has to you have evidence), manner (be clear, orderly, concise, address the question of coherence—how the different and avoid obscure language), and relation (be relevant). parts of texts hang together. One crucial aspect of Sperber and Wilson (1995) and Wilson and Sperber coherence will be relevance, how the meanings con- (1987) developed Gricean theory by subsuming all veyed by one sentence can be related to or interact with these maxims under relevance. For them, communica- information already in the reader’s/hearer’s mind, espe- tive utterances carry the presumption of optimal rele- cially information already given in the preceding text. vance. Utterances are relevant if the information they The first philosopher to emphasize the importance convey interacts with the hearer’s existing assump- of relevance to discoursal pragmatics was Grice (1975 tions. These interactions, called contextual effects, can [1967], 1989). His cooperative principle proposes four be either the strengthening of an existing assumption, maxims that govern conversation in particular, namely the weakening/cancelation of an existing assumption, quantity (give as much information as is required, no or the production of a contextual implication. In the 901 RELEVANCE IN DISCOURSE case of contextual implication, the information given above is to A, but to function as relevant in discourse by an utterance interacts with an existing assumption it has to be interpretable, which may involve implica- by deductive logic, to create a new assumption. One of tion or inferencing. So Grice and Sperber/Wilson pre- Sperber and Wilson’s examples is as follows: fer an inferential account of communication. In this, A: Would you drive a Rolls Royce? coding and decoding using conventional signs are B: I wouldn’t drive any expensive car. never sufficient for communication, and sometimes even unnecessary. If you ask whether I enjoyed my B does not obviously answer A’s question. However, skiing trip, and I raise my broken leg covered in plas- if A already knows that a Rolls Royce is an expensive ter, this might communicate the answer ‘no’. But such car, this assumption interacts with B’s reply to produce an action has no conventional coded meaning. In infer- the contextual implication ‘B would not drive a Rolls ential theories, utterances give evidence for a hypoth- Royce’. So, B’s answer is still relevant to A’s question. esis about an intended meaning, often an implied one, The assumption supplied by the hearer ‘a Rolls as in the examples above. Royce is an expensive car’ is called an implicated prem- However, relevance theory goes further than Grice, ise. That different contextual implications arise from in maintaining that pragmatic inferential processes are different implicated premises is particularly important important in making explicit the full propositional to the intertextuality of discourse. Intertextuality, put content of the utterance, even before implications are simply, means the way in which one text impinges on made. For example, imagine John has cooked dinner, others (Kristeva 1974:59–60). One aspect, viewed from Mary is sitting in the dining room reading the newspa- a relevance perspective, is that the information derived per, and John comes in and puts the two plates of food from one text can provide implicated premises and on the table. He then says effect the contextual implications during the processing It’ll get cold. of a later text. For example, take the following text: It is difficult to see how any organism could remain In order to make the prepositional content com- healthy if feedback messages to the decision-making plete, we have to infer what ‘it’ refers to (for example, center were systematically stifled for long periods. Mary’s dinner rather than the newspaper), disam- biguate ‘cold’ (to mean ‘low in temperature’ rather It will be processed and interpreted quite different- than ‘experiencing low temperatures’), and narrow ly, more metaphorically, if the preceding intertext is a) down the time reference of ‘ll’ (to mean ‘will soon’). a report showing that many politicians are devoting This process, an example of what is termed explica- less and less time to their constituency surgeries/feed- ture, will give us the relatively complete proposition back sessions with the electorate rather than b) a fea- ture article on the advances in the effectiveness of the Mary’s dinner will soon become low in temperature pain-killing sprays used by football-players/athletes. This is now intelligible as a complete message. But The relevance of utterances can be computed as a fraction: in order for it to be interpretable, it would have to gen- erate a relevant contextual implication. If it interacts contextual effects with Mary’s existing assumption processing effort. John wants Mary to eat her dinner while it is still hot. All other things being equal, the greater the contex- tual effects the greater the relevance, and the greater the the resulting contextual implication would be processing effort the less the relevance. We can see that John wants Mary to come and eat her dinner very soon contextual effects correspond to Gricean maxims of quality and relation; and processing effort corresponds (cf. Sperber and Wilson 1995:176–93). to Grice’s maxim of manner (and perhaps quantity). According to this formula, B’s reply might not be seen Relevance theory was seen by Wilson and Sperber as optimally relevant. B could have reduced the pro- as a complement to Chomskyan linguistic theory, cessing effort by simply replying ‘No’. However, B’s which deals with semantics but fails to account for actual answer is likely to produce far more contextual the pragmatic meanings generated in discourse. effects than the answer ‘No’. For example, if A knows Chomsky is famously uninterested in sociolinguistics. that a Cadillac and a Lexus are also expensive cars, But, as early critics of relevance theory pointed out then B’s actual reply will generate the contextual impli- (Clark 1987), without some theory of social purposes cations ‘B would not drive a Cadillac or a Lexus’. the notion of relevance is rather vacuous. Goatly Like Gricean pragmatics, relevance theory opposes (1994, 1997) argues that in order to develop as a tool a code-based theory of communication. A successful- for discourse analysis, relevance theory needs devel- ly decoded message may be intelligible, as B’s answer oped notions of social contexts such as the Hallidayan 902  RHETORIC AND LINGUISTICS concepts of register and genre. However, Blass (1990) Goatly, Andrew. 1994. Register and the redemption of relevance has argued that relevance theory gives a more satisfac- Theory: the case of metaphor. Pragmatics 4(2). ———. 1997. The language of metaphors. London: Routledge. tory account of cohesion and coherence than the model Grice, Paul. 1975 [1967]. Logic and conversation. Syntax and of Halliday and Hasan (Cohesion in English 1976). semantics, Vol. 3: speech acts, ed. by Peter Cole and Jerry Relevance theory has been interestingly applied to Morgan. New York: Academic Press. different discourse genres, including work by Tanaka ———. 1989. Studies in the way of words. Cambridge, MA: (1994) and Forceville (1994) analyzing advertise- Harvard University Press. Halliday, M.A.K., and R. Hasan. 1976. Cohesion in English. ments, Campbell (1992) and Moeschler (1989) on London: Longman. argumentation, Wilson (1990) on political language, Kristeva, Julia. 1974. La revolution de langage poetique. Paris: and Mayher (1990) for language in education. Seuil. Mayher, J. 1990. Uncommon sense: theoretical practice in lan- guage education. London: Heinemann. References Moeschler, J. 1989. Pragmatic connectives, argumentative Blakemore, Diane. 1992. Understanding utterances: an intro- coherence and relevance. Argumentation 3.3. Dordrecht: duction to pragmatics. Oxford: Blackwell. Kluwer. Blass, Regina. 1990. Relevance relations in discourse: a study Sperber, Dan, and Deirdre Wilson. 1995. Relevance: communi- with special reference to Sissala. Cambridge: Cambridge cation and cognition, 2nd edition. Oxford: Blackwell. University Press. Tanaka, Keiko. 1994. Advertising language: a pragmatic Campbell, J. 1992. An applied relevance theory of the making approach to advertisements in Britain and Japan. London: and understanding of rhetorical arguments. Language and Routledge. Communication 12(2). Wilson, Deirdre, and Dan Sperber. 1987. An outline of rele- Clark, H. 1987. Relevance to what? Behavioural and Brain vance theory. Notes on Linguistics 39. Sciences 10(4). Wilson, J. 1990. Politically speaking: the pragmatic analysis of Forceville, Charles. 1994. Pictorial metaphor in advertising. political language. London: Blackwell. London: Routledge. ANDREW GOATLY Rhetoric and Linguistics No discussion regarding the interrelationships ences derive meanings’ (Foss et al. 1999:6), an influ- between these two disciplines can proceed without a ence triggered by the modern rhetorician Kenneth thorough understanding of what their content encom- Burke, who envisioned rhetoric to be ‘symbolic passes, the key to the alliance. The roots of rhetoric inducement’ (Burke 1966:296) for the purposes of date back to the work of Aristotle (384–322 BCE) and either cooperation or competition. The site of location the fifth-century sophists (460–380 BCE) such as has shifted from public spheres of debate such as law, Protagoras and Gorgias, who in their dialogues about politics, and religion to more private constructions of the relationship of truth and reality, and the nature and meaning, particularly the manner in which individuals purpose of spoken discourse consistently demonstrat- make sense of their world, a shift that has resulted in ed that, inevitably, for these scholars there was a the synonymous connotation that rhetoric has acquired ‘philosophical and pedagogical concern with lan- in modern times with the term ‘composition’, the pri- guage’ (Gillam 1998:15). vate creation of meaning via the most common symbol Like linguistics, rhetoric has exhibited organic of persuasion, writing, in modern universities. In this changes in focus, orientation, and scope over time, entry, these terms are sometimes used synonymously. evident in the meaning of the word rhetoric, which has Although the focus in classical accounts is with the acquired meanings that are ‘irreducibly multivalent’ construction of truth, modern conceptualizations see (Fleming 1998:174). Often associated in modern lay rhetoric as providing an arena to question the privileg- thought with the political rhetoric of ‘deceit’, rhetoric ing of some truths over others, in short, ‘a tool of cri- in academia has moved from being a ‘pejorative to an tique’ to demonstrate that ‘reality, belief and language honorific term’ (Fleming 1998:169), a return to the are not lined up as unproblematically as people would classical art of persuasion. Modern uses of the term like’ (Fleming 1998:170), a shift prompting an addi- have broadened the scope of rhetoric to include ‘virtu- tional return to antiquarian roots espousing the sophist ally all humanly created symbols from which audi- tradition of Isocrates which stresses civic development 903 RHETORIC AND LINGUISTICS or Arete (virtue) on the part of the individual. In such rhetoric-composition was in the form of the 1974 a conceptualization, ‘rhetoric is a study that in addi- National Council of Teachers of English (NCTE) tion to imparting an art and guiding good practice, statement on ‘Student’s rights to their own language’, encourages critical and substantive reflection about a document first drafted by the Conference on College the situated relations of discourse to reason, character, Composition and Communication (CCCC). The 1970s and community in human action’ (Fleming 1998:184), split in linguistics between Cartesian/theoretical lin- a definition stressing the link with linguistics, particu- guistics and applied linguistics prompted a shift in larly sociolinguistics. influence, with most of the influence coming from The link between the two fields is often conceptu- applied linguistics. Some of the most commonly cited alized as the applications of linguistic theory to rheto- linguists at this time were Charles C. Fries, Kenneth ric, an alliance that can be broadly divided into two Pike, Paul Roberts, and Noam Chomsky. periods: optimistic applications of linguistic theory to These linguistic influences prompted a shift from rhetoric from the 1950s to the mid-1980s— a period of prescriptive accounts of written discourse to more ‘a good deal of excitement’ (Smitherman 1999:351)— descriptive orientations to student writing, as well as an to pessimistic speculations as to the actual applicabil- overt recognition that the cultural, linguistic, and ity of linguistic theory in the early 1990s to the rhetorical backgrounds of writers shaped their discur- present. It is important to stress that even though the sive practices. Other areas of interest in this period literature seems to insist that the influence was unidi- came from studies in contrastive rhetoric, whose pro- rectional, linguistic research has also been influenced pounder, Robert Kaplan (1966), was heavily influ- by studies in rhetoric. This might be attributed to the enced by the work in contrastive analysis in applied parallel development of composition-rhetoric and lin- linguistics. The aim of such early as well as later theo- guistics in the 1950s and 1960s as ‘both fields sought rists such as Ulla Connor was to demonstrate first lan- to reinvent themselves and stake intellectual claim to guage sociocognitive influences of speakers engaged in distinct identities among the established disciplines of second language rhetorical practices, work resonating the academy’ (Smitherman 1999:351). Consequently, the cultural determinism inherent in the linguistic the- the antiquarian interest in audience-influenced dis- ories of Benjamin Lee Whorf. Recognizing the socio- course styles saw linguistic applications in the work of cultural basis of writing, linguists such as Muriel Martin Joos (1961), who outlined five major audience- Saville-Troike and Donna M. Johnson have synthe- influenced styles of discourse. sized ethnography of communication approaches with Since language has and continues to be an integral such conceptualizations and devised linguistically ori- part of rhetoric, especially in approaches labeled epis- ented rhetorical approaches whose basic premise is temic rhetoric, whose primary focus is on the private ‘text as praxis’ (Cai 1998:123). Applications of con- sphere of semantic generation, particularly the extent trastive rhetoric in first language writing are manifest to which language simultaneously embodies and gen- in the works of Mina Shaughnessy and David erates knowledge, linguists have constantly influenced Bartholomae whose research stresses that student and been influenced by the field. For instance, errors in writing may be reflections of systematic soci- Kenneth Pike (1970) developed his theory of tag- ocultural patterns rather than random performance memics in response to the argument that since the slips, an approach heavily adopted by theorists in basic basis of language was the embodiment of knowledge, writing paradigms such as that developed by Patricia such a process could only proceed via access to reoc- Bizell and Bruce Herzberg who urge for the teaching of curring, repeatable units, the kind that language effective rhetorical strategies that permit both inter- and already offers. The dual relationship between these intracultural communication. disciplines emerges in the unique conceptualization of Research in the 1980s saw a deep alliance between Rogerian rhetoric, a persuasive technique based on the fields and ‘scholarship influenced by Pragmatics, conciliatory rather than adversarial argumentation. Discourse Analysis and Functionalist Linguistics Similar links can be made with the work of Kenneth appeared with some frequency in composition journals Burke (1966) whose model of language as action, ‘an and monographs’ (Kimball 2000:180). Shirley Brice equipment for living’ and ‘a strategy for coping’ Heath’s (1983) Ways with words, an ethnography of (Warnock 1998:11) had linguistic parallels in the per- speaking approach, influenced rhetoricians interested formative model of J.L. Austin’s speech act theory. in issues of literacy and written composition then and The influence in the 1950s through the 1970s came even today. Other connections to linguistics came in from theoretical linguistics, particularly in the early the alliance with stylistics. This marriage of disci- phases, structuralism and, later, generative grammar, plines saw a culmination in the publication of two as well as applied linguistics, particularly the work of books, one by Raskin and Weiser (1987) of Language sociolinguists. One of the most overt influences in and writing: applications of linguistics to rhetoric and 904  RHETORIC AND LINGUISTICS composition, a book demonstrating pragmatic applica- systemic functional linguistics. Researchers such as tions of linguistics in the teaching of rhetoric, and the Susan Jarratt and Lynn Worsham, Lester Faigley, and other by Beale (1987) of A pragmatic theory of rheto- James Berlin push for a view of written rhetoric as sit- ric, a rhetorical model based on J.L. Austin’s Speech uated social action, arguing that the only means to Act Theory (1962) since ‘classical concepts of rheto- empower neophyte writers from being silenced is to ric square nicely with speech act theory, making them educate them about the genres of powered discourse. mutually illuminating’ (Winterowd 1987:282). One Recent reapplications of early psycholinguistic and can say that this was the last happy alliance. sociolinguistic models have manifested themselves in By the early 1990s, institutional changes that sepa- whole language teaching approaches to rhetoric, rated English and Linguistics departments at a number which emphasize linguistic socialization within and of universities triggered pessimistic feelings on the not apart from reading, writing, listening, and speak- part of rhetoricians concerning the utility of linguistic ing. Other perspectives in genre theory have influ- analysis, a sentiment often voiced in articulations that enced rhetorical theories such as the literacy-orality the scientific method of linguistics, particularly ‘the theories of Walter Ong, an extension of Whorfian lin- atomistic nature of most linguistic analysis’ (Roy guistics situating the source of writer difference in oral 1991:580), was in contradiction to the humanistic vs. analytical cultures, and/or individuals, an outcome method of rhetoric, a difficulty accentuated even fur- of which may be a difference in cognitive complexity. ther by the metalanguage adopted by linguistics that Although this approach has been heavily critiqued in prompted Alice Roy (1991), in a classic review of four rhetorical and linguistic circles, applications of genre studies on the alliance, to describe the link as ‘prob- differences in spoken and written discourse have been lematic’ (580). Recent rhetoricians have commented recently cited by Kimball (2000) as an arena for that this pessimism is unwarranted, a consequence of research. Using the linguistic analyses of linguists a misapplication of linguistic theory, a view based on such as M.A.K. Halliday, Wallace Chafe and ‘limited and inaccurate views of linguistics’ (Kimball Danielwicz, and Douglas Biber, she argues that some 2000:181). Utilizing research from Gary Olson and of the difficulties that novice rhetoricians face in this Lester Faigley, who review Chomskian applications to era may be an inability to distinguish between the fea- rhetoric concluding that ‘It’s very rare that you ever tures of this genre, a knowledge of which is presumed get a free ride from some other field’ (34), prompts by the academy, and oral discourse. Returning to the Kimball (2000) to call for a renewed interest in the unity of the disciplines in classical antiquity has applicability of linguistic theories to rhetoric today. caused some researchers in linguistics to recreate an Obvious examples of such renewed interest come in alliance that benefits both pragmatics and rhetoric. discourse theories in rhetoric propounded by Kinneavy With the renewed interest in applied linguistic the- et al. (1990), as well as social constructivist research in ories, the alliance between rhetoric and linguistics the field by theorists such as Susan Jarratt and Lynn looks promising, a possibility accentuated by the fact Worsham, Lester Faigley, and James Berlin, all of that more and more students and faculty in composi- whom view language to be a complex, pluralistic tion-rhetoric and linguistics departments have some signifying practice that simultaneously constructs real- exposure to each other. What makes the synthesis even ities as it presents it—a rhetorical paradigm rooted in more viable is that the area of interest in both rhetoric Roland Barthes’s (1968) theory of social semiotics, an and linguistics from antiquity to the present has been approach emphasizing the social and ideological questions of language. In both fields therefore, we underpinnings of symbol use. These conceptualizations have to continue to see the product of language as a show heavy influences from linguistic research in crit- ‘subject of study rather than an object of intervention’ ical discourse analysis, particularly social semiotics. (Kimball 2000:188). With this said, one can only pre- Renewed interest has occurred in the area of gener- dict more of an alliance in this century. ative rhetoric, a paradigm originally devised by Christensen and Christensen (1976), who, influenced by generative linguistics, argued that texture in dis- References course proceeded via syntactic elaboration, particular- Austin, J.L. 1962. How to do things with words. Oxford: ly via modification and addition. There has been a Oxford University Press. recent pedagogical push for more overt sentence com- Barthes, Roland. 1968. Elements of semiotics, transl. by Anetto bining instruction in writing. An exponent, Kimball Lavers and Colin Smith. New York: Hill and Wang. (2000) argues for this approach to be utilized in the Beale, Walter H. 1987. A pragmatic theory of rhetoric. Carbondale: Southern Illinois University Press. revision stage of writing. Other current applications of Burke, Kenneth. 1966. Rhetoric and poetics. Language as sym- linguistics are in genre theory, a theoretical paradigm bolic action: essays on life, literature and method, 295–307. strongly influenced by the works of M.A.K. Halliday’s Berkeley: University of California Press. 905 RHETORIC AND LINGUISTICS Cai, Guanjun. 1998. Contrastive rhetoric. Theorizing composi- Kimball, Sara. 2000. Linguistics and composition. Reforming tion: a critical sourcebook of theory and scholarship in con- college composition: writing the wrongs, ed. by Ray temporary composition studies, ed. by Mary Lynch Wallace, Alan Jackson and Susan Lewis Wallace, 179–90. Kennedy, 52–6. Westport, CT: Greenwood Press. Westport, CT: Greenwood Press. Christensen, Francis, and Bonnijean Christensen. 1976. A new Kinneavy, James L., William J. McCleary, and Neil Nakadate. rhetoric. New York: Harper Press. 1990. Writing in the liberal arts tradition, 2nd edition. New Fleming, David. 1998. Rhetoric as a course of study. College York: Harper. English 61(2). 169–91. Pike, Kenneth L. 1970. Tagmemics and matrix linguistics Foss, Karen, A., Sonja K. Foss, and Cindy L. Griffin. 1999. applied to selected African languages. Norman: Summer Feminist rhetorical theories. Thousand Oaks, CA: Sage Institute of Linguistics of the University of Oklahoma. Publications. Raskin, Victor, and Irwin Weiser. 1987. Language and writing: Gillam, Alice. 1998. Classical rhetoric. Theorizing composi- applications of linguistics to rhetoric and composition. tion: a critical sourcebook of theory and scholarship in con- Norwood, NJ: Ablex Publishing Corporation. temporary composition studies, ed. by Mary Lynch Roy, Alice. 1991. Four studies of linguistics and composition. Kennedy, 15–26. Westport, CT: Greenwood Press. College English 53(5). 580–6. Heath, Shirley Brice. 1982. Protean shapes in literacy events: Smitherman, Geneva. 1999. CCCC’s role in the struggle for ever-shifting oral and literate traditions. Spoken and written language rights. College Composition and Communication language, ed. by Deborah Tannen, 91–117. Norwood, NJ: 50(3). 349–75. Ablex Publishing Company. Warnock, Tilly. 1998. Burkean theories of rhetoric. Theorizing ———. 1983. Ways with words: language, life, and work in composition: a critical sourcebook of theory and scholarship communities and classrooms. New York: Cambridge in contemporary composition studies, ed. by Mary Lynch University Press. Kennedy, 9–13. Westport, CT: Greenwood Press. Joos, Martin. 1961. The five clocks. New York: Harcourt, Brace Winterowd, Ross W. 1987. Literacy, linguistics and rhetoric. and World. Teaching composition: twelve bibliographic essays, ed. by Kaplan, Robert B. 1966. Cultural thought patterns in intercul- Gary Tate, 265–90. Fort Worth: Texas Christian University tural education. Language Learning: A Journal of Applied Press. Linguistics 17. 1–20. ANJALI PANDEY Romanian The Romanian language (Limba româna ˘) is a Romance Romanian language is a combination of the two main language that has been continuously spoken in the east- geographical varieties, the Moldavian (west of the Prut ern part of the Roman Empire, including the Romanized river) and the Wallachian (Muntenian), the latter playing provinces of Dacia, Southern Pannonia, Dardania, and the more important role in setting the standard. Given Moesia, from the moment Latin was brought there until the language’s relative isolation from the other Romance the present. The Romanian language has four main languages and its close contact with Hungarian and the dialects: Daco-Romanian (or Romanian proper), Slavic languages surrounding Romania, the develop- Macedo-Romanian or Aromanian, Megleno-Romanian, ment of Romanian followed a different path from those and Istro-Romanian. The Aromanian dialect is spoken of most other Romance languages. in Albania, Macedonia, Greece, and Bulgaria by scat- One of the most important foreign influences came tered minorities, and the main area where Megleno- from the Slavic languages. As a liturgical language, Old Romanian is spoken is in Greece. The Istro-Romanian Slav(on)ic (or Old Bulgarian) provided a relatively large dialect was spoken on the Istrian Peninsula of Croatia. number of words to Romanian. This situation, combined The Daco-Romanian dialect (‘Romanian’) is spoken by with the fact that until the nineteenth century the Cyrillic more than 25 million people in Romania, Serbia, alphabet was used to write Romanian and with the Croatia, Slovenia, Bosnia and Herzegovina, Bulgaria, geopolitical position of Romania, has misled many peo- Hungary, and Moldova, as well as in Romanian com- ple into thinking that Romanian is a Slavic language and munities in the USA, Australia, Italy, Germany, etc. The that Romania is a successor state of the Soviet Union. variety of Romanian spoken in the Republic of Moldova Another major influence in the development of the (east of the Prut river) is officially called ‘the Moldavian Romanian vocabulary came from Greek, especially dur- language’. ing the Phanariot period (1711–1821). However, very The earliest known written document in Romanian is few of the Greek borrowings are still in use today. Other a letter dating from 1521. The present spelling system foreign influences came from Turkish, Hungarian, and was settled around 1860–1880. The basis of the standard German. Turkish influence was prominent in Moldavia 906  RULES VS. CONSTRAINTS and Wallachia during the Phanariot period, and Gender in Romanian distinguishes masculine, fem- Hungarian influence was prominent in Transylvania inine, and neuter (masculine and feminine genders are while it was part of the Austro-Hungarian Empire. The inherited from Latin, while neuter is a creation within German elements used by the Romanian minorities of Romanian). German origin also played a role in the diversification of The changes from Latin are typically simplifica- the Romanian vocabulary. More recent foreign influ- tions (as in all Romance languages), which led to a ences came from French in the first part of the twentieth reorganization of the vocabulary by reducing the com- century, while today massive English influence on the plicated forms of the noun and adjective, by simplify- Romanian vocabulary can be easily observed. ing and rearranging the verbal system, and by Within the geographical area of Romania, there are simplifying the case system. three main varieties of the language: Moldavian, While the Romance languages usually have Wallachian (Muntenian), and Transylvanian. reduced nominal declensions, Romanian has three Additionally, in the Republic of Moldavia the inhabi- forms of noun inflection inherited from Latin. The tants speak the Bessarabian variety of Romanian, declension of adjectives is identical to that of nouns, which is characterized by the massive influence of varying according to gender, number, and case. The Russian. All these varieties, however, are mutually pronominal declension is the richest and the closest to intelligible. The standard pronunciation of Romanian the complex Latin system of all the Romance lan- is based on orthographic norms. In this respect, the guages. The Romanian verbal system is largely based principle generally promoted is that of identity on Latin. The most noticeable differences from the between writing and pronunciation. other Romance languages are the analytic future (voi As far as the vocabulary is concerned, Romanian citi ‘I shall read’) and the supine (de citit ‘for read- inherited c. 2,000 original Latin words, but Latin ele- ing’). Romanian is the only Romance language with a ments make up 60% of the Romanian vocabulary. free sequence of tenses and a relatively free word Most of the original Latin words refer to parts of the order. human body and have several meanings, including metaphorical ones. References In contrast to the vocabulary, Romanian grammati- Constantinescu-Dobridor, Gheorghe. 1994. Sintaxa limbii cal structure is almost entirely Latin. There are Latin române (The syntax of the Romanian language). Bucharest: prepositions (e.g. cu ‘with’) and conjunctions (e.g. s,i Editura Stiintifica. B ‘and’), Latin interrogative pronouns (e.g. cine ‘who’), Coteanu, Ion. 1993. Gramatica de baza a limbii române (The fundamental grammar of the Romanian language). Latin endings for marking the plural, Latin verbal end- Bucharest: Garamond. ings for marking person and tense, Latin pronouns, Daniliuc, Radu, and Laura Daniliuc. 2000. Descriptive and Latin numerals from one to ten. Romanian grammar. An outline. Munich: Lincom Europa. Romanian grammar involves ten parts of speech: Doca, Gheorghe. 1996. Limba Româna ˘. Probleme de sinonimie six inflected (noun, article, adjective, pronoun, numer- s, i de cultivare a limbii (Romanian language. Synonymy and style issues). Bucharest: TUB. al, and verb) and four noninflected (adverb, preposi- Harris, Martin, and Nigel Vincent (eds.) 1990. The romance tion, conjunction, and interjection). The following languages. London: Routledge. grammatical categories all stem from Latin: gender, Pop, Liana. 1993. Romanian with or without a teacher. Cluj: number, case, person, comparison, voice, mood, and Echinox. tense. However, the development from Latin to Posner, Rebecca. 1996. The romance languages. Cambridge: Cambridge University Press. Modern Standard Romanian generated several differ- Serban, Domnica. 1986. Syntactic functions in universal gram- ences. Thus, for example, Latin had six cases, while mar. A contrastive study of the subject in English and Romanian preserves only five—nominative, genitive, Romanian. Bucharest: TUB. dative, accusative, and vocative. RADU DANILIUC Rules vs. Constraints The question of whether grammar should be modeled theoretical issues. In some work, e.g. Karttunen using rules or constraints presents a morass of termi- (1993), the terms ‘(declarative) rules’ and ‘constraints’ nological confusion in addition to some legitimate are used interchangeably. There are even positions 907 RULES VS. CONSTRAINTS such as Mohanan’s (2000:146), who argues that the favor rule-based models of grammar, by pointing out a distinction is incoherent once we recognize that both number of conceptual and empirical problems with rules and constraints express propositions. This article constraints. Our first criticism of constraint-based proposes a definition of ‘rule’ vs. ‘constraint’ that will models is related to issues of learnability, acquisition, surely be inconsistent with other usages, but which and the nature of universal grammar (UG), i.e. the provides the basis for discussing competing models of human language faculty. Many linguists, especially grammar, focussing here on phonology. phonologists, have assumed that both UG and particu- The use of both rules and constraints requires a lar grammars contain constraints—either positive or representational matching procedure (RMP) to deter- negative conditions that, respectively, must or must not mine whether a given rule or constraint P applies to a be satisfied by grammatical representations. However, given representation q. Part of the statement of P must such conditions cannot be learned by positive evi- be a structural description, SDp. The RMP must eval- dence. This is because a generative grammar generates uate q to determine whether it satisfies SDp. The RMP an unbounded number of well-formed structures, but outputs two possible results: YES, q satisfies the only a finite sample can be encountered by any point structural description of P; or NO, q does not satisfy in acquisition. The learner may find a supposed ill- the structural description of P. In other words, the formed structure in the next sentence. For example, a RMP defines the domain of the function P. In many putative constraint that all syllables must have onsets theories, the RMP relies on subsumption—q is in the may be consistent with the data received up to a given domain of P if SDp subsumes q—although there are point; however, there is no guarantee that the inductive other possibilities. conclusion consistent with the constraint will turn out If the RMP determines that q is in the domain of P, to be correct. Therefore, such conditions could only be we can now distinguish between a rule and a con- learned via negative evidence. However, it is general- straint. Consider a rule of the form R(I)  O, that is, R ly accepted (see Marcus (1993) for convincing argu- maps input I to output O. If q is in the domain of R (i.e. ments) that negative evidence is neither supplied to the if the RMP determines that q matches I), then that part child with sufficient regularity, nor attended to by the of q that is identical to I is rewritten as O yielding q: child when supplied, to play a significant role in lan- R(q)  q. Thus, we define rules as functions from rep- guage learning. Thus, the constraints cannot be resentations to representations, with both input and learned via positive evidence (for reasons of logic), output built from the same representational primitives. nor through negative evidence (according to the In a constraint-based system, if the RMP deter- empirical data from language acquisition). So they mines that q is in the domain of a constraint C, then an must be innate. operation of constraint evaluation must determine This conclusion follows from the premises, and it whether q satisfies the constraint. If q satisfies C, con- has led to implausibly rich versions of UG. In straint evaluation outputs the value NOVIOLATION; if q phonology, for example, innate constraints have been does not satisfy C, constraint evaluation outputs the posited to account for the vast variety of phonological value VIOLATION. These values then serve as input to alternations seen in the languages of the world. The another component of the grammar. In some models, OT literature provides examples of constraints to if there is any constraint that q violates, then q is account for voicing of initial obstruents in the second ungrammatical. In optimality theory (OT), in contrast, member of Japanese compounds, flapping in English, the evaluations of q for each constraint are fed to and so on, thus trivializing the notion of UG. another component of the grammar, which compares q Fortunately, the conclusion of innate constraints can to competing representations. The output of the gram- be rejected: the problem lies with the assumption that mar typically does violate some constraints in OT, but UG, and also particular grammars, consist of con- they must be lower ranked than constraints violated by straints. By positing a set of representational and competing representations. computational primitives, such as syllables and feet, To reiterate: a rule is defined as a function from rep- precedence relations, identity evaluations, etc., and a resentations to representations; a constraint is defined learning mechanism that combines these primitives as a function from representations to the set into rules that capture alternations, a learner can build {VIOLATION, NOVIOLATION}. Note that a constraint- a language-specific grammar based solely on positive based system requires two evaluations of a given rep- evidence. By positing highly specific rules that con- resentation q—one corresponding to the RMP to form to the data, the learning path is constrained with- determine whether a constraint C is relevant to q and out the use of constraints. one to the evaluation of whether q violates C. A cluster of arguments against constraint-based mod- Equipped with this working distinction between els relate to the fact that they typically depend on adopt- rule and constraint, we can now offer arguments to ing a notion of (absolute or relative) ill-formedness 908  RULES VS. CONSTRAINTS of linguistic representations. Ill-formed representa- ‘individual psychology’ as the domain of inquiry for tions violate constraints, and this result is passed on to linguistics, Chomsky (1986) argued convincingly that other modules of the grammar. An obvious objection the fact that knowledge of language is instantiated in to this approach is that there is a much simpler alter- individual minds/brains means that there is necessari- native to the idea that grammars generate ill-formed ly a ‘correct’ characterization of a speaker’s grammar. representations and then reject them as possible out- Rejecting the view that the language faculty con- puts—why not assume that the grammar just generates sists of constraints has methodological implications. those representations that it outputs? The grammar Just as we do not want the language faculty to contain itself does not need the notion of ill-formedness. Of constraints against grammatical structure, models of course, a linguist may refer to a representation as ill- the language faculty should not be described via con- formed, in the sense of ‘not generatable by a particu- straints. If our current hypothesis concerning UG is lar grammar L’, but this description by a linguist stated only in positive terms, as statements of what should not be interpreted as a description of an actual grammars have access to or consist of, without prohi- property of L. Instead of accounting for the infinite set bitions or constraints, we can achieve a more econom- of representations that L does not generate, normal sci- ical model. The positive terms are just those entities entific practice seems to dictate that we account for and operations (features, syllables, deletions, inser- what L does generate. Ill-formedness, or rather non- tions (in phonology); and Merge, Move and feature- formedness, is thus a derivative notion. checking (in syntax)) which have been observed Various theories of grammar, including OT and empirically or inferred in the course of model con- some versions of minimalism and its predecessors struction. When faced with a phenomenon that is not posit a mechanism that allows unconstrained genera- immediately amenable to modeling using existing ele- tion of linguistic representations. In OT, this device is ments of the vocabulary, scientific methodology (basi- GEN which, given an input, generates the universal cally Occam’s Razor) guides us. We must first try to candidate set of possible outputs. In various syntactic reduce the new phenomenon to a description in terms theories, an analog to GEN is the ‘free’ concatenation of the vocabulary we already have. If this can be of morphemes, or the ‘free’ application of operations shown to be impossible, only then can we justify such as Move α. A derivation which is thus generated expanding the vocabulary. will either satisfy certain constraints at PF and LF, the Once one accepts that modules or processes of grammar’s interface levels, and thus converge; or it grammar must have a certain set of properties and that will not satisfy those conditions and it will crash. Both these properties ultimately must be describable in the OT approach and the free-generation-with-inter- terms of a set of positive statements (a vocabulary) face-conditions approach in syntax are flawed in the and can be incorporated into the structural descrip- following (related) ways. tions of rules, it appears to be the case that a proce- First, it is easy to proclaim something like ‘GEN gen- dural, or rule-based approach to grammar that erates any possible linguistic representation’ or ‘The generates a sequence of representations constituting a syntactic component allows Move α to apply freely’. derivation is to be preferred to a constraint-based, However, it is not clear what such statements mean. nonderivational theory. In other words, grammars can One could argue that the theory of grammar need not be understood as complex functions mapping inputs be computationally tractable, since grammar models to outputs. A rule-based model just breaks the com- knowledge and does not necessarily provide an algo- plex function into simpler components in order to rithm for humans to produce speech output. However, understand the whole. A theory that incorporates GEN there must be some mechanism that does generate par- or Move α avoids the problem of characterizing the ticular linguistic representations, and it is not unrea- function that is the grammar. sonable to ask a theory of grammar to characterize the A further objection to the notion of ill-formedness is formal properties of this mechanism. It seems that any the fact that the justification for evaluating representa- implementation of GEN or the syntactic component that tions as ill-formed is typically derived from some ver- incorporates Move α will have to be explicit about what sion of markedness theory. Depending on the it does. One way to achieve this is to be explicit about formulation of a given constraint, either matching or what the abstract grammar generates. failing to fulfill the condition specified by the constraint Second, the free generation-cum-filters model signals VIOLATION. For example, a constraint formulat- stinks somewhat of antimentalism. It basically says ed as ‘Don’t have a coda’ leads to an evaluation of ill- ‘We don’t care how the candidate forms are generated, formedness for a syllable that has a coda. A constraint as long as they are generated. One way is as good as formulated as ‘Have an onset’ leads to an evaluation of the next, as long as they are extensionally (empirical- ill-formedness for a syllable that does not have an onset. ly) equivalent.’ In defining I-language, a matter of Relative and absolute ill-formedness or markedness 909 RULES VS. CONSTRAINTS evaluations of linguistic representations are ascribed by phonological theory innate constraints that lead to linguists for grammar-external reasons, such as ease of place feature neutralization in codas. The facts of per- articulation or perception, avoidance of ambiguity, etc., ception demonstrated in the laboratory, combined with and thus have no place in a formal model of grammar— a theory of language acquisition and change, will humans know about ambiguity, and can try to avoid it, account for the attested patterns. Offering an addition- but grammars do not know anything. al cause, within the phonology, is just bad science. Marked or ill-formed structures typically are Based on learnability considerations, the weakness claimed to have at least one of the following properties: of markedness considerations, and various conceptual arguments, we conclude that a rule-based derivational ● Relative rarity in the languages of the world. model of grammar is superior, since it can be stated in ● Late ‘acquisition’ by children (typically referring purely positive terms, without prohibitions. The prob- to the recognizability of a form in child speech). lem of inductive uncertainty is avoided, by the learner ● Loss in aphasia (typically referring to the recog- and the linguist, if we adopt a rule-based model of nizability of a form in aphasic speech). grammar, as defined here. ● Relative difficulty of perception (not always experimentally validated). ● Relative difficulty of articulation (based on References impressions of what is hard to say). ● Tendency to be lost in language change and to Chomsky, Noam. 1986. Knowledge of language. Westport, CT: Praeger. not arise in language change. Epstein, S., E. Groat, R. Kawashima, and H. Kitahara. 1998. All of these criteria have been criticized by Hale A derivational approach to syntactic relations. Oxford: Oxford University Press. and Reiss (2000a,b; see references therein). These Hale, M., and C. Reiss. 2000a. Substance abuse and dysfunc- works conclude that the best way to gain an under- tionalism: current trends in phonology. Linguistic Inquiry standing of the computational system of phonology is 31. 157–69. to assume that the phonetic substance (say, the spectral Hale, M., and C. Reiss. 2000b. Phonology as cognition. properties of sound waves, or the physiology of artic- Phonological knowledge, ed. by N. Burton-Roberts, Philip Carr, and Gerry Docherty. Oxford: Oxford University Press. ulation) that leads to the construction of phonological Karttunen, Lauri. 1993. Finite state constraints. The last phono- entities (say, feature matrices) never directly deter- logical rule, ed. by John Goldsmith. Chicago, IL: University mines how the phonological entities are treated by the of Chicago Press. computational system. The computational system Mohanan, K. P. 2000. The theoretical substance of optimality treats features as arbitrary symbols. formalism. The Linguistic Review 17. 143–66. Ohala, J. 1990. The phonetics and phonology of aspects of An additional consideration in evaluating marked- assimilation. Papers in laboratory phonology I: between the ness and the notion of grammatical ill-formedness, grammar and physics of speech, ed. by J. Kingston and M. pointed out by John Ohala over the years, is that build- Beckman. Cambridge: Cambridge University Press. ing into a theory of universal grammar the articulatory Reiss, C. 2003. Deriving the feature-filling/feature-changing or perceptual basis for a recurring sound pattern leads contrast: an application to Hungarian vowel harmony. Linguistic Inquiry 34. 199–224. to an unacceptable duplication of explanatory mecha- ———. Constraining acquisition without constraints. Rules nisms. For example, if we can show in a laboratory and constraints in phonology, ed. by A. Nevins and B. Vaux. that consonant place distinctions are less salient in Oxford: Oxford University Press, to appear. codas than in onsets, we do not need to build into CHARLES REISS Russell, Bertrand Arthur William A leading British philosopher, logician, essayist, paci- only to be invested with the Order of Merit in 1949 fist, and political activist, Bertrand Russell was born in and, a year later, awarded the Nobel Prize for 1872 into a liberal and aristocratic family but was Literature. By the time of his death in 1970 at the age sadly orphaned at the age of 4. He grew up as an enfant of 97, he had authored some 70 books, and around terrible and later a tireless critic of the establishment, 4,300 articles, book chapters, reviews, pamphlets etc., 910  RUSSELL, BERTRAND ARTHUR WILLIAM besides more than 61,000 letters and other documents when there was no referent to the descriptive phrase not originally intended for publication. Along with ‘the present king of France’), was bereft of any defi- G.E. Moore, he is widely acclaimed as one of the nite truth value, because the existence of the referent is founders of analytic philosophy and, with Kurt Gödel, a precondition (or, technically, presupposition) for he shares the reputation of having made the most sig- what is asserted by the sentence to be capable of being nificant contribution to the study of logic in the twen- either true or false. When the presupposition is not tieth century. An unflinching pacifist, he was true, what is asserted by the sentence will be neither imprisoned for involvement in antiwar protests and true nor false. This meant that the time-honored prin- dismissed from his job at Cambridge University in ciple of logic known as the law of the excluded middle 1916, and in 1940 from his job at the City College, according to which, given a sentence and its (contra- New York; years later, in 1961, he was again arrested dictory) negation, one of them has to be necessarily and imprisoned for a week, this time for engaging in true no matter what the world is like would be, from antinuclear protests. As an essayist and pamphleteer, now on, subject to the additional constraint that stipu- Russell addressed a number of issues of contemporary lated ‘provided the sentence’s presuppositions are social, political, and moral concern and never shied true’. Russell realized that such a consequence was away from expressing his views boldly and openly as simply devastating for logic’s reputation as a ‘science in his book Why I am not a Christian. of pure reason’. Russell was a strong advocate of logicism—the Russell’s alternative analysis began by showing that view that all of mathematics is ultimately reducible to the logical form of the sentence under discussion was logic. To him goes the credit for having directed the much more complex than Frege had thought: it was, he attention of fellow philosophers and logicians to the argued, actually a conjunction of three separate propo- utmost importance of detecting and rooting out contra- sitions, namely: (1) There is a present King of France. dictions in set theory and elsewhere. In 1901, Russell (2) There is only one king of France. (3) Whosoever is discovered what is famously known in the literature as the King of France is bald. Now, it is commonplace in ‘Russell’s paradox’, later fully elaborated in his work logic that the negation of a compound sentence could Principles of mathematics (1903). The paradox, or be shown to be true if any one (or more) of its con- antinomy as it is sometimes called, was identified by stituents is known to be false. So, the nonexistence of imagining a set of all sets that were not members of a King of France was as strong a reason for a sentence themselves and asking, apropos of that resultant set, if affirming his baldness to be false as the fact of the king it was a member of itself, the answer being that it can being hirsute. What Frege took to be a presupposition only be a member of itself by not being a member of of the sentence was thus shown to be capable of being itself—which is a blatant contradiction. The discovery treated as one of its straightforward entailments. of this paradox stimulated intense research in set theo- Whereas the Fregean analysis demanded a thorough ry and led to several suggested solutions. Russell’s own revision of one of the fundamental claims on behalf of solution was elaborated in his theory of types, which logic (in addition to the familiar ‘true’ and ‘false’, basically consisted in postulating a hierarchy of sen- there was the need to admit of ‘neither true nor false’ tences with strict prohibition on mixing levels and thus as a third possibility), Russell’s solution required no preempting the very question that led to the discovery new apparatus other than the standard bivalent logic. of the paradox in the first place. From the 1970s onward, many linguists such as Among Russell’s most significant contributions to Deirdre Wilson (1975) and Ruth Kempson (1975) the philosophy of language is his famous ‘theory of have followed Russell’s lead and urged that many of definite descriptions’ put forward in response to the the phenomena treated by earlier researchers as pre- German logician Gottlob Frege’s theory of ‘sense’ suppositions should be better treated as semantic (Sinn) and ‘reference’ (Bedeutung). Frege had argued entailments, and any residual problems such as that two expressions, say, ‘the morning star’ and ‘the favored readings (including the intuitive feeling that evening star’, although clearly different in their sens- certain presuppositions do hold good in determinate es, can nevertheless have the same referent, namely, in circumstances) should be accounted for by a separate this particular instance, the planet Venus. Despite its pragmatic component that will have recourse to such intuitive appeal, Russell found the distinction unsatis- theoretical apparatuses as Grice’s theory of conversa- factory and was appalled by what Frege had shown to tional maxims. The decision to consign part of the be an important consequence of the distinction for the problematic to pragmatics may be seen as having as its very enterprise of logic as it had been traditionally ultimate inspirational source a key element in P.F. conceived. For, in Frege’s view, the sentence ‘The Strawson’s objections to Russell’s analysis on the present king of France is bald’, uttered at a time when grounds that it did not do full justice to ordinary lan- France was no longer a monarchy (or, equivalently, guage. (In a protracted polemic with Russell, Strawson 911 RUSSELL, BERTRAND ARTHUR WILLIAM insisted that there was a need to distinguish a sentence dox. From 1907 to 1910 he worked on Principia math- or an expression from a use and an utterance thereof.) ematica in collaboration with A.N. Whitehead. He was Russell’s theory of descriptions provided the logical elected Fellow of the Royal Society in 1908. He was basis for his epistemology centered on a distinction dismissed from Trinity College for antiwar protests between ‘knowledge by acquaintance’ and ‘knowledge (1916), and was imprisoned for six months for involve- by description’. Among the objects that could be ment in antiwar protests (1918). He was awarded the known through acquaintance Russell included what he Order of Merit (1949) and Nobel Prize for Literature referred to as ‘logically proper names’, whose function (1950). In 1955, he released the Russell–Einstein was to logically refer to certain objects such as sense Manifesto condemning the nuclear arms race. In 1958, data and universals. Ordinary names, by contrast, were he became founding President of the Campaign for regarded by Russell as functioning as definite descrip- Nuclear Disarmament and was imprisoned for one tions in disguise. Russell thus pioneered the program week in 1961 for leading antinuclear protests. He died known as ‘logical atomism’, the locus classicus of in Penrhyndeudraeth, Wales, on February 2, 1970. which is Wittgenstein’s Tratactus logico-philosophicus (translated into English under the same title in 1922), References originally prepared under Russell’s own supervision. In Ayer, A.J. 1972. Russell. London: Fontana/Collins. Russell’s view, the world is wholly made up of atomic Clark, Ronald. 1975. The life of Bertrand Russell. London: facts and every meaningful proposition must be com- J. Cape. posed of constituents with which we are acquainted. Kempson, Ruth. 1975. Presupposition and the delimitation of Among Russell’s major publications is the monu- semantics. Cambridge: Cambridge University Press. mental work Principia mathematica (1910–1913), Irvine, A.D. (ed.) 1998. Bertrand Russell: critical assessments, 4 vols. London: Routledge. written in collaboration with his Cambridge tutor Monk, Ray. 1996. Bertrand Russell: the spirit of solitude. Alfred North Whitehead. His 1912 work, The prob- London: J. Cape. lems of philosophy, has long served as one of the pop- Russell, Bertrand Arthur William, and Alfred North Whitehead. ular introductions to the subject for undergraduates. 1910, 1912, 1913. Principia mathematica, 3 vols. Cambridge: Cambridge University Press. Russell, Bertrand Arthur William. 1912. The problems of phi- Biography losophy. London: Williams and Norgate; New York: Henry Holt and Company. Bertrand Russell was born on May 18, 1872 at ———. 1914. Our knowledge of the external world. Chicago Ravenscroft, Wales. He was orphaned at the age of 4, and London: The Open Court Publishing Company. with the death of his mother and sister in 1874, and his ———. 1927. Why i am not a Christian, London: Watts; New York: The Truth Seeker Company. father in 1876. In 1876, Russell’s grandfather, Lord ———. 1929. Marriage and morals. London: George Allen & John Russell (a former Prime Minister of Great Unwin; New York: Horace Liveright. Britain), and grandmother succeeded in overturning his ———. 1945. A History of western philosophy. New York: father’s will to win custody of Russell and his brother. Simon and Schuster; London: George Allen & Unwin, 1946. With the death of his grandfather in 1878, Russell’s ———. 1959. My philosophical development. London: George Allen & Unwin; New York: Simon and Schuster. grandmother began supervising his upbringing. In Sainsbury, R.M. 1979. Russell. London: Routledge. 1890, he went to Trinity College, Cambridge, and three Wilson, Deirdre. 1975. Presuppositions and non-truth condi- years later was awarded a first class B.A. in tional semantics. London: Academic Press. Mathematics. In 1901, he discovered Russell’s para- KANAVILLIL RAJAGOPALAN Russian and East Slavic Languages The East Slavic languages—Russian (R), Ukrainian The East Slavic languages developed from the east- (U), Belarusian (Br)—constitute one of the three ern dialects of Late Common Slavic, when speakers branches of the Slavic language family. All three East migrated northward and eastward from the presumed Slavic languages are written in the Cyrillic alphabet, Slavic homeland north of the Carpathians around 500 but Belarusian has a parallel Latin orthographical tra- CE into territory sparsely populated by Finno-Ugrian dition dating from the mid-sixteenth century. and Baltic peoples. By the eighth century, a lively 912  RUSSIAN AND EAST SLAVIC LANGUAGES trade in furs, amber, wax, honey, slaves, and silver texts provide evidence of the actual speech habits of across this vast territory attracted Scandinavian war- the Rusian scribe producing them. rior-traders, the Rus’, who sought to control the routes Since the 1950s, archeological exploration has that linked the markets of Europe and Asia. revealed direct examples of written Rusian in docu- The Rus’ dominated the East Slavic tribes in the ments uncovered in predominately northwestern ninth and tenth centuries, consolidating power first in locales, especially Novgorod, Pskov, and Staraja the lands of the north around Lake Il’men’ and at Russa. Unlike their Slavonic counterparts preserved Novgorod, and then in the south at Kiev. They gradu- on costly parchment, these legal and financial docu- ally assimilated into the larger Slavic population, leav- ments or personal correspondence were scratched onto ing traces in the form of a few dozen loanwords (e.g. disposable birchbark. They provide valuable examples R jákor, U jákir, Br jákar ‘anchor’; cf. Old Swedish of the Rusian vernacular in a geographically circum- ankari), but most importantly in the very name Rus’, scribed area from the eleventh through the fifteenth ascribed first to the East Slavs around Kiev and deriv- centuries. atively to the Kievan lands. The name ultimately came From the very beginning, strictly Rusian features to refer to East Slavic lands collectively and by exten- mingled with those of Slavonic, a situation that became sion to all the early East Slavs and to their language, more common as new texts, created and borrowed, which for the early historical period (up to the four- provided fresh opportunities for linguistic choice. teenth century) can be called Rusian. Generic and thematic factors certainly influenced the Our direct knowledge of Rusian is complicated by predilection for Slavonic in religious and other solemn the fact that Rusian Church Slavonic, and not Rusian, texts and Russian in mundane ones, but they were not was the first written language, gradually introduced overriding. At times, for example, a scribe might pre- after the conversion of the East Slavs to Christianity in fer Slavonic grad ¯b over Rusian gorod ¯b simply to save 988. Rusian Church Slavonic (hereafter Slavonic) was the space of one letter on the expensive parchment at a local variant of Old Church Slavonic, the liturgical the end of a line. language devised by Saints Cyril and Methodius on Although there was dialectal divergence among the the basis of their ninth-century South Slavic East Slavs by the time of Christianization, virtually no (Macedonian) dialect for use in spreading Byzantine evidence exists to suggest that it caused any difficulty in Christianity among the Slavs. communication between residents of the Novgorodian Slavonic was basically comprehensible to the East north and the Kievan south. For this reason, the vernac- Slavs but contained a number of superficially South ular language of the early East Slavs is best considered Slavic features, primarily of a phonological and mor- as Rusian alone, with dialectal variants distributed geo- phological nature. For example, the Common Slavic graphically. Nonetheless, linguistic evolution and his- roots *gard- ‘fortified town, wall’ and *berg- ‘bank, torical circumstance helped to extend and deepen these shore’ were realized as grad-, brêg- in Slavonic, where- distinctions to a point that it is more appropriate to as the vernacular Rusian had the regular East Slavic speak of separate languages rather than dialects. pleophonic counterparts gorod-, bereg-, respectively. The effects of the phonological change known as Similarly, the Slavonic and Rusian reflexes of the the ‘jer shift’ produced dramatic differences north and Common Slavic sequence *tj were šJ (shch) and J (ch), south, alternative choices that affected consonant and respectively, and thus Slavonic svêšJ a and Rusian vowel inventory as well as morphophonemic and mor- svêJ a ‘lamp; candle’, both derived from the Common phological alternation (see below). Historical and Slavic root *svêt- ‘light’ plus the suffix -j-. Despite political factors also served to create distinct commu- such formal differences, the basic grammar and the nities of speakers whose speech patterns became majority of the most common lexical items in Slavonic increasingly divergent from one another and subject to and Rusian were identical. Some scholars prefer to influence from other languages. For example, the treat them as variants or registers of a single language, growing importance of such northeastern centers as whereas others view them as two closely related but Rostov, Suzdal’, and Vladimir (twelfth century) in the separate languages. The latter view is adopted here. face of older Novgorod and Kiev; the Mongol inva- In addition to the superficial distinctions noted sions (mid-thirteenth century); the Novgorodian colo- above, written Slavonic differed from vernacular nization of the north (thirteenth to mid-fifteenth Rusian in its richer lexicon, and in its potentially more centuries); the incorporation of southwestern Rus’ into complex syntax, both inspired by elements of the the Grand Duchy of Lithuania (thirteenth to fourteenth Bulgarian Slavonic, Serbian Slavonic, or Greek texts centuries) and Poland (fourteenth century); and the that served as sources for translation. Although the pri- ascendancy of Moscow (fourteenth to fifteenth cen- mary written record of early Rus’ appeared in Slavonic turies) all contributed to the enhancement of regional rather than Rusian, lapses in the copying and editing of linguistic difference. 913 RUSSIAN AND EAST SLAVIC LANGUAGES By the fourteenth century, two language areas of with ê (and typically ô as well) subsequently raising Rus’ had developed: a northern, Russian one destined and fronting to i (so-called ikavism), e.g. Rusian lês¯b to come under the hegemony of Moscow, and a south- ‘forest’, stol¯b ‘table’  U l’is, st’il. ern, Ruthenian one, at first found largely within the The oldest layers of dialectal differentiation in Lithuanian political orbit, and then from 1569, under Rusian follow a path from southwest to north and the rule of the Polish–Lithuanian Commonwealth. In northeast, reflecting the early direction of Slavic Ruthenian territory, the prominent distinctions development from the original Carpathian center to between Belarusian and Ukrainian were apparent by the northern and eastern periphery. Later innovation is the fifteenth to sixteenth centuries, although their stimulated by the rise of new centers of cultural diffu- beginnings are discernable three to four centuries ear- sion, as noted above. Thus, among the oldest isogloss- lier. From the mid-seventeenth century, parts of the es is one that distinguishes dialects that preserve Ukrainian and Belarusian territories gradually fell Common Slavic *g as a stop (the more ‘peripheral’ under the control of Muscovy, then the Russian North Russian) from those that have lost closure, pro- Empire, and finally the Soviet Union, a process that ducing a fricative like γ or h (the more ‘central’ South extended well into the twentieth century. Russian, Belarusian, Ukrainian). Another isogloss dis- Consideration of the structure and dialectal distri- tinguishes the development of the tense jers (in the bution of Rusian provides a common basis for com- environment before j) as mid-vowels e and o (the more paring and contrasting the development of the East ‘peripheral’ Russian) or vowels i or y (the more ‘cen- Slavic languages with respect to each other. Rusian tral’ Ruthenian), e.g. Rusian m¯bju, R moju, U myju, Br had a rich system of vowels (i,y,u,ê,e,o,ä,a,b,¯b), includ- miju [mýju] ‘I wash’. A third, involving the hardening ing two high, lax vowels, the jers (symbolized as front of palatalized labial and dental consonants before b and back ¯b, respectively), a higher-mid tense front Rusian e, separates the dialects of Russian and vowel (symbolized as ê), and a low tense front vowel Belarusian (presence of palatalization) from those of (symbolized as ä). Over the course of the eleventh and Ukrainian (absence of palatalization). Some archaic twelfth centuries, the jers were either reidentified as Carpathian dialects of Ukrainian have preserved the mid-vowels e and o, or were lost (e.g. Russian pbnb Rusian phonemic distinction between i and y, unlike ‘stump’, s¯bn¯b ‘sleep; dream’  R,Br p’en’, son, U the rest of East Slavic. pen’, son). The jer shift reduced the number of dis- Rusian, like Slavonic, had a rich system of gram- tinctive vowels by a third and nearly doubled the matical categories. Nouns, pronouns, and adjectives inventory of consonants by creating sets of distinctive- were inflected by gender (masculine, feminine, ly paired plain (hard) labial and dental consonants neuter), number (singular, dual, plural), case (nomina- (p,b,m,v,t,d,s,z,n,l,r) and palatalized (soft) counterparts tive, accusative, genitive, locative, dative, instrumen- (p’,b’,m’,v’,t’,d’,s’,z’,n’,l’,r’). Russian has largely pre- tal), animacy, and a special vocative form. Verbs were served this later Rusian phonemic inheritance, where- capable of distinguishing mood, tense, voice, and as Ukrainian, apart from the dispalatalization aspect. Participles and adjectives were inflected by (hardening) of consonants before Rusian i and e, pre- gender, number, and case according to two declen- served most soft dentals, but hardened soft labials. sions, a ‘short’ nominal declension and a ‘long’ adjec- Belarussian preserved soft labials before vowels, hard- tival declension, the latter providing expression for ened r’ to r, and affricated t’ and d’ to c’ (ts’) and ’ definiteness or generic status. All three East Slavic (dz’), the result of assibilation, so-called cekan’e and languages have retained the dynamic, free stress char- dzekan’e, respectively. After the jer shift, most new acteristic of Rusian. sequences of dentals or palatals plus j yielded soft By the fourteenth century, a number of Rusian gram- geminate (doubled) dentals and palatals (r’ and j are matical categories had been lost, no longer productive the exceptions) in Ukrainian and Belarusian, a reflex in any of the East Slavic languages. These include dual not typical of Russian, e.g. Rusian Dit b je ‘life’, U number, the aorist and imperfect tenses, and case in Dyt’t’a, Br Dic’c’o, R Dit’jo. short adjectives and participles. The perfect took over Many Rusian dialects had a seven-vowel system the function of past tense in all of East Slavic. after the jer shift: i,e,a,o,u plus the tense vowels ê and Ukrainian has preserved a separate vocative form. ô. The latter two were distributed differently, north and Ukrainian and Belarusian both have a rarely encoun- south, and were ultimately the most unstable. At the tered pluperfect tense. The morphophonemic velar ~ present time, standard Russian and Belarusian each dental sibilant alternation found in most Rusian dialects have five-vowel systems (i,e,a,o,u), whereas standard was retained on a more limited basis in Ukrainian and Ukrainian has a six-vowel system (i,y,e,a,o,u). The Belarusian, but lost in Russian (e.g. Rusian ruka/rucê higher-mid front y is a characteristically Ukrainian ‘hand, arm’ nom.sg./dat.-loc. sg., U ruka/ruc’i, Br distinctive vowel, derived from merged Rusian i and y, ruka/ruce, but R ruka/ruk’e). Of syntactic interest are 914  RUSSIAN AND EAST SLAVIC LANGUAGES the many East Slavic constructions without overt nom- Karamzin in the late eighteenth to early nineteenth cen- inative subjects (e.g. R mn’e ujt’i ‘I have to go’ [lit. to tury and then Pushkin helped to refine the Russian lan- me to go], U cerkvu bulo zbudovano ‘The church was guage by example, setting the stage for it to become a built’ [lit. churchacc. wasneut.sg. builtneut.sg.], Br m’an’e n’e vehicle capable of artistic and scientific expression on bilo [byló] doma ‘I wasn’t home’ [lit. of me not was- the highest levels of European culture. neut.sg. at home]). Russian is traditionally divided into three major dialect zones: north, south, and central. South Russian is distinguished by the presence of two innovations: Russian (1) the fricative γ instead of original stop g and (2) Spoken by approximately 275 million people, 165 akan’e. Akan’e refers to the pronunciation of million of them natively, Russian is by far the largest unstressed o and a as [a] initially or after a hard con- of the Slavic languages. The vast majority of Russian sonant in the first pretonic syllable (the syllable imme- speakers reside in the Russian Federation or constitute diately preceding the stressed one) and as central lax substantial minorities in several of the republics of the schwa elsewhere (symbolized [ə]): compare molodój former Soviet Union, including Ukraine, Belarus, [məladój] ‘young’ and nagradnój [nəgradnój] ‘per- Latvia, Kyrgyzstan, and Kazakhstan. It is the major taining to reward’. The large transitional zone between language of the Russian Federation and one of the offi- North and South Russian is composed of the central cial languages of Belarus and Kazakhstan. There are dialects (including Moscow), which show their inter- significant Russian-speaking communities in Western mediate status in favoring northern consonantism with Europe, the United States, and Canada. Russian is one g and southern vocalism with akan’e. It is the of the five official languages of the United Nations. Muscovite dialect that ultimately served as the basis of The history of written language in Russian territory the standard language: compare NoR mogú [mogú], is one of continuing tension between Slavonic and ver- SoR moγú [maγú], and standard Russian mogú nacular Russian from the fourteenth to the eighteenth [magú]. In addition to akan’e, standard Russian is also centuries. The church-supported emergence of characterized by ikan’e, the pronunciation of Moscow as the center of the evolving Russian state unstressed o,a,e as [i] in the first pretonic syllable after saw the development of Slavonic into an ever more a soft consonant, e.g. n’os’í [n’is’í] ‘carry’ (imperative ornamental, complicated medium with intentional 2sg.), zan’alá [zən’ilá] ‘occupy’ (past fem.sg.), n’emój archaization from the late fourteenth century onward, [n’imój] ‘dumb’ (adjective nom.sg.masc.) a trend that resonated with the late fourteenth-century reforms in Bulgaria. These changes had the effect of Ukrainian distancing Slavonic from the spoken language by eliminating vernacular forms and introducing artificial Ukrainian is spoken by approximately 49 million peo- elements and grammatical patterns. At the same time, ple, making it the second largest Slavic language. during the fifteenth and sixteenth centuries, the lan- Most of the speakers of Ukrainian reside in Ukraine, guage of the growing Muscovite bureaucracy—the but there are substantial communities in Russia, formulaic, vernacular-based chancery language—dis- Poland, Slovakia, Canada, and the United States. played a normative tendency that elevated the speech Ukrainian is the state language of Ukraine. In addition patterns of Moscow and avoided the dialectal features to Ruthenian (rus’ka mova) and plain talk (prosta of other areas of Russia. mova), it was also called Little Russian from the mid- The union of Left-Bank Ukraine and Muscovite seventeenth century to the early twentieth, a derivative Rus’ in the mid-seventeenth century provided from Little Russia, itself a calque of a Byzantine Ukrainian bookmen with the opportunity to bring to Greek administrative term used to distinguish the Muscovy their better developed grammatical tradition more proximate Ukrainian territory from the more dis- based on Latin and Greek models. They codified and tant Muscovite (called Great Russia). enriched the Slavonic tradition, particularly with for- From the late fourteenth century, Ruthenian Slavonic eign vocabulary from the West, including Latin, prima- in Ukrainian and Belarusian territories underwent rily through Polish and Ukrainian intermediaries. In archaization analogous to that of Slavonic in Russian their turn, Slavonic and the chancery language influ- territory. By this time, Galicia had been absorbed into enced the development of vernacular Russian. It was the Kingdom of Poland and the rest of the Ukrainian only in the mid-eighteenth century, especially through and Belarusian territories into Lithuania. This turn of the grammatical and lexical reforms of Adodurov and events had far-reaching linguistic implications for both especially Lomonosov, that there was a reasoned Ruthenian and Slavonic. In Polish Galicia the vernacu- attempt to control the distribution of foreign, Slavonic, lar-based Ruthenian chancery language was ultimately and Russian elements in the Russian written language. replaced by Latin in accordance with a 1433 decree. In 915 RUSSIAN AND EAST SLAVIC LANGUAGES Lithuania, on the contrary, Ruthenian was elevated in and Slavonic, although not to the degree seen in status to become the official language of government Russian. Harsh restrictions were imposed by the administration. Based on the East Slavic vernacular Russian government on the use of Ukrainian in 1863, spoken around Vilna (Vilnius), the chancery language leading to the ban on its public use in 1876, a prohibi- developed as a kind of koine, which favored specific tion lifted only in 1905. Meanwhile, the Ukrainian lit- Belarusian elements over Ukrainian ones, but avoided erary tradition that had continued in Polish Galicia the more remarkable characteristics of either, such as (ceded to Austria in 1772, 1795) had an impact on the Belarusian cekan’e/dzekan’e and akan’e, and Ukrainian future development of Ukrainian in eastern Ukraine. ikavism, in its various developmental stages. By the The two centers of Ukrainian culture were joined only middle of the sixteenth century, spoken and written after World War II. The Soviet period was marked by Ruthenian was making its influence felt in Slavonic in an intense government campaign of Russification. the presence of greater numbers of vernacular forms. Modern Ukrainian competes with Russian and a The 1569 Union of Lublin resulted in the merger of Russianized Ukrainian hybrid called surzhyk, espe- Poland and Lithuania as the Polish–Lithuanian cially in the capital of Kyiv and in eastern Ukraine. Commonwealth. With Poland as the dominant partner, Since independence (1991), Ukrainian, the sole state Polish and, to a lesser extent, Latin were introduced as language, has gained considerable ground in the offi- the official languages of government administration cial sphere, especially in government, communication, throughout the land. Without administrative function, and education. Ruthenian became subject to local adaptation. Ukrainian is traditionally divided into three major Features from the western and northern Ukrainian dialect zones: northern, southwestern (including the dialects began to dislodge their Belarusian counter- Carpathians, eastern Poland, and eastern Slovakia), parts in the Ukrainian territories to the south, whereas and southeastern. A typical feature of the northern more prominent Belarusian features were developed in dialects is the presence of diphthongs from older the Belarusian territories to the north. Both were sub- ê and ô under stress that correspond to monophthon- ject to heavy Polonization and influence from Slavonic gal vowels in the standard language in either the as well. southwest or southeastern variant. The southeastern After the 1596 Union of Brest brought together dialects are characterized by ikavism, whereas this Catholics and Orthodox Uniates, Polish and Latin process has been only partly realized in many south- began to challenge Slavonic as well in the spiritual west dialects. sphere. A Slavonic revival begun in the western Ukrainian town of Ostrih resulted in the publication of Belarusian a complete translation of the Bible in 1581. The revival spread later to L’viv and Kyiv and resulted in Belarusian has approximately 10 million speakers, new Slavonic dictionaries and grammars. most of whom reside in Belarus. There are small com- Ukrainian had become the administrative language munities of Belarusian speakers in Poland, the United in Ukrainian territory east of the Dnieper after an States, and Canada. One of the official languages of uprising in 1648 resulted in the establishment of a Belarus, Belarusian (also rarely Belarusan) has been Cossack state, the Hetmanate. It retained that role even called Belorussian or Byelorussian, a name influenced after the union with Muscovy in 1654. Mazepa’s by the incorporation of Belarusian territory into the unsuccessful attempt in 1709 to wrest the Hetmanate Russian Empire. The English rendering of Belarusian from Russian control led to the replacement of as ‘White Russian’ produces an unfortunate ambiguity Ruthenian Slavonic by Russian Slavonic. Late in the with the political term referring to the enemies of the century, Russian replaced Ukrainian as the language Bolshevik Revolution in Russia. of administration, both in the territories of the old Following the Union of Brest in 1596, Belarusian Hetmanate, and in the Right-Bank Ukrainian territory (plain talk) increasingly yielded to Latin and finally acquired in the 1793 and 1795 partitions of Poland. to Polish as the authoritative written language, and The emergence of the modern Ukrainian standard in 1697 it was banned from administrative use com- language begins with Kotljarevs’kyj’s translation- pletely. During the eighteenth century, Belarusian was reworking of the Aeneid (1798, 1809, 1842), in which largely limited to the burlesque interludes for religious he showed a preference for southeastern dialect fea- school dramas, themselves written in Slavonic, Latin, tures (Poltava, Kharkiv, Kyiv). He prepared the way or Polish. After Belarusian territory was incorporated for a flowering of literary Ukrainian during the period into Russia during the partitions of Poland in 1772 of Ukrainian Romanticism (1820s–1840s), chiefly and 1795, the Russian government banned the use found in the writings of Kvitka, ŠevJenko, and Kuliš, of Belarusian in schools and in publications, a pro- all of whom sought to realize a synthesis of Ukrainian scription that lasted until 1905. The publication of a 916  RUSSIAN AND EAST SLAVIC LANGUAGES normative grammar by Branislaw TaraškeviJ (1918 based on the central dialects (including Minsk), has and later editions) was a major step toward the estab- strong akan’e and jakan’e as well. lishment of a standard literary language. Belarusian was subjected to heavy Russification during the Soviet References period. Its fate remains precarious in contemporary Avanesaw, R.I. 1964. Narysy pa belaruskaj dyjalektalohii. Belarus: even in the capital of Minsk, one is more like- Minsk: Navuka i tèxnika. ly to hear Russian or a Russianized Belarusian hybrid Avanesov, R.I., and V.G. Orlova. 1965. Russkaja dialektologija. called triasanka than Belarusian itself. Moscow: Nauka. There are three major dialect areas of Belarusian: Comrie, Bernard, and Greville G. Corbett (eds.) 1993. The Slavonic languages. London and New York: Routledge. northeast, southwest, and a broad central one (includ- Issatschenko, Alexander V. 1980. Russian. In Schenker and ing Minsk), transitional to both. The northeastern Stankiewicz. dialects, themselves phonologically and morphologi- Mayo, Peter. 1993. Belorussian. In Comrie and Corbett. cally transitional to Russian, feature dissimilative McMillin, Arnold. 1980. Belorussian. In Schenker and akan’e and jakan’e. Jakan’e refers to the neutralization Stankiewicz. Schenker, Alexander M., and Edward Stankiewicz (eds.) 1980. of o,a,e in unstressed environments after soft conso- The Slavic literary languages: formation and development. nants in which one of the pronounced variants is [a]. New Haven: Yale University Press. The dissimilative principle yields two basic variants of Shevelov, George Y. 1979. A historical phonology of the neutralization in the first pretonic syllable: (1) the low Ukrainian language. Heidelberg: Carl Winter vowel [a] and (2) a nonlow vowel, commonly [ə] in Universitätsverlag. ———. 1980. Ukrainian. In Schenker and Stankiewicz. akan’e and [i] in jakan’e. The first pretonic vowel ———. 1993. Ukrainian. In Comrie and Corbett. in each case dissimilates the stressed (tonic) vowel in Strumins’kyj, Bohdan. 1984. The language question in the height: if the tonic vowel is high, the first pretonic Ukrainian lands before the nineteenth century. Aspects of variant is low, and vice versa; thus vodí [vadý] ‘water’ the Slavic language question, Vol. 2: East Slavic, ed. by (gen.sg.), but vodá [vədá] ‘water’ (nom.sg.), n’osú Riccardo Picchio and Harvey Goldblatt. New Haven: Yale Concilium on International and Area Studies. [n’asú] ‘carry’ (nonpast, 1st sg.), but n’os’éš [n’is’éš] Timberlake, Alan. 1993. Russian. In Comrie and Corbett. ‘carry’ (nonpast, 2nd sg.). The southwestern dialects Vlasto, A.P. 1986. A linguistic history of Russia to the end of are phonologically and morphologically transitional to the eighteenth century. Oxford: Clarendon Press. Ukrainian. In all but the southernmost regions, which Wexler, Paul. 1977. A historical phonology of the Belorussian are devoid of akan’e and jakan’e, strong (nondissim- language. Heidelberg: Carl Winter Universitätsverlag. Worth, Dean S. 1998. Language. Modern Russian culture, ed. by ilative) akan’e and jakan’e are predominant: the first Nicholas Rzhevsky. Cambridge: Cambridge University Press. pretonic neutralized variant of o,a,e is always realized Cylko, F.T. 1966. Narysy z dialektolohiji ukrajins’koji movy, as [a], regardless of the nature of the tonic vowel; thus 2nd edition. Kyiv: Radjans’ka Škola. [vadý], [vadá], [n’asú], [n’as’éš]. Standard Belarusian, MICHAEL S. FLIER 917  S Sandhi Sandhi rules are phonological alternations that are articulation of the consonant that follows it. The latter triggered at junctures of words or morphemes. The consonant can therefore be said to act as the trigger for name sandhi comes from Sanskrit and means ‘junc- this phonological change. ture’. It does not designate one particular phonological One of the internal sandhi rules of Sanskrit is a process. Instead, it is a nonspecific cover term for any phonological process known as voice assimilation. kind of sound mutation that occurs at the edges of The phonological process in question leads to a words and morphemes, and that is triggered in envi- change of voiceless consonants, such as /p/, /t/, and ronments created by morphological or syntactic con- /d/, turning them into their voiced counterparts /b/, /d/, catenation operations. Sandhi phenomena are very and /g/. The effect of this rule is observable in the common across the languages of the world. Some of compound sadaha ‘good day’ derived from sat ‘good’ the best-known examples are tone sandhi, which is and aha ‘day’. Compound formation leads to a config- found in Chinese for example, Welsh and Irish conso- uration in which the two segments––the devoiced nant mutations, and a set of Sanskrit sandhi rules. word final /t/ and the voiced, word initial /a/––become Two types of sandhi can be distinguished: internal adjacent, and thus creates the conditions for the appli- and external sandhi. Internal sandhi rules are triggered cation of voice assimilation. word internally, between two morphemes. External Interestingly, the same rule that operates word inter- sandhi rules operate at word edges, when two words nally in Sanskrit, between two portions of a compound, become adjacent as a result of some syntactic process also applies as external sandhi between two words of a in the language. phrase. For instance, the phrase samyag uktam ‘cor- An example of internal sandhi can be found in rectly spoken’ consists of the word samyak ‘correctly’ English words such as symbol and symbiosis. Both and uktam ‘spoken’. Again, voice assimilation of the words are derived by addition of the prefix syn to the word-final consonant, in this case /k./, can be observed stem. In its basic form, this prefix can be observed in under the influence of the following voiced segment. words like synthesis, synopsis, syntax, synchronicity, In English, external sandhi may occur in phrases synonym, etc. When combined with a stem beginning such as miss you, kiss you, caught you, etc, where the with /b/, as in the words symbol and symbiosis, the word final consonant of the verb changes to [sh] or [ch] last consonant of the prefix changes to /m/. This inter- to give mi[sh] you, ki[sh] you, and cough[ch] you. This nal sandhi rule is an instance of the phonological phonological process is known as palatalization. process known as place assimilation, in which the Another example of external sandhi is known by the place of articulation of some consonant becomes name of intrusive-R. It exists in some, but not all, dependent on the place of articulation of an adjacent dialects of English. It is a feature of the nonrhotic vari- consonant. In the case at hand, the place of articulation ants (i.e. those in which the word final /r/ as in bar, star, of the consonant /n/ becomes identical to the place of Zanzibar, etc. is not pronounced). Intrusive-R is an 919 SANDHI insertion of [r] between two words of which the first determined by the phrasal groupings of the words word ends in a vowel, while the second begins with one, involved, as shown by the data in (1a) and (1b). Since e.g. vanilla-r-ice-cream, the idea-r-of it, I saw-r-it, etc. the phrasal organization of the sentence is determined An external sandhi rule that is present in some by syntax, the conditions under which external sandhi dialects of Italian is known as raddoppiamento sintatti- rules apply are to some extent influenced by this com- co. It is triggered only when the two words between ponent of the grammar. However, this influence only which the sandhi operates are contained in the same goes some way. The constituents into which words are local domain. The phonological process involved is the syntactically organized such as syntactic phrases are lengthening (or doubling) of the word-initial consonant. not isomorphic to the groupings of words into phono- It is triggered when the word is preceded by another logical constituents such as phonological phrases, word ending with a stressed vowel, eg. città ppulita although there are certain connections between the ‘clean city’ from città ‘city’ and pulita ‘clean’, or metà two. The relationship of syntax and phonology (the llibro ‘half book’ from metà ‘half’ and libro ‘book’. syntax–phonology interface) and the field of prosodic However, for consonant doubling to take place, the phonology in general is a large and fertile field of words have to be contained within the same phonolog- research (Nespor and Vogel 1986; Selkirk 1984). ical phrase. When this requirement is met, as in (1a), and the two words are found within the relevant domain References (i.e. the phonological phrase here marked by the sub- script Φ), raddoppiamento sintattico takes place. On the Andersen, Henning (ed.) 1986. Sandhi phenomena in the lan- guages of Europe. Berlin, New York and Amsterdam: other hand, if the two words are separated by a phono- Mouton de Gruyter. logical boundary (marked as //), raddoppiamento sintat- Chen, Matthew Y. 2000. Tone Sandhi: patterns across Chinese tico fails. This is illustrated by (1b), where the dialects. Cambridge: Cambridge University Press. word-initial consonant in molto ‘old’ fails to undergo Inkelas, Sharon, and Draga Zec. The phonology syntax connec- lengthening. tion. Chicago and London: University of Chicago Press. Nespor, Marina, and Irene Vogel. 1982. ’Prosodic domains of (1a) ( Avra lletò )Φ ( il libro )Φ. external Sandhi rules’ The structure of phonological repre- sentations, Part I, ed. by Harry van der Hulst and Norval ‘He will have read the book.’ Smith, 225–55. Dordrecht: Foris Publications. (1b) (Visita la città )Φ // ( molto vechie )Φ. ———. 1986. Prosodic phonology. Dordrecht: Foris ‘He visits very old cities.’ Publications. Selkirk, Elisabeth O. 1984. Phonology and syntax: the relation An important feature of external sandhi is that an between sound and structure. Cambridge, MA: MIT Press. application of the phonological process in question is AMELA CAMDZIC Sanskrit Sanskrit belongs to the Indo-Iranian branch of Indo- ond millennium BCE. Classical Sanskrit is a highly European, whose subgroupings are Iranian and Indo- standardized language, as such used without interrup- Aryan. Indo-Aryan includes Vedic Sanskrit, Classical tion from its first written attestations, and virtually Sanskrit, and the Prakrits, or ‘vulgar’ varieties, which unchanged until the present; however, in the early have as their offshoots the modern Indo-Aryan lan- phases dialectal variation is visible, even between guages, such as Hindi-Urdu, Bengali, Gujerati, Vedic and Classical Sanskrit. The former is not the Marathi, Panjabi, Sinhalese, and Romani. direct ancestor of Sanskrit: its peculiarities include Sanskrit is one of the most anciently attested Indo- both archaisms and innovations with respect to the European languages and has played a crucial role in classical literary language, pointing toward some kind the reconstruction of Proto-Indo-European. Apart of diatopic variation. Furthermore, the Prakrits of the from some isolated forms (see below), the most Middle Indian period (300 BCE–200 CE) do not ancient literary sources are the poems collected in the derive directly from Classical Sanskrit, but display a books of Veda (hence the name ‘Vedic Sanskrit’), variety of different features, some directly connectable which record an oral tradition, going back to the sec- to Vedic. Among the Prakrits, the most important as a 920  SANSKRIT literary language is Pa¯li, the language of the Buddhist short diphthongs underwent monophthongization: Canon. The coexistence of Sanskrit with Prakrits is /ai/ > /e/, /au/ > /o/. Hence, already in Vedic there is ancient, and is mostly a matter of social variation: in again a distinction, at least for long vowels, but the classical drama, socially inferior characters (including /e/’s and /o/’s of Sanskrit do not reflect the same women) usually speak Prakrit. vowels of Proto-Indo-European. The Aryan populations reached India toward the Sanskrit belongs to the so-called satə m group within end of the second millennium BCE, and established Indo-European, i.e. it is one of those languages, along themselves in the upper Indus valley. Before that time, with the remaining Indo-Iranian languages, Slavic, and the Aryans had already built a linguistic group distinct Armenian, in which velars changed into spirants or from the Iranians, as shown by the traces left during affricates, and labiovelars changed into velars. their journey through the Middle East. In the first half A major role in Sanskrit phonology is played by of the second millennium BCE, an Aryan aristocracy sandhi, i.e. a set of phonological rules determined by dominated the Hurrian kingdom of Mitanni, located the clustering of vowels or consonants. on the upper course of Tigris and Euphrates. Names of Indic deities are mentioned in a treaty with a Hittite Morphology king, and some Indic words are found as loanwords in Sanskrit morphology is extremely rich. Inflectional a Hurrian text of the period, among which are some classes of nouns are based on the Indo-European dis- numerals, which attest a pre-Sanskrit variety (e.g. tinction between thematic and athematic declension, aika, ‘one’, in which the monophthongization /ai/ > /e/ with subclasses of the latter for stems in nasal, liquid, had not yet taken place, cf. Sanskrit eka). /i/, /u/, and long vowel. Nouns have eight cases (nom- inative, vocative, accusative, instrumental, dative, ablative, genitive, and locative) and three numbers Grammatical Sketch (singular, dual, and plural); gender distinctions include Phonology and Script masculine, feminine, and neuter. Adjectives, which Sanskrit is written in devana¯gari script, still used for also inflect for degrees of comparison (comparative the modern Aryan varieties, which derives from an and superlative), follow the inflection of nouns for earlier script, called bra¯hmÃ. The devana¯gari is a pecu- number and case, while pronouns follow special pat- liar writing system, halfway between a syllabary and terns of inflection. an alphabet, consisting of 13 signs for vowels and 35 Verbs are traditionally divided into ten classes for consonants, plus a number of diacritics and sever- (eight in Vedic), of which the first, fourth, sixth, and al compound signs used for representing consonant tenth follow the thematic conjugation, while the clusters. Consonant signs not followed by any diacrit- remaining follow the athematic conjugation. In the ic are understood as containing the consonant and the thematic conjugation, the thematic vowel -a- (< PIE * vowel /a/. -o-) is inserted between the stem and the ending; note The devana¯gari script is very well suited for the rep- that, contrary to the other Indo-European languages, resentation of Sanskrit phonemes, each sign correspon- which have-o¯ as ending of the first person singular of ding to a single phoneme (except for diphthong signs) the present indicative, Sanskrit has reintroduced the and each phoneme being represented by a sign. Thus, ending -mi of the athematic declension, so that we find Table 1 can also serve as an inventory of Sanskrit a cumulation of both endings: bhara¯mi, ‘I bring’, first phonemes. class, thematic, built with the present stem bhar-, end- An important feature of Sanskrit phonology, espe- ing-a¯ plus ending -mi (cp. with Latin fer-o– , same cially in comparison with the other ancient Indo- meaning). Verbs distinguish three persons in the sin- European languages, lies in the phonologization of gular, dual, and plural. both voiced and voiceless aspirates, the existence of a Tenses include the present, the imperfect, the aorist, series of retroflex (also called ‘cerebral’) stops, bor- the perfect, and the future; in addition, Vedic also had rowed from the non-Indo-European substrate lan- a compound pluperfect. In origin, the verb system was guages of India, and the retention of vocalic liquids, based on an aspectual opposition between present and partly with length opposition. (Length opposition is perfect. The present denoted actions or processes, only relevant for /r# / ~ /r# /; besides, /l#/ is a very rare while the perfect denoted states. The Indo-European phoneme.) From a diachronic point of view, the most opposition between perfective/imperfective is contin- significant innovations are found in the vowel system. ued only formally, in the opposition between aorist At an early stage in Proto-Indo-Iranian, the three vow- and present, but not semantically (see below). Verbs els of Proto-Indo-European, /a/, /o/, and /e/, merged in have a special stem for the present, the perfect, and /a/ (this change involved both short and long vowels). the aorist; the differences among stems are usually Later, after the separation of Iranian and Indo-Aryan, based on root gradation, which opposes a reduced 921 SANSKRIT Devana¯gar¯ alphabet Primary vowels Short Long Diphthongs Initial Diacritic Initial Diacritic Initial Diacritic Unrounded low central  a  pa Á  pÁ Unrounded high front  i  pi  à  pà Rounded high back  u  pu  Æ  pÆ Syllabic variant ~ r9  pr9  r9¯  pr9¯ Secondary vowels Unrounded front  e  pe  ai  pai Rounded back = o  po > au  pau Other symbols  an· anusvÁ ra - nasalises vowel W am¢ anunÁsika/candrabindu - nasalises vowel  ah¢ visarga - adds voiceless breath after vowel R p virÁma - mutes vowel Consonants Occlusives Voiceless plosives Voiced plosives Nasals unaspirated aspirated unaspirated aspirated Velar  ka % kha ga & gha G n˙a Palatal < ca E cha ; ja D jha A ña Retroflex ? ²a @ t ¢ha 4 °a 6 °ha B n¢a Dental 8 ta 5 tha K da C dha 3 na Labial  pa  pha ! ba F bha  ma Sonorants and fricatives Palatal Retroflex Dental Labial Sonorants ' ya 7 ra # la va Sibilants ( s´ a  s¢a H sa Other letters Variant letters used in Mumbai  ha G l ¢a n " jha c n¢a A selection of conjunct consonants | ks¢a S jña _ tta Z tra W' pya O tka ? B t k¢ a e hya A ttva Numerals ) * + / , - . 0 1 2 *) 0 1 2 3 4 5 6 7 8 9 10 (Source: http://www.omniglot.com/writing/devanagari.htm) 922  SANSKRIT grade (e.g. kr#-, ‘make’) to full grade, also called gun a adjectives, genitives, and restrictive relative clauses (kar-), to lengthened grade, also called vr #ddhi (ka¯r-). precede their head, and adpositions mostly follow the (Note that there is no one-to-one correspondence noun. OV features are all present in Vedic, too, but between vowel grade and a specific tense.) The perfect Classical Sanskrit is more consistent with respect is usually made on a reduplicated stem (e.g. ca-kar-a, to them. ‘I have made’). Subordination includes relative clauses and a vari- Already in Vedic Sanskrit, the aspectual value of the ety of adverbial clauses, with conjunctions derived aorist was given up: the aorist is used in reference to from the stem of the relative pronoun. Besides, par- recent past, and never without past reference in contrast ticiples and gerunds also contribute to the building of to the present (as was the Ancient Greek aorist). The hypotaxis. Participles (especially the past passive par- perfect added to its original stative meaning a more ticiple in -ta) and gerunds are often used without the recent resultative meaning, which also had its temporal verb ‘be’, as if they were finite verb forms. implication of remote past. Furthermore, the so-called Sanskrit displays a peculiar preference for passive, imperfect, derived from the present tense, was used in rather than active construction. The would-be-subject of narration, where it did not display any particular ten- the active is thus expressed through an agent phrase in dency toward expressing the imperfective aspect. the instrumental: tenedam uktam, ‘he said this’, lit.: In Classical Sanskrit, moods include the indicative, ‘this (was) said by him’ (tena, ‘by him’ + idam, ‘this’, the optative, and the imperative for the present tense; uktam, past passive participle of vac-, ‘say’). the aorist has an indicative and a precative, and the Impersonal passives with instrumental agent phrases are future has a conditional. Vedic had a more developed also found in intransitive verbs; the past participle, how- mood system, which included an indicative, a sub- ever, occurs more frequently with an agent–subject junctive, an optative, and an injunctive, all made on all phrase in the nominative, in which case it has active three stems (present, aorist, and perfect). From a typo- meaning (so sah gra¯mam gatah , ‘he went to town’, sah, logical point of view, the most interesting feature of ‘he’, gra¯mam, ‘to-town’, gatah , past participle of gam-, this mood system is constituted by the injunctive. In a ‘go’, rather than tena gra¯mam gatam, lit.: ‘by him (tena, way, it stands outside mood oppositions, and is used instr.) (it was) gone to town’). for general statements. Voice includes an active (parasmaipada) and a Sanskrit and Indo-European middle (a¯ tmanepada); not all verbs have both voices, but for those that do, the middle mostly has reflexive The importance of Sanskrit for Indo-European linguis- value. Furthermore, the middle can have passive tics cannot be overstated: it was only after European meaning, except for the present stem, which builds a scholars learned about it that they realized that their special passive by adding the suffix -ya-. own European languages presented striking similari- The Sanskrit verb has a variety of nominal forms: ties among each other, and set out to reconstruct a active and middle present, future and perfect partici- proto-language. Actually, for some time Sanskrit was ples, a past passive participle, from which a past active even thought to be the proto-language: but also later, participle is also made, a future passive participle, also after it lost this status, Sanskrit was regarded as the called gerundive, an indeclinable participle, also most conservative Indo-European language, especially called gerund, and an infinitive. with respect to its numerous inflectional categories, A peculiarity of Sanskrit is the large extent to which fully developed in rich and complex paradigms. composition is used for building nouns. Following the The discovery of Anatolian, at the beginning of the Indian classification, types of compound include dvand- twentieth century, gave a serious stroke to the primacy va (copulative compounds, e.g. sim  hagaja¯h, ‘lions and of Sanskrit. Some scholars started to reconstruct a elephants’, from sim  ha-, ‘lion’, plus gaja-, ‘elephant’), Proto-Indo-European with a smaller case system, with tatpurus a (determinative: tat-purus a ‘that - man’= ‘the less verbal moods, and in which some basic Greek– man of that one’, ‘his man’), and bahuvrà hi (possessive, Sanskrit isoglosses (augmented past tenses, aorist, or exocentric: bahu-vrà hi ‘much - rice’ = ‘having much dual number, etc.) played a less relevant role. Another rice’). challenge to the possible archaic character of Sanskrit has come from the so-called ‘ejective model’, accord- Syntax ing to which the stops of Proto-Indo-European include Sanskrit has free word order, but in unmarked state- voiceless/glottalized/aspirate, rather than voiceless/ ments displays a tendency toward SOV order; unac- voiced/voiced aspirate of the traditional reconstruction cented pronouns and other clitics are consistently (see Hopper 1973). A full discussion of new recon- placed after the first accented word in the sentence in struction trends can be found in Gamkrelidze and which they occur (Wackernagel’s position), modifying Ivanov (1995). 923 SANSKRIT References Gonda, Jan. 1951. Remarks on the Sanskrit passive. Leiden: Brill. Hoffman, Karl. 1967. Der Iniunktiv im Veda. Heidelberg: Winter. Allen, Sidney. 1962. Sandhi. The Hague: Mouton. Hopper, Paul. 1973. Glottalized and murmured occlusives in Bloch, J. 1965. Indo-Aryan from the Vedas to modern times. Indo-European. Glossa 7(2). 141–66. Paris: Adrien-Maisonneuve. McDonell, A.A. 1916. A Vedic grammar for students. Oxford: Burrow, T. 1965. The Sanskrit language. London: Faber & Faber. Oxford University Press. Cardona, George. 1990. Sanskrit. The world major languages, McDonell, Arthur A. 1926. A Sanskrit grammar for students. ed. by B. Comrie. Oxford: Oxford University Press. Oxford: Oxford University Press. Gamkrelidze, Th. V., and V. V. Ivanov. 1995. Indo-European Speijer, J. S. 1980. Sanskrit syntax. Delhi: Motilal Banarsidass and the Indo-Europeans. Berlin, New York: Mouton de (reprinted from 1886 edition). Guyter. SILVIA LURAGHI Sapir, Edward As a senior in the undergraduate program in German portions as some admirers attributed to him extensive language and literature at Columbia University in knowledge, and even fluent command, of languages 1903–1904, Edward Sapir enrolled in a graduate sem- which he had investigated for just a few days in con- inar in American languages taught by the distin- sultation with a single speaker. guished anthropologist Franz Boas. Sapir had already After receiving the doctorate, he took up the post as studied Latin, Greek, French, Spanish, and his major chief of the Division of Anthropology in the subject German, and now he was exposed to the diver- Geological Survey of the Canadian National Museum. sity of the native languages of North America. The This was both a hindrance and an asset to his career as effect was not immediate, for Sapir continued on to the a linguist. On the one hand, the position isolated him master of arts degree, still concentrating on Germanic from his colleagues in the United States and at times philology and literature. But his 1905 master’s thesis, directed his energies toward administration. But the on traditional German linguistics, did include data advantages were many, not least of which was control from several American languages to demonstrate the of a research staff and a budget, always too small, with grammatical complexity of all languages, those with- the authority to determine and direct fieldwork. It was out writing systems as well as those with long estab- Sapir’s intent to create a systematic survey of the lan- lished literary traditions. Soon Sapir was engaged in guages and cultures of all of Canada. Since he was fieldwork, traveling to the West Coast of the United nearly always the only staff member with formal train- States to study Wishram in Washington in the summer ing in linguistics, it was he who did most of the lin- of 1905, and Takelma in Oregon in 1906. Although his guistic work, concentrating his attention on the Germanic studies had been largely philological and languages of Western Canada. He made many field historical, his grammar of Takelma was entirely a syn- trips in the years 1910–1915, and thereafter worked chronic analysis, describing the language without frequently with speakers at his office in Ottawa. recourse to historical information, a characteristic of The use of trained native assistants for the record- most of the linguistic work done under the direction of ing of text was a common practice with Boas and his his mentor, Franz Boas. students, but Sapir was exceptional in recognizing that Sapir published his master’s thesis, in the journal native speakers’ insights were relevant to the develop- Modern Philology, but virtually all of his other early ment of linguistic theory, especially in the area of publications were on native American languages. His phonology. In the early years of Sapir’s career, lin- doctorate was awarded in 1909, with the Takelma guists were struggling with the new concept of the grammar serving as his dissertation, and by the end of phoneme, which Sapir later came to define as that year he had published nine articles, reviews, and ‘a functionally significant unit in the rigidly defined pat- monographs on the Kwakiutl, Upper Chinook, Yana, tern or configuration of sounds peculiar to a language . Yuchi, and Wishram languages, as well as Takelma. . ., as distinct from . . .the ‘sound’ or ‘phonetic element’ His reputation as a specialist in native American lan- as such (an objectively definable entity in the articulat- guages was established early, and continued to grow ed and perceived totality of speech)’. (Sapir, ‘The psy- throughout his lifetime, reaching nearly mythic pro- chological reality of phonemes’ 1985[1933]: 46) 924  SAPIR, EDWARD One of Sapir’s most important and long-lasting con- languages into types based on the structural properties tributions to the development of the phoneme concept of their units (as opposed to genetic classification was his view that the phonemic structure of a language based on shared history). Sapir’s classification system was part of the unconscious knowledge that speakers of drew heavily on his background in Germanic linguis- a language possessed, part of their ‘phonological intu- tics and can be viewed as the twentieth-century apex of itions.’ Drawing on his years of work with native inter- nineteenth-century European typology. The system preters of American Indian languages, he argued that provides categories for the morphology (the processes ordinary speakers hear phonemes, not phonetic details, of word formation) of languages along two dimen- and he offered evidence in the way that interpreters sions: the extent to which words employ affixes and the would transcribe their language, always recording dis- ease with which the root and affixes of a word can be tinctive contrasts in sounds while not noting phonetic segmented. For example, an isolating language would differences that were irrelevant to their phonological be one in which most words consist solely of roots, system. The paper in which these concepts were most with little affixing. Sapir gives Chinese as an example. clearly expressed was written in French and published A language with many affixes intertwined and difficult in Paris, France in 1933. At that time, it was little noted to segment would be labeled fusional; Sapir cites by American linguists, but when it appeared in an Salinan, a language of southwest California. This part English version in 1949, ten years after Sapir’s death, it of Sapir’s typology, based as it was on the well–known sparked something of a revival of interest in Sapir’s categories of earlier linguists, has been modified some- work. His notions of the psychological reality of the what in terminology, but it remains useful in the classi- phoneme, of native speaker intuition, and of uncon- fication of morphological types. Sapir, however, also scious linguistic knowledge were all to find their way attempted to overlay this system with another, a con- into the development of generative phonology in the ceptual classification that distinguished concrete con- late 1950s and 1960s, as did his description of phono- cepts and relational concepts, with further bifurcation logical systems as ‘sound patterns’, explicated in his into simple and complex. This aspect of his typological widely read 1925 article ‘Sound patterns in language’ classification has not been pursued, but Sapir himself published in the first issue of Language, journal of the recognized the preliminary nature of his system and the Linguistic Society of America. need to explore a greater number of more diverse lan- Also growing out of Sapir’s work in Canada was his guages than he was able to consider at the time. This is genetic classification system for American Indian lan- exactly what modern typologists have done. guages. Drawing on his background in Germanic his- Sapir’s interest in conceptual categories was mani- torical linguistics, Sapir began to apply the techniques fested in another area of research in which he engaged of the Indo-European comparative method to the orig- in the later 1920s and early 1930s––the issue of an inal languages of America. Beginning with the 1891 international auxiliary language. Beginning in 1925, at classification system developed by John Wesley about the time he accepted a position at the University Powell and the Bureau of American Ethnology, Sapir of Chicago, Sapir became involved with the took Powell’s 55 ‘stocks’ (or language families) and International Auxiliary Language Association, Inc. eventually reduced them to six groups: Eskimo-Aleut, (IALA), headquartered in New York City. Working in Algonquian-Ritwan, Na-Dene, Penutian, Hokan- cooperation with the co-founder and honorary secretary Siouan, and Aztec-Tanoan. There was a good deal of of the association, Alice Vanderbilt Morris, Sapir wrote controversy at the time over this plan, in part because a number of scholarly and popular pieces supporting Sapir provided little evidence to support it. Sapir him- research on and development of an international auxil- self considered the scheme to be probable but not at all iary language. He was not alone among linguists in this proven, given that ‘scientific comparative work on interest, which was shared most enthusiastically by these difficult languages is still in its infancy’ (‘Central the Danish linguist Otto Jespersen, a number of other and North American languages’, 1985[1929]: 169), European linguists, as well as by many of Sapir’s but his system remains the fundamental reference American colleagues. Sapir served briefly as IALA’s point for those working on the genetic classification of linguistic research director and under IALA sponsor- American Indian languages. ship produced three studies of conceptual structure, Sapir’s most widely read work is the small book which comprise Sapir’s total output on semantics. He Language, published in 1921 while he was still in made a number of programmatic statements but did no Canada. Aimed at both students of linguistics and the further work in this area and has had no lasting influ- public, the volume predated his work on phonology. Its ence. Linguists’ interest in an international auxiliary chapters focus on words and grammar, with a particu- language dissipated during the 1930s, a period of pro- lar emphasis on what is today called linguistic typolo- fessionalization within the discipline and of increasing gy, the classification of languages and of features of American isolationism on the national scene. 925 SAPIR, EDWARD In 1931, Sapir moved from the University of Sterling Professor of Anthropology and Linguistics in Chicago to Yale University. He never again conducted 1931. He was President, Linguistic Society of America fieldwork, and he largely withdrew from linguistic (LSA) in 1933; he also taught Introduction to Linguistic study to concentrate on culture and personality. But he Science and Field Methods at the 1937 summer LSA did supervise dissertations and postdoctoral work on Linguistic Institute at University of Michigan. Sapir died American languages by a number of Yale students who in New Haven, Connecticut on February 4, 1939. went on to distinguished careers in linguistics, includ- ing Mary R. Haas, Stanley Newman, Morris Swadesh, References and Charles F. Voegelin. At Yale, Sapir also came to know Benjamin Lee Whorf, a largely self-educated Anderson, Stephen R. 1985. Phonology in the twentieth centu- man interested primarily in philosophical issues and ry: theories of rules and theories of representations. Chicago and London: University of Chicago Press. the study of meaning, who associated himself with Campbell, Lyle. 1997. American Indian languages: the histori- Sapir and his circle of graduate students. Whorf’s lin- cal linguistics of native America. New York and Oxford: guistic relativity hypothesis (that grammatical struc- Oxford University Press. ture of a language influences its users’ perceptions of Cowan, William, Michael K. Foster, and Konrad Koerner (eds.) the world) became well known beyond the field of lin- 1986. New perspectives in language, culture, and personali- ty: proceedings of the Edward Sapir centenary conference, guistics. It is sometimes referred to, incorrectly, as ‘the Ottawa, 1–3 October 1984. Amsterdam and Philadelphia: Sapir–Whorf hypothesis’, but it was developed fully John Benjamins. only after Sapir’s death in 1939. Most contemporary Darnell, Regna. 1990. Edward Sapir: linguist, anthropologist, linguists found it too speculative, and the hypothesis humanist. Berkeley and Los Angeles: University of never became part of mainstream linguistics. California Press. Falk, Julia S. 1995. Words without grammar: linguists and the Edward Sapir was a key figure of American lin- international auxiliary language movement in the United guistics in the first half of the twentieth century, States. Language & Communication 15(3). 241–59. equaled in stature for that period only by his contem- Hymes, Dell. 1985. Epilogue. Selected writings of Edward porary Leonard Bloomfield. Sapir’s students protected Sapir in language, culture and personality, ed. by David G. and promoted his work and ideas and he remains ‘a Mandelbaum, paperback edition, 598–600. Berkeley and Los Angeles: University of California Press. symbol of scope and insight in the study of language’ Koerner, Konrad (ed.) 1984. Edward Sapir: appraisals of his life (Dell Hymes 1985: 598). and work. Amsterdam and Philadelphia: John Benjamins. Lee, Penny.1996. The Whorf theory complex: a critical recon- struction. Amsterdam and Philadelphia: John Benjamins. Biography Sapir, Edward. 1989. The collected works of Edward Sapir, 16 vols. Berlin and New York: Mouton de Gruyter. Edward sapir was born on January 26, 1884 in Sapir, Edward. 1921. Language: an introduction to the study of Lauenberg, Germany, now part of Poland, the son of speech. New York: Harcourt, Brace. Lithuanian parents who soon moved to England and Whaley, Lindsay J. 1997. Introduction to typology: the unity and then on to the United States. He spent his childhood in diversity of language. Thousand Oaks, CA: Sage Publications. Richmond, Virginia up to the age of ten, and then in New All the following works are reprinted in Selected writings of Edward Sapir in language, culture and personality, ed. by York City. He began the study of German, French, David G. Mandelbaum. Berkeley and Los Angeles: Spanish, Latin, and Greek at Peter Stuyvesant High University of California Press, 1949. Page references in the School, matriculating with a German major at Columbia text are to the paperback edition of 1985. University in 1901. He received his B.A. in 1904; M.A. Sapir, Edward. 1925. Sound patterns in language. Language 1. (1905) in Germanic philology and literature; and Ph.D. 37–51. ———. 1929. Central and North American languages. (1909) in anthropology, specializing in linguistics with a Encyclopaedia Britannica, 14th edition, Vol. 5, 138–41. dissertation on the Takelma language of Oregon super- London and New York: Encyclopaedia Britannica Co. vised by Franz Boas. He was Research Associate, ———. 1929. The status of linguistics as a science. Language Department of Anthropology, University of California, 5. 207–14. Berkeley, 1907–1908, and Fellow and then instructor at ———. 1931. Function of an international auxiliary language. Psyche 11. 4–15. the University of Pennsylvania, 1908–1910. He was ———. 1933. Language. Encyclopaedia of the social sciences, Chief, Division of Anthropology, Geological Survey of Vol. 9, 155–69. New York: Macmillan. the Canadian National Museum, Ottawa, 1910–1925. ———. 1933. La Réalité Psychologique des Phonèmes. Journal He went to the University of Chicago in 1925, and was de Psychologie Normale et Pathologique 30. 247–65. promoted to Professor of Anthropology and General ———. 1944. Grading: a study in semantics. Philosophy of Science 11. 93–116. Linguistics in 1927. Concurrently, he was also Director JULIA S. FALK of Linguistic Research, International Auxiliary Language Association, New York City, from October See also Boas, Franz; Sapir–Whorf Hypothesis; 1930 to July 1, 1931. He went to Yale University as Whorf, Benjamin Lee 926  SAPIR–WHORF HYPOTHESIS Sapir–Whorf Hypothesis The idea that one’s native language ‘colors’ his the designation the ‘Whorfian hypothesis’. Some, Weltanschauung has been around at least since the pointing to Sapir’s preeminent stature as a linguist, days of Johann Gottfried Herder (1744–1803) and prefer the appellation the ‘Sapir-Whorf hypothesis’. Wilhelm von Humboldt (1762–1835). Charles Sanders When viewed in terms of output, however, one could Peirce (1839–1914) postulates that man’s symbolic counter that a more appropriate label would be the universe could only make sense via language. ‘Whorf–Sapir hypothesis’. Ferdinand de Saussure (1857–1913) (Cours de lin- A rather interesting development in this debate guistique générale [1916:155]) states that: ‘No ideas over giving credit where credit is due has been the are established in advance, and nothing is distinct, attempt to disassociate Sapir from the hypothesis before the introduction of linguistic structure.’ entirely. Desirous of preventing the image of the However, the principle of linguistic relativity has great maestro Sapir from being tarnished by the taint largely become associated with Benjamin Lee Whorf of controversy, some, most notably Alfred L. (1897–1941), who along with Edward Sapir Kroeber, have claimed that Sapir’s views were not (1884–1939), his linguistic mentor at Yale University, really that (pro)Whorfian. This viewpoint is not borne used linguistics to advocate their position that lan- out by an examination of Sapir’s own writings. For guage influences the way in which a speech commu- example, as one can see in a (1929) passage (from nity perceives and conceives of its reality. the journal Language, p. 209), he equated language Part of the groundwork for the Sapir–Whorf and thinking in terms of the speech community’s hypothesis was laid by Whorf’s work as a fire insur- overall culture: ance investigator. During his career, he had the oppor- Language is a guide to ‘social reality’…it powerfully tunity to analyze many reports as to why fires broke conditions all our thinking about social problems and out. He found that workers would use extreme caution processes. Human beings do not live in the objective when around ‘full’ drums of gasoline. Just as one world alone, nor alone in the world of social activity as might expect, workers were careful not to smoke ordinarily understood, but are very much at the mercy of around ‘full’ drums. Yet, these same workers, when the particular language which has become the medium around ‘empty’ drums of gasoline, would sometimes of expression for their society. toss lit cigarettes nearby. This caused a violent explo- There are actually two different versions of the sion because an empty drum still contained volatile hypothesis. This is understandable when one considers gasoline vapor. Thus, an ‘empty’ drum was really that Whorf did all of his writing from 1925 to 1941, much more of a threat than a ‘full’ one. Using these and his ideas were continuously evolving. The strong data, Whorf concluded that the meanings of certain version of the hypothesis, which is called linguistic words had an effect on a person’s behavior. determinism, holds that language determines thinking. It was the extensive research of both Sapir and This position is most difficult to defend primarily Whorf into the grammatical systems of American because translation between one language and another Indian languages, however, that proved to have the is possible, and ‘thinking’ can take place without lan- greatest impact on the hypothesis. By predicating their guage; e.g. an artist or sculptor can and often does insights into the interrelationships of language and think with his fingers. culture on what they had learned from the structures of Mirroring Sapir’s thoughts, Whorf notes in his these languages, the basic idea of language shaping (1940) ‘Science and linguistics’: the perceptions of its speakers and providing for them a vehicle so that their experiences and emotions could We dissect nature along lines laid down by our native be placed into significant cognitive categories was languages…We cut nature up, organize it into concepts, given its scientific underpinnings. Generally, Sapir is and ascribe significances as we do, largely because we credited as giving the problem of establishing the link are parties to an agreement to organize it in this way –– an agreement that holds throughout our speech com- between language and culture its initial formulation, munity and is codified in the patterns of our language. while Whorf is honored as the one who took this idea (Carroll 1956:213). and developed it further to include grammar in addi- tion to lexis, thereby making it a bona fide hypothesis. The milder version of the hypothesis is labeled lin- Hence, the resultant supposition is commonly given guistic relativity, coined by Whorf, since he always tried 927 SAPIR–WHORF HYPOTHESIS to qualify his assertions. This claims that the native lan- truth of the matter is that while no one has proved it guage influences one’s thoughts and perceptions. wrong, neither has anyone proved it right. Linguistic relativity can be illustrated using one of Whichever end of the continuum one considers in Whorf’s favorite sources of data, (Uto-Aztecan) Hopi. the relationship between language and culture, it is With the exception of birds, there is, according to important to realize the interpenetration of the two. In Whorf, only one word in Hopi for everything that flies. areas like bilingualism, is it really possible to learn a One also notes that Bedouin Arabic dialects have more foreign language without also simultaneously learning than 6,000 different lexemes for various types of the Weltanschauung of its speakers? Not only is lan- camels and camel paraphernalia. Perhaps the most guage part of culture, it allows for its acquisition. incontrovertible piece of evidence in its favor comes Sapir maintains that ‘the worlds in which different from the realm of numbers. Some languages (e.g. societies live are distinct worlds, not merely the same Hottentot [Nama]) only have words for the numerals world with different labels attached’ (Mandelbaum ‘one’ and ‘two’ and a word roughly translatable as 1949:162). Whorf took that one step further, asserting ‘many’ for ‘three or more’ (and a few languages have that grammar not only allows for the voicing of ideas no numbers at all). Thus, such concepts as googolplex but it also ‘is the shaper of ideas’ (Carroll 1956:221). are beyond everyday verbal expression in these lan- guages. Mathematics (as we know it) is thus not pos- sible, i.e. people are indeed ‘at the mercy of’ their References native tongues. Carroll, John B. (ed.) 1956. Language, thought, and reality: In ‘An American Indian Model of the Universe’ selected writings of Benjamin Lee Whorf. Cambridge, MA: (International Journal of American Linguistics MIT Press. Darnell, Regna. 1990. Edward Sapir: linguist, anthropologist, [1950]), Whorf argues that since there is neither an humanist. Berkeley: University of California Press. explicit nor an implicit reference to time in Hopi and Gumperz, John J., and Stephen C. Levinson (eds.) 1997. thus no tenses for its verbs, time and space cannot be Rethinking linguistic relativity. Cambridge: Cambridge the same concepts for Hopis as they are for English University Press. speakers. The Hopi Weltanschauung is different from Hoijer, Harry (ed.) 1954. Language in culture. Chicago, IL: University of Chicago Press. that of an SAE (Standard Average European, a Koerner, Konrad (ed.) 1984. Edward Sapir: appraisals of his life Whorfian term) speaker. Many of his other essays and work. Amsterdam and Philadelphia: John Benjamins. present data for his basic contention that Hopi meta- Lee, Penny. 1996. The Whorf theory complex. Amsterdam and physics, which underlies its cognition, is different Philadelphia: John Benjamins. from that of an English speaker; i.e. the Hopis cali- Mandelbaum, David G. (ed.) 1949. Selected writings of Edward Sapir in language, culture, and personality. Berkeley: brate the world differently because their language University of California Press. defines experience differently for them. As more infor- ALAN S. KAYE mation has surfaced on Hopi, some of Whorf’s specif- ic grammatical points have not held up well. This See also Humboldt, Wilhelm von; Peirce, Charles; explains why most linguists today believe the Sapir, Edward; Saussure, Ferdinand de; Whorf, Sapir–Whorf hypothesis to be invalid. However, the Benjamin Lee Saramaccan Saramaccan (or Saamáka) is the name of the creole lan- and eighteenth centuries. These runaway slaves and guage spoken by the Saramaka people, who live along their descendants are often referred to as ‘Maroons’, a the Suriname River in central Suriname (in the northeast word derived from Spanish cimarron, meaning ‘stray of South America). The names ‘Saramaka’ and animal’. The creole languages spoken by Maroon com- ‘Saamáka’ derive from the fact that the first settlements munities may be referred to as ‘maroon creoles’, to dis- of these people were located along the Saramacca River, tinguish them from (former) ‘plantation creoles’, such also in central Suriname. The c. 50,000 Saramaccans as Sranan. There are two Maroon creoles (or rather: living today are the descendants of African slaves who groups of Maroon creoles) in Suriname: Saramaccan escaped from the plantations to create their own com- (including three different dialects) and Ndyuka (used munities in the Suriname rain forest in the seventeenth here as a cover term for four different dialects). Maroon 928  SARAMACCAN creoles outside of Suriname that are still spoken today words could lead one to view Saramaccan as a creole include Palenquero (Colombia) and Angolar (São with two lexifier languages—English and Portuguese–– Tomé, an island off the coast of West Central Africa). the fact that most function words are from English sug- The main reason for distinguishing Maroon creoles gests that the Portuguese element was added later. For as a separate category is the fact that, due to their rel- this reason, Saramaccan is generally categorized as an ative isolation from outside influence, they are English-lexicon creole, albeit one with a strong assumed to be more ‘radical’ than (former) plantation Portuguese element. creoles, which have remained in contact with the lan- A second difference between Saramaccan and guage from which they have derived their lexicon (the Sranan is related to the fact that the former has a high- ‘lexifier language’). The term ‘radicality’ refers to the er percentage of words derived from African lan- typological distance between a creole and its lexifier guages. This is probably due to the fact that there was language. Although until now very little comparative much less contact with other languages in the case of research regarding the degree of radicality of different Saramaccan than in the case of Sranan. As to the ori- creoles has been done, it seems clear that the typolog- gins of these words, three African languages (or rather: ical distance between, say, Saramaccan and its (main) language clusters) have been especially important: lexifier, English, is larger than that between, say, Cape Kikongo (a Bantu language), Gbe, and Akan (both of Verdean Creole and Portuguese. Therefore, Maroon them Kwa languages; both Bantu and Kwa belong to creoles are assumed to be closer to creoles at the time the family of Niger-Congo languages). Although noth- of their formation (say, 300 years ago) than (former) ing is known about the specific African origins of the plantation creoles. Among the Maroon creoles, individual runaway slaves who formed the ‘founder Saramaccan has acquired a special status, especially in population’ of the Saramaka people, we do have reli- research associated with Bioprogram Theory (see e.g. able information about the origins of the African Bickerton 1984), to the extent that it has come to be slaves in general, who were brought to Suriname in the considered as the most radical creole. According to 1675–1700 period (Arends 1995a:243). In this period, this view, Saramaccan provides a rare opportunity to roughly speaking, half of all Suriname slaves came get a closer look at the process of creolization. from an area where Bantu languages such as Kikongo, While there are still important gaps regarding the were spoken, while the other half came from an area history of the Saramaka and their language, the fol- where Kwa languages, such as Gbe and Akan, were lowing seems to be clear (Smith 2002). The origins of spoken. The connection between the ethnolinguistic the Saramaka people and their language go back to origins of the Suriname slaves and the traces that were escapes from plantations that took place before 1700, left in the Suriname creoles by their native languages at a time when the formation of a plantation creole receives further support from the fact that Saramaccan must have been well on its way. Assuming that most of exhibits some rather marked phonological features, the runaway slaves had been on the plantations for such as lexical tone and nasal and complex stops, some time before making their escape, they took at which are characteristic of these three languages least some knowledge of this evolving plantation cre- (Bruyn 2002). ole with them. This explains the structural similarities Like its sister language Sranan, Saramaccan is well between Saramaccan and Sranan, both of which are documented in its early stages. In the case of descendants of the earlier plantation creole. Saramaccan, however, the early documentation is lim- In spite of these similarities, however, there are also ited to a very short period, roughly 1780–1820. This a number of differences between the two languages. has to do with the fact that the Moravian Brethren, to One of these is the proportion of Portuguese-derived whom we owe these early writings, stopped their mis- words, which is much larger in Saramaccan than it is in sionary activities among the Saramaka around 1820. Sranan. In the former, one third of the basic vocabulary These early documents, which together number well (basic words such as ‘water’, ‘go’, ‘big’) is derived over 2,000 manuscript pages, consist mainly of reli- from Portuguese, while this is much smaller in Sranan. gious texts, such as Bible translations, although some The remainder of the basic vocabulary is largely linguistic descriptive works, such as dictionaries, are derived from English (Suriname was an English colony included as well (see Arends (1995b) for further infor- before it became Dutch), while there are also a few mation). Unfortunately, however, up to now only a few items from West African languages. The presence of of these documents have been made available for lin- these Portuguese-based words is explained by the fact guistic research (Arends and Perl 1995). that, as has been established by historical research (e.g. As to the major structural features of Saramaccan, Price 1976, 1983), many of the first Saramaka came many of these are also found in Sranan. Here, we will from plantations owned by Sephardic Jews, who were only mention and illustrate some of the features in Portuguese- (and to some extent Spanish-) speaking. which the two languages differ. (This section draws Although the presence of many Portuguese-derived heavily on Bruyn 2002; see also Bakker et al. (1995.) 929 SARAMACCAN Lexicon. Some examples of Portuguese-derived Verbal arts. In the literature on the ‘verbal arts’ basic vocabulary items are búka (<boca) ‘mouth’ and (story telling, song, etc.), Saramaccan is often dis- dá (<dar) ‘give’. In both cases, the equivalent word in cussed together with Sranan (Herskovits and Sranan is derived from English: mofo (< mouth) Herskovits 1936; Voorhoeve and Lichtveld 1975). A ‘mouth’ and gi (<give) ‘give’. Some examples of work that is entirely devoted to storytelling in African-derived words are katangá ‘cramp’ from Saramaccan is Two evenings in Saramaka by Richard Kikongo nkatangá, azε ‘magic’ from Gbe àze, and and Sally Price (1991). gongosá ‘gossip’ from Twi ŋkɔŋkɔnsá. Phonology. For a language to have lexical tone, as does Saramaccan, means that some words differ only References in their tonal pattern, for example in having a low tone Arends, Jacques. 1995a. Demographic factors in the formation of in a syllable where the other word has a high tone, Sranan. The early stages of creolization, ed. by Jacques Arends, 233–85. Amsterdam and Philadelphia, PA: Benjamins. while being completely identical in every other ———. 1995b. Early Saramaccan texts in the Moravian respect. An example is the pair ná~nà, with the first archives in Paramaribo, Herrnhut and Utrecht. The Carrier having a high tone where the second has a low. Pidgin 23(3). 1–6. Despite this (seemingly) small difference, the mean- Arends, Jacques and Matthias Perl. 1995. Early Suriname ing of the two words is completely opposite: ná creole texts: a collection of 18th-century Sranan and Saramaccan documents. Frankfurt: Vervuert. means ‘be’, while nà means ‘be not’. Examples of Bakker, Peter, Norval Smith, and Tonjes Veenstra. 1995. nasal stops are /mb/ and /nd/; examples of complex Saramaccan. in Pidgins and creoles: an introduction, ed. by stops are /kp/ and /gb/. Note that /mb/, etc., refer to Jacques Arends, Pieter Muysken, and Norval Smith, phonemes, not combinations of phonemes. In other 165–78. Amsterdam and Philadelphia, PA: Benjamins. words, a word like mbéti ‘meat, animal’ consists of Bickerton, Derek. 1984. The language bioprogram hypothesis. The Behavioral and Brain Sciences 7. 173–88. four phonemes, not five. Bruyn, Adrienne. 2002. The structure of the Surinamese creoles, Morphology. Apart from other functions, reduplica- Carlin and Arends. 153–82. tion is used in Saramaccan to derive adjectives from Carlin, Eithne, and Jacques Arends (eds.) 2002. Atlas of the lan- verbs, e.g. nákináki ‘beaten’, derived from náki ‘beat’. guages of Suriname. Leiden: KITLV Press. These reduplicated forms are used both attributively, Herskovits, Melville, and Jean Herskovits. 1936. Suriname folk-lore. New York: Columbia University Press. as in dí nákináki miíi ‘the beaten child’, and predica- Price, Richard. 1976. The Guiana Maroons: a historical and tively, as in dí miíi dε nákináki ‘the child has been bibliographical introduction. Baltimore: The Johns Hopkins beaten (is in a beaten state)’. Saramaccan also differs University Press. from Sranan in that the agentive suffix -ma (-man in ———. 1983. First-time: the historical vision of an Afro- Sranan) may follow an entire verb phrase, which may American people. Baltimore: The Johns Hopkins University Press. itself even contain a subordinate clause. This may Price, Richard, and Sally Price. 1991. Two evenings in result in quite complex agentive nouns, such as seti-u- Saramaka. Chicago: Chicago University Press. kanda-ma ‘precentor’ (lit. ‘start-to-sing-suffix’). Smith, Norval. 2002. The history of the Suriname creoles: ori- Syntax. The syntactic differences between gin and differentiation, Carlin and Arends. 131–51. Saramaccan and Sranan have not been investigated in Veenstra, Tonjes. 1996. Serial verbs in Saramaccan: predication and creole genesis. Unpublished Dissertation. University of sufficient detail to allow them to be discussed here. One Amsterdam. feature that has drawn considerable attention, however, Voorhoeve, Jan, and Ursy Lichtveld (eds.) 1975. Creole drum: is the serial verb construction, which is also found in an anthology of Creole literature in Suriname. New Haven Sranan. For an elaborate discussion of serial verbs in and London: Yale University Press. Saramaccan, the reader is referred to Veenstra (1996). JACQUES ARENDS Saussure, Ferdinand de Ferdinand de Saussure is known as one of the founders retained a major interest in historical linguistics, his of contemporary linguistics and as the father of student dissertation on the a vowels of Indo-European European Structuralism. Throughout his life he (1879) containing the solution to a problem that was to 930  SAUSSURE, FERDINAND DE be empirically confirmed many years later. Certain a new and coherent theory. There is no doubt about the concepts for which he was to become known were influence of the CLG on European Structuralist present in his early works, but would be more fully Linguistics and on Structuralist thought more general- developed in his Geneva lectures, notes from which ly. Firstly, through his lectures in Paris, Saussure exert- were published posthumously in 1916 as the Cours de ed a major influence on future generations of French linguistique générale (Course in general linguistics linguists, through Meillet for historical linguistics and (CLG)). In the first three sections of the Cours, Passy and Grammond for phonetics, to Benveniste and Saussure attempts to come to grips with the kind of Guillaume; in Switzerland, the Geneva School contin- phenomenon that we are studying when we analyze ued his work, with Bally working on a ‘linguistique de language. In the face of the many possible approaches la parole’. Through his writing and through the pres- (historical, sociological, etc.) to the heterogeneous phe- ence of Karcevski, he influenced the Prague Circle nomenon that is ‘langage’ (translated as ‘language’ including Jakobsen, as he did the Copenhagen Circle. or as ‘language faculty’), Saussure tries to focus on In the English-speaking world, his influence existed, its essence through a series of binary divisions. First, but in a weaker form: Bloomfield published a positive it is necessary to distinguish the underlying sys- review of the Cours in 1924, and it is an important tem (‘langue’) from its individual representations point of reference in Chomsky’s Cartesian Linguistics (‘parole’). Second, Saussure differentiates between the (with langue/parole being in some ways a precursor of diachronic (evolution of a language) and the synchron- competence/performance). For the Structuralists of the ic (a snapshot of a language at a particular time), the mid-twentieth century in other disciplines, as for the latter being the proper object of study in general lin- postmodernists, the CLG is the fundamental text. guistics. ‘Langue’ is a system in which the whole is Saussure’s work has been continuously promoted by more than the sum of the parts (Saussure himself did the Cahiers Ferdinand de Saussure and there has been not use the term structure). The system is made up of a regeneration of interest with the ongoing publication units, each of which is defined by its difference from of new manuscript material and the creation of an other units, whether words or phonemes (it should be Institut Ferdinand de Saussure. When reading the noted Saussure generally used ‘phonology’ where Cours, it is important to remember that Saussure him- today we would use ‘phonetics’ and vice versa). Each self did not write it, and also that the English transla- unit functions as a linguistic ‘sign’ composed of two tions of some of his terms are misleading. Thus, the inseparable faces: the ‘signifiant’ (signifier or acoustic ideas of a major twentieth-century thinker come down image) and the ‘signifié’ (signified or mental represen- to us in a tantalisingly uncertain form. tation of external reality). It is a characteristic feature of sign systems, of which language is the principal one, Biography that the link between the sign and the reality to which it refers is arbitrary. Linguistic signs are linked togeth- Saussure was born in Geneva in 1857 into a family of er either paradigmatically (one is selected from a distinguished scientists and scholars. In 1876, he went series, e.g. grammatical cases) and syntagmatically (in to study in Leipzig, where he encountered the a chain within the utterance). ‘neogrammarian’ movement. In 1877, he presented his In talking about ‘langue’, Saussure refers to a men- first paper to the Société de Linguistique de Paris, and tal reality, and to a collective and therefore social phe- in 1879 published a dissertation entitled Mémoire sur le nomenon. That he considers linguistics proper to be a système primitif des voyelles des langues indo- discipline in its own right and as such separate from européennes, which immediately brought him renown. psychology and sociology does not, however, mean The following year, his doctorate on the genitive in that he denies the links between language and history Sanskrit was awarded summa cum laude and published or society. He has been criticized for neglecting in 1881. After a field trip to Lithuania in 1880, Saussure ‘parole’, syntactic analysis, and units beyond the sen- went to Paris to study at the Ecole Pratique des Hautes tence. It appears from the CLG that he intended a ‘lin- Etudes under Bréal and from the following year began guistique de la parole’ to follow the ‘linguistique de la to teach there. In 1891, Saussure returned to Geneva to langue’, and that he considered syntax to fall within take up a Chair in Indo-European linguistics, and later the remit of ‘parole’. The discipline––then still to be in General Linguistics. In three series of lectures (1907, established––of which Saussure considered linguistics 1908–1909, and 1910–1911), he covered historical lin- to be the most important part is semiology (otherwise guistics and started to lay the foundations for the sci- known as semiotics). ence of linguistics. In contrast to the brillance of his There has been much debate about influences on early years, in later life Saussure published relatively lit- Saussure’s theories, but it is generally agreed that he tle, apparently feeling convinced of the need to forge the provides key insights that interlock to form the basis of basic concepts of linguistic analysis but daunted by this 931 SAUSSURE, FERDINAND DE prospect. He developed related interests, writing about by C. Sanders and M. Pires. Oxford: Oxford University the possible widespread presence of anagrams in Latin Press. Saussure, Ferdinand de. 1916. Cours de linguistique générale, poetry and about the German epic Niebelungenlied. In publié par C. Bally et A. Sechehaye avec la collaboration de addition to this crisis of confidence, Saussure suffered A. Riedlinger. Lausanne/Paris: Payot. from ill health, and died in 1913. ––––––. 1966. Course in general linguistics, transl. by Wade Baskin, 2nd edition. New York and London: Harper & Row. ––––––. 1968. Cours de linguistique générale, édition critique References par R.Engler, Vol. I. Wiesbaden: Otto Harrassowitz, II. Arrivé, M. 2000. Saussure. Paris: Les Belles Lettres. ––––––. 1983. Course in general linguistics, transl. by Roy Bouquet, Simon. 1997. Introduction à la lecture de Saussure. Harris. London: Duckworth. Paris: Payot. ––––––. 1993. Troisième Cours de linguistique générale Fehr, Johannes. 2000. Saussure entre linguistique et sémiologie. (1910–1911) d’après les cahiers d’Émile Constantin, ed./ Paris: Presses Universitaires de France. transl. by E. Komatsu and R.Harris. Oxford: Pergamon. Gadet, Françoise. 1989. Saussure and contemporary culture. ––––––. 1995. Cours de linguistique générale, publié par London: Hutchinson. Charles Bally et albert Séchehaye; avec la collaboration de Gordon, W. Terence. 1996. Saussure for beginners. New York: Albert Riedlinger; édition critique préparée par Tullio de Writers and Readers. Mauro. Paris: Payot. Harris, Roy. 1987. Reading Saussure. London: Duckworth. ––––––. 1996. Premier Cours de linguistique générale (1907) ––––––. 2001. Saussure and his interpreters. Edinburgh: d’après les cahiers d’Albert Riedlinger, ed./transl. by E. Edinburgh University Press. Komatsu and G.Wolf. Oxford: Pergamon. Holdcroft, David. 1991. Saussure. Signs, system and arbitrari- ———. 1997. Deuxième Cours de linguistique générale ness. Cambridge/New York: Cambridge University Press. (1908–1909) d’après les cahiers d’Albert Riedlinger et Koerner, E.F.K. 1999. Linguistic historiography projects and Charles Patois, ed./transl. by E. Komatsu and G.Wolf. prospects, Chapters 6, 8, 10. Amsterdam/Philadelphia: John Oxford: Pergamon. Benjamins. Thibault, Paul J. 1997. Re-reading Saussure. London/New Normand, Claudine. 2000. Saussure. Paris: Les Belles Lettres. York: Routledge. Sanders, Carol. 2000. Saussure translated. Historiographia ––––––. 2001. Ferdinand de Saussure: critical assessments of Linguistica XXVI(2/3). 345–58. leading linguists. 6 vol. London/New York: Routledge. ———. 2005. The Cambridge Companion to Saussre. CAROL SANDERS Cambridge: Cambridge University Press. ———. 2005. Writing in general linguistics, by Ferdinand See also Jakobson, Roman; Meillet, Antoine; Struc- de Saussure, ed. by S. Bouquet and R. Engler, and transl. turalism Scandinavia Scandinavia usually refers to the mainland Scandinavian also spoken in Finland, Norway, and Sweden, and countries Denmark, Norway, and Sweden. Denmark Swedish is spoken in Finland. includes the Faeroes and Greenland, which are Despite many types of language contacts through- autonomous regions of the Danish kingdom. Finland out history, the Nordic countries have all attained a includes the Swedish-speaking Åland Isles in the Baltic proportion of 90% or more speakers, for their domi- Sea. Finnish and Sami––the other main indigenous lan- nant national languages, at the cost of inherent cultur- guages in the region––neither belong to the Nordic al and linguistic pluralism. Most speakers of the (Swe. nordiska språk) nor the Scandinavian languages. indigenous minority and secondary official languages The general term for Denmark, Finland, Iceland, of the Nordic countries are bilingual, and many are Norway, and Sweden is the Nordic countries (Swe. dominant in their second language, the majority lan- Norden or de nordiska länderna; Fi. Pohjoismaat, lit. guage of their country. ‘the North countries’). When discussing Scandinavia and its languages, it is for historical and political reasons Sweden (Sverige) necessary to include all Nordic countries and the Samis. The Samis, the aboriginal people of the North, are inhab- Sweden has a parliamentary and democratic itants of Finland, Norway, and Sweden (and Russia). monarchy, a population of nine million, and no official They form a supranational Sami homeland, Sapme, with language de jure. Until 1809 it included Finland, and its own linguistic and cultural characteristics. Finnish is from 1814 to 1905 it formed a union with Norway. 932  SCANDINAVIA Swedish is the first language of about 8.3–8.5 million Greenland (an Eskimoic/Inuit language; Kalaalit people, Finnish (Sweden Finnish) has about 230,000 Nunat) and to Faeroese in the Faeroes. The Faeroes speakers (no official language censuses exist); received autonomy in 1948 and Greenland received Meänkieli (closely related to Finnish) has about autonomy in 1979. Danish was temporarily a high 50,000 speakers, mainly in the North-Swedish variety in Norway during the era of Danish dominance Tornedalen region; a Finnish-speaking population in over Norway (1380–1814). the central forest regions (Värmland, Dalecarlia) The main migrant languages are Turkish (50,000), became extinct in the late 1960s, after four centuries. German, English, the Scandinavian languages, Bosnian/ An equal amount of Finland Swedish descendants Croatian/Serbian, Punjabi, Vietnamese, Persian, and (migrants from Finland), Norwegian, and Danish Somali. speakers live in Sweden: about 50,000 each. For Finland, until recently, Sweden has been a buffer zone Norway (Norge, Noreg) for cultural influence from Southern and Western Europe. Similarly, Finland has functioned as a buffer Norway has a parliamentary and democratic zone for Swedish contacts––peaceful and monarchy, autonomous since 1814 and independent hostile––with Russia and its historical predecessors. since 1905, with a population of 4.6 million. The cap- The Samis inhabited the Swedish mainland down to ital Oslo area has about 750,000 inhabitants. Two offi- central Sweden, when the Scandinavians arrived. In cial varieties of Norwegian are spoken: Nynorsk and Sweden, Sami has about 20,000 speakers, divided into Bokmål (about 95% of all Norwegians speak one of several (five or more) languages. Some archaic them as their first language; Bokmål is chosen by 80- Swedish varieties are occasionally held to be lan- 85% ). Bokmål is mainly spoken in the urban and guages, e.g. Älvdalsmål in the province of Dalecarlia northern parts, and Nynorsk in the western/central (Swe. mål, old form for dialekt). Capital Stockholm rural regions. Of the 440 municipalities, 114 have (metropolitan Stockholm area with 1.6 million inhabi- declared themselves as officially Nynorsk-speaking, tants) is multilingual, with some suburbs almost exclu- 175 have declared Bokmål as their official choice, and sively inhabited by non-Swedish speakers. the rest, 151, are ‘neutral’. The larger cities (Oslo, The main migrant languages are Albanian, Arabic, Bergen, Trondheim) have declared themselves ‘neu- Bosnian/Croatian/Serbian, English, Finnish/Meänkieli, tral’, but in practice there is a clear dominance for Greek, Kurdish, Persian, Spanish, Turkish, and Polish. Bokmål. About 120 languages are taught in hour-based mother The largest Nordic population of Sami speakers is tongue instruction in the Swedish basic compulsory found in Norway (30,000). Most of them belong to school. Since 2000, Sami, Finnish, Meänkieli, Romani the North Sami language group. It is a regional offi- Chib, and Yiddish are official minority languages. cial language in Finnmark and Tromsø provinces, Northern Norway, where some of the municipalities are bilingually administrated. Among the indigenous Denmark (Danmark) Kvens (who speak a Finnish-related language/vari- Denmark has a parliamentary and democratic ety), in and east of the town of Tromsø, about 3,000 monarchy, with a population of 5.3 million. Danish is know the language. Several main migrant groups are the official language. It is the first language of about ‘nonvisible minorities’. The main migrant languages 4.9 million Danes. The capital Copenhagen area are the Scandinavian languages (100,000), English (København) has about 1.7 million inhabitants. (70,000), Punjabi (24,000), German, Bosnian/ German is an official regional language in Southern Croatian/Serbian, Vietnamese, Turkish, Persian, and Denmark (20,000 speakers in Northern Schleswig), Finnish. and a notable population of speakers of Danish lives in Northern Germany/Schleswig, as a result of border Finland (Suomi) treaties between Denmark and Germany. Denmark’s long-term enemy was Sweden, with which Denmark Finland was a part of the Swedish kingdom for fought about the dominance over the Baltic Sea area. about 700 years, until 1809. It was the Autonomous Denmark lost three provinces on the Swedish main- Grand Duchy of the Russian Empire from 1809 to land (Scania, Halland, and Blekinge) and Gotland to 1917, and became an independent republic in 1917. It Sweden in the mid-seventeenth century. The has a population of 5.1 million, officially bilingual in Norwegian provinces of Härjedalen and Jämtland, Finnish and Swedish (nationally). Sami is an official then parts of Denmark, were lost to Sweden simulta- regional language since 1991. Romani, Tatar, and sign neously. Danish is a former high variety and a second- language have been given minority status since 1994, ary school language, in addition to Greenlandic in and Russian is on the verge of becoming acknowledged 933 SCANDINAVIA as one in 2001. In all, 93% speak Finnish as their first Language Contact Types in language. Finland Swedish (finlandssvenska) is spoken Scandinavian/Nordic Past and Present in the southern and southwestern coastal areas. The Finnish Åland Isles in the Baltic Sea have a specific Several main types of language contacts have exist- legal and linguistic position; they are officially mono- ed in Scandinavia. The oldest known one is that of the lingual in Swedish (League of Nations and the UN, Sami and Scandinavian (in Sweden and Norway), and 1921). Finland Swedish, with 297,000 speakers Sami and Finnish settlers (in Finland; 2500–200 BC). (5.7%), is a fully protected official language (the In older place names and special vocabulary, the Sami Constitution and the Language Act, 1919, 1921 and impact is still obvious. The impact of the majority lan- 2004). Finland Swedish is referred to as the East guages is more apparent in Sami, however. High pro- Swedish dialect. It has an overt (Sweden Swedish) and portions of Sami speakers live in the three capitals: a covert norm (Finland Swedish standard). It is also Stockholm, Oslo, and Helsinki. seen as an archaic Swedish variety, influenced by The second type consists of Scandinavian expan- Finnish. Of the municipalities, 22 are officially sionism, which includes the Viking era (800–1100). Swedish-speaking (2001), 20 are bilingual in Finnish- During this period, the Faeroes, Iceland, and Greenland Swedish, three are bilingual in Finnish-Sami, and 389 received their Scandinavian settlers and later, domi- are monolingual in Finnish. The capital of Helsinki nance. The Scandinavian/Swedish settlement of south- (Helsingfors; one million inhabitants in the metropoli- ern and western Finland, a third type, is both older and tan area) is bilingual. The vast majority of the Finland simultaneous with this. These two types of migration Swedish speakers are bilingual, who function as lan- formed the basis for Scandinavian “semicommunica- guage and culture mediators from and to the other tion” and later pan-Nordism (Swe. nordism; see Scandinavian countries. Most Finns learn another below). Such receptive bilingual communication also national/‘domestic’ language in school, by force of the developed due to the Danish dominance over Norway Educational Act. English competes with Swedish for and Iceland. The changing borders between Sweden Finnish-speaking pupils today. The main migrant lan- and Denmark–Norway reinforced this; when the for- guages are Russian (20,000 speakers), Estonian, mer dialect areas of Denmark – Norway were Somali, Vietnamese, English, Kurdish, and Karelian. Swedicized from the mid-1600s, new overlapping lin- guistic continua developed. The fourth type is that of continuous migration and Iceland (Island) settlements of Finns in Sweden and Norway, from Iceland is a republic, which became independent about the fifteenth century CE. Vast areas were settled in 1944, with a population of about 280,000. The and farmed by Finnish speakers. These so-called for- capital area of Reykjavík has 150,000 inhabitants. est Finns in central Sweden and southeastern Norway Iceland was under Norwegian rule from 1380 to have recently experienced language death. A separate 1550, and under Danish rule from 1550 to 1848. migration was that of Finns to Stockholm, which has Icelanders retained Icelandic as their lingua sacra been continuous for about 800 years. Migration to during the Reformation. Icelandic is the official lan- Sweden was internal until 1809; later, it reemerged as guage. Iceland is one of the few practically monolin- emigration from Finland to Sweden. Fewer Finns have gual nation-states of today. In school, Danish and left for Norway and even less for Denmark. Migration English are secondary languages. An American air- in the other direction has been rare. base, Keflavik, is the only additional major language Fifth, Latin was the scholarly language for the representation of Iceland, but it lives a separate life. mainland Scandinavian countries and Finland for sev- Icelandic pupils learn an adapted variety of eral centuries, until the early nineteenth century. The “Scandinavian” based on modifications of pronunci- former countries shifted to their national languages ation and a common Scandinavian vocabulary, for earlier than did Finland to Finnish, for cultural, public their contacts with Scandinavians, since Icelandic and educational purposes. differs considerably from the mainland Scandinavian All of the countries faced extensive migration to languages. North America, especially to the regions close to the Language and corpus planning play an active role Great Lakes, during the latter half of the nineteenth in the everyday lives of Icelanders. Language contact century. Scandinavians on an average moved earlier effects are, however, inevitable, since quite a few than Finns. Icelandic students temporarily study abroad (North More recently, there have been sporadic migration America, Scandinavia). Today, modern media and waves to all of the Scandinavian countries, during and communication technology also bring English into the after World War II. Both labor migration and refugee everyday life of Icelanders. waves have normally been punctual in kind, compared 934  SCANDINAVIA to the continuous migration between the Nordic coun- Norn. Initially, speakers of Insular Nordic languages tries. Due to the booming post-World War II economy presumably had contacts with Celtic speakers, who of the Nordic countries, first in Sweden and later in had inhabited the same areas temporarily. Norway, Denmark, and Finland, all countries have Scandinavian raids and trade created external lan- received notable numbers of migrants from first, the guage islands in vast areas of Northern and Western neighboring countries, then from Southern Europe, and Europe. The Viking age (c. 800–1100 CE) started with more recently globally. The most significant of these tribal chieftains’ small-scale raiding. Later, this turned migration waves was that of Finns to Sweden from the into a large-scale warfare led by powerful Viking late 1950s to the late 1980s. The free Nordic labor mar- dynasties. The improved shipbuilding techniques con- ket was one important contributing factor. Today, there tributed to this. The Viking period thus changed the are groups of equal size of speakers of Swedish in language history of e.g. the British Isles. A landmark Finland, as there are speakers of Finnish in Sweden. was the Danish conquest of Britain in the eleventh A last type of language contact is a matter of cul- century CE. The Danes later continued their expansion tural and linguistic transfer, rather than the transfer of in more peaceful ways. The Norwegians concentrated people. Earlier, Latin, German, and French played their raids to the islands north and west of Scotland such roles. The increasing presence of English in and England, including Ireland, Isle of Man, and Scandinavia has been seen as an asset since the 1950s, Scotland itself. Dublin was founded by a Norwegian especially in Sweden, but has recently become a mat- chieftain in 841 CE. In Normandy, a Danish Duchy ter of concern. English is diminishing the language use was founded in the tenth century by King Rollon. The domains of the national languages. English is by far Swedes established their rule in Kiev, Ukraine, before the most popular foreign language in school, and it is the Viking age and later in the Baltic countries. Russia compulsory. Its impact differs from earlier foreign lan- was a target for early Swedish Viking trade and raids. guages, which were transplanted for specific purposes Later, it became the target of the unified Swedish king- or by socially defined groups. English is now used as dom’s attempts both to expand and protect itself. a language of wider communication, and English Finland was Christianized by the Swedes and thus expressions occur in all types of domains and in all came under its religious and cultural impact from the social groups. twelfth century, even if Scandinavians had lived in the The common linguistic and cultural roots, in combi- coastal areas for centuries by then. nation with the later separate historical and political When the dominance of Denmark weakened, the developments in the three countries, make the prob- Viking rule in the Baltic Sea was soon replaced by that lems of the concept of language evident. If languages of others. The commercial union of the Hanseatic are defined as language forms characterized by mutual League, especially under the leadership of Lübeck in intelligibility, Swedish and Norwegian or Norwegian the fourteenth and fifteenth centuries, was particularly and Danish could be defined as dialects rather than sep- successful. For some time, the League ruled the whole arate languages. On the other hand, speakers of some Baltic Sea and the North Sea. The establishment of Low of the Swedish dialects, such as Scanian (skånska) in German merchant layers in the main cities, Stockholm, the south of Sweden, which until 1645 belonged to Kalmar and Visby in Sweden, Copenhagen in Denmark, Denmark, the northernmost Swedish dialects, which and Bergen in Norway, paved the way for an extensive bear evidence of language contacts with Sami and linguistic influence on the local languages. In Sweden, Finnish, or the archaic Gutnish dialects of Gotland, do this created a new prestige language; Low German not share full intelligibility with all other dialects. became the language of the merchants, and the towns- people became highly influenced by its cultural impact. About one-third of modern Swedish consists of Low Scandinavian Expansionism and Retreat German vocabulary, for example, släkt ‘family, rela- A division into West vs. East Scandinavian started tives’, handel ‘trade’, krig ‘war’, möjlig ‘possible’, in Old Scandinavian in the sixth or seventh century, bliva ‘be, become’, smaka ‘taste’, skriva ‘write’, and when Norse farmers started sailing westward and men ‘but’. This indicates a high degree of bilingualism inhabited the Faeroes and Iceland. The process was among important sections of Swedish speakers. Another completed in the ninth and tenth centuries CE. This central contribution from Low German was the borrow- contiguous migration also reached Greenland. From ing of derivational prefixes, like an-, be-, för-: använda Greenland, Leif Eriksson sailed westward and “dis- ‘use’, behöva ‘need’, förstå ‘understand’, and suffices covered” North America in the eleventh century CE. like -era, -het, -inna: studera ‘study’, svaghet ‘weak- New varieties of Old Norse developed. The daughter ness’, lärarinna ‘female teacher’. languages Icelandic and Faeroese developed, in addi- Scandinavian ‘semicommunication’ refers to the tion to two extinct languages, Orkney and Shetland possibility to use one’s own language, and to understand 935 SCANDINAVIA and be understood by speakers of other Scandinavian form of mutual understanding of each other’s lan- languages. It functions thanks to the common back- guages. Today, Nordism is getting competition from ground of the Scandinavian languages. Finns have in EU- level cooperation. To some extent, it is also chal- addition learned some Swedish in school or needed it lenged by the attempts to integrate the three Baltic for their civil careers, for several centuries. Icelanders countries, Estonia, Latvia, and Lithuania, into the have learned both Danish and, more recently, adapted Nordic sphere, in which cooperation in English is ‘Scandinavian’. With some minor phonological modifi- preferred. cations, the use of common vocabulary, and avoidance of ‘difficult’ language-specific features, most Nordic citizens knowing a Scandinavian language may com- References municate among themselves. The demand to modify Aikio, Samuli, et al. 1994. The Sami culture in Finland. speech is the highest for Danes and Icelanders, who are Helsinki: Lapin Sivistysseura. the furthest away from modern, common Scandinavian. Att förstå varandra i Norden, (To understand each other in the Nordic countries). Stockholm: Nordiska rådet/Nordiska The common historical basis of Danish, Norwegian, ministerrådet/Nordiska Språkrådet, 1997. Icelandic, and Swedish is one prerequisite of this type Blom, Gunilla, et al. (ed.) 1992. Minority languages: the of communication. Another is the idea and philosophy Scandinavian experience. Oslo: Nordic Language Secretariat. of a shared political agenda, which to a high degree was Börestam Uhlmann, Ulla. 1994. Skandinaver samtalar, created in the latter half of the 1800s. This pan-Nordic (Scandinavians in conversation). Skrifter från institutionen för nordiska språk, 38. Uppsala: Institutionen för nordiska idea means that the countries do not only share histori- språk vid Uppsala universitet. cal and cultural features, but they have also attempted to Hasselmo, Nils. 1974. Amerikasvenska (American Swedish). teach each other’s languages and about each other’s Skrifter utgivna av Svenska språknämnden, 51. Lund. countries in school. Even today, Nordic people may Stockholm: Svenska språknämnden. present a common face to other countries and in inter- Haugen, Einar. 1969. The Norwegian language in America: a study in bilingual behavior, Vols. I–II, 2nd edition. national cooperation, based on shared values. During Bloomington: Indiana University Press. the last decades, this linguistic understanding and com- Melander, Björn (ed.) 2000. Svenskan som EU-språk (Swedish municative potential have weakened, and so has the as a EU Language). Uppsala: Hallgren & Fallgren. striving for a common political agenda. Denmark, Simonsen, Dag F. 1996. Nordens språk i EU:s Europa (The lan- Finland, and Sweden are members of the European guages of the Nordic countries in the Europe of the European Union), Rapport 22. Oslo: Nordisk Språksekretariat. Union (EU); Norway and Iceland are not. Nevertheless, Vikør, Lars. 1993. The Nordic languages: their status and inter- in practical administrative, political life and in the atti- relations. Oslo: Novus Press. tudes of people from the Nordic countries, there is a Virtaranta, Pertti, et al. 1993. Amerikansuomi. (American sense of togetherness and regional identity, which goes Finnish.) Tietolipas 125. Helsinki: Suomalaisen Kirjallisuuden beyond ordinary neighborhood feelings. The Samis Seura. were more or less excluded from such cooperation ear- Web-sites: lier, but they have now created their own Nordic net- works, both between Samis in different countries, and www.dst.dk (Statistics Denmark) www.ssb.no (Statistics Norway) between different Sami groups in the same country. www.scb.se (Statistics Sweden) The original Nordism of the 1860s has later been JARMO LAINIO extended in many formalized ways. The exchange of ideas, culture, and ordinary people between the coun- See also Swedish and Scandinavian languages; tries has continued and enabled the historically based Finnish and Finnic Languages; Language: Con- Scandinavian communication to be prolonged in the tact—Overview; Bilingualism; Language Planning Searle, John R. John Searle, the contemporary analytic philosopher, this book, Searle systematized a number of ideas that rose to fame with the publication in 1969 of Speech were originally developed by J.L. Austin whom he acts: An essay in the philosophy of language (SA). In knew as a graduate student at Oxford in the 1950s. 936  SEARLE, JOHN R. Searle’s reputation as Austin’s main intellectual heir, nally picked up from his Oxford mentor who was self- while no doubt right as far as it goes, must nonetheless declaredly an Aristotelian. be properly qualified. On the one hand, it should not There can be little doubt that the theory of speech be allowed to eclipse his own important contribution acts, in the form in which it is known to scholars out- to not only the theory of speech acts in particular and side the disciplinary bounds of analytic philosophy, is philosophy of language in general but also to the phi- largely due to Searle’s decisive intervention. This is losophy of mind, a branch of philosophy in which he especially the case in areas such as linguistics, psy- has established himself as a key, although contentious, chology, sociology, and anthropology where speech figure. On the other hand, it is not entirely true either act theory has had a tremendous impact. In the early that his chief merit consisted in simply systematizing 1970s, for instance, there was a very strong dissident and bringing to fruition what his teacher Austin would movement called Generative Semantics within the have done all by himself, had an untimely death not Generative paradigm in linguistics that, spurred on by prevented him from doing it. Unfortunately, many Searle’s innovative idea of prepositional acts, sought writers have helped create the widespread mispercep- to incorporate the concept of speech acts into the tion that Searle simply took over from where Austin semantic component of grammar by stipulating that left off or, equivalently, that, except for some minor every sentence had an underlying form with a highest differences here and there, the work of the two clause that specified its speech act status. philosophers constitutes a smooth continuum. With the publication of Expression and meaning Searle’s originality as a philosopher was announced (EM) in 1979, Searle marked the consolidation of a on the very first pages of SA, where he insisted that his theory that, after two decades of rapid expansion and work was, as the subtitle proudly proclaimed, an exer- intensive research, had important things to say about a cise in the philosophy of language rather than Austin’s range of interesting phenomena that included own narrower ‘linguistic philosophy’. In fact, Searle metaphor, fiction, and the so-called indirect speech has ever since shown little interest in some of the acts (such as, say, the performance of an act of request major tenets of linguistic philosophy, notably the idea by asking one’s interlocutor a question––as in saying that a careful analysis of the ordinary, everyday use of ‘Can you please pass the salt’ to someone sitting next language can help solve a great number of philosoph- to you at a dinner table). Searle also indicated that his ical puzzles. Searle’s central concern was announced own future work in the philosophy of language was in SA in the form of two aphorisms: (a) Speaking a going to be strongly influenced by the concept of language is engaging in a form of rule-governed intentionality. In fact, along with his interest in bring- behavior.(b) The basic unit of linguistic communica- ing the speech act theory in line with the thesis of tion is the speech act. Searle also made his differences intentionality, Searle also announced (in the preface to with Austin very clear in his early essay ‘Locutionary EM) yet another task for the years ahead: that of pro- and illocutionary acts’, wherein he argued that viding an adequate formalization of the theory using Austin’s concept of a ‘locutinary act’ should be the resources of modern logic. replaced by his own notion of a ‘propositional act’. In 1983, Searle published Intentionality and two For Austin, locutionary acts were the acts of producing years later Foundations of illocutionary logic (FIL), in utterances with a certain sense and meaning and these collaboration with the Canadian logician Daniel were opposed to illocutionary acts (performing certain Vanderveken. Subtitled An essay in the philosophy of acts such as promising in saying certain words) and mind, Intentionality marked a decisive turn in Searle’s perlocutionary acts (doing certain things such as philosophical enterprise and prepared the ground for threatening by saying certain words). By introducing his later distinctive and often polemical interventions the concept of proposition into speech act theory, in the then burgeoning new discipline called ‘cognitive Searle put Austin’s philosophical reflections back on science’. As regards the second research direction the beaten track of analytic philosophy. He also pre- announced in EM, FIL did fulfill the promise. But the pared the grounds for locating certain universals in scholarly community has been rather critical of the speech acts across the multitude of world’s languages, line of inquiry announced in FIL and, to judge from a quest that Austin would in all likelihood have dis- Searle’s own lack of enthusiasm in pursuing it, the missed as premature at best and a wild goose chase at program no longer seems to enjoy the support of the worst. For Searle, illocutionary acts had the form ‘F Berkeley philosopher (although his erstwhile collabo- (p)’ as their canonical structure––where ‘F’ denoted an rator Vanderveken has single-handedly carried the Illocutionary Force Indicating Device (IFID) and the torch since then). ‘p’ represented the ‘propositional content’ of the act in Scholars are also divided in their full support for the question. As many scholars have pointed out, Searle turn marked by Intentionality, especially insofar as it is gave a distinctive Platonic twist to ideas he had origi- meant to be a complement to his earlier work on speech 937 SEARLE, JOHN R. acts. Many resent what they see as an exaggerated men- Oxford University (1952–1959), where he took his talist swing in conceptualizing speech acts, accompa- B.A. First Class Honours (1955), M.A. (1959) and nied by Searle’s interest in looking for a universal base D.Phil. (1959). He was a Lecturer in Philosophy at to what many see as acts tethered to specific cultures. Christ Church (Oxford) (1957–1959). In 1969, he pub- Some, notably certain scholars with an ethnomethod- lished Speech acts: An essay in the philosophy of lan- ological orientation, regard Searle’s mentalism as a step guage. The same year, he took up a teaching position in the wrong direction away from Austin’s more cultur- at UC Berkeley. Searle was elected for Distinguished ally sensitive approach to language. Teaching Award by UC Berkeley (1999) and was From the early 1980s onward, Searle has been very awarded Honorary Degrees by the Universities of active in the philosophy of mind. He has been a tire- Wisconsin (1994) and Adelphi (1993), among others. less critic of those cognitive scientists who have He has also been a visiting professor at many univer- equated the working of a human brain with that of a sities across the five continents. digital computer. In the philosophy of mind, Searle is also known for his work on the problem of conscious- References ness. His famous Chinese Room Argument, put for- Burkhardt, A. (ed.) 1990. Speech acts, meaning and intentions. ward in 1980, according to which a mere computer Berlin: Walter de Gruyter. model of associating symbols is no guarantee of true Lepore, E., and R. Van Gulik (eds.) 1991. John Searle and his knowledge of the language in question (say, Chinese), critics. Oxford. Blackwell. has been discussed over and over again in the literature Love, N. 2000. On Searle on language. Language & and has influenced discussion in the area ever since. Communication 19. Parret, H., and J. Vershueren (eds.) 1992. (On) Searle on con- Besides his intensive work in academic philosophy, versation. Amsterdam: John Benjamins Co. Searle has been an active contributor to periodicals Rajagopalan, K. 2000. On Searle [on Austin] on meaning. such as the New York Review of Books, where he has Language & Communication 20. published scores of reviews and write-ups sharply crit- Searle, John R. Speech acts: an essay in the philosophy of lan- ical of such contemporary trends as poststructuralism, guage. New York: Cambridge University Press. 1969. ———. 1971. The campus war, New York: The World pragmatism, and deconstruction. He has also taken Publishing Co. highly polemical positions on student politics, and ———. 1979. Expression and meaning: studies in the theory of politicization of life on campuses. speech acts. New York: Cambridge University Press. ———. 1984. Minds, brains and science. London: Penguin Books. Biography ———. 1983. Intentionality: an essay in the philosophy of mind. New York: Cambridge University Press. John R. Searle has been Mills Professor of the ———. 1992. The rediscovery of the mind. Cambridge, MA: Philosophy of Mind and Language at the University of MIT Press. California at Berkeley since 1959. He was born in ———. 1995. The construction of social reality. London: Penguin. Denver, Colorado (USA) in 1932. His father was an ———. 1997. The mystery of consciousness. New York: New AT&T executive and his mother a medical doctor. As York Review Press. a child, he moved to New York, then to New Jersey, ———. 1998. Mind, language and society, philosophy in the and finally to Wisconsin, where he graduated from real world. high school. He began his university work at the Tsohatzidis, S.L. (ed.) 1994. Foundations of speech act theory. London: Routledge. University of Wisconsin in 1947, where he stayed till KANAVILLIL RAJAGOPALAN 1952. At 19, as a Rhodes Scholar, he went to England and studied under P.F. Strawson and J.L. Austin at See also Austin, John Langshaw; Speech Acts Second Language: Acquisition It has been said that multilingual human brains are major insight from the field of second-language acqui- more common than monolingual ones; most of the sition (SLA) has been that the grammars of second- people on this planet know more than one language. A language learners are systematic and rule-governed. 938  SECOND LANGUAGE: ACQUISITION These so-called ‘interlanguage’ grammars contain ele- Therefore, this is one possible source of error, and ments from the first language (L1)—e.g. German final hence one way in which people can sound nonnative- devoicing or French verb raising—and elements from like; they can get the stress placement wrong. This the second language (L2)—e.g. Canadian vowel rais- would be a phonological problem. They could, how- ing or Spanish null subjects. A learner’s production ever, put the stress on the right syllable but still not (even if it includes errors) is the product of the mental sound like a native speaker, Maybe in their L1, they representation of the grammar. There has been debate indicate stress by loudness or vowel length, whereas in as to whether interlanguage grammars are subject to English the main means of indicating stress is by a the same constraints as the grammars of primary lan- pitch change. This would be an example of a phonetic guages—the ‘Access to UG’ debate. Related to this is marker of their accent. The question of why some the question of whether L2 learners can acquire a lin- sounds are harder to learn than others has been guistic structure (e.g. gender, tone, [strident]) if that addressed from typological (Eckman), phonological structure is completely absent from the L1. (Brown), and phonetic perspectives (Flege). A basic insight from cognitive theory that helps us Languages also vary in terms of their syntactic to understand SLA involves knowledge vs. skill. The structures. In English, you need to have a subject for knowledge that we store in our heads is a relatively your sentences. You can say ‘She speaks French’, but stable trait. You either know the word ‘cat’ or you do not ‘*Speaks French’. In languages like Spanish, not. You either know that the sentence, ‘They is inable though, you can drop your pronominal subject and say to speaked French’ is ungrammatical or you do not. either ‘Yo hablo español’ or ‘Hablo español’. Hence, Clearly, you have to acquire knowledge of your L2, Spanish speakers of English are going to have to learn but you also have to acquire skills. You have to be able that they cannot drop their subjects in English. L2 to comprehend fast speech, or carry on a conversation. learners will also have to acquire the grammatical Proficiency in a second language is a complex con- structures necessary for making questions, forming struct that includes a range of knowledge (from passives, or embedding clauses (just to name a few). grapheme to phoneme to sentence to text) and a range They will have to learn to interpret quantifier scope of abilities (from politeness routines to appropriate and other subtle semantic properties. These particular register). Can adults attain nativelike proficiency in a structures can be used to identify two main approach- second language? Often, this question is investigated es to explanation in SLA. under the rubric of the critical period hypothesis. There are some who argue that structures that are There are undeniably age-related effects in L2 learn- common (or natural) in the world’s languages will be ing; people who start acquiring their L2 early are less easy to learn (regardless of your L1). For example, likely to have an accent. However, as we have seen, every language in the world that has null subjects also accent is a small part of L2 proficiency. Is there some allows sentences with pronominal subjects. The pres- point after which nativelike proficiency is impossible? ence of null subjects implies the presence of subjects. The answer to this question depends on which ele- There are no languages that allow only null subjects. In ments of proficiency one is looking at. White and this framework, then, it is assumed that null subjects are Genesee (1996) have argued that nonnative speakers more marked than pronominal subjects. Furthermore, it who started their L2 learning later in their lives can is assumed that structures that are more marked will be evidence grammatical knowledge and performance more difficult to acquire (Eckman 1991). that is statistically indistinguishable from native Generative linguists have also been interested in speakers. Bongaerts (1999) has shown this for nonna- explaining L2 acquisition. Consider the above exam- tive accents as well. ples. Linguists might invoke learnability issues to One of the most recognizable characteristics of explain these facts. An English speaker trying to learn adult SLA is that a great many of the people who Spanish will hear Spanish sentences that have null attempt it do not sound like native speakers of the lan- subjects; there will be positive evidence in the linguis- guage; they may have an accent. This has led some to tic environment that the learner’s current grammar (or argue that SLA is fundamentally different from first- hypothesis) is wrong, and they must change it. The language acquisition in that the end point (or ultimate Spanish speaker trying to learn English, on the other attainment) is so divergent. At the sound level, L2 hand, will begin the process with a grammar that learners can sound nonnativelike for two reasons. Let allows both null and overt subjects. When listening to us take an example from stress. Many languages mark the English sentences of the linguistic environment, certain syllables as prominent by stressing them. The they will never hear any sentences that are inconsistent second syllable in ‘banána’ is more prominent than the with their grammar. There is no reason for them to first or third syllables. Languages vary in where they change. It has been argued that this may explain some put the stress, so it is something that has to be learned. Directionality of Difficulty effects observed in SLA. 939 SECOND LANGUAGE: ACQUISITION English speakers have little difficulty in learning to Brown, C. 2000. The interrelationship between speech percep- drop their pronouns in Spanish, while Spanish speak- tion and phonological acquisition from infant to adult. In Archibald, pp. 4–63. ers can have trouble learning to always provide a Eckman, F. 1991. The structural conformity hypothesis and the pronominal subject in English. acquisition of consonant clusters in the interlangauge of ESL learners. Studies in Second Language Acquisition 13(1). 23–42. References Flege, J. E. 1992. Speech learning in a second language. Archibald, J. (ed.) 2000. Second language acquisition and lin- Phonological development, ed. by C. Ferguson, L. Menn, guistic theory. Oxford: Blackwell. and C. Stoel-Gammon, 565–604. Timonium: York Press. Bley-Vroman, R. 1989. What is the logical problem of foreign Flynn, S., G. Marthohardjono, and W. O’Neil. 1998. The gen- language learning? Linguistic perspectives on second lan- erative study of second language acquisition. Mahwah, NJ: guage acquisition, ed. by S. Gass and J. Schachter, 41–68. Lawrence Erbaum and Associates. Cambrdige: Cambridge University Press. White, L. 1989. Universal grammar and second language acqui- Bongaerts, T. 1999. Ultimate attainment in L2 pronunciation: sition. Amsterdam: John Benjamins. the case of very advanced late L2 learners. Second language White, L., and F. Genesee. 1996. How native is near-native? acquisition and the critical period hypothesis, ed. by D. The issue of ultimate attainment in adult second language Birdsond, 133–159. Mahwah, NJ: Lawrence Erlbaum and acquisition. Second Language Research 12(3). 233–265. Associates. JOHN ARCHIBALD Second Language: Learning The ability to learn another communication system, I work at home). According to this view, UG is a the- subsequent to a first, acquired during infancy and early ory about the initial state of language development in childhood, is unique to the human species. In their use that it specifies the kind of implicit, innate knowledge of second languages, humans easily outperform even the child brings to the task of acquiring a first lan- those species endowed with highly elaborate commu- guage, preceding experience. Moreover, UG con- nication systems, among them apes trained to mimic strains the developing grammatical systems so that at human language. Yet, the question of whether it is pos- every point of development, children’s immature sible to learn a second language just like the first, i.e. grammars UG-conform and are thus ‘possible human to proceed through similar developmental phases and grammars’. Consequently, L1 acquisition consists of attain the same type of knowledge and skills, is con- both learning from experience (e.g. lexical learning, troversial. Answers to this and related questions prom- learning of language-specific properties) and the acti- ise to shed light on the nature of the human language vation (‘triggering’) of principles of UG. When lan- capacity, which may explain why much of the research guages vary, children are said to set parameters on one on second-language learning addresses problems that of the given options (‘values’), based on their experi- involve comparisons between first- and second-lan- ence with target input. Children growing up with two guage acquisition. or more languages can set a parameter on different First languages (L1) are generally understood to be options, if required by the target systems. Bilingual the end products of developments guided by the acquisition during childhood is thus an instance of human language faculty. The core properties of its simultaneous acquisition of two ‘first’ languages. grammatical components are captured in terms of Second languages (L2), on the other hand, result principles of Universal Grammar (UG). Some of these from learning processes initiated once a first language principles apply invariantly to all languages, whereas has already been acquired, either fully, in the case of others are ‘parameterized’, i.e. they offer two (or adult L2, or partially, in the case of child L2 acquisi- more) options (‘parameter values’) in which languages tion. Distinguishing between various acquisitional may differ. An invariant principle specifies, for exam- types implies that substantive differences exist between ple, that sentences need to contain a subject position; L1 and L2, and less important ones between child and a parameterized option allows for this position to be adult L2. Although much research supports this view, lexically empty (e.g. Spanish––trabajo en casa) or it the issue is far from settled. Particularly, the critical needs to be lexically filled (e.g. its English equivalent age range separating L1 from L2 is a matter of much 940  SECOND LANGUAGE: LEARNING controversy; it is clearly not the same for phonological, acquisition proceeds through an ordered sequence of syntactic, and morphological properties of the target stages, it is tempting to conclude that both L1 and L2 language, with the ability to acquire a native phonolo- are guided by the same underlying principles. gy fading out first, possibly at around age 6. A conser- However, differences also manifest themselves with vative hypothesis consists in classifying as instances of respect to the developmental sequences. The mor- ‘L1’ those cases in which the child has been exposed to pheme order studies showed that the sequences are not one or more languages before age 3, as ‘child L2’ cases exactly the same in L1 and in L2 acquisition, a finding in which the onset of learning of one or more addition- confirmed by subsequent research for a broad range of al languages lies between the age range 5 and 8 years grammatical phenomena. In sum, although both types of age, and as ‘adult L2’ those cases in which the onset of acquisition exhibit a course of development that is of learning lies after 10. Further research is needed in uniform across individuals, these invariant aspects are order to fill the gaps in the given age ranges and to not the same in L1 and L2 acquisition, thus raising the decide whether the critical distinction is indeed the one question of whether the underlying principles and between L1 and child L2 acquisition. mechanisms are identical. Descriptively, L1 and L2 acquisition differ in In L1 acquisition, developmental sequences are important aspects, namely in the course of acquisition commonly explained through UG, i.e. the succession as well as in the final state attained by learners. As for of L1 developmental phases can be understood as a explanation, the question is whether L2 acquisition is reflection of the fact that children increasingly gain guided by the human language faculty and, conse- more complete access to, or become able to use, the quently, whether it eventually leads to a kind of knowl- principles of UG and set parameters for the values edge of the target language that is identical to or at required by the target grammar. Prerequisites for this least resembles that of native speakers of the language. development are, most importantly, a growing lexicon This amounts to asking to what extent L2 acquisition and the ability to deal with longer and more complex is determined by principles of UG. If the approxima- constructions, as a result of a more powerful working tive systems underlying L2 learners’ language use memory and increasing processing capacities. should turn out not to be constrained by UG principles, This kind of explanation cannot simply be carried it would be questionable to refer to them as to ‘gram- over to L2 acquisition; older learners can rely on fully mars’, in the sense defined for L1. Second languages developed working memories and on mature language might then possibly be defined not as ‘natural lan- processing systems. While explaining L1–L2 differ- guages’, contrary to our initial assumption. ences, it also suggests that adult L2 learners should In at least one regard, L2 acquisition does indeed outperform child L2 and L1 learners––hardly a satis- resemble L1 development: one finds strictly ordered fying result in view of the fact that L1 development in developmental sequences. Many, although not all, nonpathological cases leads to remarkably uniform grammatical characteristics of a language emerge in a results (despite considerable differences in learning fixed order across learners, independently of varying environments), whereas this is only true, at best, for a learning contexts, individual properties of the learners, small group of L2 learners. A possible solution for this or structural features of their first languages. In the problem is that mechanisms of language use play dif- 1970s, the ‘morpheme order’ studies showed that cer- ferent roles in L1 and L2 acquisition. Whereas in all tain grammatical elements like verb endings (-s, -ed, types of acquisition, they influence the choice of con- -ing, etc) tend to appear in the same order in the speech structions, resulting in preferences of use for particu- of different child and adult L2 learners of English. lar forms, in L2, they might determine the internal Subsequent research analyzing a larger variety of order of acquisitional sequences. This proposal is grammatical phenomena (particularly word order) in a based on the insight that processing of linguistic forms number of languages revealed that structural properties and constructions requires a certain cognitive capacity. of the target languages emerge crossindividually in an Given that the total amount is limited, learners need to invariant order, although the later ones do not depend, keep low the cognitive costs required by formal in principle, on the earlier ones. Moreover, the order of aspects of language in order to deal with problems of acquisition and use found in naturalistic L2 acquisition content and with communicative tasks. Consequently, proved to be difficult to manipulate by means of for- they first acquire those constructions requiring the eign language instruction. It seems, thus, that there least amount of processing capacity. If correct, exists a hidden logic by which learners of a given tar- processability helps to determine the underlying logic get language are guided. A major goal of L2 research of L2 acquisitional sequences. In L1, they follow is, therefore, to uncover the underlying logic, for this essentially a grammatical agenda. should be crucial in deciding if second languages are Assuming that the explanation of L2 sequences in indeed natural languages. In view of the fact that L2 terms of processability of linguistic forms is correct 941 SECOND LANGUAGE: LEARNING does not entail that the human language faculty has no principles of UG are operative in L2 acquisition, where- role to play in L2 acquisition. A plausible scenario pre- as parameter values not instantiated in the L1 grammar dicts that the various cognitive resources compete, and are not available any more, and the respective proper- that it is due to the fact that some of them, e.g. memory ties of the target grammars need to be learned empiri- and processing, are more developed that they interfere cally. Although some evidence supports the latter with the grammatical module in a way that is not possi- hypothesis, the controversy is far from settled and this ble during early childhood in L1 development. The issue is likely to occupy L2 research in the future. well-foundedness of this competition hypothesis is dif- A further important source of knowledge, in addi- ficult to assess, however, since not much empirical work tion to UG and transfer, is the learners’ linguistic envi- has been dedicated to its scrutiny. Instead, much ronment, i.e. input and feedback offered by their research since the mid-1980s has tried to find empirical interlocutors. Their fully developed cognitive capaci- evidence to show that UG is still accessible in L2 acqui- ties and communicative skills may be expected to sition. Further possible sources of knowledge for L2 enable them to make better use of this source than L1 learners are their respective first and other previously children. Numerous studies indeed confirm that native learned languages. To the extent that this type of knowl- interlocutors tend to modify their language use when edge interferes with what is offered by UG, this might addressing L2 learners who, themselves, can actively contribute to an explanation of L1–L2 differences. engage in negotiations to provide them with a possibly The contribution of principles of UG, on the one more comprehensible input, corrections, etc. All this hand, and transfer of previously acquired knowledge, happens in naturalistic acquisition, which has been the on the other, can be illustrated by considering the initial object of the preceding remarks since there is no evi- state of the learning process. In L1, the initial state is dence that acquisitional processes differ significantly determined by the human language faculty and thus by in tutored learning. But addressee-adapted speech, UG. As for L2, there is broad consensus that learners, especially rejection and correction (negative feedback) at the onset of acquisition, make use of knowledge of target-deviant learner utterances, is of special sources not available to the L1 child, for example, importance in classroom learning. The question, how- grammatical knowledge or mechanisms of language ever, of whether these and other types of input modifi- use previously acquired. The exact nature and amount cations have substantial and long-term effects for the of transfer, however, is a particularly controversial course and the result of learning processes is difficult issue. In the vast amount of publications dedicated to to answer. From all we know, developmental this topic over the last 50 years of research, radically sequences are not altered, but the rate of acquisition different hypotheses have been defended, from ‘full can be speeded up; moreover, certain acquisitional transfer’ to attributing a marginal role to syntactic and tasks (phonological and lexical learning) seem to be morphological transfer. It is accepted that previous lin- easier to influence in this way than others (grammati- guistic experience does influence L2 acquisition, result- cal morphology and syntax). ing in a different approach to the learning task by L2 as A characteristic feature of L2 learning is its compared to L1 learners. A frequently postulated, interindividual variability. In spite of the fact that although contested, hypothesis claims that parameter one finds invariable developmental sequences, individ- settings are transferred from the L1 grammar to the L2 uals and groups of learners differ considerably. This competence. Consequently, learners need to make refers not only to acquisition rates, use of transfer, etc. appropriate changes in order to attain target-conform- but, most significantly, also to ultimate success. Un- ing knowledge of their L2 system, if source and target doubtedly, some learners acquire the knowledge and grammars differ with respect to particular parameter skills enabling them to perform in a manner that settings. In principle, this can be achieved either by makes their language production indistinguishable inductive learning or by ‘resetting’ the parameters in from that of native speakers. But we have no satisfy- question. Parameter ‘resetting’, however, presupposes ing answer, as yet, to the question of why only a few that learners continue to have access to the various learners achieve nativelike grammatical knowledge. A parameter values provided by UG, and this issue there- possible solution is that differences across learners are fore represents the core problem in the debate about due to sociopsychological factors, most notably their continuous access to UG. Whereas some researchers attitudes and motivation toward the learning task, the claim that L2 learners have ‘full access’ to UG princi- target language, etc. But although this is certainly a ples and parameters, others argue that direct access is plausible hypothesis, evidence demonstrating a cause- ruled out for principled reasons and that UG shapes the and-effect relationship between such factors and par- L2 system merely via the L1 grammar. An intermediate ticular types of learning is scarce. position is defended by those who favor the idea of par- In conclusion, second-language learning, whether in tial access to UG, meaning that only nonparameterized a naturalistic or in a classroom setting, proceeds 942  SECOND LANGUAGE: TEACHING through acquisitional sequences, largely independent language acquisition, ed. by Susan M. Gass and Jacquelyn of individual factors or contextual influences. These Schachter, 41–68. Cambridge: Cambridge University Press. Carroll, Susanne E. 2000. Input and evidence: the raw material sequences are not identical, however, to those found in of second language acquisition. Amsterdam: John L1 development. Moreover, L2 learners vary consider- Benjamins. ably in the kind of use they make of their linguistic Clahsen, Harald, and Pieter Muysken. 1986. The availability of knowledge and in their ultimate attainment. For the universal grammar to adult and child learners: a study of the time being, only partial and temporary answers are acquisition of German word order. Second Language Research 2(2). 93–119. possible to the questions of whether L2 acquisition is Dulay, Heidi C., Marina K. Burt, and Stephen Krashen. 1982. guided by UG principles and whether second lan- Language two. Oxford: Oxford University Press. guages are ‘natural languages’. It is not even clear Felix, Sascha W. 1987. Cognition and language growth. whether ‘perfect learners’ have acquired a competence Dordrecht: Foris. of the same nature as the grammatical knowledge of Long, Michael H. 1990. Maturational constraints on language development. Studies in Second Language Acquisition native speakers. Overall, the observable L1–L2 differ- 12(3). 251–85. ences suggest that the underlying systems exhibit qual- Meisel, Jürgen M., Harald Clahsen, and Manfred Pienemann. itative differences. This is supported by neurolinguistic 1981. On determining developmental stages in natural sec- evidence, which indicates that grammatical informa- ond language acquisition. Studies in Second Language tion from second languages is, at least partially, Acquisition 3(2). 109–35. Pienemann, Manfred. 1998. Language processing and second processed differently and in different areas of the brain, language development: processability theory. Amsterdam: when compared to first languages. One may thus spec- John Benjamins. ulate that L2 knowledge should be characterized as a Ritchie, William C., and Tej K. Bhatia (eds.) 1996. Handbook hybrid system, combining grammatical principles and of second language acquisition, San Diego, CA: Academic mechanisms constrained by the human language facul- Press. Schwartz, Bonnie D., and Rex A. Sprouse. 1996. L2 cognitive ty with others learned by means of general problem- states and the full transfer/full access model. Second solving capacities of the cognitive system. Language Research 12(1). 40–72. White, Lydia. 1989. Universal grammar and second language acquisition. Amsterdam: John Benjamins. References Wode, Henning. 1981. Learning a Second Language: An Bley-Vroman, Robert. 1989. What is the logical problem of for- Integrated View of Language Acquisition. Tübingen: Narr. eign language Learning? Linguistic perspectives on second JÜRGEN M. MEISEL Second Language: Teaching Second-language teaching (SLT) involves teaching one Key factors include the learners’ age and proficiency in language to speakers of another language. A distinction the target language; their linguistic, cultural, and edu- is often made between SLT and foreign-language teach- cational experience; the time and money available for ing (FLT). The latter refers to teaching a language not instruction and self-study and associated learning widely used in the community in which the language is resources; and their reasons for learning the language. being learned. For example, teaching English to stu- Some of these purposes may be quite general. Others dents in Italy or Japan is often termed FLT, whereas may be very specific: for example, airline pilots need teaching English to learners in the United States and in English for occupational purposes, British students countries such as Nigeria or India, where English is an wishing to attend university in Paris may study French official language, is usually referred to as SLT. for academic purposes, or Japanese engineers may For both SLT and FLT, several issues have to be need to read scientific and technological reports written addressed. These include the purpose for which the in English. The analysis of backgrounds, needs, and target language is taught (aim), what is taught (syl- resources facilitates decisions about the composition of labus), how the content is taught (methodology), and classes, the length and frequency of instruction, the how learning is assessed (evaluation). type of teacher deemed most suitable, and the syllabus. Consideration of the learners’ background is very Most SLT syllabuses in Britain and the United important when considering the aim of the program. States 40 years ago were structural in orientation––that 943 SECOND LANGUAGE: TEACHING is, they were based on the formal elements of phonolo- may lead to error. By contrast, learners in a typical CLT gy, syntax, and morphology in the language. Learners classroom are expected to be mentally proactive in were guided to master the formal content of the syl- forming and testing their own hypotheses about lan- labus at one level of difficulty before moving on to the guage and socially active in using the language with next. Many SLT programs are still designed in this fellow learners. The audiolingual teacher models the way. Since the 1970s, however, a wider view of the language input, controls the direction and pace of prac- authentic use of language and increasing attention to tice, judges the accuracy of the learners’ language the needs of learners have led to the adoption of alter- behavior, and corrects errors as they occur. The CLT native approaches to syllabus design––especially teacher is expected to participate in, rather than domi- regarding the teaching of English. Predominant among nate, the process of learning––and to exemplify a these is communicative language teaching (CLT), model communicator, to determine and respond to the which sees the communication of information and culture and needs of the learners, to facilitate group ideas as the primary goal of second-language learning. learning (for example, by identifying and enhancing The emphasis is less on what learners know about the the students’ learning styles and strategies), and to target language at the sentence level (linguistic compe- organize a variety of resources, among which interac- tence) and more on what they can do with the language tive computer-assisted language learning is becoming in longer stretches of spoken or written discourse increasingly salient. (communicative competence). Instead of a linear pro- The emphasis on communication has led to a shift gression of items presented in order of structural diffi- in how to evaluate second-language learning. culty, language points are sequenced in terms of their Knowledge of the discrete phonological, syntactical, perceived usefulness, and they are integrated with other and lexical contents of a structural syllabus may be elements and constantly revisited. In the 1980s, there reliably measured by test formats such as multiple- were ‘strong’ and ‘weak’ CLT syllabuses; proponents choice questions, gap-filling tasks, sentence comple- of the strongest versions argued that the explicit teach- tion, and structural transformation exercises. However, ing of grammatical structures hinders the acquisition the validity of such formats to assess communicative and communication of meaning. Weaker versions of ability is questioned. Rather than measuring knowl- CLT incorporated structural elements within a broader edge in terms of pass/fail marks, the focus is now on framework of communicative functions, language skill providing descriptive profiles on the basis of criteria- development, and interactional learning strategies. referenced performance in authentic oral and written More recently, there has been a tendency to reconsider interaction. Also, based on the assumption that candi- how to focus on form in communicative syllabuses. dates may operate at differential levels of productive The emergence of alternative approaches to syllabus or receptive competence, there is now a tendency to design led to a review of how second languages should evaluate language skills separately, sometimes by be taught. The methodology most often associated with portfolios of work and computer-adaptive assessment structural syllabuses––the audiolingual method––is rather than pencil-and-paper tests. The contexts, mate- underpinned by a behaviorist view of learning. rials, and tasks used to evaluate oral or written com- Typically, this involves the presentation of paradigm munication are increasingly designed to reflect the grammatical structures, followed by repetition and drill specific purposes of the learners. with a view to the production of error-free sentences: practice makes perfect. The written language is used References largely to reinforce or consolidate language learned orally. In CLT, by contrast, the emphasis is on learning Brown, H. Douglas. 1980. Principles of language learning and teaching. Englewood Cliffs, NJ: Prentice-Hall Regents; 4th by doing––on communicating fluently rather than edition, 2000. accurately, and appropriately in terms of the social con- Cohen, Andrew D. 1994. Assessing language ability in the text in which the language is used. The four language classroom, Boston, MA: Heinle & Heinle. skills of speaking, listening, reading, and writing are ––––––. 1998. Strategies in learning and using a second lan- interwoven in meaningful classroom activities. The guage. London and New York: Longman. Harmer, Jeremy. 1998. How to teach English: an introduction to topics, contexts, and interactional patterns adopted in the practice of English language teaching. Harlow: Longman. CLT classrooms reflect the learners’ backgrounds and Hinkel, Eli (ed.) 1998. Culture in second language teaching and learning needs. The roles of learner and teacher are cor- learning. Cambridge: Cambridge University Press. respondingly different. In audiolingualism, learners are Levy, Michael. 1997. Computer-assisted language learning: reactive: they listen to the input and respond by repeat- context and conceptualization. Oxford: Clarendon Press. Odlin, Terence (ed.) 1994. Perspectives on pedagogical gram- ing, manipulating, transforming, and memorizing the mar. Cambridge: Cambridge University Press. linguistic data under the direction of the teacher. They Paul, Peter. 1993. Linguistics for language teaching. Melbourne: are not encouraged to initiate interactions, because this Macmillan. 944  SEMANTIC AND DISCOURSE TYPOLOGY Richards, Jack C., and Theodore R. Rodgers. 2001. Approaches Tomlinson, Brian (ed.) 1998. Materials development in and methods in language teaching, 2nd edition. Cambridge: language teaching. Cambridge: Cambridge University Cambridge University Press. Press. Richards, Jack C., and Richard Schmidt. 2002. Longman dic- van Lier, Leo. 1996. Interaction in the language curriculum. tionary of language teaching and applied linguistics, 3rd edi- London and New York: Longman. tion. Harlow: Pearson Education Ltd. ROGER BARNARD Semantic and Discourse Typology Semantic typology deals with the different types of The Longman lexicon of contemporary English semantic categories that can be found in natural lan- uses the following semantic fields, which are further guages, whereas discourse typology deals with differ- subdivided: ent types of discourse. A very appealing approach to the classification of semantic categories that can be Life and living things found in the lexicon is Anna Wierzbicka’s proposal, The body: its functions and welfare which is based on the assumption of the existence of a People and the family closed set of semantic primitives that can be used for Buildings, houses, the home, clothes, belongings, defining any concept in any language. According to and personal care this proposal, the meaning of a word does not depend Food, drink, and farming on the meaning of other words, but rather on a config- Feelings, emotions, attitudes, and sensations uration of the semantic primitives. The list she pro- Thought and communication, language and grammar poses is the following: Substances, materials, objects, and equipment Arts and crafts, science and technology, industry Substantives: I, you, someone/person, something/ and education thing, body Numbers, measurement, money, and commerce Determiners: this, the same, other Entertainment, sports, and games Quantifiers: one, two, some, all, many/much Space and time Evaluators: good, bad Movement, location, travel, and transport Descriptors: big, small General and abstract terms Mental predicates: think, know, want, feel, see, hear Speech: say, word, true These and other groupings can be proposed for con- Actions, events, motion: do, happen, move, touch tent vocabulary, in contrast to grammar words, which Existence and possession: there is, have are used only for expressing grammatical relations Life and death: live, die within the sentence. Another approach to a semantic Time: when/time, now, before, after, a long time, a classification takes into account the semantics of the short time, for some time, moment different elements that make up the sentence. It is Space: where/place, here, above, below, far, near, assumed that in any sentence, there is a predicate (the side, inside verb) and several arguments that take up semantic Logical concepts: not, maybe, can, because, if roles. These semantic roles are verb specific (giver, Intensifier, augmentor: very, more thinker, hearer, etc.), but they can be grouped Taxonomy, partonomy: kind of, part of into more general relations like Agent (entity that per- Similarity: like forms an action), Experiencer (entity that experiences something, as a ‘cognizer’, ‘perceiver’, or ‘emoter’), Other classifications are based not on the existence Recipient (entity that receives something), Stimulus of semantic primitives that can be combined to form (entity that is the object of experience), Theme (entity definitions, but rather on the existence of semantic that is located or that undergoes a change of location), fields that group the different concepts that can and Patient (entity that undergoes a process or an be expressed through the vocabulary of a language. action initiated by another entity). These semantic 945 SEMANTIC AND DISCOURSE TYPOLOGY roles can be reduced to only two: Actor and point for a discourse classification would be to take Undergoer. Sentential semantic roles can be found in into account the different functions that any discourse any language; they only differ in the way in which they may have, which depend on the objectives. The objec- are expressed in the grammar of every particular lan- tive of discourse may be information (which is the guage. There exist many classifications of semantic basis of informative discourse), expression (the basis roles, and each grammatical theory has its own list, so of narrative discourse), and persuasion (the basis of that we may find other role names, like Source, Path, argumentative discourse). All three types normally Goal, Location, Object, etc. appear in combination, but they serve to give some In contrast to semantic typologies, discourse foundation to a more sophisticated classification of typologies are more concerned with broader units, discourse types. One of those classifications is related to discourse. ‘Discourse’ is a term that covers Werlich’s, in which there are five basic forms underly- many kinds of communication. According to the pos- ing all discourse types: descriptive, narrative, explana- sibilities available, we can establish different general tory, argumentative, and instructive. These basic forms types. A first distinction can be made between oral and are related to innate categorization possibilities that written discourse. Oral discourse is characterized by human beings have, and they are characteristically the direct contact made by the participants in the inter- expressed by different types of sentences. action and its immediacy. Oral communication takes These five basic forms have two different methods place ‘in real time’, that is, the message is decoded by of presentation: subjective, which shows the writer’s the receiver as soon as it is encoded into the oral mes- perspective, and objective, which is subject to verifi- sage by the speaker. In such a communicative situa- cation by the readers. This gives the following dis- tion, the message is also conveyed with the help of course types with subjective presentation: means other than language, such as hand gestures, impressionistic description (descriptive), report (nar- intonation, posture, etc. In contrast, written discourse rative), essay (explanatory), comment (argumenta- is not so immediate and there is no simultaneous par- tive), and instructions (instructive). If the method of ticipation of the writer and the reader in the commu- presentation is objective, then we have the following nicative situation. It may take a very long time since types: technical description (descriptive), news story the writer writes the message until the reader reads it. (narrative), explication (explanatory), argumentation Everything should be encoded into the text and may be (argumentative), and directions, rules, regulations, and subject to a much closer scrutiny than the oral mes- statutes (instructive). All of these types have charac- sage, which means that it requires a more careful pro- teristic sentence structures. A further subdivision of duction. Another distinction that should be taken into discourse types can be carried out according to the account is the one between everyday and literary lan- ‘channel’ involved (oral channel vs. written channel). guage. It is obvious that literary language involves a Another type of classification is based on the more careful discourse, which affects organization and discourse situation. Steger et al. distinguish six dis- the use of vocabulary and discourse markers. course types according to six types of discourse situa- However, this is still a basic kind of grouping, which tion: presentation, message, report, public debate, would be previous to a more detailed classification. conversation, and interview. These types are character- Closer to the semantic typologies presented are ized by certain features present in the discourse situa- some classifications that have been proposed for ana- tion. For some of them (presentation, message, report) lyzing the internal structure of discourse, such as a there is only one speaker, whereas for others (public classification of different semantic relations within debate, conversation, interview) there are multiple discourse. Discourse fragments can be related accord- speakers. The rank of the actors involved can be equal ing to temporal relations (chronological sequence and (in public debate and conversation) or unequal (in the temporal overlap), matching relations (contrast and rest). The theme can be predetermined (in all except comparison), cause–effect relations, truth-and-validity conversation) or not predetermined (in conversation). relations (statement-affirmation, statement-denial, Finally, the method of theme treatment can be descrip- denial-correction, concession-contraexpectation, etc.), tive (in message and report), argumentative (in presen- alternation relations (contrastive and supplementary), tation, public debate, and interview), or associative (in bonding relations (coupling, exemplification, excep- conversation). tion), paraphrase, amplification (specification, exem- plification), setting relations, etc. References As for the external characterization of discourse, Chafe, Wallace L. 1982. Integration and involvement in speak- there have been many different attempts to establish an ing, writing and oral literature. Spoken and written lan- exhaustive classification of discourse types, but all of guage. Exploring orality and literacy, ed. by Deborah them accomplish this task partially. A good starting Tannen. Norwood, NJ: Ablex. 946  SEMANTICS Crombie, Winifred. 1985. Process and relation in dis- Forschungshypothese. Jahrbuch Gesprochene Sprache, ed. course and language learning. Oxford: Oxford University by H. Moser. Düsseldorf: Schwann. Press. Van Valin, Robert D. 2001. An introduction to syntax, [Chapter Goddard, Cliff. 2001. Lexico-semantic universals: a critical 2, Grammatical relations.] Cambridge: Cambridge University overview. Linguistic Typology 5–1. Press. Renkema, Jan. 1993. Discourse studies, Amsterdam and Werlich, Egon. 1982. A text grammar of English. Heidelberg: Philadelphia: John Benjamins. Quelle and Meyer. Steger, Hugo, et al. 1974. Redekonstellation, Wierzbicka, Anna. 1996. Semantics: primes and universals. Redekonstellationstyp, Textexemplar, Textsorte im Rahmen Oxford: Oxford University Press. eines Sprachverhaltensmodells. Begründung einer CARLOS INCHAURRALDE Semantics Semantics is the study of the meaning of signs and is especially the case in contemporary semantic theo- representations, both mental and linguistic. The ulti- ries, which often invoke causal (e.g. Fodor 1990) and mate target of semantics is the construction of a gen- perceptual factors in order to explain the meaning of eral theory of meaning. symbols ––and not just indexical and iconic signs. Following a traditional taxonomy, we can distinguish From a philosophical point of view, symbols con- three different kinds of signs depending on the nature of stitute the fundamental category within a semantic the relationship between the sign form and its meaning: theory. This is the case, at least, within the so-called indexical, iconic, and symbolic signs. Indexical signs contemporary analytic tradition in philosophy. If we are those in which the sign or representation is causally specify the range of phenomena that a semantic theo- connected to its meaning. The redness of a face, for ry addresses following the scheme ‘A interprets B as instance, is an indexical sign of anger or irritation. representing C’, where ‘A’ is the interpreter, ‘B’ is an Certain skin spots are indexical signs of an illness such object, event, symbol or representation, and ‘C’ is the as measles. This connection is sometimes labeled, fol- meaning of B, we can establish some important theo- lowing Grice (1957), the natural meaning of a sign. In retical demarcations. The formal structure of B is stud- the case of icons, the relationship between the sign and ied by syntax. The relationship between B and its its meaning is based on a perceptual similarity. The map interpreter(s) is the domain of pragmatics. A semantic of a geographical area is thus an iconic sign of the terri- analysis would shed light on the nature and structure tory it represents. Symbols, finally, are signs whose of C as well as on the nature and structure of the rela- relationship with their meaning is established through a tionship between B and C. conventional rule. A red light symbolically, i.e. conven- Despite these basic differences, there are multiple tionally, represents the obligation to stop. Individual and complex relationships between these three differ- expressions and sentences of natural languages are par- ent disciplines. Thus, the array of meanings that C can adigmatic examples of this semantic category. Thus, the represent is constrained by the syntax of B within a sentence ‘snow is white’ is the symbol we use in particular linguistic or representational system. In English to represent the fact that snow is white. order to represent an event, for instance, the type of The appeal to this taxonomy implicitly involves the symbol should be a sentence, not a name. In this sense, idea that the different relationships that connect the logic, an essentially syntactic discipline, is sometimes meaning of a sign to its form should be explained fol- considered a part of semantics because it aims at lowing different theoretical principles. Thus, the explaining the structural properties of the expressions semantics of indexical signs would depend crucially of a language. on a theory of causal relationships; the semantics of If the boundary between syntax and semantics can icons would depend on the similarity principles that be fuzzy, the demarcation between semantics and determine perceptual structures; and the semantics of pragmatics is even more problematic. The most radical symbols would be characterized in terms of the princi- conception in this sense, namely a conception accord- ples that regulate and constrain the conventional rules ing to which the notion of meaning of a symbol can be that help individuate the symbols’ meaning. However, reduced to the conditions of its use (Wittgenstein these demarcations are not as clear as they seem. This 1953), denies that there is a distinction at all. 947 SEMANTICS These difficulties notwithstanding, there are indeed other kinds of expressions, such as declarative sen- certain questions that are the proper domain of seman- tences. In this case, the meaning (sense) of a sentence is tics. Among these questions, the most prominent is per- characterized as its truth conditions, i.e. as what the haps that of clarifying and determining the very nature world would be like if the sentence were true. The ref- of the notion of meaning. In principle, the meaning of erence of the sentence is its truth value. There are only an expression could be thought as just what the expres- two truth values in classical semantics: truth and falsity, sion stands for, i.e. its reference. This is a technical and establishing the truth value of a sentence, unlike sense of reference understood as a function that assigns, determining its meaning, is an empirical matter. The to the symbols of a language, different extralinguistic idea of the meaning of a sentence being its truth condi- entities taken from a particular domain of interpretation. tions has to be amended to cover nondeclarative sen- This definition coincides with the characterization of tences such as imperatives or questions, but it has been reference as a semantic value. Its only peculiarity con- a central notion in all contemporary semantic theories. sists in including––as an additional layer of complexi- Thus, Donald Davidson’s primary insight, and his major ty––the idea of a domain of interpretation. In the case of contribution to semantic theory, was to adapt a theory of natural languages––i.e. languages such as English or truth developed by the mathematician Alfred Tarski Spanish––this interpretation is already given and the (1944) to the problem of developing a theory of mean- task of a semantic theory is to specify the interpretation ing for natural languages. Tarski assumed the idea of already imposed by the way in which the language is meaning––the idea that a sentence could be the transla- used. However, in the case of nonnatural or artificial tion of another sentence––to get at the notion of truth. languages (e.g. the language of logic or mathematics), Davidson inverts Tarski’s scheme and uses the idea of we have to fix a nonempty domain of objects, if we truth as basic in order to develop a characterization of want the symbols of those languages to mean anything meaning in terms of truth conditions (Davidson 1984). at all, i.e. if we want to interpret those languages or––as Such a characterization of meaning also helps is sometimes said when talking about formal lan- establish what inferences can be legitimately drawn guages––if we want to have a model of those languages. from each sentence, since a valid inference or argu- Thus, this characterization of the notion of reference ment is one in which the truth of the conclusion fol- responds to the basic requirements of what is known, in lows from the truth of the premises. The notion of mathematics, as model theory (see e.g. Hodges 1993). truth––the core notion in logic––and the notion of Although this, in itself, is no more than a branch of meaning—the core notion in semantics––are thus inti- mathematics that studies the relationships between dif- mately related. In fact, mathematical logic has provid- ferent types of languages and different types of mathe- ed the basic theoretical tools for the development of matical structures, it has strong implications and some of the most interesting semantic theories of this applications in the territories of semantics, linguistics, century. In particular, the development of first-order and the philosophy of language. predicate logic by Gottlob Frege opened the door for The notion of reference, however, does not com- many applications of semantics within areas such as pletely exhaust the meaning of a symbol. Frege (1892) computer science and artificial intelligence. introduces an additional and even more fundamental Despite this initial success, contemporary semantic semantic concept to complement that of reference: the theorists tend to question the plausibility of applying the concept of sense. The need to invoke this new seman- model of first-order logic to natural languages. Natural tic dimension when trying to determine the meaning of languages have a much more complicated syntax than a symbol becomes clear whenever we address the the language of first-order logic: expressions are often issue of explaining why true identity statements ambiguous, and there are many kinds of sentences other involving different proper names (e.g. ‘Hesperus is than declarative sentences. A first attempt at providing Phosphorus’) which refer to the same object (Venus) an alternative account was developed by Montague differ from identity statements that involve the same (1974), who advocated the use of a more powerful logic name (e.g. ‘Hesperus is Hesperus’). If the meaning of (intensional logic) as a way of solving some of the prob- a proper name (e.g. ‘Hesperus’) were just its reference lems encountered within the original Fregean treatment. (i.e. the planet Venus), the two kinds of identity state- Montague’s proposal allows us to characterize the ments should have the same cognitive value. Since this meaning of sentences involving the idea of necessity or is not the case––since we might learn something we possibility and also of those sentences that express did not know when we learn that Hesperus is propositional attitudes such as belief or desire. Phosphorus––we have to admit that the meaning of a However, even Montague’s treatment remains heavily proper name cannot be just its reference. embedded in the formalist approach set up by the math- The same two concepts—sense and reference— have ematical view of semantics and it too has been called been used to analyze not just proper names but also into question by more recent semantic theories. 948  SEMIOTICS The characterization of the notion of sense, espe- properties as properties of some other (more primitive, cially when applied to common names, i.e. expressions less problematic) kind since semantic properties are such as gold, tiger, etc., has also been harshly criticized not commonly held to be part of the basic ontological over the last few decades. This notion was traditionally furniture of the world. Most theorists engaged in natu- characterized in terms of a set of necessary and suffi- ralization projects for semantics thus assume that the cient conditions that determine the reference of an program of naturalization demands a higher-to-lower, expression. The basic idea of the recent critics is that top-to-bottom, kind of explanatory strategy. They we can ascribe to the average speaker of a language believe, in other words, that the nonsemantic proper- knowledge of the meaning of an expression without ties on which semantic properties depend, belong to assuming thereby that the speaker knows the set of nec- what are intuitively lower levels of description than the essary and sufficient conditions that determine the ref- semantic level itself (contenders include biological erence of such an expression. The alternative proposal properties; see Millikan 1984). The reductionist flavor is to substitute the notion of stereotype for the idea of is unmistakable here. Achieving a clear understanding sufficient and necessary conditions. A stereotype is a of ‘naturalism’ as applied to semantics and deciding mental representation of a paradigmatic instance of an whether a naturalistic semantics is possible are proba- object or property with which the subject has had some bly two of the most important issues for contemporary kind of experience (Putnam 1975). philosophy of language and philosophy of mind. In contemporary semantics and philosophy of lan- guage, the notions of narrow and wide content have References come to substitute those of sense and reference. Davidson, Donald. 1984. Inquiries into truth and interpretation. Although intimately related to the original ones, the Oxford: Clarendon Press. new notions have a psychological slant which was not Fodor, Jerry. 1990. A theory of content and other essays. present in earlier versions of the same idea, and are Cambridge, MA: MIT Press. basically used in the context of specifying the meaning Frege, Gottlob. 1982. Über Sinn und Bedeutung. Zeitschrift für Philosophie und philosophische Kritik 100; Also in of a subject’s mental representations. Translations from the philosophical writings of Gottlob On the one hand, the narrow content (or meaning) Frege, ed. by Peter Geach and Max Black. Oxford: of a mental state is constituted by the intrinsic (i.e. Blackwell, 1960. context-independent) properties of the individual who Grice, Paul. 1957. Meaning. Philosophical Review 64. is in that mental state. Internalism is the point of view Hodges, Wilfrid. 1993. Model theory. Cambridge: Cambridge University Press. characterized by the thesis that such contents are indi- Millikan, Ruth. 1984. Language, thought, and other biological viduated without essential reference to the subject’s categories: new foundations for realism. Cambridge, MA: physical and social environment. On the other hand, MIT Press. the externalist claims that contents are individuated Montague, Richard. 1974. Formal philosophy: selected papers ‘widely’, i.e. by reference to the subject’s environ- of R. E. Montague, ed. by R.H. Thomason. New Haven: Yale University Press. mental or social context. Putnam, Hilary. 1975. The meaning of meaning. Minnesota To begin with, we said that the aim of semantics is studies in the philosophy of science, Vol. VII: language, the development of a theory of meaning. It is impor- mind and knowledge, ed. by Keith Gunderson, Minnesota: tant to stress now that one of the most important top- University of Minnesota Press. ics in this discipline is the study of the empirical Tarski, Alfred. 1944. The semantic conception of truth and the foundations of semantics. Philosophy and Phenomenological evidence that verifies––or falsifies––such a theory. Research 4. This empirical dimension brings semantics into the Wittgenstein, Ludwig. 1953. Philosophical investigations, domain of naturalism, i.e. the domain of empirical sci- transl. by Elisabeth Anscombe, Oxford: Blackwell. ence. Dominant naturalistic theories depict semantic JOSEFA TORIBIO Semiotics Semiotics is the science of signs. Even though the the modern world to the ideas of the Swiss linguist study of signs has been a constant in the history of Ferdinand de Saussure and the American philosophers thought, we owe the rise of semiotics as a discipline in Charles S. Peirce and Charles W. Morris. 949 SEMIOTICS In the tradition of Saussure, a sign is composed of signified. An iconic sign is one in which the signifier the signifier, i.e. the form that the sign takes, and the physically or perceptually resembles the signified (e.g. signified, i.e. the concept or object it represents. The a portrait or a sculpture). Indexical signs are those in signifier is characterized as the physical form of the which the signifier is physically or causally connected sign. It is important to keep in mind the distinction to the signified, such as smoke to fire, facial spots to between the sign vehicle, which is just the form in measles, etc. which the sign appears (spoken or written form, for When we talk about signs in this context, we are instance), and the sign itself, which includes both the thus referring not just to linguistic symbols. A sign is signifier and the signified. The signified has, within anything that stands for something else. Thus, semi- this tradition, not a physical but a psychological status. otics can be characterized as a theoretical approach to It is interpreted as some kind of mental construct, the such diverse phenomena as images, sounds, paintings, psychological impression that the signifier makes on rituals, gestures, etc., as well as languages. The pres- the interpreter of the sign. Given this characterization ence of the theoretical apparatus of linguistics is, how- of how signs represent, the idea of signs referring to ever, quite significant since linguistic symbols are still objects in the world is excluded from this particular considered paradigmatic within this discipline. In this view of semiotics. Signs do refer, but only to mental sense, Claude Lévi-Strauss, a contemporary anthro- concepts and psychological entities constructed by the pologist who sees anthropology as a branch of semi- interpreter. otics, claims that ‘language is the semiotic system par This mentalist understanding of the notion of refer- excellence; it cannot but signify, and exists only ence of a sign (the signified) seems to support an ide- through signification’ (Lévi-Strauss 1972: 48). alist philosophical interpretation of reality, i.e. an Semiotics enjoyed its golden era within the philosoph- interpretation according to which there are no ultimate ical movement known as Logical Positivism in the real objects and events in the world regardless of how 1930s. C.W. Morris’ Foundations of the theory of signs we construct them throughout the use of signs. It is probably provides the most complete characterization important to notice, though, that this rather psycholo- of this theory, presenting it as a science whose main gized view of the referent or signified of a sign is not aim is to provide the foundations for any other partic- shared by all semioticians. Charles Peirce’s proposal, ular science about signs. for instance, views the referent of a sign as something Having such a wide range of phenomena as its sub- with its own ontological status, something external to ject makes it hard to establish a precise demarcation the sign’s interpreter. Peirce’s model, by characteriz- between semiotics and the other disciplines with ing the referents of signs as something external to the which it shares some of the same objects of study, e.g. interpreter, as something with an independent ontolog- disciplines such as linguistics, logic, mathematics, ical weight, seems to suggest a more realist philo- esthetics, and anthropology—all of which address, in sophical interpretation of the world according to one way or another, the general phenomenon of signs. which reality is composed of objects and properties Nor is it obvious that the proper theoretical status of which stand on their own and independently of our semiotics is that of a science, i.e. that we should con- means of describing them or of knowing them via a sider semiotics as having the same theoretical status system of signs. as other empirical sciences, such as biology or chem- Saussure calls the relationship between the signifi- istry. Semiotics sometimes appears as a style of philo- er and the signified the signification. For Saussure, the sophical investigation, sometimes as a theory, and relationships between signifiers and their signifieds some other times as a methodology. In order to gain a are ontologically arbitrary, but once these relationships better understanding of this discipline, we shall focus have been established by the use of a community here on two issues: the domain of semiotics and its throughout history, such arbitrary connections cannot theoretical status. be changed. But regarding this issue of arbitrariness, Semiotic processes or events involve, according to not all semioticians agree. Most would nonetheless Charles S. Peirce, three factors: a sign, an object, and allow that there are degrees of arbitrariness in how the an interpreter. Depending on which factor gets isolated signifier and the signified relate to each other. Thus, for study, semiotics splits into three different branches: following Charles Peirce’s model, we can establish pure grammar (the study of signs themselves), logic three different kinds of signs depending on how the (the study of the object of the sign), and rhetoric (the signs vehicles relate to their referents. From the most study of the sign’s interpreter). But what characterizes to the least conventionally based relationships, we find semiotics and makes it different from the other disci- symbols, icons, and indexes. Symbols are purely con- plines is that the interrelation between these three fac- ventional, i.e. there is no perceptual or naturally rec- tors is absolutely essential for the individuation of ognizable relationship between the signifier and the semiotic phenomena. Charles W. Morris (1938, 1946) 950  SEMIOTICS likewise stresses the dynamic interrelation between syntax or semantics, we can ignore the role of the sign, denotation, and interpreter. From Morris’ point of interpreter. However, such a suggestion is highly con- view, no object, event, or particular action can be con- troversial, especially if we realize that the complex sidered, in itself, a sign, a denotation, or an interpreter. notion of the meaning of a symbol seems to include a The properties that characterize each of these entities variety of aspects related to the interpreter’s use of are relational properties. Such properties are acquired those symbols, thereby blurring the boundaries only through a semiotic process, i.e. a process through between semantics and pragmatics. Morris himself which something becomes a sign for someone (cf. downplayed this taxonomy in his later work Signs, Morris 1946: 353). Morris establishes four components language and behavior, where he introduces it just as in this process: the sign (vehicle), the denotation of the a general suggestion and only in the last chapter. The sign, the organism for which something is a sign (the problem of distinguishing pragmatics and semantics interpreter), and the effect the sign produces in the has been and still is one of the central problems in lin- interpreter. Among the multiple examples that Morris guistics and the philosophy of language. provides to illustrate the role played by each of those So far, we have been talking about the object of factors, perhaps the clearest is that of a traveler who, semiotics. But what can we say about its theoretical ready to go to a specific geographical area, receives the status? For some authors, semiotics ought to be placed letter of a friend describing that region. The traveler at the same level of philosophy of language (where then packs his luggage in accordance with the informa- ‘language’ includes nonverbal languages). Umberto tion provided by his friend in the letter. In that exam- Eco (1984) is perhaps the most fervent advocate of this ple, the letter is the sign, the geographic conditions are view. For him, the philosophical character of semiotics the denotation of the sign, the traveler’s preparations turns it into a discipline without any predictive power, regarding what to pack are the effect that the sign pro- and in which the object of study is also always medi- duces in the interpreter, and the traveler, of course, is ated by the general presuppositions of the theory itself the interpreter. The most important idea is that none of (Eco 1984: 10–3). However, this is not a widely shared those factors would belong to these categories were it view. Especially when we talk about specific semiotics not for the interrelationships among them (cf. Morris (e.g. narrative semiotics, folk semiotics, animal behav- 1938: 3–4). ior, and human institution semiotics, etc.), the descrip- Despite this theoretical interdependence, Morris, as tive slant of the discipline is stressed and, with it, its well as Peirce, concedes that we can isolate the rela- empirical character. Finally—and without considering tionships between such categories and define different these distinctions exhaustive—semiotics is sometimes branches within semiotics. Thus, if we are mainly considered more as a general methodology or as the interested in the relationships between the signs them- sum of different semiotic methods. From this point of selves, we focus upon the syntactic dimension of semi- view, the notion of sign is applied not only to linguis- otics. We focus on the semantic dimension if we are tic objects but also to objects such as buildings, rituals, mainly interested in the relationships between the landscapes, and star formations. Among the semiotic signs and their denotations and upon the pragmatic methods, we can distinguish at least three: formaliza- dimension if our focus is upon the relationships tion, linguistic analysis, and method of interpretation. between signs and their interpreters (Morris 1938: The method of interpretation—especially the for- 6–7). These partitions can, in principle, help establish malization of natural languages—stresses the syntac- clearer theoretical connections between semiotics and tic properties of signs and what is sometimes called other disciplines. Thus, linguistics and logic are about ‘operational meaning’: ‘operational’ because it is nor- the syntactic semiotic dimension. The object of mally characterized in terms of rules that specify the semantics is the relationship between signs and their admissible forms for manipulating signs. Once signs objects—the semantic dimension of semiotics. are stripped of their semantic and pragmatic proper- Pragmatics, finally, is the most problematic discipline ties, they constitute a purely syntactic skeleton, ready whenever we try to assign to it a specific semiotic to be dressed with different interpretations. dimension. The reason for this is that the study of the The method of linguistic analysis consists in the relationships between signs and their interpreter(s) reconstruction of the deep structure of the analyzed covers a complex field, which includes psychological, expression and in the enumeration of all their seman- biological, and sociological aspects of sign manipula- tic and pragmatic properties. The use of this method is tion. Psychoanalysis, the sociology of knowledge, and thus complementary to the method of formalization even some approaches within esthetics or anthropolo- and it allows us to find out what an expression or a gy all touch upon this pragmatic dimension of semi- particular piece of text really means and what infor- otics. Furthermore, the problem with this taxonomy is mation can be derived from its meaning. The use of that it seems to suggest that whenever we are doing this method lies at the heart of so-called ‘structural 951 SEMIOTICS linguistics’ (Saussure 1916). It has also been widely References applied to the sciences, paying special attention to the Bouissac, Paul (ed.) 1998. Encyclopedia of semiotics. Oxford: process of elaboration of particular scientific lan- Oxford University Press. guages and to the creation of new terms. Eco, Umberto. 1984. Semiotics and the philosophy of language. The method of interpretation consists fundamental- London: Macmillan Press. ly in the symbolic treatment of objects, events, or phe- Jackobson, Roman. 1980. The framework of language. Ann Arbor, MI: Slavic Pub. nomena that are not themselves signs (e.g. food, Lévi-Strauss, Claude. 1972. Structural anthropology. rituals, human faces, etc.). When we interpret them in Harmondsworth: Penguin. this way, we direct our attention beyond the object or Morris, Charles W. 1938. Foundations of the theory of signs. event itself. In so doing, objects, events, and phenom- Chicago: University of Chicago Press. ena that are not in principle symbolic at all become an Morris, Charles W. 1946. Signs, language, and behavior. New York: Prentice-Hall. instrumental part of our system of communication. Nöth, Winfried. 1990. Handbook of semiotics. Bloomington: This is probably the most familiar version of what is Indiana University Press. often understood by ‘semiotics’, namely, a treatment Peirce, Charles S. 1966. Selected writings, New York: Dover. that can be considered an extension of hermeneutics in Saussure, Ferdinand de. 1916. Cours de linguistique générale the sense of being applied to phenomena that are not (Course in General Linguistics). Paris: Payot and London: Fontana/Collins, 1974. intrinsically symbolic. This method is highly relevant Sebeok, Thomas (ed.) 1977. A perfusion of signs, for disciplines such as psychology or sociology since Bloomington: Indiana University Press. one characteristic type of explanation within these sci- JOSEFA TORIBIO ences consists in bringing to light intentions and motives that are only indirectly expressed in human See also Levi-Strauss, Claude; Peirce, Charles; behavior and human intellectual creations. Saussure, Ferdinand de Semitic Languages Four of the five literary Semitic languages that underlie 1934, an Indo-Europeanist, Louis H. Gray, essayed a great civilizations of Asia, Africa, and Europe––Arabic, description of Semitic that eschewed the root-and-pat- Aramaic, Ge‘ez or Ethiopic, and Hebrew––are covered tern model and observed verb roots of the form elsewhere, so here we concentrate on the fifth, CCVC––like many other languages––rather than the Akkadian; on the less well-known Semitic languages of unique CCC. (See Table 1. This view has recently been antiquity; and on what they have in common and how adopted independently by several scholars.) they relate to each other. Semitic itself is one of five or Typical of the Semitic languages is their way of six branches of the Afroasiatic phylum, the only one expressing a relationship between two nouns. Rather spoken principally in (southwest) Asia. than marking the possessor, as in the king’s peace, Semitic puts the possessed into the CONSTRUCT STATE, as in Akkadian šulum šarri, literally ‘peace-of king’; General Characteristics cf. šulmu ‘peace’, šarru ‘king’. These words also For a thousand years or more, linguists (beginning illustrate two of the three noun cases: nominative with the Arab grammarians) have taught that words in ending in -u, genitive (possessive or adnominal) end- Semitic languages are formed from ‘roots’ consisting ing in -i, and accusative (objective or adverbal) -a. of three consonants and no vowels at all, and only Nouns in Semitic have two genders: masculine or ‘modifications’ of the ‘root meaning’ are indicated by feminine. vowel ‘patterns’. But this seems to be an artifact of the The morphological typology of Semitic is inflect- writing systems used by the Arab (and Hebrew and ing; the syntactic typology is basically Verb–Subject– Syriac) grammarians, which notated only the conso- Object, even in those languages where the verb usual- nants, with the short vowels included only optionally ly comes at the end of a clause: Ethiopic, presumably and with incidental-looking marks. Moreover, it under the influence of Cushitic languages, and relates only to verbs, and not to ‘primary nouns’. In Akkadian, under the influence of Sumerian. 952  SEMITIC LANGUAGES TABLE 1 Consonant Correspondences, Prefix Conjugation, and Verb Stems1 Translation Akkadian Hebrew Syriac Ge‘ez Arabic Proto-Semitic I guard as.s.uru ’es.s.or et..t ur *’ns.r ’anz.uru ’a-nθ· ur We grind nit. ên nit. h.an net. h.an n.th.an nat. h.anu na-t. h.an You (m.sg.) are lame te¯gir tah.˘agor teh.gar cf. th.gl tah. juru ta-h.gur You (m.pl.) harvest te¯s.ida¯ ˘ s.ad ‘axe’) (ma‘a teh.s.du¯ n t‘ddu¯ tahC d.idu¯ na ta-‘/hC ð. id-u¯ You (f.sg.) seize ta¯ hC uz¯ι to¯h.˘az¯ι te(’)h.z¯ιn t’hC az¯ι ta’hC uðι¯ na ta-’hC uð -¯ι You (f.pl.) enter/set te¯ruba¯ (‘ereb ‘evening’) te‘rba¯n t‘raba¯ tag'rubna ta-g'rub-na He binds ¯ιsir ye’e˘sor ne(’)sor y’sr ya’siru ya-’sir They (m.) go illiku¯ yahalku¯ *nehlku¯n *yhlku¯ yahliku¯na ya-hlik-u¯ She places taš¯ιm tas'¯ιm¯ι ts¯ιm tι¯ m taš¯ιmu ta-s'yim They (f.) hunt tis.u¯ da¯ tas.o¯ dna¯ ts.u¯ du¯n *ts.u¯da¯ tas.u¯du¯na ta-s.wud-na To sing (D) zummuru zamme¯r mzamma¯ ru¯ zammro¯ tazm¯ιrun zmi/ur To cause to hear (Š) šušmû hašm¯ι˘a‘ mašma¯ ‘u¯ asm‘o¯ ’isma¯ ‘un šma‘ To be twisted (N) naptulu hippa¯ te¯ l — ( ptl) — ( ftl) infita¯ lun ptil To be sought for (Gt) pitqudu — ( pqd) metpqa¯du¯ tafaqdo¯ iftiqa¯ dun pqid To find one’s match (Dt) mutaššulu hitmašše¯ l metmatta¯ lu¯ tamasslo¯ tamaθθ ulun mθ ul 1 The correspondences yielding the 29 reconstructed Proto-Semitic consonants can be read in the rows, as can the prefixes. The last five rows show the most important derived stems: G(round) or Basic stem (in the top 10 rows), D(ouble)Intensive, S Causative, NPassive, tPassive/Reciprocal. * marks roots not attested in this language or this stem. — ( …) marks stems not used in this lan- guage, with the consonant correspondences for the language. Classification would develop independently). This approach helps explain how some characteristics (like noun case end- The sole representative of East Semitic is Akkadian, ings) can be found in the earliest (Akkadian) and latest which includes the dialects Assyrian, Babylonian, and (Arabic) forms of Semitic, but not in the intermediate the recently discovered Eblaite. West Semitic was forms (taken to be Ethiopic/South Arabian, Canaanite, understood to comprise Northwest (divided into and Aramaic in order). Canaanite and Aramaic) and Southwest (the languages Although Akkadian was deciphered in the late of South Arabia, both Ancient and Modern, and 1840s, its grammar was not well understood before the Ethiopia); Arabic used to be included with the insights of the Czech-born Leipzig/Ankara/Chicago Southwest group, but in recent years arguments were Assyriologist Benno Landsberger, which were codified advanced that it belonged better with the Northwest first in a chapter by Gotthelf Bergsträsser (1928/1995) group. The criteria for the classification are presented and then by Landsberger’s student Wolfram von Soden below, following Faber in Hetzron (1997). (1952/1995); the earliest form of Akkadian received Recently, however, Stammbaum or ‘family tree’ dia- due attention only with the work of I.J. Gelb in the grams of language family relationships have fallen out 1950s. Consequently, Akkadian was not well assimilat- of favor, as linguists have recognized the importance of ed into the early comparative sketches and collections social interaction in the shaping of languages, and a new of data such as those by William Wright (1890), model for the development of the Semitic languages has Theodor Nöldeke (1889/1911), and Carl Brockelmann been offered (Kienast 2001). It suggests that the succes- (1908, 1908–1913) and even that by Landsberger’s sive appearances of the languages in the historical, i.e. own teacher Heinrich Zimmern (1898). In these written, record represent successive attestations from a works––and in subsequent manuals, notably the wide- single, continually developing speech community (and ly used Moscati et al. (1964)––the reconstructed Proto- after they make their entries, of course, each language Semitic bears a very close relationship to Classical 953 SEMITIC LANGUAGES Arabic. (Lipin´ski (1997) is not suitable for the beginner related words like learn and teach. But in Semitic, because, although detailed and up to date, it does not ‘teach’ is literally ‘cause to learn’: Akk. lama¯du, šul- distinguish between the generally accepted opinion and mudu. Besides this CAUSATIVE, Semitic languages the author’s own; but its bibliography is extensive.) have an INTENSIVE––the derivation marked by the Only with the work of Gelb himself (1969) and of fel- lengthened middle consonant––and one or two kinds low Assyriologist Burkhart Kienast (2001) does a of PASSIVE (and additional derivations appear in Proto-Semitic emerge that offers a more credible individual languages). Some examples are shown abstraction that might foreshadow East Semitic as well in Table 1. as West Semitic. The forms in Table 1 also exemplify the sound cor- Even before the decipherment of Mesopotamian respondences that both demonstrate the unity of the cuneiform had progressed very far, the language writ- Semitic family and characterize its different branches. ten with that script could be identified as Semitic In East Semitic, the five gutturals merge into ’ [ʔ], but (which in turn proved very useful in the decipherment). the three most highly marked ones, g´ h. ‘ [γ  ʕ], color The clue was the identification of sets of words exhibit- an adjacent a to e, while h and ’ do not. ing perhaps the most distinctive trait of Semitic mor- In Central Semitic, the EMPHATIC series (tran- phology: alternations of the type iprus–iparras–paris. scribed with the underdot) is realized with pharyn- The precise meanings of these verb paradigms turned gealization, rather than glottalization as in South and out not to be exactly those familiar from Hebrew, (probably) East Semitic. The nonpast yaqattal is Arabic, and Ge‘ez, but the forms were identical. replaced by yaqtulu. Within Central Semitic, In fact, the divergence in meaning of the verb forms Northwest Semitic is distinguished from Arabic by is taken as the most salient characteristic for distin- the change of almost every initial w to y. Within guishing West Semitic as a whole from Akkadian. All Northwest Semitic, Canaanite shows several common Semitic languages have just two basic verb forms: in morphological innovations, and the change of a¯ to o. Akkadian iprus, the ‘preterite’ or past, and iparras, the Aramaic displays a simplification and ‘neatening’ of ‘present’ (persons are identified by prefixes; Table 1); the derived verb stem system, but its typical phono- paris, the ‘stative’, stands somewhat outside the sys- logical change, the interdentals [θ ð] becoming stops tem (persons are identified by suffixes; Table 2). But [t d], is not found in the earliest attested texts. In West Semitic has turned the equivalent of paris, e.g. Canaanite, the interdentals become sibilants [ʃ ]. Arabic qatala, into the ‘perfect’, denoting completed (Arabic preserves all the ancient distinctions except action (forms of qatala ‘kill’ are traditionally used to one of the sibilants. Semitists usually write š s´ g´ hC for illustrate verb forms; this word is not found in [ʃ  γ x].) Akkadian, and para¯su ‘divide’ serves instead). The South Semitic’s diagnostic morphological innova- forms with a lengthened middle consonant, e.g. qatta- tion is seen in Table 2. The k of the old first-person sin- la, are reserved for a quite different function, but one gular has spread to the second person (while in that (with a different selection of vowels) it already Northwest Semitic, the t of the second person has had in Akkadian. spread to the first). Southeast Semitic (Modern South This different function is another signal characteris- Arabian) is not the direct descendant of Old (or tic of Semitic: rich derivational morphology of the Epigraphic) South Arabian; the latter is closer to verb. English, for example, has pairs of semantically Ethiopic Semitic than the former is to either of them. TABLE 2 Person, Gender, and Number Suffixes on paris/qatala Forms Person Akkadian Hebrew Aramaic Ge‘ez Arabic Proto-Semitic 1 sg. -a¯ku -t¯ι -et -ku¯ -tu -ku 1pl. -a¯nu -nu¯ -n -na -na¯ -na 2m.sg. -a¯tu -ta¯ -t -ka -ta -ta 2m.pl.. -a¯tunu -tem -tu¯n -kmmu¯ -tum -tumu 2f.sg. -a¯ti -t -t -k¯ι -ti -ti 2f.pl. -a¯tina -ten -te¯n -kn -tunna -tinna 3m.sg. — — — -a -a — 3m.pl. -u¯ -u¯ -u¯ -u¯ -u¯ -u¯ 3f.sg. -at -a¯ -at -at -at -at 3f.pl. -a¯ -u¯ -a¯ -a¯ -na -a¯ 954  SEMITIC LANGUAGES Specific Languages illuminating the religious background of the Bible. In 1975, farther east in Syria, a large number of clay Akkadian was spoken in Mesopotamia from before tablets were found at Tell Mardikh, ancient Ebla. At 2500 BCE until the early to mid–first millennium first their Eblaite language, written in the familiar BCE and survived as a literary and then scholarly lan- cuneiform of c. 2500 BCE, was interpreted as also rep- guage into the early centuries CE (Geller 1997). (The resenting a very early relative of Hebrew, but it was latest known dated Akkadian text is from 75 CE.) soon recognized as a form of Akkadian or as a separate Akkadian was written by impressing groups of East Semitic language used far to the west. wedges in soft clay shaped into ‘tablets’, which when The South Arabian languages, Modern (Mehri, they dried (or got baked when a building burned H. arsusi, Bat.h. ari, Hobyot, Jibbali, and Soqot. ri) and down) were quite permanent, and thousands were Old (Sabean, Qatabanian, Hadramauti, and Minean), recovered when explorers began poking into the were the first languages (after Phoenician) to be ‘dis- remains of ancient cities. Its decipherment in the late covered’ in modern times, in the 1830s. They occupy 1840s revealed the existence of an unsuspected the southern fringes of the Arabian peninsula and near- Mesopotamian civilization of which bare hints were by islands. All the proto-Semitic consonants are pre- found in the Bible; early on, it was found to have a lit- served, and the Old South Arabian script records the erature, a religion, and a mythology (including a 29, and no vowels at all. The modern languages are not Flood story) that illuminated not only the Bible but its written; Mehri and Soqot. ri may have enough speakers own society. Even more, though, emerges from the not to be immediately endangered. tens of thousands of non-literary tablets recording ordinary transactions of the most varied kind. There are two scholarly dictionaries: Gelb et al. (1956–) and Semitic Writing von Soden (1958–1981); Black et al. (2000) is a con- The scriptures of the three Western monotheistic reli- densation of the latter. gions are written in Semitic languages (Hebrew, The second half of the nineteenth century witnessed Aramaic, and Arabic). But the Semitic languages the discovery of a wide variety of inscriptions in hosted the development of the writing systems them- Hebrew and other, closely related Canaanite lan- selves that serve virtually every language outside guages, the first and foremost, and for a long time the East Asia. Mesopotamian cuneiform, after being earliest (c. 850 BCE), being the stela of King Mesha, adapted from its Sumerian original, was used for in Moabite. Its script is very similar to that of many languages throughout the ancient Near East Phoenician, which had been deciphered in the 1760s; (with the exception of Egyptian) over at least 2,000 unfortunately, the large number of inscriptions in years, first of all. Phoenician (c. 1000 BCE–500 CE), found scattered all Cuneiform was cumbersome, though, requiring the round the Mediterranean, including its late variety, the memorization of several hundred characters during Punic of Carthage and environs, are almost entirely years of study. It was when the basic principle of brief and uninformative, so that little can be recon- Egyptian hieroglyphics––write only consonants––was structed of Phoenician political history. The sole mon- applied in a highly simplified manner to an early West ument in Moabite, however, is a historical inscription Semitic language that the sequence leading to the resembling contemporary accounts in Aramaic, alphabets of the modern world got going. Little mate- Akkadian, and Hebrew. In recent decades, it has been rial survives from the earliest stages; by the time of joined by similar but less extensive materials in Ugaritic, we see an adaptation of the letters to writing Ammonite and Edomite. All three languages are from on clay. One variety is found to the south, used in Old east of the Jordan River. Akkadian tablets preserve ref- North Arabic graffiti and Old South Arabian monu- erences to thousands of people whose names are in mental inscriptions, and then in Ethiopic inscriptions Amorite, the West Semitic language of the famed King and manuscripts. Hammurapi and his dynasty. The other variety emerges in Phoenician and An epochal discovery in 1929 at Ras Shamra on the Aramaic versions; Phoenician survives almost exclu- coast of Syria was clay tablets in what soon proved to sively in its offshoot, the Greek alphabet, and its be a Northwest Semitic language, Ugaritic. Its script is descendants including the Roman and Cyrillic. Scions consonantal like Phoenician’s, and not derived from of the Aramaic version serve Hebrew, Arabic, and Mesopotamian cuneiform despite the use of wedge- Aramaic––and spread all across South, Southeast, and made letters. Dating from the fourteenth–thirteenth Inner Asia. The variety of ways in which vowel nota- centuries BCE, Ugaritic is too early to display specif- tion came to the originally consonantal scripts is a ically Canaanite features, but again it preserved texts saga in itself. 955 SEMITIC LANGUAGES References Kienast, Burkhart. 2001. Historische Semitische Sprach- wissenschaft [Historical Semitic linguistics]. Wiesbaden: Bergsträsser, Gotthelf. 1983. Introduction to the Semitic lan- Harrassowitz. guages, transl. by Peter T. Daniels. Winona Lake, In: Lipin´ski, Edward. 1997. Semitic languages: outline of a com- Eisenbrauns; revised, 1995 (German original, 1928). parative grammar. Louvain: Peeters. Black, Jeremy, Andrew George, and Nicholas Postgate 2000. A Moscati, Sabatino, Anton Spitaler, Edward Ullendorff, and concise dictionary of Akkadian. Wiesbaden: Harrassowitz. Wolfram von Soden. 1964. An introduction to the compara- Brockelmann, Carl. 1908. Kurzgefasste vergleichende Grammatik tive grammar of the Semitic languages, ed. by S. Moscati. der semitischen Sprachen [Brief comparative grammar of the Wiesbaden: Harrassowitz. Semitic languages]. Berlin: Reuther & Reichard. Nöldeke, Theodor. 1889. Semitic languages. Encyclopædia ––––––. 1908–1913. Grundriss der vergleichenden Grammatik Britannica, 9th edition, revised 11th edition, 1911. der semitischen Sprachen [Foundations of the comparative von Soden, Wolfram. 1952. Grundriss der akkadischen grammar of the Semitic languages], 2 vols. Berlin: Reuther Grammatik [Foundations of Akkadian grammar]. Rome: & Reichard. Pontifical Biblical Institute, 3rd edition, 1995. Gelb, I. J. 1952. Old Akkadian writing and grammar. Chicago: ––––––. 1958–1981.Akkadisches Handwörterbuch [Akkadian University of Chicago Press, 2nd edition, 1961. desk dictionary], 3 vols. Wiesbaden: Harrassowitz. ––––––. 1969. Sequential reconstruction of Proto-Akkadian. Wright, William. 1890. Lectures on the comparative grammar Chicago: University of Chicago Press. of the Semitic languages. Cambridge: Cambridge University Gelb, I. J., et al. (eds.) 1956–. The Assyrian dictionary of the Press. Oriental Institute of the University of Chicago, 21 vols. Zimmern, Heinrich. 1898. Vergleichende Grammatik der semi- Chicago: Oriental Institute. tischen Sprachen [Comparative grammar of the Semitic lan- Geller, M. 1997. The last wedge. Zeitschrift für Assyriologie 87. guages]. Berlin: Reuther & Reichard. Gray, Louis H. 1934. Introduction to Semitic comparative lin- guistics. New York: Columbia University Press. PETER T. DANIELS Hetzron, Robert (ed.) 1997. The Semitic languages. London: Routledge. See also Afroasiatic; Arabic; Aramaic; Philology Serbo-Croatian and South Slavic Languages The South Slavic languages, Serbo-Croatian, Slovene, which, together with the Black Sea, separates South Bulgarian, and Macedonian, descend from Slavic Slavic from Ukrainian. Within the South Slavic dialects that were brought to the sub-Alpine and branch, two subgroups are distinguished: Western Balkan regions of southwestern Europe c. 500 CE by South Slavic, constituted by Slovene and Serbo- waves of westward migration along and across the Croatian, and Eastern South Slavic, constituted by Danube, Drava, and Sava river systems. In their new Macedonian and Bulgarian. The languages are also territory, the South Slavs encountered and undoubted- divided along cultural and religious lines: Slovene and ly mixed with Latin-speaking peoples, probably Croatian are spoken predominantly by Catholics, descendants of older Indo-European-speaking peo- whereas Serbian, Macedonian, and Bulgarian are spo- ples, for example, Illyrian and Thracian. ken by Eastern Orthodox Christians. The exact relationships among the dialects at the These divisions have determined the choice of time of settlement are uncertain, but it is not the case alphabet, Latin being chosen in Catholic areas, and that there were already nascent Slovene, Serbo- Cyrillic (a modified variety of the Greek alphabet) in Croatian, Macedonian, and Bulgarian dialects. Rather, Eastern Orthodox areas. Bosnia, which has been reli- these formed over the subsequent millennium. The giously and ethnically mixed and also includes a sig- South Slavic group may now be defined by its geo- nificant Muslim population, had vacillated among graphical discontinuity to the remainder of the different alphabets. Since the disintegration of Slavic-speaking world. To the north, Slovene is bound- Yugoslavia, the standard Bosnian of Muslims is written ed by Friulian and Italian in Italy, by German in in the Latin alphabet, whereas the Bosnian Serbs use Austria, and by Hungarian in Hungary. Croatian and Cyrillic. As with most Indo-European languages, the Serbian are also bounded by Hungarian and Romanian South Slavic group is characterized by many grammat- (Romania). Bulgarian is bounded by Romanian, ical endings, with nouns and verbs changing form 956  SERBO-CROATIAN AND SOUTH SLAVIC LANGUAGES depending on their position in the sentence or on their The Serbo-Croatian speech territory is character- function as subjects or objects, singulars or plurals. ized by three distinct dialect areas, each labeled by Slovene and Serbo-Croatian go with the rest of the both professionals and the laity by the word meaning Slavic-speaking world in having preserved most of ‘what’. A transitional zone called the Torlak group dis- these endings in nouns, but verbs have become some- plays features of both Štokavian and neighboring what simplified. Macedonian and Bulgarian have the Macedonian and is thus arguably within the scope of opposite: simplified nouns but more complicated verbs. the Balkan Sprachbund, an area of linguistic conver- Serbo-Croatian is spoken by approximately 16 mil- gence among distantly related or even unrelated lan- lion people. It is the state language of the Republic of guages caused by long-term contact, which also Croatia (where it is called Croatian), Bosnia and includes Albanian, Aromanian, Greek, Romanian, Herzegovina (where it is called Bosnian), and Serbia Romany, and, to some extent, Turkish. and Montenegro (where it is called Serbian); minority Generally speaking, linguists’ attention has been speakers are also found in Italy, Hungary, Austria, drawn to Serbo-Croatian (as well as Slovene), espe- Romania, Bulgaria, and Macedonia. cially for its phonological (sound pattern) features, The Serbo-Croatian standard was formed in the high degree of dialect variation, and preservation of nineteenth century as a compromise between Serbs key archaisms that aid in the reconstruction of Proto- and Croats, whose major dialect divisions and corre- Slavic, the prehistorical language thought to have been sponding divergent literary traditions, particularly in spoken by all Slavs before 500 CE. Standard Slovene the Croatian case, had fostered disunity. The Hakavian and Serbo-Croatian, as reflected in many of their and Kajkavian dialects, both spoken in Croatian ethnic dialects, contrast long and short vowels, and, along territory, and which had developed into sophisticated with stress, have rising and falling tones (similar to literary vehicles during the Renaissance and Chinese), e.g. Slovene brá:t(i) ‘to read’ (long low Reformation, respectively, were abandoned as models pitch), brà:t ‘to go read’ (long high pitch), and bràt for the standard language in favor of the Štokavian ‘brother’ (short high pitch). Other features are of inter- dialect, spoken in Croatia and all of Serbia, as well as est, particularly word and sentence structure, e.g. in Bosnia and Herzegovina and Montenegro. In Serbo-Croatian has begun to simplify its nouns, as has Serbia, the new Štokavian-based standard replaced the occurred more radically in Macedonian and Bulgarian, artificial Slaveno-Serbian literary language, which by reducing the number of grammatical endings was based largely on Old Church Slavic. The compro- (‘cases’), especially in the plural. mise, which was engineered by intellectuals around Structurally, Slovene is closest to Serbo-Croatian the Croat Ljudevit Gaj and the Serb Vuk KaradDic´, was and is spoken by approximately two million people, codified in the Literary Agreement of 1850. The stan- largely in the Republic of Slovenia, where it is the dard had two varieties, the Croatian (or Western), writ- primary official language (alongside regionally offi- ten in a modified Latin alphabet, and the Serbian (or cial Italian and Hungarian). It is also spoken by sig- Eastern), written in a modified Cyrillic. This standard nificant minorities in neighboring Italy, Austria, and persisted officially as the language of the Croats, Hungary. Serbs, and (Bosnian and SandDakian) Muslims, as well Modern standard Slovene, which began its develop- as the de facto lingua franca of Yugoslavia, until the ment with the religious translations of the Protestant disintegration of the state in 1991. Since then, separate Primus Truber (PrimoD Trubar in Slovene) in the mid- Croatian, Serbian, and Bosnian state languages (the sixteenth century, was established in largely its current latter using the same alphabet as Croatian and having form toward the end of the nineteenth century. It is a relatively higher number of Turkish and other based on the urban speech of the capital, Ljubljana, Islamic cultural borrowings) have been cultivated, and the surrounding central dialects, although it also each continuing from their inherited Štokavian-based has features selected from its highly variegated precursor; all three standard languages remain almost dialects. It is written in a modified variety of the Latin completely mutually intelligible. (For this reason, alphabet, similar to Croatian. ‘Serbo-Croatian’ persists as a linguistically valid term, With its relatively small speech territory, Slovene referring to the speech territory and the common base has seven dialect bases and greater internal differenti- of the separate language collectively. However, it is no ation than any of the South Slavic languages. Speakers longer considered an acceptable term to most lay from the most extreme dialects (e.g. Rezija, speakers or the governments of the successor states.) Prekmurje) generally cannot be understood by stan- Other regional movements, including notably a dard speakers. Slovene preserves archaic features that Montenegrin one, suggest the possibility of forming have been lost in Serbo-Croatian. For example, it dis- further standard languages in the future. tinguishes not just singular and plural but also dual 957 SERBO-CROATIAN AND SOUTH SLAVIC LANGUAGES number (pogovarjava se ‘we two are conversing’); it zhenata ‘I saw the [a certain] woman’ vs. Vidov zhena makes the future tense with an auxiliary verb and a ‘I saw a woman’; Serbo-Croatian makes no such dis- participle (bom sedela ‘I shall sit’); and it preserves a tinction, having only Vidjela sam Denu ‘I saw the/a special ‘supine’ form of the verb that signals intention woman.’ A distinction expressed by choices among (kupovat bom šel ‘I shall go to shop’). In contrast to alternative verb forms is made between witnessed and Serbo-Croatian, Slovene has a relatively significant nonwitnessed events, e.g. Bulgarian Toj napisa pis- number of borrowings from German (e.g. farba moto ‘he wrote the letter [I know so because I saw ‘color’ from Farbe), Italian (fant ‘boy’ from fante), him do it]’ vs. Toj napisal pismoto ‘he wrote the letter and Friulian (kriD ‘cross’ from a seventh-century [so it is said—I did not see him do it]’; Serbo- Friulian form kroDe). Croatian makes no such distinction, having only Bulgarian is spoken by approximately nine million Napisao je pismo ‘He wrote the/a letter.’ The inherit- people, predominantly in the Republic of ed infinitive has been lost and replaced by a subordi- Bulgaria, where it is the primary state language, as well nate clause, e.g. Bulgarian Iskam da otida na maJ ‘I as by minority speakers in Serbia and Macedonia. want to go [literally ‘that I go’] to a game’ vs. Serbo- Structurally, Bulgarian is closest in type to Macedonian. Croatian Hoc´u i_c´i na utakmicu ‘I want to go to a Modern Bulgarian dates to the seventeenth century game.’ The origin of such convergence features is and developed substantially into its current form in the much debated: they may be a continuation of struc- middle of the nineteenth century. It is based on the tures from languages that have disappeared (substra- TaK rnovo dialect of northeastern Bulgaria, but with ele- tum languages)—Illyrian and Thracian—or a result of ments from various dialect areas. Medieval varieties of language contact itself and diffusion of linguistic fea- Bulgarian served as the primary examples of Slavic tures, although the working of both explanations writing, with prominent writing centers located in together is not excluded. Preslav and TaK rnovo. Modern Bulgarian is written in a The South Slavic languages represent a picture of modified variety of Cyrillic. great diversity among the Slavic languages, and, Macedonian is spoken by approximately two mil- because they are located at a crossroads of European lion people, primarily in the Republic of Macedonia. languages and cultures, they have been affected by Significant groups of Macedonian speakers are also contacts with numerous languages. The volatile polit- found in northern Greece, western Bulgaria, Serbia, ical fortunes of the region promise to push the devel- and in some villages in Albania. opment of the languages, especially the newly Macedonian was codified as a standard language in differentiated Bosnian, Croatian, and Serbian, toward 1944, although the beginnings of the contemporary ever greater diversity. language may be traced to the middle of the nineteenth century. Macedonian is written in a modified variety of References the Cyrillic alphabet. The language of the Macedonian speech territory can be traced back organically to the Browne, Wayles. 1993. Serbo-Croat. In Comrie and Corbett. Comrie, Bernard, and Greville G. Corbett (eds.) 1993. The speech that gave rise to the first Slavic written language Slavonic languages. London: Routledge. in the ninth century CE, known today as Old Church Friedman, Victor A. 1993. Macedonian. In Comrie and Slavic. Corbett. Linguists have tended to concentrate on the structure Greenberg, Marc L. 2000. A historical phonology of the Slovene of Macedonian and Bulgarian words and their relation- language ( Historical phonology of the Slavic languages XIII). Heidelberg: Carl Winter Universitätsverlag. ship to syntax and meaning, as well as the interaction of Huntley, David. 1993. Old Church Slavonic. In Comrie and the languages with others in the Balkan linguistic con- Corbett. vergence area (or Sprachbund). For the period from the Ivic´, Pavle. 1958. Die serbokroatischen Dialekte: Ihre Struktur tenth to the twelfth centuries, the textual evidence of und Entwicklung. The Hague: Mouton & Co. Proto-Macedonian and Bulgarian is important for the KaticK ic´, Radoslav. 1976. Ancient languages of the Balkans. The Hague: Mouton. earliest body of attestations of Slavic in general, known Koneski, BlaDe. 1983. A historical phonology of the as the canonical period of Old Church Slavic. For this Macedonian language ( Historical phonology of the Slavic reason, Indo-Europeanists have made substantial use of languages XII), transl. by V. A. Friedman. Heidelberg: Carl older Macedo-Bulgarian material. Winter Universitätsverlag. Because of their participation in the convergence Lehiste, Ilse, and Pavle Ivic´. 1986. Word and sentence prosody in Serbocroatian. Cambridge, MA: The MIT Press. area, Macedonian and Bulgarian display features not Lencek, Rado L. 1982. The structure and history of the Slovene found elsewhere in the Slavic-speaking world. For language. Columbus: Slavica. example, the category of definiteness is marked by the Priestly, T. M. S. 1993. Slovene. In Comrie and Corbett. presence (vs. absence) of an article after the first Scatton, Ernest A. 1993. Bulgarian. In Comrie and Corbett. member of a noun phrase, e.g. Macedonian Ja vidov MARC L. GREENBERG 958  SERIAL VERB CONSTRUCTIONS Serial Verb Constructions Some languages—especially those of West Africa, Some insist that serial verbs must be distinguished Southeast Asia, and New Guinea—allow several from verbal compounding. However, the literature on verbs to appear within a clause to express a single nominal compounding is replete with statements of dif- meaning, as illustrated from Kalam (Papua New ficulties in distinguishing between phrases such as Guinea), where ‘massage’ is expressed by the follow- wood turner and words such as woodlice. Verbal com- ing string of verbs: pounding is no less problematic, and any attempt to categorize verbal compounds in many languages typi- pk wyk d ap tan cally runs up against the problem of deciding where the ‘strike rub hold come ascend boundary between compounding and serialization lies. d ap yap g- However, this has the potential to divert attention hold come descend do’ away from the real issues. The relevant structures could simply be regarded as complex verbs composed Serial verbs are so prevalent in Kalam that most of more than one grammatical element, avoiding a verbal meanings can be expressed by combining fewer distinction between verbal compounding and verb than 100 verb roots. serialization. The two, in fact, exhibit a variety of In other languages, just two or three verbs are seri- similarities crosslinguistically, with the only real dif- alized. In Paamese (Vanuatu), vul- means ‘break’, but ference being that we are dealing with either words this verb is seldom used, the strong preference being to or phrases. serialize it with another verb indicating the precise Verb serialization can be distinguished from activity that caused the breakage, for example: subordination and coordination in that there should be no sentence-level connectives between the various Ni-matil vul a:i. verbal elements. If there is any evidence of subordina- 1SING.FUTURE-sleep break plank tors (such as that or which) or coordinators (such ‘I will sleep on the plank, thereby breaking it.’ as and or or)—even if only as an optional variant, as we find with English Go (and/to) get the book—then These constructions represent single clauses, in we are dealing with a complex sentence rather than a contrast to both subordinate and coordinate multi- single clause. clause constructions. However, it is difficult to clearly There is considerable diversity within the cate- and consistently define the term ‘serial verb’, and gory of serial verbs among different languages and there is also the related problem of the wide variation even within individual languages, with some kinds in terminology used to refer to structurally similar or of serial verb constructions behaving in ways that historically related phenomena in different languages. more closely resemble verbal compounds, and oth- Different writers have referred to ‘verbids’, ‘locative ers sharing features with subordinate or coordinate verbs’, ‘post verbs’, ‘postverbal prepositions’, ‘loca- constructions. tive copulas’, and ‘coverbs’ to refer to essentially the Many verb-final languages have clause chains that same patterns in West African languages. Many are not are structurally midway between serial verbs and com- explicit about what they mean by serial verbs, some plex sentences. There may be extensive juxtaposition treating any verb–verb sequences as serial verbs as of verb phrases without any subordinating or coordi- long as the second verb is not marked as an infinitive, nating markers, but with considerable contextual ellip- e.g. Go get the book. Sebba (1987) provides a more sis of shared arguments. This means that we are explicit set of criteria for recognizing serial verbs: dealing with a structural continuum as follows: verbal (1) Both verbs must be able to function independ- compounds  serial verbs  clause chains  subordi- ently within a clause as verbs in their own right. nate clauses  coordinate clauses. It is probably wish- (2) Both must be interpreted as having the same ful thinking to assume that we can produce a categories of tense–aspect–mood. universally applicable definition of verb serialization: (3) There must be no marking of a clause boundary. rather than a separate universal category, serialization (4) There should be no conjunction appearing is more accurately characterized as a syndrome of fea- between the two. tures and phenomena. 959 SERIAL VERB CONSTRUCTIONS References Lord, Carole. 1993. Historical change in serial verb construc- tions. Amsterdam: John Benjamins. Alsina, Alex, Joan Bresnan, and Peter Sells (eds.) 1997. Pawley, Andrew. 1993. A language which defies description Complex predicates. Stanford: Centre for the Study of by ordinary means. The role of theory in language Language and Information. description. ed. by William A. Foley. Berlin: Mouton de Baker, Mark C. 1989. Object sharing and projection in serial Gruyter. verb constructions. Linguistic Inquiry 20. Sebba, Mark. 1987. The syntax of serial verbs: an investigation Crowley, Terry. 2002. Serial verbs in Oceanic: a descriptive into serialisation in Sranan and other languages. typology. Oxford: Oxford University Press. Amsterdam: John Benjamins Publishing Company. Durie, Mark. 1997. Grammatical structures in verb serialisa- tion. In Alsina, Bresnan, and Sells. TERRY CROWLEY Sign Relationships A signifier, or ‘sign’ for short, is any x that stands for ed as signifying God’s wrath, this is not owing to its or represents some y, that calls some y to mind (or is having aimed at that result (or at any other). In sum, all thought of as capable of so doing). For example, a that is necessary for an act of signification to take place south-pointing weathervane can be a ‘sign’ that the is one mind’s perceiving some x, whatever its source, local wind is a southerly, if so interpreted by a well- and tying it to some y, however legitimately. disposed observer (Watt 1993:432). By the same Language is the supreme signifier. Its morphemes, token, the word ‘unicorn’ is a sign of, or signifies, a words, phrases, clauses, sentences, paragraphs, and mythical equine creature with a single helical horn whole discourses are almost entirely emitted in the peculiarly susceptible to the importunities of virgins. hope of signifying, whether to ourselves or to others; In the latter case, ‘signifies’ can be replaced by and we listen and read in the wan hope of being ‘means’, but this is not always so. For instance, it informed. Homo sapiens is, above all, homo signifex would be odd to say that the letter ‘A’, which signifies (homo the sign-maker). the set of sounds comprising the phoneme /a/, ‘means’ Signs of whatever sort are often bound into a set, a those sounds. And whereas it is safe to say that the system, characterized by its members’ widespread word ‘thwirk’, which has never existed till now, has no sharing of attributes. Many such attributes are uncon- meaning at all in the ordinary way, it still signifies sciously known to the signs’ users, since often they can something (e.g. that its creator knows how to observe reliably judge whether or not a proposed new member the rules of English phonology). Hence, ‘signifies’ is of the set is sufficiently like the old ones (i.e. shares the more general term, since ‘means’ will sometimes enough of their attributes) to win entry. Some such seem misused when neither y (‘a mythical equine. . .’) judgments are quite subtle. For instance, a speaker of nor x (‘unicorn’) is a linguistic expression. English will accept ‘thwill’ as a possible (or unknown) All living things, from microbes to elephants, and English word, while rejecting ‘thpill’ (despite the cur- including plants and humans, process signs of one sort rency of ‘thpill’ in the speech of lispers); users of the or another, whether they be words and sentences or, for Romani English alphabet are more likely to accept ‘0’ plants, the signals of chemicals and light (Sebeok as a possible new letter than ‘0’ (Jameson 1994). 1989:12–24). Restricting our purview to humans, how- Besides being related to others within their set, signs ever, it is safe to say that every signification, every are also related to what they signify. Thus, a sign may tying of some signified y to some signifier x, must at a be physically associated with what it signifies, as the minimum happen in the mind of a human recipient who weathervane is to the wind that actuates it; or it may be perceives the x and ties a y to it. Some other human perceived as being physically similar to what it signi- may have sent the x in the hope of its being properly fies, as the word ‘whinny’ is to the sound made by a interpreted; but many x’ s are interpreted in such a way horse; or it may have neither of those ties, being pure- as to defeat any such hope. And many x’s are not sent ly arbitrary or conventional, as the word ‘horse’ is. at all, by anybody anywhere. Should a sender use the Thus, in Charles S. Peirce’s well-known terms, a sign word ‘livid’, he or she is likely to find it misinterpreted can be, respectively, ‘indexical’, ‘iconic’, or ‘symbol- as meaning ‘furious’; when an earthquake is interpret- ic’ (Peirce 1935–1966:2.275). Many signs in their 960  SIGNED LANGUAGES actual occurrence, including words or morphemes, The signs of language differ but little, at the lower partake of two or even three of these qualities: the levels (phonemes; morphemes), from the signs of stick figure representing ‘MAN’, for instance, conveys some other systems, either in complexity or even in ‘Here’s a men’s room’ only when it is on or near an sheer numbers. Morphemes and the cattle brands of expectably placed door, thus being indexical; it looks the American West, for instance, have comparable a little like a man, thus being also iconic; but it is still morphologies. On the other hand, the signs of lan- highly abstracted or conventional, thus being symbol- guage at the higher levels (phrases, clauses, sentences, ic as well. And, turning now to words, since ‘whinny’ discourses) differ in kind from those of any other sign is but a poor imitation, it is also partly arbitrary, i.e. system known, both in their formal complexity and, of symbolic. Even the structure and superficial word course, in their number, which is unbounded. order of sentences, although in large part arbitrary However, despite their differences, all signs, of what- (which is why different languages can order things dif- soever kind, lend themselves to description in terms of ferently), can even then be said to be also somewhat how they are related to what they signify, to each iconic, as when, in English, the order of cardinal sen- other, and to those who send and receive their physical tential elements seems to mirror a feeling that one is manifestations and interpret them. better understood when new or indefinite information is withheld till after a given or definite context has References been established (Halliday 1967; Clark and Haviland 1977). (Thus, although ordinarily an active sentence is Clark, H.H., and S.B. Haviland. 1977. Comprehension and the given-new contract. Discourse production and comprehen- more felicitous than a passive one, this preference is sion, ed. by R. Freedle. Norwood, NJ: Ablex. reversed when the underlying subject is indefinite and Halliday, M.A.K. 1967. Notes on transitivity and theme in the object is definite: the passive ‘The General’s been English: ll. Journal of Linguistics 3. 199–244. hit by an arrow’ is likely to strike most speakers as Jameson, K. 1994. Empirical methods for evaluating generative more natural than the active ‘An arrow’s hit the semiotic models: an application to the Roman majuscules. Writing systems and cognition: perspectives from psycholo- General’.) gy, physiology, linguistics, and semiotics, ed. by W.C. Watt. The signs habitually used by humans are also, natu- Dordrecht: Kluwer. rally enough, related to their human users. The sounds Peirce, C.S. 1935–1966. Collected papers of Charles Sanders of a natural language are those the human vocal tract Peirce, ed. by C. Hartshome, P. Weiss and A.W. Burks (eds.). can easily make and the human ear can distinguish, and Cambridge, MA: Harvard University Press. Sebeok, T.A. 1989. The sign & its masters. Lanham, MD: the letters of the Romani English alphabet are those the University Press of America. human hand  pen (the ‘scribal tract’) can easily draw Watt, W.C. 1993. Signification and its discontents. Semiotica and the human eye can discriminate—an obvious 97. 427–37. point, but one not noted enough. WILLIAM C. WATT Signed Languages Signed languages are a class of languages in which guages in manual gestures. The use of the term here speech is replaced by manual and facial gesture and will be restricted to those natural signed languages that auditory reception is replaced by vision. Natural evolved within deaf communities. A natural sign lan- signed languages have evolved within the boundaries guage is learned as a first language by children in the of subcultures of deaf communities throughout the community, is not related to the spoken language of world. The term ‘signed language’ is often used more the broader culture, and possesses the complexity and generally to include alternate sign languages, manual types of structural features found in spoken languages. sign systems developed in certain hearing communi- The use of manual signs dates back at least to the ties as alternatives to spoken language that serve early Greeks; Plato mentioned the use of signs by deaf social, religious, or occupational functions. The term individuals in the Cratylus. Signed languages are, is also used to describe manual language codes, sign however, a recent addition to the study of language systems invented by educators to represent spoken lan- and language structure. Serious study of signed 961 SIGNED LANGUAGES languages began only in the last half of the twentieth partial intelligibility indicates that these three lan- century. Prior to that time, signs used in deaf commu- guages belong to the same language family. nities were believed to be rooted in a biological ges- In many cases, historical and social forces have tural system that was universal and, to a great extent, influenced deaf communities differently from hearing iconic; and signed communication was thought to lack communities in the same regions. Geographical bound- the complexity of spoken languages. Two events aries sometimes follow those for spoken languages, but occurred in the mid-1900s that changed the status of they often do not. ASL and British Sign Language are signed languages. The first systematic analysis of a not mutually intelligible, for example, although signed language became available in 1960. In that English is the spoken language of both geographic year, William Stokoe published a monograph on regions. Instead, there are some similarities between American Sign Language (ASL). Stokoe demonstrat- ASL and French Sign Language (FSL) because mod- ed that the signs of ASL possessed an abstract sublex- ern ASL is derived from contact between one or more ical structure analogous to the phonological structure indigenous sign languages in the United States and of words in spoken languages. Five years later, he and FSL, brought to the United States in the early 1800s. two colleagues (Dorothy Casterline and Carl Circumstances promoting the emergence of a new Croneberg) published the first dictionary of a signed signed language have recently provided linguists with language (ASL) based on linguistic principles. In this the opportunity to examine first hand the forces driving same time period, rapidly developing video and film the creation of a new language. The birth of technology provided an economical way to capture Nicaraguan Sign Language has been extensively docu- and preserve dynamic properties of signed languages mented over the past 25 years. When the Sandinistas for analysis. These two developments, video tech- came into power in Nicaragua in 1979, formerly isolat- niques to record basic language data and Stokoe’s ed deaf children from all parts of the country were demonstration that signed languages were subject to brought together in a new residential school for deaf the same analytical techniques as spoken languages, children. The children who arrived possessed only provided the basis for the linguistic study of signed rudimentary idiosyncratic home sign systems. Home languages throughout the world. sign systems are created by deaf children who live in isolation and have little or no conventional language input. Upon arriving at the school, the Nicaraguan Distribution school children immediately began to use signs from Currently, at least preliminary descriptions exist for their home sign systems with each other. The result was well over 100 signed languages, spanning the major the development of a pidgin language between home geographical regions of Africa, Asia, Australasia, sign systems. Exposure of the next generation of school Europe, and North and South America. The names of children to the pidgin resulted in the development of a these languages typically reflect the geographical more complex signed language through a gradual area(s) where they are used (e.g. Kenyan Sign process of creolization. The resulting language is the Language, Japanese Sign Language, Australian Sign language now called Nicaraguan Sign Language. Language or Auslan, Italian Sign Language, ASL, and Brazilian Sign Language, respectively). Many signed Language Structure: Universals and Modality languages developed independently within the deaf Effects communities of somewhat isolated regions. Providence Island Sign Language, for example, is the Abstractness is a defining quality of human language. primary language of fewer than two dozen deaf people The addition of signed languages to the inventory of and a second language for some hearing people on world languages has raised the level of abstractness. Providence Island, Columbia, a small island off the Languages are no longer restricted to the vocal–audi- coast of Nicaragua. Providence Island Sign Language tory channel. Elimination of this one-time universal developed independently of the spoken language appears to have no effect on other proposed universals. (English and English Creole) of the island, although Signed languages satisfy most proposed universals, deaf residents had no exposure to other signed lan- and, where some do not, spoken language counterex- guages. Signed languages have also developed amples have also been discovered. Signed languages through processes of migration and cultural contact. allow complexity of expression equivalent to that of With time, historical change produces distinctly dif- spoken languages with complexity of grammar that is ferent signed languages. For example, three distinct equivalent to grammars of spoken languages. ASL, for signed languages have been identified in Vietnam: Ho example, shares features with spoken Navajo and Chi Minh City Sign Language, Hanoi Sign Language, Japanese. The form and complexity of grammatical and Haiphong Sign Language. Available evidence of properties in signed languages do not differ from those 962  SIGNED LANGUAGES of spoken languages, but the means by which those Classifier constructions are ubiquitous in signed lan- grammatical properties are implemented differ by guages. Classifier systems take advantage of the spatial modality. medium to express types of motion through space and Stokoe’s analysis of the sublexical structure of ASL position in space, as well as stative-descriptive and signs provided the first example of how a universal handling information. Complex predicate classifier grammatical property can be implemented in the man- constructions are universal to all signed languages ana- ual modality. Stokoe demonstrated that ASL signs lyzed to date. They also exist in evolving languages could be described by three phonological categories of like Nicaraguan Sign Language and in idiosyncratic arbitrary components: 19 handshapes, 12 locations, home sign systems. In classifier predicates, handshapes and 24 movement patterns. Minimal pairs of signs are that function as classifier morphemes combine with identical except for a change of one component from different types of movement morphemes to form com- one phonological category. Orientation of the hand is plex predicates. Categories of handshape and move- also needed in some cases to discriminate certain min- ment morphemes found in both ASL and Danish Sign imal pairs of ASL signs. Components from the phono- Language (taken from Emmorey 2002) illustrate the logical categories were originally thought to occur properties of signed language classifier systems. simultaneously in sign formation. Later analyses, Categories of handshape and movement mor- however, have demonstrated that signs, like spoken phemes for ASL and Danish Sign Language include words, have linear segmental structure. Analyses indi- the following: cate that sign languages throughout the world use sim- ilar components at the phonological level, but the Handshape Morphemes specific inventory of handshapes, locations, and move- Whole entity classifiers: refer to an object as a whole ments differs from language to language. Handling/instrument classifiers: represent a whole enti- The use of the gestural–visual channel provides ty, but imply an agent handling the entity both advantages and constraints for the development Limb classifiers: represent limbs of animals and humans of grammatical mechanisms in signed languages. The Extension and surface classifiers: represent depth and use of manual articulators results in gross motor width of an object or width of a surface movements that take more time to produce than the fine motor movements involved in speech. Time con- Movement Morphemes straints on linear sequencing are complemented, how- Position morphemes: indicate ‘to be located’, posture, ever, by the availability of two articulators, the greater or a change in object orientation simultaneity of components in word formation in Motion morphemes: indicate movement along a path signed languages, the use of three-dimensional space, Manner morphemes: indicate manner (e.g. speed) of and the coordinated use of nonmanual and manual motion without specifying a path Extension morphemes: depict the outline or configura- components. Some language processes based in the tion of an entity or mass gestural–visual channel appear to be universal to signed languages, whereas others are language specif- Each language imposes certain constraints on pos- ic or are found in a subset of signed languages. sible morpheme combinations. Whole entity classi- All signed languages analyzed to date possess com- fiers combine with extension movement morphemes to plex systems for the creation of multimorphemic signs. indicate spatial arrangement of multiple objects. For In spoken languages, complex words are typically cre- example, the ASL handshape classifier for vehicles ated by processes of concatenation. Complex word for- combined with the linear extension movement mor- mation processes differ for signed languages in that pheme indicates ‘cars in a row’. The instrument hand- simultaneous morphological components take the shape classifier used to represent ‘knife’ could not, place of linear strings of affixes. Temporal aspect however, be substituted for the whole entity ‘vehicle’ inflection is carried out through changes in the move- classifier to indicate ‘a row of knives’. ment characteristics of sign stems in many signed lan- Signed languages, but not spoken languages, allow guages, including ASL, British Sign Language, and for certain types of simultaneous constructions. These Swedish Sign Language. Derivational processes creat- constructions involve the use of the two hands to ing nouns from verbs in ASL also involve changes in express two different predicates at the same time. In the movement patterns of the verb forms. In general, ASL, for example, a simultaneous construction can be the linear affixation processes prominent in spoken lan- used to express, ‘The cat jumps on a shelf’. In this con- guages are rare or nonexistent in signed languages. struction, the nondominant hand predicate indicates the Instead, these examples of inflectional and derivational background element, ‘the shelf’. The handshape classi- morphology involve the simultaneous occurrence of a fier for narrow surfaces (a flat hand) is combined with base sign and morphological markings. the ‘hold’ morpheme, a static movement morpheme 963 SIGNED LANGUAGES that indicates static motion or holds the motion of an movement components that are individually used to entity in place for purposes of the construction. The convey different grammatical distinctions that may dominant hand predicate combines a classifier hand- include person, number, and case. Pointing is used to shape used for small animals with an ‘arc’ motion mor- refer to entities, whether present or not. When the ref- pheme. The end of the arc motion makes contact with erent is present, the signer points to the actual position the nondominate hand, which is located above the level of the referent. When the referent is absent, the signer of the starting point of the arc motion. establishes a spatial location for the referent and points Manual signs are normally expressed in a signing toward that spatial location. Signed languages have space that extends vertically from just below the waist conventionalized ways of assigning spatial locations to the top of the head and horizontally from just over for different types of referents. the signer’s shoulder outward a little more than a foot. All signed languages examined to date use modifi- Lines defined in this space are used to express different cations of the verb to mark grammatical agreement forms of temporal reference. The abstract concept of with subject and object for some classes of verbs. time is overlaid on these time lines. The use of time Agreement is marked through modification of the lines to convey temporal information is found in lan- sign’s hand orientation and location and direction of guages as diverse as ASL, Argentine Sign Language, the movement. The movement of these verbs is direct- Australian Sign Language, British Sign Language, ed from the spatial location marking the subject toward Jordanian Sign Language, and the Sign Language of the spatial location marking the object. In Italian Sign the Netherlands. At least two of these time lines may be Language, for example, the sign for TEACH is com- universal to signed langauges. One of these time lines posed of an initial handshape that is closed such that is related to immediate context. This time line extends the fingertips touch the tip of the thumb. The hand from the top of the dominant shoulder outward, per- opens to a flat handshape with the fingers spreading as pendicular to the body. The signer’s body provides the the movement of the sign is produced. In the sentence reference point for ‘the present’, with the past behind ‘I teach you,’ the movement of the sign begins near the the signer and the future in front. Temporal adverbs signer’s dominant shoulder and moves outward toward make use of this time line. For example, the sign for the spatial location of the addressee. The front/palm of YESTERDAY moves backward from the side of the the hand faces the addressee throughout the production cheek, and TOMORROW moves forward. Adverbial of the sign. The location, movement, and hand orienta- modification of other signs is used to express the past tion of the sign TEACH are modified to mark subject or future. In ASL, for example, the signs for days of the and object agreement in the sentence, ‘Elena teaches week are made near the body. These signs can be mod- Louis.’ If the spatial position marking Elena is to the ified to indicate the next occurrence of that day by signer’s right and the spatial position marking Louis is changing the location of the sign to a point forward on to the left, the movement of the verb TEACH is from the time line (e.g. MONDAY modified to represent right to left, beginning at the spatial location marking NEXT-MONDAY). A second time line represents suc- the subject, Elena, and moving toward the spatial posi- cession or duration of time. This time line extends hor- tion marking the object, Louis. The hand is oriented izontally, parallel to the front of the signer’s body. Sign toward the spatial location marking Louis as the hand- movement from left to right along this time line estab- shape changes from closed to open. lishes temporal reference in discourse from early to late In addition to the manual channel, signed languages in many signed languages. The British Sign Language make use of a second, nonmanual, channel to convey sign CONTINUE uses this kind of movement to grammatical information. Facial expression and express the persistence of an event or state over a long changes in head and body position are used to provide period of time. Jordanian Sign Language also makes distinct linguistic contrasts simultaneously with the use of the sequential time line, but time from early to production of manual signs. Nonmanual markers are late is expressed in right to left movement. used grammatically to distinguish similar structures Another common, probably universal, mechanism across signed languages. These structures include found in signed languages is the use of spatial location negation, Wh- questions, Yes–No questions, condi- to refer back to something that was previously men- tionals, relative clauses, and rhetorical questions. In tioned and to indicate which nouns in the sentence are general, the duration of the nonmanual marker indi- subject and object, respectively (agreement). In signed cates to which part of the sentence it applies. The non- languages, personal pronouns take the form of point- manual component used to mark negation, for ing gestures. Pointing has been co-opted from the ges- example, typically includes a specific facial expres- tural realm to form the basis for linguistic pronominal sion with accompanying head movement. The gram- systems in signed languages. Points in signed lan- matical facial expression is maintained for the entire guages are categorized into handshape, location, and duration that it takes to produce the phrase being 964  SINDHI negated. Nonmanual components appear to be similar modality-independent nature of language, both in lan- in both form and function across signed languages. guage form and in functional neuroanatomy, has rekindled debate about the origin of language. Neural Systems for Language References In the same way that the form of language is not Ahlgren, Inger, Brita Bergman, and Mary Brennan (eds.) 1994. dependent on language modality, the neural systems Perspectives on sign language usage, Vols. 1–2. Durham: underlying language processing are not dependent on International Sign Linguistics Association, University of language modality. Brain-damaged signers display Durham. similar types of error patterns as brain-damaged Brentari, Diane. 1998. A prosodic model of sign language pro- cessing. Cambridge, MA: MIT Press. speakers. Language is typically processed in the left Emmorey, Karen. 2002. Language, cognition, and the brain. hemisphere for both signed and spoken language and Mahwah, NJ and London: Lawrence Erlbaum Associates. specific types of language disorders are localized in ———. (ed.) 2003. Perspectives on classifier constructions in the same neuroanatomical structures of the left hemi- sign languages. Mahwah, NJ and London: Lawrence sphere. Classic patterns of Broca’s and Wernicke’s Erlbaum Associates. Klima, Edward S., and Ursula Bellugi. 1979. The signs of lan- aphasia have been described for deaf signers. Prosodic guage. Cambridge, MA: Harvard University Press. disturbances and reduced, ‘telegraphic’ syntactic con- Meier, Richard P., Kearsy Cormier, and David Quinto-Pozos structions are associated with damage to Broca’s area (eds.) 2002. Modality and structure in signed and spoken of the left hemisphere for both spoken and signed lan- language. Cambridge: Cambridge University Press. guage. Errors in word or phoneme choice and the pro- Neidle, Carol, Judy Kegl, Dawn MacLaughlin, Benjamin Bahan, and Robert G. Lee. 2000. The syntax of American duction of jargon (grammatically correct but sign language: functional categories and hierarchical struc- unintelligible sentences) are associated with left hemi- ture. Cambridge, MA: MIT Press. sphere damage in Wernicke’s area for both spoken and Poizner, Howard, Edward S. Klima, and Ursula Bellugi. 1987. signed language. Wernicke’s area is also involved in What the hands reveal about the brain. Cambridge, MA: language reception. Specific sites in Wernicke’s area MIT Press. Stokoe Jr., William C., Dorothy C. Casterline, and Carl G. have been associated with the processing of phonet- Croneberg. 1965. A dictionary of American Sign Language ic–syllabic units, independent of language modality. on linguistic principles. Washington, DC: Gallaudet College The left hemisphere appears to be specialized for Press. language, independent of modality. The abstract, PATRICIA SIPLE Sindhi Sindhi is a language that originated in the Lower Indus and Hyderabad, and alongside Hindi, Gujarati, and Valley region of the Indian subcontinent, spoken by other regional languages in India. Sindhi is closely over 40 million people in present-day Pakistan and related to Siraiki, spoken in the north of Sindh India and by a large diaspora community around the Province, and to Kachhi, spoken to the south in the world. Sindhi belongs to the Indo-Aryan language Kacch region of Gujarat in India. family within Indo-European, and is classed with Kashmiri in the Northwestern subgroup. Sindhi is the Background primary language of the province of Sindh in Pakistan, and is spoken along the Indus River Valley stretching Sindhi shares many features in common with related into the Thar Desert to the east, and bounded by the Indo-Aryan languages. The sound inventory includes Sukkur Dam to the north, the Kirthar Mountain Range the distinctive voiced and voiceless aspirated obstruents to the west, and the Arabian Sea to the southwest. To at five places of articulation (labial, dental-alveolar, the south, the Sindhi region extends into the Rann of postalveolar, palato-alveolar, and velar), and a full set Kacch in India. While Sindhi is used exclusively or as of paired long and short vowels, all of which can occur the primary language in most rural areas within this with nasalization. Common features in the morphology region, it exists alongside Urdu and English in the include number, gender, and case marking for nouns, urban centers of Sindh Province, including Karachi a rich system of verb inflection, and a productive 965 SINDHI process of compound verb formation. The basic syntac- sounds are genuine glottalic ingressives (Nihalani tic structures of Sindhi are also common to Indo-Aryan, 1986). The postalveolar consonants are apical and the including a canonical Subject–Object–Verb (SOV) palato-alveolar consonants are laminal; these are word order that is subject to permutation (i.e. so-called termed retroflex and palatal in traditional nomencla- free word order), question formation with no preposing ture. Sounds that are shown in parentheses in the chart of question words (i.e. question words in situ), and below are either restricted to borrowings (/z, , x, /) or dependent clauses involving the parallel relative–corel- occur only as allophonic variants. For instance, the ative construction. postalveolar rhotic tap [] is in complementary distri- Sindhi also has several features that differentiate it bution with the homorganic stop []. The aspirated from other Indo-Aryan languages. Its consonant inven- sonorants /m, n, , l ,/ occur only in intervo- tory includes four implosive stops, unique to Sindhi calic position and are never distinguished from a (and its close relatives such as Kacchi and Siraiki) sequence of sonorant + /h/, although speakers will among all Indo-European languages. Sindhi has describe them as single sounds, even though they are retained from Sanskrit the full set of five phonemic written as sequences in the Arabic orthography. The nasal stops, but reduces all historical geminate conso- vowel inventory consists of the standard symmetrical nants -CC- in the word-medial position to a singleton Indo-Aryan system, comprising five pairs of long and consonant -C-. Although the vowel inventory in Sindhi short vowels. Phonetic values shown below are as iden- is common to Indo-Aryan, the status of the word-final tified by Nihalani (1995). In addition, all the vowels short vowels is remarkable. Final short /i,u,a/ express can occur with nasalization; the long nasal vowels have grammatical information such as number, gender, and phonemic status, but nasalization on short vowels can case on nouns, and yet they are produced with extreme- be ascribed to the context of a following tautosyllabic ly short phonetic duration. These final vowels are typi- nasal stop (see Table 1). cally not recognized by nonnative speakers, despite their important morpho-syntactic function, and are not retained in Sindhi words (e.g. proper names) that are Morphophonology adopted into Urdu, Hindi, or English. Although sylla- Sindhi has a complex system of morhophonological ble structure in Sindhi is similar to other Indo-Aryan vowel alternations that affect stem-final vowels in cer- languages, Sindhi does not allow word-final conso- tain morphological constructions. These changes are nants. Indeed, many Sindhi words that end in a final highly idiosyncratic to the specific construction. For super-short vowel have cognate forms in related lan- example, nouns are marked for plural number by mod- guages that end in a consonant. With respect to mor- ification of the stem-final thematic vowel alone or in phology, Sindhi differs from most Indo-Aryan combination with a plural suffix, resulting in plural languages in its use of pronominal clitics that attach to endings marked variously by /-a, -a, -u, -iu/, or nouns, postpositions, and verbs. These clitics take the unmarked, depending on the gender class of the noun place of full noun phrases to express e.g. the possessor and on its lexically determined stem-final thematic of a noun (hata-mi ‘my hands’), the complement of a vowel: ar-a ‘houses’, pit-a ‘wheels’, ka-u  ‘cots’, postposition (i-mi ‘belonging to me’), and with dif- ili- u  ‘cats’, rati ‘nights’. ferent clitic sets, the subject, object, or other comple- There are also very many verb stems that undergo ment of a verb (udaya -a ‘I shall tell you’). irregular allomorphy in the formation of the Linguistic studies of Sindhi are few in number, but Unspecified Perfective (i.e. simple past) tense. A com- include instrumental phonetic studies (Nihalani 1986, prehensive discussion of these alternations is provided 1995), sociolinguistic and dialect studies (Rohra 1971, in Trumpp (1872) and Grierson (1919). Bhugio 2001), and contemporary grammatical analysis (Khubchandani 1961). Published Sindhi grammars Morphology include Trumpp (1872), Grierson (1919), and Yegorova (1971). Nouns, Adjectives, and Pronouns Nouns are classed by grammatical gender, and this classification determines the declension pattern for the Key Structural Features marking of number and case, expressed primarily Sound Inventory through stem vowel alternations. The noun stem ends The Sindhi sound inventory is shown below, and is in a thematic vowel, which for most nouns serves to notable for its inclusion of four implosive stops, which mark the gender class, although there are numerous derive from Middle Indo-Aryan geminate voiced stops exceptional stems whose gender is not predictable on in medial position, and from singleton voiced stops in the basis of the final vowel. All nouns are marked for the initial position. Phonetic studies show that these number and case through noun stem modification, 966  SINDHI TABLE 1 Sindhi Consonant Inventory Labial Dental Alveolar Post alveolar Palatoalveolar Velar Glottal Stop p b t d   k  p b t d   k  Implosive   ! Affricate t t d d Nasal m n   " (m) (n) () Fricative f s (z) () (x ) h Rhotic r ()  Lateral l (l) Glide  y () Sindhi Vowel Inventory i [i] e [e] a [#] o [o] u [u] i [$] e [] a [] o [] u [%] involving a change in the thematic vowel, sometimes Genitive accompanied by suffixation. Patterns of noun stem (masc.) gara jo (daru) ‘(door) of the modification that mark the plural and the nominative, house’ oblique, ablative and vocative cases are illustrated in house-masc.,sg.,Obl.|Gen- the example noun paradigms as shown in Table 2 for masc.,sg.,Nom.|door-masc.,sg.,Nom. distinct thematic vowels, grouped by gender class. (fem.) gara ji (dari) ‘(window) of the The dative, ablative, comitative, and locative cases house’ are marked through the use of a postposition following house-masc.,sg.,Obl.|Gen- the noun in the oblique singular form. The genitive post- fem.,sg.,Nom.|window-fem.,sg.,Nom. position has eight variants (of which only two are shown below); it is declined like an adjective and agrees Adjectives agree with the noun they modify in gender, in number and gender with the possessed noun, but number, and case, marked through the same kind of bears the case specification of the possessor. There is no alternation of the thematic (final) vowel: nano garu accusative postposition, and instead the nominative or ‘small house, Nom.’ nana gara ‘small houses, dative form is used (depending on the animacy of sub- Nom.’, nandi meza ‘small table, Nom.’, nandiyu  ject and object). Also, the oblique case forms are used mezu  ‘small tables, Nom.’. Pronouns are marked for to mark ergative subjects (see Syntax section below). number, but case is marked only for nominative and oblique (no ablative or vocative pronoun forms). Gender Case/number marking with postpositions in exam- is distinguished only in the nominative, third-person sin- ple phrases with g‫ܒ‬ar- ‘house’ gular forms. The third-person pronouns are also used for demonstratives, for which there are distinct proximal Dative gara ke ‘the house (direct and distal forms: hiu ‘this, masc.’, hu ‘that, masc.’, hia object)’ ‘this, fem.’, hua ‘that, fem.’ The personal pronouns Ablative gara ka ‘from (the direction shown in Table 3, including dialectal variants, and simi- of) the house’ lar patterns of declension occur with the distinct relative gara ma ‘from in the house’ and correlative pronouns (jo
relative ‘who/that- gara ta ‘from on the house’ masc.,sg.,Nom.’; so
co-relative ‘who/that-masc., Comitative gara sa ‘with the house’ sg., Nom.’), and the indefinite pronouns (ko
indefinite Locative gara me ‘in the house’ ‘anyone/someone-masc.,sg.,Nom.’). Of the interrogative gara te ‘at/on the house’ pronouns, only ‘who’ is marked for number and case gara madi ‘within the house’ (e.g. keru ‘masc.,sg.,Nom.’, ka h ‘masc./fem.,sg.Obl.’). 967 SINDHI TABLE 2 Masculine: - u ‘flower’ - o ‘boy’ - u ‘resident’ Nom.sg. ulu tokiro rahaku Nom.pl ula tokira rahaku Obl.sg. ula tokire rahakua, rahakui Obl.pl. ulani tokirani rahakuani, rahakuni Abl.sg. ula tokira rahakua Abl.pl. ulania tokirania rahakuania Voc.sg. ula tokira rahaku Voc.pl. ula tokira rahakua Feminine: - a ‘table’ - i ‘eye’ - a ‘world’ - i ‘notebook’ Nom.sg. meza aki duniya kapi Nom.pl mezu aki duniyau kapiu Obl.sg. meza akiu duniya kapia, kapia Obl.pl. mezuni akiuni duniyauni kapiuni Abl.sg. meza akia duniya kapia Abl.pl. mezunia akiunia duniyaunia kapiunia Voc.sg. meza aki duniya kapi Voc.pl. mezu akiu duniyau kapiu TABLE 3 Personal Pronouns First Person Second Person Third Person (Proximal) Singular Nom. au, a ; ma, mu tu hi, hiu (masc.); hi, hia (fem.) Obl. a, mu, mu ; ma to hina Gen. a-jo, muhu-jo, mu-jo tuhu-jo, tuh&'-jo hina-jo Plural Nom. as&' tavah&', tav&', tah&' ; hi, he avah&' Obl. asa, asah&', asahu tavaha, taha ; avaha, hinani aha Gen. asa-jo, asah&'-jo tavha- jo ; avha-jo hinani-jo Verbs Syntax The verb complex in Sindhi consists of a primary verb, The neutral word order in Sindhi is Subject–Object– alone or followed by an auxiliary verb. An operator or Verb, although these elements can be permuted in modal verb element may also occur, placed in between any order, in which case the first element is typically the the primary verb (in participle form) and the auxiliary Topic. Within phrases, word order is fixed, with verb. These parts combine in various ways to produce the head element always at the end. Thus, the determin- 17 distinct finite verb forms that encode aspect (per- er and adjective precede the head noun in a noun phrase: fective, imperfective, unspecified), tense (past, pres- hia suhii tokiri this|beautiful|girl ‘this beautiful girl’, ent, future), mood (subjunctive, imperative, asa ja a akra our|two|goats ‘our two goats’. presumptive, counterfactual), and concordance (gen- Similarly, the verb appears at the end of the verb phrase, der and number). Aspect is expressed in the choice of following any (nonsentential) arguments: ama-ke tii the primary verb form, and is marked by suffixation to liki mother-to|letter|wrote ‘wrote a letter to mother’. the verb stem. In most finite verb forms, tense, mood, The subject of the sentence appears in the nomina- and concordance features are expressed on the auxil- tive case with two exceptions: “experiencer” subjects iary verb, through suffixation or auxiliary verb stem of verbs expressing physical sensation, psychological allomorphy. There are also six nonfinite verb forms state or kinship are in the dative case, and the ergative that function as nominal, adjectival, and adverbial par- subject of a transitive verb in the perfective aspect is in ticiples. Examples of finite and nonfinite verb forms the oblique case. The verb agrees with the subject if it are shown in Table 4. Each finite and nonfinite verb is nominative, and otherwise agrees with a nominative form can undergo further modification, not shown (inanimate) object if present. If there is no nominative here, to express voice (active/passive) and valence subject or object, the verb takes as default the third (transitive/causative) distinctions through the use of person, masculine, singular agreement. An indirect suffixes that attach directly to the verb stem. object appears with the dative postposition. There is 968  SINDHI TABLE 4 Examples of Finite Verb Forms with the Verb Stem hal- ‘to go’ Contingent future hale ‘if he goes’ Present unspecified hale to ‘he goes’ Contrafactual unspecified hale ha ‘had he gone’ Definite future halando ‘he will go’ Present habitual halando ahe ‘he goes’ Past habitual halando huyo ‘he used to go’ Presumptive imperfective halando hundo ‘he is probably going’ Subjunctive imperfective halando huje ‘(perhaps) he goes’ Present continuous hali rahyo ahe ‘he is going’ Past continuous hali rahyo huyo ‘he was going’ Unspecified perfective halyo ‘he went’ Present perfective halyo ahe ‘he has gone’ Past perfective halyo huyo ‘he had gone’ Presumptive perfective halyo hundo ‘he must have gone’ Subjunctive perfective halyo huje ‘he may have gone’ Past iterative halyo the ‘he would go’, ‘he often went’, ‘he used to go’ Imperative halu ‘Go!’ (informal) halo, halije ‘Go!’ (polite) Examples of Nonfinite Verb Forms with Various Stems Infinitive halau ‘to go’ aiu ‘to sing’ Adjectival unspecified mario ‘about to be struck’ (masc.,sg.) waio ‘about to be taken’ Adjectival imperfective halando ‘going’ (masc.,sg.) marindo ‘striking’ Adjectival perfective halyalu, halyo ‘went’ (masc.,sg.) maryalu, maryo ‘struck’ Adverbial imperfective halande ‘(as he was) going’ marinde ‘(as he was) striking’ Adverbial perfective hali, hali kare ‘(as he) went’ mare, mare kare ‘(as he) struck’ no distinct accusative case; a direct object appears Historical Development and Sociopolitical Factors with the dative postposition for animate objects, and is Sindh is home to the ancient Indus Valley civilization in the nominative case for inanimate objects. of Mohen-jo-daro, but a historical link between Sindhi Examples of verb agreement with nominative case and the language of Mohen-jo-daro has not yet been subjects firmly established. Sindh has been subject to foreign rule at many times in its history, and there is evidence hua hale ti of language contact in numerous borrowings from they,fem.,Nom.| go| Aux.,3pl.fem. Persian, Arabic, and English. The formation of ‘They (fem.) go.’ Pakistan in 1947 had a tremendous impact on Sindhi, hu tii paanda as millions of mostly Urdu-speaking immigrants came they,masc.,Nom.| letter,fem.,sg.,.Nom.| will to Sindh from India. The immigrants were under no read,3pl.masc. pressure to learn Sindhi, and enjoyed the advantage of ‘They (masc.) will read the letter.’ speaking Urdu, the language chosen as the national language of Pakistan. On the other hand, Sindhi speak- Examples of verb agreement in sentences with ers, especially those in urban areas, were compelled to oblique case ergative and experiencer subjects learn Urdu and, in recent years, English. Bughio’s hunani tii pai ahe (2001) sociolinguistic study reports census data that they,masc.,Obl.| letter,fem.sg.,Nom.| paint a bleak picture of the diminishing status of read,perf.,3sg.,fem.| Aux,3sg. Sindhi. Census data from 1941 show that 82% of the ‘They (masc.) have read the letter (fem.).’ Sindh population claimed Sindhi as their mother tavaha ke hunani sa aa"o pavando tongue, while in 1981 this number fell to 52%. Sindh you,pl.masc.,Obl.| Dat.| them,pl.,Obl.| with| has a large rural population, with 57% of the total pop- go,Inf.,Obl.| must,3sg,masc. ulation, and in rural areas Sindhi is the mother tongue ‘You (pl.) must go with them.’ for 78% of the population, compared to only 18% in 969 SINDHI urban areas. The rural and urban populations of Sindh mens of Sindhi and Lahnda. Calcutta: Superintendent also differ in literacy, with 51% literacy rates in urban Government Printing. Khubchandani, Lachman M. 1961. The phonology and mor- areas compared to only 16% in rural areas of which phophonemics of Sindhi. Unpublished M.A. thesis females constitute only 5%. The decline in Sindhi liter- University of Pennsylvania. acy has implications for the survival of the rich body of Masica, Colin P. 1991. The Indo-Aryan languages. Cambridge: Sindhi literature, dating from the sixteenth century Cambridge University Press. (Ajwani 1970). Nihalani, Paroo. 1986. Phonetic implementation of implosives. Language and Speech 29. 253–62. Nihalani, Paroo. 1995. Sindhi. Journal of the IPA 25(2). 95–98. Rohra, Satish K. 1971. Sindhi, Kacchi, and emigrant Sindhi. References Indian Linguistics 32(2). 123–31. Ajwani, Lalsingh H. 1970. History of Sindhi literature. New Stack, Captain George. 1849. Dictionary, English and Sindhi. Delhi: Sahitya Akademi (reprinted 1994). Bombay: American Mission Press. Bughio, Qasim. 2001. A comparative sociolinguistic study of Trumpp, Ernest. 1872. Grammar of the Sindhi language. rural and urban Sindhi. Muenchen, Germany: Lincom London: Tru(bner and Co. and Leipzig: FA Brockhaus. Europa. Yegorova, R.P. 1971. The Sindhi language. Moscow: Naukua Grierson, Sir George A. 1919. Linguistic Survey of India, Vol. Publishing House. VII, Part I: Indo-Aryan family: northwestern group, speci- JENNIFER COLE Sino-Tibetan Languages The Sino-Tibetan languages are spoken by (a) approx- in Northern Siberia, e.g. Sergei Starostin (1984). imately 1.2 billion speakers of Chinese and (b) over 30 These speculations seem to be based on incomplete or million speakers of Tibeto-Burman. The language inappropriate evidence, which cannot be substantiated family extends over a vast linguistic area stretching at the present state of knowledge. some 3,500 miles across much of eastern Asia, the The internal composition of Sino-Tibetan, as nar- Tibetan plateau, and southeast Asia — north- and east- rowly defined, appears to be less problematic, despite ward from the northeastern corner of China along the lack of general consensus. The two major divisions of eastern coasts, southward along the Irrawaddy and Sino-Tibetan — Chinese and Tibeto-Burman — can Salween Rivers to the Bay of Bengal, and westward be divided into seven subgroups each. Figure 1 illus- along the Himalayas, extending further down to Nepal trates the genetic schema. and India. Besides, Chinese is spoken by several mil- The exact number of Tibeto-Burman languages or lion overseas Chinese in communities across the dialects is not known, partly because access to some of world. Sino-Tibetan languages comprise the world’s them by scholars from outside is difficult, and partly second largest language family after Indo-European. because the names and designations of a number of As the preceding figures show, Chinese (Sinitic) languages are quite confusing. Thus, some languages accounts for most of the family’s total speakers. listed under different names may turn out to be the Despite recent progress, opinions are still sharply same language, while others that are lumped together divided as to the membership of Sino-Tibetan. Many under one name may in fact be different languages Chinese scholars hold that Hmong-Mien (previously altogether. Besides, a number of languages share the called Miao-Yao) and Tai-Kadai families should be same name with non-Tibeto-Burman languages. For included, a quarto-partite scheme proposed by Fang- example, the Nung language of Tibeto-Burman must Kuei Li (1973 [1937]). This hypothesis is generally be distinguished from the Nung language of the out of favor with other Sino-Tibetanists, who, influ- Tai-Kadai family in Guangxi and North Vietnam. enced by Paul Benedict, believe that the large-scale Similarly, the Kiao/Kiu language of Tibeto-Burman in resemblances of structure and vocabulary among the China is not the same as that of the Dravidian language languages in question are due to diffusion rather than in India. Ethnologue lists some 300 languages, a sig- genetic inheritance. Some linguists connect Sino- nificant number of which are found across national Tibetan with Austronesian, e.g. August Conrady boundaries. Many still await description. A list of (1916), Laurent Sagart (1993). Others link Sino- Tibeto-Burman languages and dialects with subgroup Tibetan with Yeniseian and North Caucasus languages affiliation is given in James Matisoff (2003:695–704). 970  SINO-TIBETAN LANGUAGES Sino-Tibetan Tibeto-Burman Chinese Kamarupan Baic Yue Mandarin Himalayaish Karenic Kejia Wu Qiangic Min Xiang Lolo (Yi) - Jinppho-Nungish-Luish Burmese-Naxi Gan Figure 1. Proposed Sino-Tibetan family tree. In terms of diversity, India has the largest number of The assumed ancestral language Proto-Sino- languages listed (137), followed by Myanmar (around Tibetan is thought to have split some 6,000 – 7,000 80) and China (around 60). Of these, several languages years ago, with the homeland being somewhere on the number over a million (Burmese, some 25,000,000; Himalayan plateau. Proto-Tibeto-Burman is estimated Tibetan, Yi [previously called Lolo], Kham, Tujia, to have a time depth of around 6,000 years (Matisoff each with over a million). A small number of lan- 1991). After their split with Tibeto-Burman, Han guages have probably less than a hundred speakers or Chinese at first flourished in the Middle Kingdom semispeakers (Xiandao or Samtao, Lui, Samei). (roughly present Shaanxi, Henan, Hebei, Shandong Sinitic languages appear to be more homogeneous and vicinities) before they established a strong pres- in mutual intelligibility than their Tibeto-Burman ence south of the Yangtze River around the beginning counterparts. This is partly because they use a common of the current era (221 BC) (Chang 1999). In the writing system as a lingua franca. Also, it is customary course of intermingling with the coterritorial non- to talk about Chinese as having seven main dialect Chinese populations, Chinese speakers have absorbed groups (q.v.), each with several subdialects, but the elements of material cultures and vocabulary from actual number of languages or dialects could be much their neighbors. An example is the Yue and Min higher. One might even count them in the hundreds, as dialects in south China which possess features that are mutual intelligibility can be low (e.g. Mandarin vs. found to have been borrowed from Kam-Tai, indicat- Cantonese; Wu vs. Kejia) and they are quite diversified. ing intense contact with the latter. In religion and phi- Indeed, dialect differences can be so drastic that, in losophy, the influence of Indian culture is prevalent, some cases, varieties could reasonably be considered particularly in Tibeto-Burman. distantly related languages. This diversity probably The Sino-Tibetan languages share a set of typolog- speaks for dual (Altaic and Austronesian) or even mul- ical features. The vast majority of lexical roots in tiple origin of the Chinese language, as suggested by a Sino-Tibetan languages are monosyllabic. Most Sino- number of scholars in the field. Tibetan languages use intonational pitch (tone) to dis- Several writing systems are used for Sino-Tibetan tinguish word meanings, with the exception of the languages. Separate traditions characterize Chinese, Bodo-Garo languages of Assam in India and a few Tibetan, and Burmese. The Chinese writing system Tibetan dialects such as Purik and Ambo, which lack goes back at least 3,000 years. It contains pictograph- tones. The overwhelming majority of Sino-Tibetan ic elements that are preserved in the oracle and animal languages possess noun classifiers. Derivational and bone inscriptions and in bronze inscriptions. The morphological affixes can be shown to have existed at Tibetan writing system was derived from the seventh- one time in Sino-Tibetan, for which reconstructions century Gupta script in a form representing the north- have been proposed at the Proto-Sino-Tibetan level: s- western variety of the ancient Nam and Zhang-zhung (causative), -s (nominalizing), -t (transitivizing). languages. The Burmese writing system dates back to Morphological processes are characteristically repre- at least the eleventh century. It was based on Old Mon sented in initial consonant alternation and vowel alter- and Pyu, an extinct Sino-Tibetan language found to nation to produce ‘word families’, that is, variant have existed in Burma since AD 500. The Tomba forms that are partially related to each other phono- script of Naxi (Moso) was developed around the logically and semantically, comparable to ‘doublets’ twelfth–thirteenth century; this pictographic script was in Indo-European. used mainly for religious purposes. Xixia (His-Hsia or However, the Sino-Tibetan languages are quite dif- Tangut), a Tibeto-Burman language developed in the ferent in word order. Sinitic languages have the eleventh to thirteenth century in northwestern China, S(ubject)–V(erb)–O(bject) and modifier–modified now extinct, was based on the Chinese model. word order, Tibeto-Burman are basically SOV and 971 SINO-TIBETAN LANGUAGES modified–modifier, except for the Karenic languages research collaboration, the Sino-Tibetan Etymological of Burma and Thailand, the Bai language of Yunnan Dictionary and Thesaurus (STEDT) Project, was initi- and Tujia of Hunan and Hubei, which have the SVO ated in Berkeley in 1987 under the directorship of order like Chinese. James Matisoff, culminating in an 800-page mono- The idea of Sino-Tibetan as a language family is graph, Handbook of Proto Tibeto-Burman (Matisoff based on the assumption that Sino-Tibetan languages 2003). This impressive volume contains several hun- are all related and are derived from a common source. dred cognate sets for Tibeto-Burman. It offers many Since the late nineteenth century, particularly the early insightful observations and discussions of morphologi- twentieth century, scholars have made various attempts cal and derivational processes, which firmly establish to verify the Sino-Tibetan hypothesis. This typically Tibeto-Burman as a major branch of Sino-Tibetan in its involves searching for cognates (words derived from own right. However, due to insufficient data, the hand- the same historical source) by comparing reconstructed book is able to provide only partial reconstructions of forms between Chinese and Written Tibetan. Among Proto-Tibeto-Burman forms. Problems involving the important successes were Robert Shafer’s multivol- reconstructed tones for Tibeto-Burman have yet to be ume Introduction to Sino-Tibetan (1966–1974), as well resolved. A number of proposed protoforms still cannot as Benedict’s (1972) Sino-Tibetan: A conspectus, be verified due to incomplete data. Several etyma are where a sizeable number of basic words are shown to set up provisionally without having established step- be related, such as the following: by-step sound correspondences in advance. These inef- ficiencies can only be overcome when solid work on English Chinese Tibetan the reconstructions of individual subgroups is carried sun njit nyi-ma out. In addition, the issue of loan contact is becoming moon nguat s-ngw(y)at more acute as it is often exceedingly difficult to sepa- eye myak mig rate early loans from inherited cognates. The position fire hwər me of Tujia and Bai within Tibeto-Burman needs to be fur- fish ngjag nya ther examined. I ngag nga Sino-Tibetan represents one of the oldest and most diverse civilizations and cultures in the world. Subsequent efforts include Coblin (1986), where Although remarkable progress has been made, the some 500 cognate sets are proposed. Austin Hale field is still in its infancy. Thus, most proposals for (1982) offers a critique of the research situation up to external affiliations and internal subgroupings must be the early 1980s. A state-of-the-art survey is provided considered working hypotheses rather than firm con- in Thurgood and LaPolla (2003). clusions. Deeper investigations, in collaboration with On the Chinese front, the great Swedish sinologist allied fields of archeology, history, anthropology, and Bernhard Karlgren conducted systematic studies of human genetics, will allow verification of these Chinese dialects and the phonetic series from histori- hypotheses, and will enhance our understanding of the cal and philological sources. He also considered complexities of cultural history in this region. Chinese loanwords in neighboring languages such as Japanese. On the basis of this combined evidence, he was able to propose several thousand reconstructed References forms for Old Chinese. These serve as the basis for subsequent enquiries. Further work along these lines Baxter, William. 1992. A handbook of old Chinese phonology. has helped to refine the reconstructions (William Trends in Linguistics, Studies and Monographs No. 64. Berlin/New York: Mouton de Gruyter. Baxter 1992). Although much remains to be done, a Benedict, Paul K. 1972. Sino-Tibetan: a conspectus (Contri- basic phonological system of Proto Chinese can be buting editor: James Matisoff) (Princeton–Cambridge established on the basis of the available materials. Studies in Chinese linguistics, Vol. 2.). Cambridge and New Similar achievements characterize Tibeto-Burman York: Cambridge University Press. studies. The 1980s and the early 1990s witnessed an Bradley, David. (ed.) 1997. Tibeto-Burman languages and clas- sification. Tibeto-Burman languages of the Himalayas, 1–72 influx of new data, with detailed descriptions of a sig- (Papers in Southeast Asian Linguistics, No.14). Pacific nificant number of individual languages. Particularly Linguistics, Series A, No. 86. Canberra: Department of worth noting are two major comparative dictionaries by Linguistics, Research School of Pacific and Asian Studies, Chinese researchers (Sun et al [eds.] 1991; Dai, Huang Australian National University. et al. [eds.] 1992), which contain valuable information Chang, Kwang-Chih. 1999. China on the eve of the historical period. The Cambridge history of ancient China: from the on some 50 Tibeto-Burman languages in China. origins of civilizations to 221 B.C., ed. by Michael Loewe Bradley (1997) reports on research in Tibeto-Burman and Edward L. Shaughnessy, 37–73. Cambridge: Cambridge languages of the Himalaya region. In America, a large University Press. 972  SLOBIN, DAN ISAAC Coblin, W. South. 1986. A sinologist’s handlist of Sino-Tibetan Matisoff, James A. 2003. Handbook of Proto Tibeto-Burman: sys- lexical comparisons, Monumenta Serica Monograph Series tem and philosophy of Sino-Tibetan reconstruction. xviii. Nettetal: Steyler Verlag. Berkeley/Los Angeles/London: University of California Press. Conrady, August. 1916. Eine merkwürdige Beziehung zwis- Sagart, Laurent. 1993. Chinese and Austronesian: evidence for a chen den austrischen und den indochinesischen Sprachen. genetic relationship. Journal of Chinese Linguistics 21. 1–65. Aufsätze zur Kultur und Sprachgeschichte, vornehmlich des Shafer, Robert. 1966–1974. Introduction to Sino-Tibetan, 5 Orient, Ernst Kuhn zum, vol 70, 475–504. Breslao Marcus. parts. Wiesbaden: Harrassowitz. Hale, Austin. 1982. Research on Tibeto-Burman: languages. Starostin, Sergei A. 1984. On the hypothesis of a genetic con- Trends in Linguistics, State-of-the-art Reports, 14. Berlin: nection between the Sino-Tibetan languages and Yeniseian Mouton. and North Caucasian languages. In: Linguisticheskaja Karlgren, Bernhard. 1957. Grammata Serica Recensa. BMFEA rekonstrukcija i drevnjaja istorija Vostoka, Vol 4, 19–38. 29. 1–332. Moskow: Nauka. Li, Fang-Kuei. 1973. Languages and dialects of China. Journal Thurgood, Graham. 2003. Sino-Tibetan subgroupings. Sino- of Chinese Linguistics 1. 1–13. Tibetan languages, ed. by Graham Thurgood and Randy Matisoff, James A. 1991. Sino-Tibetan linguistics: present LaPolla, 3–21. London and New York: Routledge. states and future prospects. Annual Review of Anthropology 20. 469–504. YONGXIAN LUO Slobin, Dan Isaac Dan Isaac Slobin is a prolific data-gatherer and avid tion, Grammatical transformations in childhood and investigator of child language acquisition. Growing up adulthood (Harvard University, 1963), used reaction in the polyglot, immigrant world of Russian Jewish time to study syntactic processing in both adults and immigrants in the US Midwest, and hearing English, children. His dissertation, in which he found an inter- Yiddish, Russian, and Hebrew around him, Dan action between semantics and syntax, was one of the learned the languages, music, and literature of these first counterthrusts to purely syntactic processing. The cultures. With a father who was a high school teacher dissertation was also significant as it incorporated the of history who loved to travel, Dan and his family diverse interests of his three mentors: George Miller lived in Europe and traveled all over the United States (experimental psychology), Jerome Bruner (cognitive and Mexico. As a result, Dan Slobin became hooked psychology), and Roger Brown (child language devel- on language, languages, and cultures. Philosophically, opment). Under George Miller, Dan Slobin looked for he was fascinated by the question of how a child influences of grammatical transformations of reaction becomes a member of a culture, focusing especially on time. From Jerome Bruner, Dan considered whether problems of unconscious learning, underlying struc- people would pay attention to meaning, as well as ture, and mental organization. When he was entering form. Additionally, under the influence of Roger academia in the early 1960s, language was the most Brown, he added a developmental dimension to his obvious and accessible domain in which to look for dissertation project by including children as well as underlying patterns and their development from a psy- adults. This early interdisciplinary perspective would chological point of view. set the tone for his research career. Being at the right place and in the right time to take After leaving Harvard with a Ph.D. in Social part in the emergence of cognitive science from the Relations, Dan Slobin took his first job in the world of behaviorism, Dan Slobin’s research career Psychology Department at the University of California has attempted to synthesize the perspectives and at Berkeley, where he has remained throughout his insights of the diverse teachers and colleagues from illustrious career. At Berkeley, he was influenced by a whom he learnt at Harvard and MIT: George Miller, group of colleagues who added social, interpersonal, Jerome Bruner, Roger Brown, Eric Lenneberg, Noam and cultural dimensions to his work, which has grown Chomsky, and Roman Jakobson. to include crosslinguistic, cognitive, sociolinguistic, Dan Slobin was a graduate student at Harvard and psycholinguistic perspectives. In California, he (1960–1963) when the cognitive revolution was in its worked with Susan Ervin-Tripp, John Gumperz, John infancy. Under the influence of the first phase of Searle, Irving Goffman, Dell Hymes, and Charles Chomsky’s transformational grammar, Dan’s disserta- Ferguson. 973 SLOBIN, DAN ISAAC Dan Slobin is credited with focusing the field on Psychology at the University of California (Berkeley) the importance of addressing crosslinguistic data since 1964 (Professor since 1972), and also Research when exploring universals in child language acquisi- Psychologist, Institute of Cognitive Studies and tion by proposing that different types of languages Institute of Human Development. His other positions pose different types of acquisition problems. His include: Visiting Professor, Stanford University research over the past 35 years has been especially (1998–1999); Visiting Professor, Summer Linguistics concerned with processes of language acquisition and Institute, University of Goettingen, Germany (1992); their relations to cognition and thinking on the one Visiting Professor, University of Santiago de hand, and to culture and communication on the other. Compostela, Spain (1989); and Visiting Professor, Throughout his career, Dan has used a crosslinguistic Summer Linguistics Institute of the Linguistic Society research design in an attempt to identify universal of America, University of New Mexico, Albuquerque developmental patterns, and has formulated a general (1980, 1995). His research positions include: Institute model of ‘operating principles’ for the development of of Human Development, University of California at grammar. His multivolume Crosslinguistic study of Berkeley (1988); Institute of Cognitive Studies, language acquisition summarizes this work and pro- University of California at Berkeley (1964): Visiting vides a comprehensive contribution to the field of Scholar, Laboratoire de Psychologie, CNRS, Paris, child language acquisition. France 1964–1985; Visiting Scholar, Max-Planck- Since the late1990s, Dan Slobin has focused on the Institute for Psycholinguistics, Nijmegen, The effect of different languages on psycholinguistic Netherlands (1981, 1984–1985, 1992–1999); Visiting processes, including development and adult behavior, Scholar, Department of Linguistics, Tel-Aviv in order to account for specific semantic notions that University, Israel (1983–1985); and Visiting Scholar, are encoded in the morphology and syntax of the Department of Psychology, Bogazici University, world’s languages (e.g. the origins of grammar). Istanbul, Turkey (1969–1970, 1972–1973). He has Slobin’s most recent work is on early child language held numerous administrative and editorial positions, development — in the age range from one and a half and has received awards and grants including the to five — in a crosslinguistic framework. His interest Harvard Foundation for Advanced Study Prize is in developing relations between language and Fellowship (1960–1961); Guggenheim Fellow (1984– thought, looking for universal and language-specific 1985); and The New York Academy of Sciences Award patterns. The current focus in the Child Language in the Behavioral Sciences. He has also been a recipi- Research Laboratory is deaf children’s acquisition of ent of grants from The Ford Foundation, The W.T. sign language as a first language, observing a Grant Foundation, National Science Foundation, American Sign Language and Sign Language of the National Institutes of Mental Health, American Netherlands (SLN), comparing families in which par- Council of Learned Societies, Social Science Research ents are deaf and use a native sign language with the Council, The Sloan Foundation, Israel–United States children, and families in which parents are hearing, Binational Science Foundation, Institute of Turkish and are acquiring sign language as a second language. Studies, American Psychological Association, Thus, in his most recent work, Dan Slobin is again International Research and Exchange Board (IREX), attempting to synthesize research from historical lan- and the Mellon Foundation. guage change and grammaticization, linguistic typolo- gy, and universals, pidgin and creole linguistics, sign References language, and the study of various discourse forms, especially narrative (both oral and written). This Berman, R.A., and Slobin, D.I. 1994. Relating events in narra- tive: a crosslinguistic developmental study. Hillsdale, NJ: research incorporates experimental and naturalistic Lawrence Erlbaum Associates. methods, as well as computer-aided text analysis. Johnston, J.R., and Slobin, D.I. 1979. The development of loca- tive expressions in English, Italian, Serbo-Croatian and Turkish. Journal of Child Language 6. 531–47. Biography Slobin, Dan Isaac. 1973. Cognitive prerequisites for the devel- opment of grammar. Studies of child language development, Dan Isaac Slobin was born in Detroit, Michigan on ed. by C.A. Ferguson and D.I. Slobin, 175–210. New York: May 7, 1939. He received his B.A. in psychology Holt, Rinehart & Winston. (1960) from the University of Michigan, Phi Beta ———. 1977. Language change in childhood and in history. Kappa, M.A. (1962), and Ph.D. (1964) (Harvard Language learning and thought, ed. by J. Macnamara. New University) for his dissertation on grammatical trans- York: Academic Press. ———. 1982. Universal and particular in the acquisition of lan- formations in childhood and adulthood. He was men- guage. Language acquisition: the state of the art, ed. by E. tored by George Miller, Jerome Bruner, and Roger Wanner and L.R. Gleitman, 128–72. Cambridge: Cambridge Brown. He has held positions in the Department of University Press. 974  SOCIOLECT ———. (ed.) 1985, 1992, 1997. The crosslinguistic study of ———. 1997. The origins of grammaticizable notions: beyond language acquisition: (1985) Vol. 1. The data; (1985) Vol. 2. the individual mind. The crosslinguistic study of language Theoretical issues; (1992) Vol. 3; (1997) Vol. 4; (1997) Vol. acquisition: Vol. 5. Expanding the contexts, ed. by D.I. Slobin, 5. Expanding the contexts. Mahwah, NJ: Lawrence Erlbaum 265–323. Mahwah, NJ: Lawrence Erlbaum Associates. Associates. Slobin, D.I., and Kuntay, A. 1995. Nouns and verbs in Turkish ———. 1985. Crosslinguistic evidence for the language-mak- child directed speech. Proceedings of the 19th annual ing capacity. The crosslinguistic study of language acquisi- Boston University Conference on Language Development: tion: Vol. 2. Theoretical issues, ed. by D. Slobin, 1157–256. Vol. 1, ed. by D. MacLaughlin and S. McEwen, 323–34. Hillsdale, NJ: Lawrence Erlbaum Associates. Somerville, MA: Cascadilla Press. ———. 1996. From ‘thought and language’ to ‘thinking for KAREN WOODMAN speaking.’ Rethinking linguistic relativity, ed. by J.J. Gumperz and S. Levinson, 70–96. Cambridge: Cambridge See Also Acquisition; Miller, George; Psycholin- University Press. guistics Sociolect Sociolinguistics deals with linguistic behavior in soci- constitutes a particular problem for schools, which ety and is especially concerned with the language sit- must cope with the various linguistic forms and must uation in such social institutions as political parties, provide opportunities for different parts of the popula- governmental administration, and other nationwide tion to use the various linguistic forms in appropriate institutions. Sociolinguistics is also interested in lan- situations. Such ‘diglossic’ situations occur in many guage situations in small groups, families, and all societies and may be interpreted as a form of social other types of social organizations, addressing the dis- stratification. tribution of language varieties in such institutions. It is the scientific study of language as it varies with social differentiation. Diglossia The distribution of languages (and dialects) in a ‘Diglossia’ may be defined as the complementary use society is normally described in terms of the age, sex, of two different languages (or varieties) in different education, occupations, and ethnic membership of the situations of language use (contact) in a society. Each speakers who make up the society. Links among vari- language is appropriate in particular situations. The ous aspects of the sociolinguistic situation derive from labels ‘High’ and ‘Low’ for the varieties is probably sociological constructs that underlie the social group more an idealization than a characteristic of any real or social institution and serve as the context for the sta- social environment, but it is reasonable to use them to ble arrangement of the various factors in language suggest social relations: that is, the use of certain High environments. They are thought to provide an individ- and Low forms serves to provide an orientation for ual with the orientation needed for acceptable social speakers, a mechanism for them to confirm their behavior in any particular context. expectations. Language use can be studied in the home, the Diglossia is related to prestige factors in society: school, the church, the government, and so on. Its that is, high social prestige may accord with relative behavioral characteristics are then described and ana- competence (or lack of competence) in a High variety, lyzed. Forms of linguistic behavior that depend on and vice versa. This is a function of the standardization social factors are called ‘sociolects’. A sociolect is of the High variety. Consequently, mastering the High thus a form of language depending on age differentia- variety can also serve as a vehicle for social mobility. tion, occupational distribution, sex differentiation, Diglossic differentiation, therefore, constitutes a regional differentiation, etc. sociological model that parallels other types of social For example, in parts of Germany (e.g. Swabia), a stratification in a society: that is, it reflects general local dialect is used at home, in contact with a peer differentiation in a society. A number of sociolinguis- group, and in other private conversation, but it is nec- tic studies show that this diglossic language distribu- essary to switch to the standard language in more for- tion exists, and examine its effects on the direction mal circumstances and in contact with strangers. This and rate of development of the society. However, 975 SOCIOLECT while such studies describe the stratification in socie- nate minorities with respect to social, cultural, and lin- ty, they do not explain it; this may require a purer soci- guistic phenomena, as, for example, can be seen with ological study. respect to the Bretons, the people of Alsace-Lorraine, and the people of Provence in France, all of whom are rapidly losing their own languages in favor of French. Stratification Planning and stabilizing the minority/majority situa- Social stratification in a society is reflected in its lin- tion cannot subordinate the minority group to majori- guistic stratification, and sociolinguistics tends to ty interests. Language is a very important indicator of describe the social structure as the consequence of the interethnic relationships within a social group. linguistic stratification (i.e. the existence of sociolects) Language marks both the social identity of the indi- and its effects on the behavior of individuals. Thus, vidual and social status in society. Ethnic groups, like linguistic behavior becomes a measure of the structure other social groups (e.g. age, sex, and religious of social institutions such as governmental administra- groups), rely heavily on language to mark their sepa- tion, education, political parties, etc. Furthermore, ration from other groups — their distinctness. The lan- there appears to be a tendency to measure differences guage factor not only involves the recognition of other among ethnic, religious, and cultural institutions groups as different and the accompanying develop- exclusively along linguistic grounds. The underlying ment of cultural stereotypes; it also takes in the organ- hypothesis, then, seems to be that differences in lan- ization of all of society: thus, language is not to be guage behavior are the most important differences in interpreted as merely the medium of communication; social behavior. However, there is growing evidence it is in fact the ‘symbol of ethnicity’ par excellence. that complex social relationships cannot be described Intergroup relationships become more complex and solely in terms of their correlation with linguistic more important as ethnic groups experience a greater behavior; rather, they must be described through mod- need to assert their distinctiveness within the larger els of social function, i.e. sociological models. This is society. This phenomenon is observable in Great reflected in sociolinguistic surveys concerning main- Britain with respect to the Welsh and Scots or in Spain taining a language within a society vs. shifting from with respect to the Catalans and Basques. The notion one language variety to another. There is also a notion that equates culture, society, and language is belied in that a particular language may constitute a hindrance many instances by the efforts of various (minority) to modernization or development, but this notion must language groups to achieve self-representation and be questioned in light of the revitalization of a lan- distinctness. The best way to achieve a stable situation, guage (e.g. Hebrew in Israel) or of the extinction of a in most instances, would simply be to adopt greater functioning language (e.g. Gaelic in Ireland). The fate tolerance, because pressure on the minority to accept of any particular language or dialect in a given society majority linguistic and social norms tends to create is the result of the needs of that society to organize resistance on the part of the minority. Therefore, the social behavior. There is evidence for such a view both majority must tolerate the language of the minority as in stable multilingual situations (e.g. Switzerland, a sociolect within society. It is apparent that the argu- where there are four official languages) and in situa- ment for linguistic unity within a society is an ideo- tions where multilingualism gives way to a stable logical argument and that societies are in fact monolingual situation and/or standardization (e.g. structured organisms accommodating various lan- Malaysia, where Malay replaces local languages, and guage groups and sociolects. in many other developing countries around the world). Thus, the language situation is a result of a variety of Education social efforts involved in planning a stable society, and language change is the result of the need to change A further area of investigation concerns the intercon- society and to provide what is necessary for the func- nection between education and sociolinguistics. tional organization of that society. And such changes Research like Basil Bernstein’s (1971) generated inter- may be results of changes in the rural–urban popula- est in the systematic investigation of the educational tion distribution, in industrial environments, or in the problems of individuals and groups identified as some- integration of immigrants into society. how ‘disadvantaged’; outside Britain, more general questions are asked. Although Bernstein’ s ‘code’ model provided an insight into the relationship Minority/Majority Group Relations between language and social situation, and thus could Another aspect of research into sociolects is the be interpreted as a model for the sociolectal organiza- description of the relationship between minority and tion of society, further research has questioned both the majority groups in society. Majorities tend to domi- theoretical model and the adequacy of the empirical 976  SOCIOLINGUISTICS data. Sociolinguists are increasingly interested in the References overall awareness of language function in education, Bernstein, Basil. 1971. Class, codes, and control. London: not only in foreign language teaching but also in moth- Routledge and Kegan Paul. er-tongue teaching. Sociolinguistics can help to answer Ferguson, Charles A. 1959. Diglossia. Word 15. 325–40. relevant questions regarding language use in other soci- Fishman, Joshua A. 1972. The sociology of language: an inter- eties, and mother-tongue education can profit from the disciplinary social science approach to language in society. Rowley, MA: Newbury House. sociolinguistic differentiation of components of the Giles, R., R.Y. Bourhis, and D.M. Taylor. 1977. Towards a the- teaching process including both teaching materials and ory of language in ethnic group relations. Language, ethnic- techniques. Foreign language teaching must be highly ity, and intergroup relations, ed. by H. Giles. London: aware of the sociolectal system of a society as well as Academic Press. changes in it. Foreign language teaching was long Gumperz, John J. 1958. Dialect differences and social stratifi- cation in a North American Indian village. American dominated by psychological models of learning, partic- Anthropologist 60. 668–82. ularly the behaviorist model, but there has been grow- Hartig, Matthias. 1998. Soziolinguistik des Deutschen, 2nd edi- ing interest in the function of social rules in foreign tion. Berlin: Weidler. language learning and use. Such areas as code-switch- Hartig, Matthias, and Ursula Kurz. 1971. Sprache als soziale ing and the study of the forms and functions of cultur- Kontrolle. Frankfurt: Suhrkamp. Haugen, Einar. 1968. Language planning in modern Norway. ally dependent language use (e.g. types of address and Readings in the sociology of language, ed. by J. Fishman. the differentiation of varieties with respect to polite- The Hague: Mouton. ness) are also receiving greater attention. In turn, these Labov, William. 1966. The social stratification of English in New concerns have led to greater interest in contrastive York City. Washington, DC: Center for Applied Linguistics. analysis (across language systems, within language Rubin, Joan. 1968. National bilingualism in Paraguay. The Hague: Mouton. systems, and across cultures). It is increasingly evident Shuy, R.W. 1974. Sociolinguistic strategies for teachers in a that success in foreign language teaching is in part a southern school system. Proceedings of the 3rd International function of the ability to use the target language socio- Congress of Applied Linguistics, ed. by A.Verdoodt. functionally: mere accurate manipulation of the vocab- Heidelberg: Julius Gross. ulary and grammar of the target language is Tanner, Nancy. 1967. Speech and society among the Indonesian elite: a case study of a multilingual community. insufficient; it is also necessary to know the sociolectal Anthropological Linguistics 9(3). 15–40. system of the foreign society. MATTHIAS HARTIG Sociolinguistics Contemporary sociolinguistics is a diverse and thriv- Linguistics emerged as an academic discipline in ing field of linguistic inquiry. Its major branches (vari- the nineteenth century. Some of the founders of the ationist sociolinguistics, the sociology of language, discipline were dialectologists, such as Jules Gilliéron and the ethnography of communication/interactional in France or Georg Wenker in Germany. These men sociolinguistics) share an interest in the ways in which were interested in describing regional variation in lan- human societies organize their social lives in and guage in terms of pronunciation, word structure, syn- through language. In recent years, the theoretical mod- tax, and vocabulary, and they produced elaborate els have shifted from static ones, which assumed that dialect atlases for the areas they studied. Their concern the language use of a person was a straightforward with the ways the identities of speakers influence the expression of their social identity (social class, sex, ways they speak foreshadows the central interest of age, ethnicity, nationality, professional status, and so sociolinguistics. However, sociolinguistics in its mod- on), to social constructionist ones, which view lan- ern sense is usually dated from the 1950s. Although guage use as a practice that creates social identity. work in dialectology continued, mainstream linguis- Furthermore, the ways in which power relationships in tics had marginalized interest in social questions of society constrain the linguistic expression of its mem- language use by that time. Language as a system and bers and the interpretation of their utterances have the rules that govern that system had developed as the become a central concern. central research focus of linguistics, and interest in the 977 SOCIOLINGUISTICS speakers had largely fallen by the wayside. Increasing Eckert (2000) represents a significant step forward by frustration with this view of language, which seemed viewing linguistic variation as social practice. In this very remote from real life, led to at least three new view, speakers draw on linguistic resources to project developments, which today make up the three main certain identities while rejecting others. Linguistic vari- branches of the diverse field of sociolinguistics. ation is no longer predetermined by a speaker’s place in Dialectology had set out to describe speech differ- the social hierarchy, but has become a resource they ences between speakers of different regional back- use to stake their claims to a particular place in their grounds. It was a central assumption of that work that communities. such differences could best be observed in rural areas Another strand of sociolinguistics has been inspired and that NORMs (nonmobile, older, rural males) by widespread bi- and multilingualism. For a long would speak a dialect in its purest form. Speakers who time, mainstream linguistics took monolingualism as did not fit this bill, such as city dwellers, were thought its paradigm case, despite the fact that the majority of to vary their speech haphazardly. New Yorkers, for the world’s population is bi- or multilingual. In the instance, were known to sometimes pronounce the [r] 1950s, Einar Haugen was one of the first to study lan- in words such as card, and sometimes not. The general guage choice by bilinguals in any detail. His central assumption was that there was no pattern to this what- questions continue to be central to the sociology of lan- soever, and the phenomenon was termed ‘free varia- guage: why and how do bilinguals choose one lan- tion’ until William Labov’s work showed otherwise. In guage over the other in a particular situation? Why and the 1960s, William Labov got sales clerks in three New how do they sometimes abandon one of their languages York department stores (one up-market, one medium, and shift to another? Why and how do others maintain and one down-market) to pronounce ‘fourth floor’ for their languages? It is usually the use of minority lan- him by asking where certain products were sold. By guages that is threatened by languages of wider com- pretending to be hard of hearing, he then got them to munication. Minority status may be the result of repeat it. In this way, he got a sample of the [r] use of internal colonization (e.g. Welsh in the United upper-class, middle-class, and working-class-oriented Kingdom), external colonization (e.g. Aboriginal lan- speakers, in two different styles (casual and careful). guages in Australia), or migration (e.g. Italian in the He found that the variation was not haphazard at all, United States). State policies have a crucial influence but that the working class speakers used [r] in relevant on language maintenance or shift. Such policies may words with the lowest frequency and upper-class-ori- range from genocide and violent repression (e.g. the ented speakers with the highest. Furthermore, all Aboriginal language of Tasmania in colonial Australia; groups used more [r]s in careful pronunciation than in Kurdish in Turkey) to active promotion of bilingualism casual pronunciation. These findings of patterned (e.g. Singapore or contemporary Australia). Many social variation have since been confirmed and elabo- nations have negative attitudes toward bilingualism, rated in a number of studies of urban speech variation, and public discourses about the deficiency of bilinguals mainly in the English-speaking world. All these studies negatively affect bilingual usage (e.g. discourses about collected speech samples, often through interviews, Spanish in the United States). from a random selection of speakers, representative of Paralleling developments in variationist soci- social class, ethnic group, sex, age, style (formal vs. olinguistics, in the 1960s and 1970s, sociologists of casual speech), and so on. Whatever the variable that language sought to develop taxonomies of social situ- was studied in a specific community, this research ations that would result in specific language choices. showed that certain variables are indicative of the The concept of ‘domain’ was introduced by Joshua social background of a speaker. Many of these studies, Fishman, who showed that Puerto Ricans in New such as Labov’s (1972) work in New York or Horvath’s York consistently chose English or Spanish, depend- (1985) research in Sydney, have become sociolinguis- ing on whether their linguistic activities occurred in tic classics. However, variationist sociolinguistics has the domain of family, friendship, religion, employ- also been heavily criticized for lacking a sophisticated ment, or education. Since then, code-switching, or the social theory and for reifying social identity by saying, mixing of two or more languages, as often occurs in for instance, that working-class women speak in such bilingual speech, has been acknowledged more wide- and such a way. Particularly, language and gender ly. Furthermore, with the effect of social construc- researchers have been concerned with the static picture tionism on sociolinguistics as a whole, bilingualism of female vs. male speech that work tended to paint. researchers have also become wary of seeing language The early variationist model did not allow for individ- choice as the result of a particular social situation that ual agency, contestation, or change. Consequently, the speakers find themselves in. Rather, they have begun most exciting work in variationist sociolinguistics is to view language choice as a resource to construct currently inspired by work in language and gender. speakers’ social identities (similar to the ways in 978  SOCIOLINGUISTICS which monolinguals use different styles to project a tion to the ability to produce grammatically acceptable particular social identity). Youths in contemporary utterances, speakers also need to know when to speak multiethnic cities around the world have been found to and when to stay silent, or what is appropriate to say develop their emerging hybrid identities through elab- in a particular situation. Dell Hymes was the first to orate language choices. The choices speakers have at introduce the term ‘communicative competence’ for their disposal are crucially influenced by power rela- these culture-specific speaking rules. Parents spend tions in society, and many researchers have found it much more time to socialize their children into social- useful to integrate social constructionist approaches ly appropriate speaking rules than teaching them with views that conceptualize languages as cultural explicit grammatical rules (e.g. Don’t speak with your currencies—symbolic capital. In this view, language mouth full!). Similarly, sharing a common grammati- choice is governed by the laws of a market economy: cal system does not necessarily guarantee smooth an international language such as English is a hard- communication, as countless anecdotes about currency language, and access to this language is high- American–British misunderstandings demonstrate. ly desirable but also restricted by its symbolic owners Ethnographic research is carried out through partic- (education boards, television networks, and so on) to ipant-observation (i.e. researchers become participants safeguard the strength of their currency. in the community they wish to observe). The central An unresolved (and irresolvable) question for the goal of this data collection method is the collection of sociology of language is what a language actually is. speech that occurs naturally in a social context, rather Most people think that sociolinguists should at least be than being elicited for research purposes. The work of able to identify a language unambiguously—that they Dell Hymes also provides a framework for the com- should be able to draw clear boundaries between lan- parison of speaking practices across communities: the guages and be able to say how many languages there so-called SPEAKING grid. SPEAKING stands for are in the world today. They cannot. Even a seemingly setting, participants, ends, act sequence, key, instru- unambiguous test such as mutual intelligibility fails: mentalities, norms, and genres. speakers of Danish, Norwegian, and Swedish can Interactional sociolinguistics developed from the understand one another, yet they are said to speak three ethnography of communication and is closely associ- different languages. By contrast, speakers of Mandarin ated with the pioneering work of John Gumperz. A Chinese and Cantonese Chinese cannot understand one central idea is the one of ‘contextualization’, which another, despite the fact that they supposedly speak a posits that certain linguistic cues (e.g. prosody) are single language: Chinese. Thus, languages are defined used to create a social context and a framework for politically rather than linguistically. Max Weinreich is interpretation. Many such cues are taken for granted, often quoted as having said: ‘A language is a dialect and speakers assume them to be linguistic universals, with an army and a navy’. Indeed, the development of which in fact they are not. If speakers do not share pidgin and creole languages is another good example such presuppositions for the framing and interpreta- of this. Sociolinguists have come to recognize pidgin tion of communicative interactions, misunderstand- and creole languages as fascinating instances of lan- ings and even communication breakdown may result. guage birth. However, outsiders and even speakers Consequently, cross-cultural communication has been themselves often deride these languages as ‘not really of central interest to interactional sociolinguists. a language at all’ or ‘broken English’ (in the case of an Communication in institutional settings has been a fur- English-based creole). Only when such languages, ther central site of inquiry. Such work, whether on which typically evolved in the extreme contact situa- communication in the classroom, in the courtroom, in tions of colonialism and slavery, become the official medical settings, or in the workplace, often has practi- language of a state, such as Tok Pisin in Papua New cal applications. It seeks to uncover communication Guinea, do they get linguistic recognition. patterns that may give some interactants an advantage The third major branch of contemporary sociolin- over others in such encounters. guistics is closely connected to anthropological lin- From its inception, sociolinguistics has been con- guistics and also overlaps with conversation and cerned with power relationships and the differential discourse analysis. It is the aim of the ethnography of access to material and symbolic resources that these communication to uncover, describe, and compare the relationships result in. Many applications result from speaking practices of specific communities. Speakers’ this fundamental concern. In the field of education, communicative competence is of central importance in dialect and minority speakers have consistently been this endeavor. In general linguistics, the concept of found to be less successful than their standard-speak- competence is used to capture speakers’ subconscious ing peers. A number of court cases in the United knowledge of linguistic rules. Dell Hymes observed States have established that such a disadvantage con- that grammatical rules are only half the story. In addi- stitutes a form of discrimination. An early one of 979 SOCIOLINGUISTICS these, in Ann Arbor, Michigan, had William Labov Eckert, Penelope. 2000. Language variation as social practice: testify to the structural equality of African-American the linguistic construction of identity in Belten High. Oxford and Cambridge, MA: Blackwell. vernacular English to other languages. Outside educa- Fasold, Ralph. 1986. The sociolinguistics of society. Oxford: tion, sociolinguistic expertise finds applications in Basil Blackwell. cross-cultural and workplace communication, forensic ———. 1990. The sociolinguistics of language. Oxford: Basil contexts, doctor–patient interactions, and language Blackwell. policy and planning. Holmes, Janet. 2001. An introduction to sociolinguistics. London and New York: Longman, 1st edition, 1992. Horvath, Barbara. 1985. Variation in Australian English. References Cambridge: Cambridge University Press. Ammon, Ulrich, Norbert Dittmar, and Klaus J. Mattheier. (eds.) Meshtrie, Rajend, Joan Swann, Andrea Deumert, and William 1987–1988. Sociolinguistics: an international handbook of L. Leap. 2000. Introducing Sociolinguistics. Edinburgh: the science of language and society, 2 vols. Berlin and New Edinburgh University Press. York: Mouton de Gruyter. Romaine, Suzanne. 1994. Language in society: an introduction Chambers, J.K. 2003. Sociolinguistic theory. Oxford and to sociolinguistics. Oxford: Oxford University Press. Cambridge, MA: Blackwell, 1st edition, 1995. Wardaugh, Ronald. 1986. An introduction to sociolinguistics. Coulmas, Florian (ed.) 1997. The handbook of sociolinguistics. Oxford and Cambridge, MA: Blackwell, 3rd edition, 1997. Oxford and Cambridge, MA: Blackwell. INGRID PILLER Coupland, Nikolas, and Adam Jaworski. (eds.) 1997. Sociolinguistics: A reader and coursebook. New York: St. See also Fishman, Joshua A.; Gumperz, John Martin’s Press. Joseph; Hymes, Dell H.; Labov, William Soundwave Analysis Sound is a form of energy. It is generated by distur- time (a). The period of a wave is the duration of one bance of particles, and it travels in waves at a certain cycle, and the frequency is the reciprocal of the peri- speed, depending on the compressibility and the resist- od, expressed in Hertz (Hz). The lower the repetition ance of the medium (e.g. 344 m/s in air, 1,500 m/s in rate, the lower the perceived pitch. Amplitude repre- water). Most sounds are complex by nature, i.e. they sents the extent of vibration (b) and is correlated with consist of more than one tone, each tone with its own loudness: the lesser the amplitude, the softer the sig- frequency, amplitude, and phase. Over the years, sever- nal. The phase of the signal is the point in time at al tools have been developed to analyze speech sounds. which the waveform starts (c). Most sound analysis The first tools were based on analog techniques, but tools do not take phase differences into account, as since the 1950s digital techniques have advanced great- they hardly affect the identity of the different sounds. ly, and today software packages exist that can be used A complex sound consists of different sine waves, at home. Digital techniques are more accurate, and are each with its own frequency, amplitude, and phase. In repeatable. One of the most important approaches speech, there are two kinds of complex signals: peri- regarding the analysis of complex sounds (like speech) odic and aperiodic ones. In periodic signals, the vibra- was proposed by Joseph Fourier, a French mathemati- tion repeats itself; in aperiodic signals, the vibration cian in the second half of the nineteenth century. He has no repeatable pattern. Vowels usually have a peri- showed that a complex speech signal like speech can be odic structure due to the continuous vibration of the considered as consisting of a combination of sine vocal folds during speech production, while conso- waves (also called sinusoids or pure tones). The sine nants exhibit an aperiodic, noise-like, structure. wave is the simplest of all sounds, and it seems appro- priate to discuss its properties first. Oscillogram One of the oldest sound analysis tools is the oscillo- Sine Waves graph ( 1930). It displays sound as pressure variations Figure 1 illustrates the three main properties of a sine over time and allows the determination of relative wave: frequency, amplitude, and phase. The frequency amplitude, duration, and fundamental frequency. An of a wave refers to the repetition of a cycle per unit example of a digital oscillogram of the word ‘soldier’ 980  SOUNDWAVE ANALYSIS 0.9989 (b) (c) f = 1/T (a) 0 −0.9992 0 0.00210884 Time (s) Figure 1 is given in Figure 2A. Note that the vibration of the /s/ measurements are useful quantities because they relate is aperiodic, while that of the vowel nucleus of the /o/ directly to the power in the signal. is highly periodic. By selecting a portion of the vowel, preferably somewhere in the middle, it is possible to Spectral Analysis determine its periodicity and hence its fundamental fre- quency by taking the reciprocal of the period (f  1/T). In order to examine the characteristics of speech The fundamental frequency is the lowest frequency and sounds in more detail, (parts of) the waveform can be is the physical measure of vocal fold vibration. This submitted to a spectral analysis, also called Fourier physical property is perceived as the pitch of a signal. analysis. Recall that Fourier showed that periodic It is the primary acoustic cue to intonation and stress in waveforms, no matter how complex, can be analyzed speech and is crucial to phoneme identification in tone as the sum of sine waves varying in frequency, ampli- languages. The determination of F0 is usually done for tude, and phase. The resulting power spectrum shows several numbers of periods to minimize measuring arti- the amplitude of each different component (sine wave) facts as much as possible. The main drawback of this as a function of its frequency. These components are method is that the measurements, especially determi- also called harmonics in periodic sounds, because the nation of duration, strongly depend on the placing of frequencies of the components are multiples of the the cursors. For many speech sounds, it is very difficult fundamental frequency or the first harmonic (F0). An to determine exactly where a speech segment begins example of a spectrum of the vowel /i/ is given in and where it ends. This is illustrated in Figure 2B for Figure 3. The shape of the spectrum is given by the the /o/ of /soldier/. The end of the /s/ is easily deter- (imaginary) line connecting the amplitudes of the mined (a), but the end of the /o/ and beginning of /l/ are harmonics. Different (speech) sounds have different not marked clearly (b), so that the duration of /o/ can- spectral envelopes, because different vocal tract con- not be determined reliably. In addition to this, oscillo- figurations result in different resonant frequencies. graphic analysis does not suffice to describe some of The first component corresponds to the fundamental the important differences among vowels. Observation frequency (and hence to vocal fold vibration). of these differences requires the generation of spectral Components yielding the highest amplitude corre- representations, that is, plots of signal energy vs. fre- spond to the resonant frequencies of the sound, here at quency (see further). approximately 300, 2,000, and 3,100 Hz. Spectral analysis has proved to be a very valuable tool in dis- Root-Mean-Square Amplitude tinguishing the characteristics of the different vowels and consonants (e.g. different frequency regions of /s/ In order to determine the amplitude of the (speech) and /f/) (see Figure 3). signal, both the negative and positive parts of the waveform must be taken into account. This is done by Linear Prediction calculating the root-mean-square (RMS) amplitude of a certain number of samples. First, the samples in a A different method for estimating formant frequencies waveform window are squared, so that the portions is linear predictive coding (LPC). Like the Fourier above and below the line are positive. The squared val- transform, it relates a representation in time to one in ues in the window are then averaged and squared again frequency. However, it is not based on spectral analy- (to compensate for the initial squaring), after which sis. The hypothesis behind LPC is that a sample is another set of samples is selected. RMS amplitude partly predictable from its immediate predecessors. 981 SOUNDWAVE ANALYSIS 0.8741 0 −0.8545 0 /s/ /o/ /L/ /d/ / ier /0.383628 (a) Time (s) 0.8741 (b?) (a) 0 −0.8545 0 0.140272 (b) Time (s) Figure 2 Fourier spectrum of /i/ Sound pressure level (dB/Hz) 40 20 0 0 5000 Frequency (Hz) Figure 3 This is true under the assumption that speech does not not completely accurate, so the transmitted speech is vary too rapidly from sample to sample (which is not not perfect. Both the Fourier spectrum and LPC have always the case). LPC was originally developed as a proved to be good methods for estimating formant fre- method of compressing the signal (e.g. for speech syn- quencies. Note, however, that the two methods yield thesis). The algorithm calculates ‘predictor’ coeffi- different types of information. The Fourier spectrum cients that correspond to the vocal tract filter function shows the harmonics of the fundamental frequency. in a linear way. It does not transmit the individual sam- Formant frequencies are estimated from the harmonics ples, but determines weights and errors that are updat- with the highest amplitude. The LPC spectrum does ed every 10 or 20 ms, depending on the chosen not show harmonics, but (automatically) yields the window size. This is an immense saving compared to formant frequencies and bandwidths from the filter the original speech signal. Of course, the prediction is functions. In general, both methods compare well. 982  SOUNDWAVE ANALYSIS There is no ‘best’ method, and errors occur frequently on band-pass filtering. This method determines the at high fundamental frequencies. The higher the fun- energy of each frequency component as a function of damental frequency, the wider apart the harmonics in time and, by doing so, clearly illustrates the variation the spectrum, and the more difficult it becomes to esti- in the acoustical properties over time. While the mate the resonant frequencies. Moreover, most LPC Fourier spectrum yields information of the frequency models only account for the resonances (they are all- components of a certain speech sound in a rather stat- pole models), not the anti-resonances (zeroes) in ic way, the spectrogram displays the dynamical nature speech sounds. This means that LPC is not suitable for of a stretch of speech. In the analog spectrograph, determining formant frequencies in speech sounds that which was developed in the 1940s, the speech signal contain antiresonances, such as the nasals and laterals is analyzed by a set of band-pass filters, each of which (e.g. /m, n, l/), and some fricatives. Another disadvan- only responds to the energy within its frequency band. tage of LPC is that the number of formant frequencies In the digital version, the spectral amplitude results has to be specified before the analysis. For every two from a series of different Fourier analyses. The num- coefficients, the algorithm has to find a peak in the ber of samples in the analysis window determines the spectrum. If the number of anticipated peaks is larger width of the analysis. A short analysis window corre- than the actual number of peaks, the algorithm will sponds to a wide bandwidth, and a long analysis yield values that do not correspond to true peaks in the window to a narrow bandwidth. Figures 4A and B spectrum. Similarly, if very few peaks are specified, illustrate a wideband and narrowband analysis of ‘The some of the true peaks will not be indicated either. All soldier saluted the flag.’ In both plots, time is given on in all, it is best to use both methods on the same speech the horizontal axis, frequency on the vertical axis, and sounds, and to compare results. In this way, errors, intensity by the blackness of the pattern. However, as such as, for instance, confusing a harmonic with a for- a result of the different filter settings, there is a trade- mant frequency, can be avoided. off between temporal and spectral information. Wideband spectrograms (Figure 4A) yield high tem- poral resolution. They display the individual voicing Filtering pulses (vertical striations) as well as the resonant fre- Filters are commonly used in sound analysis. Filters quencies (broad bands of energy) of the speech pass energy at certain frequencies but not at others and sounds. Spectral changes over relatively brief inter- therefore allow a selective look at the energy in a cer- vals of time are clearly visible. Note the high-fre- tain frequency region. A low-pass filter removes all quency turbulence of the /s/, the dynamic formant frequencies above a given cut-off frequency; a high- patterns, the silence preceding unvoiced plosives, etc. pass filter removes all frequencies below a given cut- In a wideband analysis frequency resolution is low, off frequency. A band-pass filter allows frequencies and the fundamental frequency is difficult to deter- between two cut-off frequencies to pass (our telephone mine, especially when it is smeared together with the network between 300 and 3,400 Hz !), while a band- lowest formant frequency. It is possible to determine stop filter removes the information between two given the average fundamental frequency by calculating the frequencies. In speech, a preemphasis filter is often number of glottal pulses per second (vertically), but applied to boost the high-frequency components in this is a time-consuming and not very reliable method. amplitude relative to the low-frequency components. It The fundamental frequency can be better determined is actually a high-pass filter with a response that usu- from a narrowband spectrogram (Figure 4B), where ally increases at 6 dB per octave. A preemphasis filter the fundamental frequency and its harmonics are dis- is necessary, because most of the energy in speech is played as equally-spaced horizontal lines within the in the lower frequency range. This energy will tend to broader formants. A narrowband spectrogram has a dominate the analysis if there is no equalization of high-frequency resolution, more fine-grained infor- energy across frequency. Also, before sampling, the mation, but low temporal resolution. Usually, the input signal must be low-pass filtered at the value of average fundamental frequency is determined by the maximum frequency of interest (usually 10 kHz counting the first ten harmonics (see arrow in Figure when the sampling rate is 20 kHz). This antialiasing 4B), and by dividing the corresponding frequency on filter sharply attenuates frequencies above half the the y-axis by ten (see Figure 4). sampling rate to avoid aliasing effects. Instead of spectrographic analysis, it is also possible to make use of special algorithms to track the fundamen- tal frequency in the speech signal. However, these pro- Spectrographic Analyses grams also make characteristic errors, such as confusing Spectrographic analysis, an important method for the first formant (resonant) frequency with the funda- visualizing and examining (speech) sounds, is based mental frequency, doubling the fundamental frequency, 983 SOUNDWAVE ANALYSIS Broadband spectrogram 4 10 Frequency (Hz) 0 0 The s o l d ier s a l u t e d the f l a g 1.39243 Time (s) Narrowband spectrogram 5000 Frequency (Hz) 0 0 The s o l d i er s a l u t e d t he f l a g 1.39243 Time (s) 0.8741 0 −0.8545 0 1.39243 Time (s) Figure 4 finding a fundamental frequency in unvoiced parts of the will be lost or errors can occur in the analysis. The signal, or failing to find it in voiced sections. higher the sampling frequency the better the sound quality, but also the larger the required storage capac- ity and computer memory. For instance, compact Digitalization, Sampling, and Quantization disks are sampled at a rate of 44.1 kHz, meaning that Once (personal) computers became popular, many the signal is sampled 44,100 times per second. It is software programs became readily available for digi- not necessary to sample speech at such a high rate, tal (speech) sound analyses. Digital techniques can because the main acoustical properties range do what the analog systems did, only more precisely between 0 and 10,000 Hz. The Nyquist theorem and faster. Moreover, digital signals can be stored states that the sampling rate should at least be twice forever without quality degradation and the stored the maximum frequency of interest; hence, a sam- signal can be reanalyzed, edited, and displayed as pling rate of 20,000 Hz is appropriate. Together with often as necessary. Before analysis, however, the sampling (discretization as a function of time) the speech signal must be transformed to a discrete rep- signal must also be quantized (discretization as a resentation. This means that the speech signal must function of level). Quantization is expressed in bits: be sampled at certain, regular, intervals of time. But an 8-bit signal has 256 different levels (28), and a 15- what should the sampling rate be? If the rate is too bit one has 32,768 different levels (215). Although the low, much of the information of the original signal increase in the number of quantization levels results 984  SOUTH AFRICA in a better sound quality (due to the fact that there is period. Moreover, formant frequency information can a more accurate resemblance between the discretized affect pitch extraction, to the extent that the output is and original signal) it also implies more memory to not the glottal period, but the glottal period plus the store the data. For the analysis of speech sound, 12- period of the first or second formant. Autocorrelation bit quantization suffices. However, for the analysis of is one of the more reliable methods of determining F0. music, finer gradations of amplitude are required, since the dynamic range, i.e. the difference between Cepstral Analysis the quietest and loudest sound of interest, is greater. Another fairly accurate pitch extraction method is cep- stral analysis. This method makes use of the periodic Autocorrelation structure of the harmonics in a Fourier spectrum. To Instead of determining the fundamental frequency determine the interval between subsequent harmonics, a (F0) from the time signal (Figure 2) or from a (nar- Fourier transform of the log power Fourier spectrum of rowband) spectrogram (Figure 4B), it can also be the original speech signal is taken. The result is a spike detected automatically by a computer algorithm. Over at a component that corresponds to the fundamental fre- the years, many computer algorithms have been devel- quency. Because of the Fourier transformation on a oped, none of which yield a 100% accurate result (one Fourier transformation, the frequency axis of the spec- should always check!). This is because of the nonsta- trum is transformed back to a time axis. Note that ‘cep- tionary nature of speech, irregularities in vocal cord strum’ is actually ‘spectrum’ with the first syllable vibration, the wide range of possible F0 values, and reversed. Similarly, the ‘quefrency’ peak ( ‘inverse of’ degraded speech in noisy environments. In general, the frequency) represents the fundamental period. autocorrelation algorithm yields a higher accuracy than the time-domain techniques, but also requires a References higher computation load. In the autocorrelation Fant, G. 1973. Speech sounds and features. Cambridge, MA: method, the average correlation of part of the speech MIT Press. signal is computed, usually a Hanning window of a Flanagan, J. 1972. Speech analysis, synthesis and perception. constant length, and a delayed copy of itself ( auto). Berlin, Heidelberg, New York: Springer-Verlag. The ‘lag’ with the highest correlation is then taken as Furui, S., and M.M. Sondhi (eds.) 1992. Advances in speech the pitch of the period. The correlation is highest when signal processing. New York: Marcel Dekker. Johnson, K. 1997. Acoustic and auditory phonetics, 169pp. the delay is close to one pitch period. Pitch extraction Oxford: Blackwell Publishers. with the autocorrelation algorithm works well, Kent, R.D., and Ch. Read. 1992. The acoustic analysis of because the spectro-temporal changes in voiced seg- speech, 238pp. San Diego, CA: Singular Publishing Group, ments are usually relatively slow compared to those in Inc. the more dynamic speech segments (e.g. voiceless plo- Ladefoged, P. 1994. Elements of acoustic phonetics. Chicago: University of Chicago Press. sives). However, the algorithm also has its drawbacks. Pickett, J.M. 1999. The acoustics of speech communication: A common error produced by the autocorrelation fundamental, speech perception theory and technology, method is pitch doubling, i.e. when a subcomponent of 404pp. Newton, MA: Allyn & Bacon. the pitch period is mistakenly identified as the pitch ASTRID VAN WIERINGEN South Africa South Africa, nestled in the southern tip of the (7.2%), Southern Sotho (6.9%), Northern Sotho African continent, covers 1,211,037 square kilome- (9.8%), Shangaan (Tsonga) (4.2%), Swazi (2.6%), ters. The population of 45.98 million (mid-2000 esti- Ndebele (1.5%), and Venda (1.7%). The white popu- mate) includes approximately 26.98 million Africans, lation is two-thirds Afrikaans-speaking (a form of 5.99 million Whites, 3.49 million ‘Coloreds’, and Dutch) and one-third English-speaking. The 1.03 million Asians. The African population is ‘Coloreds’ are of mixed race, and the Asians are made up of Zulu (18.8%), Xhosa (18.3%), Tswana chiefly Indians. The linguistic diversity and complex 985 SOUTH AFRICA history of South Africa poses interesting issues for the fifteenth century, the Nguni occupied predomi- language students. nantly the northern, eastern, and south-eastern areas of South Africa contains two of the four language fam- South Africa, while the San and the Khoi occupied the ilies (phyla) in Africa: Khoisan and Niger- southwestern regions from the Cape to the Orange Kordofanian languages. The Bantu (or Sintu) River. By the nineteenth century, the Nguni and languages are spoken in South Africa, which are the European settlers in South Africa pushed the Khoisan most widespread and best-known subgroup of Niger- speakers into the Khalahari Desert. Members of the Congo, a subfamily of Niger-Kordofanian family. isinguni (the Nguni language group) include the Sintu reflects the Nguni prefix isi-, denoting language Ndebele, Ngoni, Swazi, Xhosa, Zulu, and some small- and culture. Sintu languages in South Africa include er peoples, and speak closely related languages. Nguni languages (Ndebele, Swazi, Xhosa, and Zulu), Currently, there are 31 main languages in South Sotho languages (North Sotho, South Sotho, and Africa, of which 27 are living. The extinct languages, Tswana), Tsonga, and Venda. The Nguni, whose all Khoisan languages, are Korana, Seroa, Xam, and descendants now comprise most of the country’s black Xegwi. The nearly extinct languages are Ng’huki and majority, were established in South Africa by the third Xiri; both these are Khoisan languages. Fanagolo is a century AD. The Nguni languages in South Africa now second language without mother-tongue speakers. The comprise Xhosa, Zulu, Swazi, Ndebele, Ponso, and living languages include Afrikaans, Birwa, Chinese Tembu. They are closely related and spoken by most (Mandarin), English, Fly Taal, Gujarati, Hindi, Nama, African people in South Africa. Ndebele, Oorlans, Ronga, Northern Sotho, Southern Greenberg (1950, 1963) further developed and pop- Sotho (Sesotho), South African Sign Language, ularized the concept of Khoisan as a language phylum. Swahili, Swati, Tamil, Tsonga, Tswa, Tswana, Urdu, The most prominent feature of the Khoisan languages Venda, Xhosa, and Zulu. Other minor languages is the sound system. It contains a large number of include Dutch, German, Greek, Italian, Portuguese, ingressive sounds known as clicks, which usually French, and others. Few people in South Africa are occur word-initially, and are often believed to be rem- monolingual, most are bilingual, and many are multi- nants of the most ancient sound inventory of human lingual to varying degrees. Several languages are par- language (Vossen 2000). Also, most Khoisan lan- tially intelligible with other languages. For example, guages make use of various levels of tone, both lexi- Tswa is partially intelligible with Ronga and Tsonga. cally and grammatically, and tone-bearing units are Population distribution is extremely uneven—more syllabic segments such as vowels and nasal conso- than two thirds live in the eastern third of the republic nants. Vowels can occur as oral, nasal, pharyngealized, and in the southern Cape. Europeans have a wide- glottalized, breathy vowels, or combinations thereof. spread geographical distribution, but more than 80% Several millennia ago, the Khoisan languages pre- reside in towns. Relatively few Africans are resident in dominated all across southern Africa (including most the western Cape, and more than 60% continue to of modern-day South Africa), from the Cape of Good reside in those rural areas that comprised the formal Hope to Somalia and Kenya (Guldemann and Vossen tribal reserves. These extend in a great horseshoe 2000). The Khoisan peoples—Bushmen, Cape along the southeastern coast and up to the Northern Khoikhoi (the so-called Hottentots), and Bergdamara Province and then southwestward to the northeastern —were the original inhabitants of South Africa, prima- Cape. The ‘Colored’ population are mainly resident in rily hunters and gatherers, but few now remain. In the the Cape, and the Asian population is concentrated seventeenth century, there were still perhaps 200,000 largely in KwaZulu/Natal and the Witwatersrand. speakers of the Cape Khoikhoi dialects between the African languages are unevenly distributed through Cape and the Transkei, but disease and expropriation of the population (see, for example, ANC Constitutional grazing land by European settlers destroyed the Cape Committee 1992). Ndebele is spoken mostly in north- Khoikhoi’s economic base and language. Now, a large ern and southern Transvaal. Setswana is the language number of Khoisan languages with usually a few hun- of the Tswana people, spoken in parts of the Western dred speakers each are scattered across the Kalahari Transvaal and Northern Cape, as well as a small area region of southern Africa, including northern South within the Orange Free State. Shangaan is spoken in Africa. At present, the total population may be estimat- Northern and Eastern Transvaal. Northern Sotho is the ed to be around 200,000 (Vossen 2000:131). group of Sotho dialects spoken in the northern parts of While the San (Bushmen) and the Khoi South Africa. South(ern) Sotho is the group of Sotho (Hottentots) lived in southern Africa possibly as far dialects spoken in Lesotho and Orange Free State, as back as the Stone Age, the Nguni, whose descendants well as western Transvaal. Siswazi is the language of now comprise most of the country’s black majority, the Swazi people, spoken mostly in the Kingdom of arrived in South Africa by the third century AD. Up to Swaziland and in the Eastern Transvaal. Tshivenda or 986  SOUTH AFRICA Chivenda is the language of the Venda people, spoken from Telugu, an Indian language, to English. He argues mainly in the Soutpansberg region of the Northern that the main causes of this shift are the dominant offi- Transvaal. Isixhosa is the language of the Xhosa peo- cial status of the English language, the government lan- ple traditionally living in what is now the Eastern Cape guage policy, the assimilation between the Andhras and Province. IsiZulu is the Zulu language; Zulus live the Tamils, and political factors such as the uprooting mainly in KwaZulu-Natal, but isiZulu is spoken in the of the well-settled Andhra community by the Group Orange Free State and the Eastern Transvaal, as well Areas Act. English remains the main means of com- as Natal. munication in South Africa’s urban areas. English was declared the sole official language of Estimates based on the 1991 census (Schuring 1993) the Cape Colony in 1822 (replacing Dutch). Then, on indicate that approximately 45% of the South African the formation of the Union of South Africa in 1910, population have a speaking knowledge of English. The English was made the official language together with majority of the population speak an African language Dutch (which was replaced by Afrikaans in 1925). as home language. Only about 10% of the population Afrikaans is a variant of Dutch spoken by the seven- cite English as a home language. Of this figure, about teenth-century colonists, with some lexical and syntac- one in three English speakers come from ethnic groups tic borrowings from Malay, Bantu languages, Khoisan other than the white one (in proportionally descending languages, Portuguese, and other European languages. order, from the South African Indian, Colored, and After the establishment of the republic of South Africa Black ethnic groups) (Lanham 1996). There is present- in 1961, this policy of two official languages (English ly considerable (and overlapping) variation in the man- and Afrikaans) continued. The first democratic elec- ifestation of English in South Africa. The variety of tions were held in 1994. The Interim Constitution of accents is a consequence of both the varied regional the new South Africa was adopted in 1994, establishing origins of groups of native English speakers who came a federal, democratic, multiparty system, and identify- to Africa at different times, and the variety of the moth- ing several aspects of education and language issues. In er tongues of the different ethnic groups. For example, section 3, subsection 9c and 10d, the Interim one can distinguish between ‘ethnic varieties’ such as Constitution made special mention of the 11 official ‘Colored’, Black, South African Indian, Afrikaans languages. Eleven languages were accorded official English, and White South African English (Lanham status. The official languages in the ‘new South Africa’ 1996). And among white English speakers, there has are Afrikaans, English, isiNdebele (Ndebele), Sesotho been a traditional threefold distinction between ‘con- sa Leboa, Sesotho (southern Sotho), seSwati, Xitsonga servative’, ‘respectable’, and ‘extreme’ South African (Tsonga), Setswana (Tswana), Tshivenda (Venda), English (Lanham 1982). isiXhosa (Xhosa), and isiZulu (Zulu). Language planning, and particularly the role of the The new constitution of May 8, 1996 guarantees that English language in South Africa’s linguistic future, the government will promote and protect African lan- receives much debate. Discussions continue on the guages. It stated that the Pan South African Language practicality of the language policy, the multiple vs. sin- Board be established to promote the 11 official lan- gle language debate, ‘tribalism’, and the meaning of guages as well as minority languages such as German, language and its role in identity (Finchilescu and Portuguese, and the Indian languages. President Thabo Nyawose 1998). For example, Titlestad (1996) argues Mbedi notes the progress in the establishment of the that the English language will inevitably play a power- Commission for the Protection and Promotion of ful, leading role as an important lingua franca because Language, Cultural, and Religious Rights as enshrined of its role internationally. Since English is the interna- in South Africa’s constitution (Mbeki 1999). Several tional and intranational language of communication researchers (e.g. Crawhall 1999) criticize the current and many South Africans already speak English, he government language policies for not sufficiently pro- argues, market forces will play a role in ensuring the moting and protecting the indigenous languages of the growth of English in South Africa. However, Webb San, Nama, and the Khoe. The few surviving San and (1996) maintains that in the interests of serving the Khoi languages are all at risk of dying out in the next needs of speakers of all languages in terms of the con- generation; land and natural resource access are essen- stitutional principle of multilingualism, language plan- tial for maintaining the languages of hunter–gatherers ning should be conducted from an Afrocentric view and traditional pastoralists (Crawhall 1999). where all languages have equal rights. While linguistic pluralism is a national objective in democratic South Africa, many people engage in the References monolingual practice of English only (Webb 1999). Alexander, N. 1990. The language question. Critical choices: an Prabhakaran (1998), for example, describes the com- agenda for South Africa in the 1990s, ed. by R. Schrire. plex process of an intergenerational language shift Cape Town: Oxford University Press. 987 SOUTH AFRICA ANC Constitutional Committee. February 1992. Ten proposed Lanham, L.W. 1982. English in South Africa. English around regions for a united South Africa. Cape Town: CDS. the world: focus on Southern Africa, ed. by R. Bailey and M. Crawhall, N. 1999. Going to a better life: perspectives on the Gorlach. Ann Arbor: University of Michigan Press. future of language in education for San and Khoe South Lanham, L.W. 1996. The pronunciation of English in South Africans. International Journal of Educational Development Africa. A dictionary of South African English on historical 19 (4–5). 323–35. principles. Oxford: Oxford University Press in association Finchilescu, G., and G. Nyawose. 1998. Talking about with the Dictionary Unit for South African English. language: Zulu students’ views on language in the New Mbeki, Thabo. 1999. Protection and promotion of languages. South Africa. South African Journal of Psychology 28(2). Presidents and Prime Ministers 8(4). 26. 53–61. Titlestad, Peter. 1996. English, the constitution, and South Greenberg, Joseph H. 1950. Studies in African linguistic classi- Africa’s language future. In de Klerk. fication. VI: the Click languages. Southwestern Journal of Vossen, Rainer. 2000. Khoisan languages with a grammatical Anthropology 6(3). 223–37. sketch of Ani (Khoe). Areal and genetic factors in language ———. 1963. The languages of Africa. The Hague: Mouton. classification and description: Africa south of the Sahara, Grobler, E., K.P. Prinsloo, and I.J. Van der Merwe (eds.) 1990. ed. by Petr Zima. Muenchen: Lincom Europa. Language atlas of South Africa. Pretoria: HSRC. Webb, Vic. 1996. English and language planning in South Guldemann, Tom, and Rainer Vossen. 2000. Khoisan. African Africa: the flip-side. In de Klerk. languages: an introduction, ed. by Bernd Heine and Derek Webb, Vic. 1999. Multilingualism in democratic South Africa: Nurse. Cambridge: Cambridge University Press. the overestimation of language policy. International Journal de Klerk, Vivian (ed.) 1996. Focus on South Africa. Amsterdam: of Educational Development 19(4–5). 351–66. John Benjamins Publishing Company. CAMILLA M. COCKERTON South America: Argentina Argentina, at the southern end of South America, has Tupi-Guarani an area of 3,761,274 square kilometers and a popula- There are three Tupi-Guarani languages currently spo- tion of 36,260,130 (2001 census). It is estimated that at ken in Argentina: Chiriguano, Tapiete, and Guaraní. least 35 different languages were spoken in this territo- Chiriguano is spoken by 15,000–21,000 people in the ry at the time of the first European contact in the early province of Salta, in northern Argentina. (There are 1500s. In addition to Spanish, the official language, approximately 50,000 speakers of Chiriguano in approximately 12 indigenous languages and a number Bolivia.) This language is also spoken by a group called of immigrant languages are also spoken at present. Chané, who at present speak a variety called Chiriguano-Chané. It is claimed that the Chané Indigenous Languages originally spoke Chané, a language believed by some to be of the Arawakan family. They were enslaved The indigenous languages of Argentina are classified by the Chiriguano people, and eventually shifted to into eight linguistic groups: Quechua, Tupi-Guarani, Chiriguano. The variety spoken by the Chané shows Guaycuruan, Matacoan, Lule-Vilela, Araucanian, traces of the original Chané language, although their Chon, and the unclassified language Yahgan. (The language has not been well studied yet. Tapiete (or Ñan- numbers of speakers of these languages are estimates, deva) is spoken by approximately 400 people in since no official record for the number of speakers for the province of Salta. (There are approximately 3,200 any of the languages exists.) speakers of Tapiete in Paraguay.) It is sometimes described as a variety of Chiriguano with a considerable Quechua amount of borrowings from other languages. Guarani is Two varieties of Quechua are spoken in Argentina. mostly spoken in Paraguay and Eastern Bolivia, but One is the variety spoken mostly by migrants from there are at least three varieties of Guarani spoken in Bolivia, with an estimated 500,000–850,000 speakers. Argentina mostly in the provinces of Formosa, The second is the Quechua spoken in Santiago del Corrientes, and Misiones. Guaraní Correntino or Estero, a province in the center of Argentina, known Corrientes Guaraní is spoken mostly in the northeastern there as Quichua Santiagueño. Santiago del Estero part of the country, with a considerable number of Quechua was brought to this area by Quechua speak- migrants in larger urban areas such as Rosario, ers who accompanied Spanish expeditions in the early Córdoba, and Buenos Aires. A second variety is known years of conquest and who settled in the area. There as Mbya, and is spoken in the province of Misiones are currently an estimated 60,000–100,000 speakers. mostly by people in rural areas. The third variety is 988  SOUTH AMERICA: ARGENTINA Paraguayan Guaraní, spoken by Paraguayan immigrants Salta and Chaco by a small number of speakers until in northeastern Argentina, particularly in provinces the mid-1960s, now believed to be extinct. closer to the Paraguayan border, and in large urban cen- ters. The estimated number of speakers of the three vari- Araucanian eties of Guarani ranges from 100,000 to 1,000,000. This One language belongs to the Araucanian family, is due partly to the lack of data on the Argentinian vari- Mapudungun (also known as Mapudungu, Mapuzungun, eties, and in part due to the constant, undocumented Araucano, or Mapuche, although the last term refers influx of illegal immigrants from Paraguay. to the speakers, from mapu ‘earth’ and che ‘people’). It is spoken by approximately 40,000 people in Argentina Guaycuruan in the provinces of Neuquén, Río Negro, La Pampa and There are four languages that belong to the Guaycuruan Chubut, and various communities in Buenos Aires language family in Argentina: Mocoví, Pilagá, Toba, and province. It is also spoken in Chile. Mapudungun the now extinct Abipón, in the Chaco region. (There is spread to Argentina from Chile through the southern also one other Guaycuruan language, Kadiwéu, spoken Andes before European arrival. There was sporadic in Matto Grosso do Sul, Brazil.) Mocoví is spoken by contact between the earlier groups in Patagonia and approximately 5,000 people in the provinces of Santa Fé the Mapuche, mostly for commercial purposes, but and Chaco. Pilagá has an estimated 4,000 speakers in in the early seventeenth century, large numbers of Formosa, mostly along the Pilcomayo River. Toba is the Mapudungun speakers crossed the Andes spread- Guaycuruan language with the greatest numbers of ing into Patagonia and the central plains, known as speakers, 25,000, mostly in the provinces of Chaco and the Pampas, forcefully occupying Tehuelche terri- Formosa, but also as migrant groups in Santa Fé, Buenos tory, imposing their language and culture. This Aires, and Salta. The Abipón were the first Guaycuruan ‘Araucanization of the Pampas’ lasted until the late Indians to settle in missions, known as reducciones, nineteenth century, with the violent imposition by the mostly in Santa Fe province; their language became Argentinian government of the authority on the indige- extinct in the early twentieth century. Two other lan- nous groups of the area. guages mentioned in several colonial sources are also believed to belong in the Guaycuruan family: Payaguá Chon and Charrúa. However, the only extant records of these The languages of the Chon family were spoken in languages are brief word lists, and the evidence for their Patagonia in southern Argentina and the southern classification as Guaycuruan languages is very weak. islands. They can be divided into two groups: Island Chon and Tehuelche, reflecting more a geographic Matacoan grouping than a linguistic one. Two languages belong The Matacoan family (also known as Mataguayo or in the Island Chon group: Haush and Selk’nam, which Mataco-Mataguayo) has three languages, spoken in were spoken in the island of Tierra del Fuego. Haush, northern Argentina: Wichí, Chorote, and Chulupí. or Manek’enk, was spoken in the southeastern tip of (The fourth Matacoan language is Maká, spoken in the island, and became extinct in the early twentieth Paraguay.) Wichí is commonly known as Mataco, but century. Only brief vocabularies of Haush remain, this is a pejorative term rejected by speakers of the lan- dating from the late nineteenth century. Selk’nam, or guage. It has between 40,000 and 60,000 speakers in Ona was spoken in the central and northwestern parts Chaco, Salta, and Formosa provinces. Chorote is spo- of the island. Although the language was not exten- ken in Salta by approximately 2,000 people (and an sively studied, it is better documented than Haush. By estimated 500 others in Paraguay). Chulupí (also 1973 there were only three speakers of Selk’nam, and known as Nivaclé and Ahluhlay or Ashlushlay) is spo- the language is now extinct. Some believe Haush to ken by anywhere from 200 to 1,200 people in Salta — be but a variety of Selk´nam, and not a distinct lan- estimates vary widely (and perhaps as many as 18,000 guage. The second group, Tehuelche, was spoken by in Paraguay). In Argentina, they mostly live in com- groups in Patagonia and the Pampas. Four varieties of munities mixed with Chorote and Wichí speakers. Tehuelche are reported: Northern Tehuelche, also known as Gününa-küne, was spoken from the Lule-Vilela Colorado River to the Chubut River. It became extinct This is a small proposed language family, with only in the 1960s. A second variety of Tehuelche is two languages: Lule and Vilela. Lule, believed to be a believed to have been spoken in the southern Pampas, dialect of the Tonocoté language referred to in early from Eastern Neuquen to the Colorado River, but very documents, was spoken mostly in Tucumán, western little documentation exists for the language (mainly Salta, and northwestern Santiago del Estero. It became toponyms and a few phrases from travelers). Southern extinct in the nineteenth century. It is believed to be Tehuelche was spoken from the Chubut River to the related to Vilela, a language spoken in the provinces of Magellan Strait. Two varieties are reported: Teushen, 989 SOUTH AMERICA: ARGENTINA now extinct and of which very little is known (briefly gradually shifting to Spanish, after decades of pressure reported in vocabularies dating from 1780 to 1900), from the Argentinean authorities, so that by the 1940s was spoken between the Chubut and Santa Cruz Welsh was no longer spoken among members of the Rivers. The second variety, known as Tehuelche or Welsh community. Korean speakers represent the lat- Aonek’, was spoken from the Santa Cruz River to the est ‘wave’ of immigration, mostly as a result of a Straits of Magellan. Only two speakers of Tehuelche diplomatic agreement between the Argentinean and are alive today, but they do not use the language reg- Korean governments. The largest number of Korean ularly and only remember it partially. speakers came to Argentina between 1965 and 1989, with most between 1984 and 1989. Patterns of lan- guage maintenance among Korean speakers reflect the Yahgan same trends as other immigrant languages, with sec- Yahgan, also known as Yámana, was spoken in the ond- and third-generation speakers already shifting to southern coast of Tierra del Fuego and the southern Spanish. The immigrant languages with a considerable islands of the area. The language is not part of the number of speakers include Italian, German, English, Chon family, and its genetic affiliation is still unclear. Welsh, Ukrainian, Polish, Lithuanian, Vlach Romani, The language became extinct in Argentina in the early Arabic, Japanese, and Korean. twentieth century. Today, there is now only one speak- er of Yahgan left in Chile, an elderly woman. The only documentation of the language was provided by the References first missionary to settle in the area, the Englishman Bridges, Esteban Lucas. 1948. Uttermost part of the earth. Thomas Bridges. London: Hodder & Stoughton. Censabella, Marisa. 1999. Las lenguas indígenas de la Argentina, Una mirada actual. Buenos Aires: EUDEBA. Other languages Cooper, John M. 1946. The Patagonian and Pampean tribes. In Steward. Argentina was first settled by Spanish colonizers in the Kaufman, Terrence. 1994. The native languages of South early 1500s, and while the population grew slowly in America. Atlas of the world’s languages, ed. by Christopher the first two centuries, the late nineteenth and twenti- Moseley and R.E. Asher. London, New York: Routledge. eth centuries witnessed a surge in immigration, most- Mera, Carolina. 1998. La inmigración coreana en Buenos Aires: ly from Europe. Although the Argentinean government multiculturalismo en el espacio urbano, Buenos Aires: EUDEBA. provided incentives for the immigrants to settle in the Métraux, Alfred. 1946. Ethnography of the Gran Chaco. In interior of the country, the majority settled in Buenos Steward. Aires and surrounding areas. Most first-language Serrano, Antonio. 2000. Los Aborígenes Argentinos, Síntesis speakers of these immigrant languages are first-gener- Etnográfrica. Córdoba: Ediciones Librería Paideia. ation immigrants, and by the second or third genera- Steward, Julian H. (ed.) 1946. Handbook of South American Indians, Vol. 1. Washington: Smithsonian Institution, Bureau tion speakers shift to Spanish, learning the immigrant of American Ethnology, Bulletin 143. language (if at all) only as a second language. Two Williams, Glyn. 1991. The Welsh in Patagonia, the state and the particularly interesting cases are those of Welsh and ethnic community. Cardiff: University of Wales Press. Korean. A large group of Welsh immigrants settled in the Patagonia, on the Chubut River in 1865 as a result VERONICA M. GRONDONA of an agreement between the Argentinean government See Also Quechua and Andean Equatorial Lan- and English representatives. They were able to main- guages; Wayampi and Tupí-Guaraní Languages; tain their language well into the twentieth century, Spanish South America: Brazil Occupying nearly half the continent of South America, Amazon Basin adds to its enigmatic and diverse char- Brazil is one of the largest and most populous countries acter. Brazil is home to a large number of little-known in the world, and its inclusion of a major portion of the indigenous languages from a variety of language 990  SOUTH AMERICA: BRAZIL families, as well as to several genetic isolates—lan- has been very common among the Indian tribes of the guages that do not seem to be related to any other Amazon basin. known languages. The status of Brazil is also unique in Despite some effort to maintain indigenous lan- South America as the only country recognizing guages, many languages have been lost due to assimi- Portuguese rather than Spanish as its official language. lation into mainstream society. Languages continue to Finally, the importation of African slaves into the be lost in Brazil as indigenous peoples are assimilated region, as well as continued immigration by a variety into the mainstream culture and younger generations of ethnic groups, has created an interesting and com- become monolingual speakers of Portuguese. While plex linguistic situation in Brazil. the linguistic diversity of Brazil, as with South America as a whole, is widely recognized, description of these languages has been insufficient for a number Indigenous Languages of reasons such as the bureaucratic difficulties of With no written records documenting precolonial working with indigenous peoples, inadequate funding, Brazil, estimates of the indigenous population at the and limited linguistic training. time of the Portuguese arrival in 1500 have ranged between one and six million. As a consequence of war Portuguese and the introduction of such European diseases as smallpox and influenza, the indigenous population The Portuguese were the first Europeans in Brazil, was drastically reduced. Today, there are about arriving in 1500, several years after Christopher 200,000 Indians in Brazil. Columbus’ first voyage to the Caribbean. Although the Despite the loss of languages that certainly Spanish, led by Columbus, were the first Europeans to occurred due to the loss of human life, about 170 dif- arrive in the New World, the Portuguese had always ferent Amerindian languages remain in the region dominated the seas in terms of long-distance travel. By today. Because little is known about many of these the end of the fifteenth century, the Portuguese had languages, and because contact among the various lan- already succeeded in sailing around Africa to reach guages has been great, a great number of classifica- India, establishing an important trade route. The dis- tions have been proposed for the languages of Brazil. covery of the Americas by the Spanish no doubt threat- One of the most commonly accepted proposes five ened Portugal’s dominance of the seas, and the major families—Arawak, Carib, Tupi, Macro-Je, and Portuguese soon sought to establish their own colonies Pano—as well as several smaller families and some in the New World. genetic isolates. The highest density of indigenous lan- The status of Brazil as the only Portuguese-speaking guages and indigenous speakers in South America is in country in South America has to do with the ways in the Amazon basin in the northern region of Brazil. which the Portuguese and the Spanish made use of their Tupi and Arawak are two of the most important of colonized land. A treaty between Portugal and Spain in these languages today, and Tupi has contributed great- 1494 divided South America into two parts, giving ly to the word stock of Brazilian Portuguese. Sharing Spain control of the western side and Portugal control a number of similarities with Tupi, Carib is spoken pri- of the eastern side. In Brazil, the main source of wealth marily in the northern region of Brazil as well as in the became the production of sugar cane, while in the northern South American countries of Colombia, Spanish colonies, the main source of wealth was the Venezuela, Guyana, Belize, and Guatemala. The two extraction of precious metals, most notably gold and language families are not only thought to have so silver. Sugar cane production required a highly organ- many similarities due to genetic affiliation, but also ized society in Brazil, with large pieces of land and because of a great deal of contact among some of the numerous slave workers imported from Africa. The languages of the two families. Although some of the search for gold and silver, on the other hand, could be indigenous languages of South America build sen- done by smaller groups of people, and its success often tences from very simple words, most tend to favor relied on chance. Thus, while the Portuguese coloniza- highly complex word structures and most tend to pre- tion of Brazil was fairly organized, the Spanish colo- fer suffixation over prefixation. In addition to interest nization of the rest of South America was more in how language families relate to one another geneti- fragmented and lacked a single centralized government. cally, there is also great interest in how languages There was never any question that Portuguese would within families are related to one another geographi- be the official language of Brazil; what was under cally, especially those associated with the Amazon debate was which Portuguese would be used. The pop- Basin, as many families are distributed disparately ular Portuguese of Brazil has always been markedly throughout the area. Additionally, due to a history of different from European Portuguese. The Brazilian slavery among the tribes of the region, multilingualism people do not consider themselves direct descendants 991 SOUTH AMERICA: BRAZIL of the Portuguese colonists but rather as a mixture of ending with the end of the slave trade in 1850. More Portuguese, Indian, and African people, and the lan- than two million slaves were imported into Brazil to guage of Brazil has always reflected this mixture of provide labor on the sugar plantations and in the gold cultures. By the nineteenth century, people began to mines, with the greatest numbers being brought in dur- debate the existence of a ‘Brazilian language’, and ing the first part of the nineteenth century. This popu- writers such as José de Alencar reacted by using words lation greatly increased that of Brazil during this and expressions they felt reflected Brazilian rather than period. The varieties of African languages that existed Portuguese culture. This trend continued into the twen- in Brazil were influential in the formation of various tieth century, especially in the modernist movement of Brazilian creoles and had some effect on Brazilian the 1920s. Nevertheless, Standard Brazilian Portuguese Portuguese; however, these languages have generally tends to follow norms inherited from European been lost in Brazil over the years. Yoruba and other Portuguese, even while allowing for some differences African languages, however, continue to be used in the in vocabulary and spelling. For instance, in European religious ceremonies of the Candomblé societies. Portuguese, the subject (pronoun) of the sentence is frequently dropped, and phrases generally avoid begin- Creole Languages ning with an object pronoun. Rather, the object pro- noun is placed after the verb, as in esperou-me ‘he or The great variety of indigenous languages in Brazil she waited for me’ (lit. ‘awaited-me’). In colloquial and language contact between them and the early Brazilian Portuguese, however, it is more common to Portuguese, as well as the importation of African use subject pronouns, and object pronouns are almost slaves, resulted in the development of a variety of pid- always placed before the verb (or not used at all in the gins and creoles in Brazil. Lingua Geral, a trade lan- case of the third person); thus, ele/ela me esperou ‘he guage created by Tupi Indian nations along the coast, or she waited for me’. (lit. ‘he or she me awaited’) for instance, was popular until the end of the seven- However, forms such as esperou-me are frequently teenth century and was still used in some isolated encountered in texts, and such forms are the ones areas, particularly in the Amazon Basin, until the late taught in schools. Although phrases beginning with an nineteenth century. Brazilian Creole Portuguese is still object pronoun, such as me esperou, are not uncommon spoken to some extent, mostly by rural Brazilians of in speech, grammarians consider them incorrect usage. African descent in the north. Although these various Differences in vocabulary and pronunciation contact languages were tolerated and even viewed as between European and Brazilian Portuguese tend to be useful during the early colonization of Brazil, the rise less stigmatized than grammatical differences. Many of Portuguese as the official language of the country, lexical items are different in the two varieties, such as and as the lingua franca in contact situations, has done Brazilian meia and European peúga ‘sock’; others a great deal to reduce the creole-speaking population. have different spellings and/or pronunciations, such as Brazilian fato and European facto ‘fact’. It is pronun- Other European Languages ciation that most clearly marks speech as either Immigration into Brazil by non-Portuguese Europeans Brazilian or European, regardless of the level of for- began around 1850, primarily involving Italians and mality. Some highly noticeable differences include the Germans settling in the subtropical regions near the pronunciation of final -e in words like nove ‘nine’ as Atlantic coast, especially in the states of São Paulo, [i] in Brazilian Portuguese and as either schwa or Paraná, Santa Catarina, and Rio Grande do Sul; immi- silent in European Portuguese and the palatalization of gration by these groups generally ended by the time of /t/ and /d/ before /i/ in most regions of Brazil (i.e. ti is World War II. These languages are still used in some of pronounced as [tʃi], di as [di]). the more homogeneous of these communities to some There are also many varieties of Portuguese in extent, particularly in religious ceremonies. The Brazil with every major city noted for having its own Japanese began arriving and settling in approximately accent. Rio de Janeiro, for example, is characterized the same region around the turn of the twentieth centu- by its use of alveopalatal s and z in the syllable-final ry, and their language is still used to a great extent in position. A word such as nós ‘we’, then, is normally Brazil. English and French, and to a lesser extent [nɔʃ] in Rio and surrounding areas, while in most German and Italian, are taught as second languages in other cities in Brazil, it is typically [nɔs]. schools. African Languages Illiteracy The ancestors of African Brazilians arrived as slaves In addition to the loss of indigenous languages in beginning in the middle of the sixteenth century and Brazil, another significant problem of note is the high 992  SOUTHEAST ASIA illiteracy rate, estimated in 2002 as 16.7% of the popu- Greenberg, Joseph H. 1987. Language in the Americas. lation. Although there are numerous literacy programs Stanford: Stanford University Press. Key, Mary Ritchie. 1979. The grouping of South American available in Brazil, especially in churches, and literacy Indian languages. Tubingen: Narr. has become much more widespread since the early Melo, Gladstone Chaves de. 1971. A Lingua do Brasil. Rio de 1900s, illiteracy is in reality only a small part of a larg- Janeiro: Fundacao Getulio Vargas. er problem. People in the lowest classes are not always Payne, Doris L. (ed.) 1990. Amazonian linguistics: studies in encouraged or even expected to go to school, and the lowland South American languages. Austin: University of Texas Press. drug trade and other crimes often provide a much faster Rodrigues, Aryon D. 1985. The present state of the study of way to make money than getting an education. Brazilian Indian languages. South American Indian lan- guages: retrospect and prospect, ed. by Harriet E. Manelis Klein and Louisa R. Stark. Austin: University of Texas References Press. Aikhenvald, Alexandra Y., and R.M.W. Dixon. 1999. Other Thomas, Earl W. 1974. A grammar of spoken Brazilian small families and isolates. In Dixon and Aikhenvald. Portuguese. Nashville: Vanderbilt University Press. Bortini-Ricardo, Stella Maris. 1985. The urbanization of rural dialect speakers: a sociolinguistic study in Brazil. LAMONT ANTIEAU AND MICHAEL COLLEY Cambridge: Cambridge University Press. Dixon, R.M.W., and Alexandra Y. Aikhenvald (eds.) 1999. See also Arawak; Carib and Cariban Languages; Introduction. The Amazonian languages. Cambridge: Pidgins and Creoles; Wayampi and Tupí-Guaraní Cambridge University Press. Languages Southeast Asia A region comprising both the Indochinese and Malay languages in the region, particularly with respect to peninsulas, as well as several island groups in the area, vocabulary. Due to China’s long domination of Southeast Asia is home to a great number of people Vietnam from 111 BC to AD 939, Chinese has proba- and cultures. Bordered on the north by the Yangtze bly had the greatest influence on Vietnamese, most River in China, on the east by the South Pacific Ocean, notably in its legal and medical terminology. A num- on the south by the Indian Ocean, and on the west by ber of other Sino-Tibetan languages are spoken almost the Indian Ocean, the Bay of Bengal, and the subcon- exclusively in countries of Southeast Asia, including tinent of India, Southeast Asia includes the countries Burmese, the official language of Myanmar, a country of Brunei, Myanmar (formerly Burma), Cambodia of over 35 million people. (Kampuchea), Indonesia, Laos, Malaysia, Singapore, The Mon-Khmer language family is another impor- Thailand, Vietnam, and the Philippine Islands, as well tant family in Southeast Asia and includes Khmer, as the southernmost region of China. Owing in part to which is the official language of Cambodia and has the diverse geographical makeup of the area and the approximately seven million speakers, and Mon, which lengthy and intricate historical relationships of its was an important language in earlier times, but has inhabitants, Southeast Asia is a region of great linguis- been reduced to about one million speakers. Just as tic variety and complexity. Chinese has had a great influence on the languages of Although more than a thousand languages are spo- Southeast Asia, so too has Khmer, particularly during ken in Southeast Asia, five language families are par- the Angkor period, which lasted from approximately ticularly well represented in the region: Sino-Tibetan, the ninth to the fifteenth centuries AD. As a result of Mon-Khmer, Tai, Miao-Yao (also known as the Khmer dominance during this period, Khmer vocabu- Hmong-Mien), and Austronesian. While such major lary is well attested in the lexicons of many languages languages of the Sino-Tibetan family as Chinese and of Southeast Asia, including Thai and Lao. Vietnamese, Tibetan are spoken primarily in China and the which is the official language of Vietnam and is spoken Himalayas, speakers of these languages can also be by about 80 million, is also classified by most scholars found in great numbers throughout Southeast Asia, as a member of the Mon-Khmer family; however, this particularly in those countries sharing a border with classification has been somewhat controversial, mostly China. In addition to being widely spoken in Southeast due to the phonemic status of tone in Vietnamese, a Asia, Chinese has also had a great influence on many trait not shared by other Mon-Khmer languages. This 993 SOUTHEAST ASIA difference led earlier scholars to classify Vietnamese as nominalization, compounding, and the use of classi- a Tai language; however, Haudricourt (1954) accounts fiers rather than morphemes to mark gender. Additional for the use of tones in Vietnamese as the result of an support for this position is that discourse patterns tend evolutionary process in which Vietnamese acquired to be similar in various Southeast Asian languages. tones from other languages in Southeast Asia, notably However, the often intense and diverse contact in the Chinese, via language contact. Linguists often classify region makes it as difficult for linguists to determine the Mon-Khmer family as a member of the Austro- whether characteristics have been borrowed by lan- Asiatic family of languages together with the Munda guages as it is to identify with precision the classifica- family of India. tion of the various languages. In addition to the Another language family of great importance in problem created by contact among neighboring lan- Southeast Asia is the Tai family, a group of tonal lan- guages for hundreds of years, the great influence over guages spoken primarily in Thailand, Laos, and southern the entire region by such historical powers as Chinese China. The two major languages of the Tai family are and Khmer, as well as the influence of languages and Thai, which is spoken by about 50 million people and is cultures from outside Southeast Asia, tends to blur the the official language of Thailand, and Lao, which is spo- distinction between languages. Numerous words, for ken by about 10 million and is the official language of instance, have been borrowed into many Southeast Laos. Other languages in the Tai family include Shan, a Asian languages from Sanskrit, while others have been language spoken by over two million in Myanmar, and borrowed from Pali due to the adoption of Buddhism Chuang, which, with over 15 million speakers, is the by some Southeast Asian cultures. language of the largest ethnic minority in China. Languages well outside Southeast Asia, especially a Languages in the Miao-Yao family are spoken in number of European languages, have also had an isolated areas of southern China and in the northern impact on the languages of the region. French has been regions of Laos, Thailand, and Vietnam. The main lan- an extremely influential language, particularly in guages of the group are Miao, which is spoken by Vietnam and Cambodia, countries that were under the about two and a half million people, and Yao, spoken rule of the French from the mid-nineteenth to the mid- by about one million. Many scholars consider the twentieth century. During this time, French was recog- Miao-Yao family a branch of the Sino-Tibetan family, nized as the official language of both countries, serving although it has also been proposed that the family is a primarily as the language of government and education. branch of either the Tai or the Mon-Khmer family. Although both Vietnam and Cambodia became inde- The Austronesian languages, formerly called pendent of French rule shortly after World War II, the Malayo-Polynesian, are spoken on the Malay bonds between France and Southeast Asia have not Peninsula and on most of the islands to the southeast been completely severed as evidenced in the continuing of mainland Asia. Distribution of the family’s lan- migration of natives of Southeast Asia to France. The guages actually extends far beyond the region of influence of the French on various Southeast Asian lan- Southeast Asia with member languages spoken from guages is most readily apparent in their lexicons, espe- Madagascar to Easter Island and from Taiwan to cially with respect to words relating to the government Hawaii, as well as in and around New Zealand. The and the military. Spanish is also of some significance in family comprises approximately 200 million speakers Southeast Asia due to its long domination of the of at least 500 languages. In Southeast Asia, the prin- Philippines from the mid-sixteenth century until the end cipal Austronesian languages are Indonesian, which is of the nineteenth century; the best evidence of Spanish the official language of Indonesia and is spoken by is found in the lexicon of Tagalog, which includes hun- about 20 million, and Malay, spoken by about 10 mil- dreds of Spanish words. English is yet another lion and the official language of Malay. Some linguists European language that has had some influence on the have claimed that Indonesian and Malay are essential- region, primarily due to the American presence in the ly the same language, differing only in orthography Philippines during the first half of the twentieth centu- and political identity. Another important language of ry, American involvement in the Vietnam War, and the the Austronesian group is Tagalog, which has been the role of English in international trade and diplomacy. national language of the Philippine Islands since 1962 Contact between Southeast Asians and others has and has about 15 million speakers. resulted in the formation of a great number of pidgin Despite the great number of differences between the and creole languages in the area. As in other parts of various languages spoken in Southeast Asia, some the world, Portuguese played a crucial and early role in scholars have pointed out a number of commonalities, the creation of pidgins and creoles in Southeast Asia, suggesting that the languages share a genetic relation- particularly in Indonesia, which hosted a number of ship as well as an areal one (see e.g. Matisoff 1978). Portuguese-based creoles that are now all considered These shared characteristics include similar patterns of extinct. The Portuguese presence in the region, which 994  SOVIET UNION dates back to the fifteenth century, also led to the devel- Southeast Asian languages that use the Brahmi script opment of a number of other Portuguese-based creoles, have modified it to some degree to reflect the sounds including Burma-Siam Creole Portuguese, of which a of their own languages, the Indian language that it was small number of speakers can be found in Penang, and borrowed from differed significantly from the Asian Papia Kristang, which is spoken in Malaysia and languages it is now used to represent. The result has Singapore. Perhaps an even greater number of pidgins been the extensive use of subscript and superscript and creoles in Southeast Asia developed due to contact symbols to represent vowels. Although some have between speakers of Spanish and speakers of various pointed to it as an example of successful writing languages on the Philippine Islands, where the number reform, Vietnamese shares in this problem and makes of Creole Spanish speakers stands at around 280,000. such wide use of diacritics that some scholars have These contact languages include Bamboo Spanish and remarked that it hardly constitutes true reform (see Kitchen Spanish, which are used primarily by Chinese Hannas 1997). The problem that these systems share is shopkeepers in the Philippine Islands. Another contact the difficulty they create for printing processes, both language that developed in Southeast Asia, this time as because of the amount of paper required for printing a result of contact between speakers of French and the great number of diacritics as well as the extra time Vietnamese, was Tay Boy, a Vietnamese-French pidgin required to keyboard the extra symbols. These prob- spoken during the French domination of Vietnam from lems have typically been at the core of the debate over approximately 1860 to 1960. Contact between speakers writing reform for Southeast Asian languages in mod- of English and Vietnamese during the Vietnam War ern times. Another has been that there are many minor- resulted in Vietnam Pidgin, which is now largely ity languages in Southeast Asia, often spoken in the extinct; in Singapore, contact between English speak- mountains of the region as opposed to the plains and ers and native inhabitants resulted in the formation of valleys, for which there are no writing systems at all. Singlish. There is also a mixed language in Southeast Asia called Davaoeno, which is a fusion of Tagalog and Visayan and is spoken by approximately 125,000 References speakers in Davao. Dinh-Hoa, Nguyen. 1980. Language in Vietnamese society: some One linguistic issue of great importance in articles by Nguyen Dinh-Hoa. Carbondale, IL: Asia Books. Southeast Asia, as it is in Asia in general, concerns the ———. 1997. Vietnamese. Amsterdam: John Benjamins. Hannas, Wm. C. 1997. Asia’s orthographic dilemma. Honolulu: various writing systems that have been implemented to University of Hawaii Press. represent the languages of the region. Many Southeast Haudricourt, André-Georges. 1954. De l’origine des tons in Asian languages, including Burmese, Cambodian, viêtnamien. Journal Asiatique 242. 68–82. Lao, and Thai, are written in a Brahmi script that was Huffman, Franklin E. 1970. Cambodian system of writing and imported from South India in the fourth or fifth centu- beginning reader with drills and glossary. New Haven: Yale University Press. ry AD and has undergone some modification. Several Jacob, Judith M. 1968. Introduction to Cambodian. Oxford: languages, including Vietnamese and Indonesian, are Oxford University Press. now written in the Roman alphabet, which was intro- Matisoff, James A. 1978. Variational semantics in Tibeto-Burman. duced by missionaries in the region in the seventeenth Philadelphia: Institute for the Study of Human Issues. and eighteenth centuries, and, in the case of Parkin, Robert. 1991. A Guide to Austroasiatic speakers and their languages. Honolulu: University of Hawaii Press. Vietnamese, replaced an earlier system based on Smith, Norval. 1995. An annotated list of Creoles, Pidgins, and Chinese orthography. Still other languages, such as mixed languages. Pidgins and Creoles: an introduction, ed. Malay, have incorporated some Arabic characters into by Jacques Arends, Pieter Muysken, and Norval Smith, their writing systems. Each of these writing systems 331–74. Amsterdam: John Benjamins. presents a number of problems. Although the LAMONT ANTIEAU Soviet Union From its inception in 1917 to its ultimate breakup in encompassed some 8,649,490 square miles with a total 1991, the Soviet Union was a multilingual and multi- population of just under 286,000,000. The population ethnic state. At its point of greatest expansion, it is unevenly distributed among the ethnic groups and 995 SOVIET UNION languages, with some very large and some quite small tinct from Romanian); Iranian (Tadzhik) as well as in number. Russians comprise the largest single ethnic Germanic (German and Yiddish), and a few others. Of group, constituting just over 50% of the total popula- these genetic groups, Indo-European is the most well tion (or slightly more than 145,000,000) in the 1989 defined. The genetic relations of the Uralic languages census. Russian was one of several Slavic languages (e.g. Finno-Ugric languages such as Estonian, Permic, spoken in the USSR, although by far the most widely Samoyedic, and Volgaic) are also clear. The Caucasian used and it alone had the status of a national lingua languages are divided into three or four subgroups, franca. The Russian language has by far the largest whose relations to one another are disputed. The Altaic number of speakers, with 81% of the population see- languages are commonly divided into three language ing itself as fluent in Russian, either as a first or sec- families (Mongolian, Tungusic, and Turkic), and ond language. although these three families constitute clear genetic groups, their relations to one another are the matter of much debate. Paleo-Siberian is an even more loosely Language and Ethnicity defined category, as it groups together the languages of The 1989 Soviet census cites approximately 130 ethnic Siberia that do not belong to any of the other genetic groups (including indigenous and immigrant popula- groups (e.g. Altaic, Uralic, etc.). Two relatively small tions) speaking a total of 150 languages. (Linguists, language families (Chukotko-Kamchatkan and Eskimo- however, estimate that there were actually closer to 200 Aleut) are usually placed in the Paleo-Siberian group. languages spoken in the former USSR.) There is not, The Soviet Union was organized into 15 union however, any one-to-one correspondence between lan- republics (or SSR, for Soviet Socialist Republic), guage and ethnicity, as these figures show. In the USSR, which can be grouped into categories according to all citizens aged 16 years and higher were required by geography: the Baltics (Estonian SSR, Latvian SSR, law to declare their ‘ethnicity’ (natsional’nost’), which and Lithuanian SSR); the Caucasus (Armenian SSR, was officially recorded in each individual’s (internal) Azerbajdan SSR, and Georgian SSR); Central Asia passport. The census regularly asked questions about (Kazakh SSR, Kirgiz SSR, Tadzhik SSR, Turkmen language use and ethnicity as well, but because lan- SSR, Uzbek SSR); Slavic territory and Moldova guage/dialect boundaries were often determined by (Belorussian SSR, Moldavian SSR, Russian SFSR, social and political factors rather than linguistic ones, Ukrainian SSR); and the Russian Far East (RSFSR, and the number of official languages and ethnic groups primarily Siberia). Each Republic bears the name of changed in accordance with changing political goals. its majority ethnic group, but it should be stressed that For example, three Tungusic languages—Even, Evenki, these are political, not ethnolinguistic divisions. and Negidal—were long considered to be dialects of a Ethnic Russians lived in all republics in varying num- single language, although all are linguistically distinct, bers. The Russian Republic was the largest in terms of so that when they each acquired official recognition, territory and also had the highest percentage of ethnic there was a net increase in the number of ethnic groups. Russians. Moreover, varying numbers of other ethnic Accordingly, official Soviet statistics should be taken as groups lived in each of the republics, and some were providing only partial information about the linguistic dispersed over several. The Tatars, for example, with a map. Speaker population size and density vary signifi- total population count of 5.5 million in 1989, live scat- cantly: some languages have millions of speakers, while tered in 80 different regions of the former Soviet others have only thousands or even hundreds of speak- Union, with only 26.6% of their total numbers living ers. Some are in dense speaker communities in urban in Tatarstan (Russian SFSR), another 17% in other areas, while others are traditionally nomadic and have regions of Russia, 7% in Uzbekistan, and so on. dispersed populations. Soviet Language Policy Linguistic Diversity The newly formed Soviet state faced illiteracy rates of The many languages of the USSR can be classified into nearly 100% in some areas, with an overall average lit- five large language families: Indo-European, Caucasian, eracy rate of only 28.4% for the entire country. In Altaic, Uralic, and Paleo-Siberian. In addition, there are order to achieve their goal of rapid industrialization, a number of language isolates (e.g. Ket, Nivkh, the Bolshevik leaders required a fairly educated work Yukagir). A number of Indo-European languages are force, which made a rapid increase in literacy one of native to territories of the USSR: the East Slavic lan- their priorities. The sincerity of early Soviet language guages (Russian, Belorussian, and Ukrainian); the policy has been the subject of much debate, but Baltic languages (Latvian and Lithuanian); Armenian; a Lenin’s basic principle was unambiguous: no one lan- Romance language (Moldovan, which is minimally dis- guage should be given the status of a state language 996  SOVIET UNION but rather, national equality and self-determinism must Ultimately, this did not occur, as the spheres of usage be promoted. The Declaration of Rights of the People of many of the national languages were quite limited, of Russia (November 2, 1917) proclaimed equality for making this native technical terminology superfluous. all people, with ‘the free development’ of the national Where loans were concerned, the early policy was to minorities and ethnic groups inhabiting the USSR. All maintain the pronunciation of the lending language. Soviet citizens were guaranteed education in their This policy was changed by the 1950s, when all loan- native language. words were written in the original Russian form. The literacy campaign was a fundamental part of Despite the many controversial aspects of the Soviet the larger nativization or, literally, ‘rooting’ (kor- literacy campaign, it did achieve rapid results. The lit- enizacija) policy, a policy intended to educate the eracy rate climbed dramatically from 44.1% in 1920 to indigenous peoples and move them into the workforce, 87.0% in 1939 to an official rate of 99.7% in 1979. especially into the Soviet administrative workforce. A The mid-1950s witnessed yet another major change major obstacle to achieving this goal was the overall in language policy. Prior to this period, the national lan- low educational levels nationwide. Thus, the success guages were the focus of Soviet language planning. of the nativization policy depended upon the ability of From this time onward, the goal of Soviet language pol- the government to educate its people. At the time of icy was to establish Russian as the language of the the formation of the Soviet Union, the majority of lan- Soviet Union. The Khrushchev era (1953–1964) intro- guages needed linguistic description and codification, duced the vision of a new Soviet people, united not only and creation of a written form. Some linguistic regions politically but also through the use of one language. (such as Georgia and Armenia in the Caucasus, and the Khrushchev declared Russian to be ‘the second nation- Turkic-speaking Central Asia) had long-standing liter- al language’. Although bilingualism was openly pro- ary traditions, as did the Georgian and Armenian lan- moted, the very need for national languages (any guages in the Caucasus and the Turkic languages in language other than Russian) was questioned. The Central Asia. At the same time, many of the languages Khrushchev introduced the notion of the ‘relative’ of the newly formed empire lacked written forms. It is importance of languages, and it became officially estimated that at the time of the Bolshevik Revolution, acceptable to view some languages as less viable than only 13 languages on Russian territory had a literary others: some languages were deemed unsuitable for norm, and only 19 had any kind of written form at all. development. Yet, Lenin’s policies of language equality The Roman alphabet was initially chosen as the were not officially repudiated, and the Communist Party basis for all new writing systems and was also used for could invoke them to justify its own policies at any time. those languages not using Cyrillic, such as those using This occurred along with a major change in educa- an Arabic or Mongolic script, although Georgian and tion policy. Clause 19 of the Education Reforms of Armenian were allowed to maintain their own alpha- 1958–1959 stated that education in the mother tongue bets. The use of Romanized script was short-lived, was no longer compulsory. By this time, instruction in being phased out in favor of Cyrillic in the mid- to late the native language was offered for most languages 1930s and completed in the early 1940s. This transi- with a written form at an elementary level, and at a tion to Cyrillic-based alphabets was part of a shift in secondary level for some. One impact of the Education language policy, which began after Lenin’s death in Reforms was pressure to begin instruction in Russian 1924. A key date is 1934, when Stalin effectively from the earliest grades, and the native language was ended the nativization campaign in an address to the replaced by Russian in many schools. Even where XVII Party Congress. In March 1938, an official native-tongue education continued, Russian was a decree made the study of Russian compulsory. compulsory subject. Alphabet creation was just one step in the literacy The results of the Education Reforms varied campaign. Equally important was the creation of a throughout the country. Despite the shift in emphasis standardized literary form for each of the targeted lan- on Russian, languages spoken by larger populations guages. Originally, emphasis was placed on phonetic may have actually gained some ground, in part due to spelling, which created problems due to vast dialecti- the lessening of cultural restrictions under cal differences for some languages, and due to the Khrushchev. This was the case in Central Asia, for large influx of Russian borrowings as a result of the example. In contrast, minority languages became seri- sociopolitical and economic changes. The creation of ously threatened as they were no longer used in a new lexical inventory was an equally vital compo- schools, and publications in these languages were seri- nent of this larger campaign. Initially, it was argued ously cut back. Regardless of the local-level particu- that each of the national languages should have a com- lars, the major change of this period is that Russian plete inventory of all technical terms, created using became the official language of the USSR and occu- language-internal resources wherever possible. pied a central position in education and government. 997 SOVIET UNION This process continued under Brezhnev (1964–1982), in fact, they received little attention from the central in a greater move toward total Russification, with government in Moscow. In many regions, the linguis- increasing pressure to make Russian the ‘second mother tic situation was essentially stable throughout this tongue’. Official statistics show Russians to be largely period; yet, a number of Republics began to change monolingual (97%), while over 40% of the non-Russian language policies within their own territories. By the population claimed itself to be bilingual, and by 1979 a end of the Soviet era, most of the territories that total of 82% of the population claimed some knowledge remained in the Russian Federation had adapted legis- of Russian. Translation work was primarily unidirec- lation to place the titular language, the language of the tional, from Russian into the native language. The ethnic majority of their region, on an equal level with Brezhnev period is characterized by a steady increase in Russian. both the sheer volume of instruction in Russian, which was progressively replacing the national languages in non-Russian schools, and a continuous growth in the References number of institutions where Russian was the sole oper- Comrie, Bernard. 1981. The languages of the Soviet Union. ative language. Party rhetoric proclaiming the impor- Cambridge: Cambridge University Press. tance of Russian increased. The official view of a single Comrie, Bernard, Gerald Stone, and Maria Polinsky. 1996. The Soviet ethnic group, the result of ‘the convergence and Russian language in the 20th century. Oxford: Clarendon Press. fusion of peoples’ (sblizhenie i slijanie narodov), was Kirkwood, Michael. (ed.) 1989. Language planning in the becoming a reality. The sphere of Russian usage spread Soviet Union. London: Macmillan. beyond education to many administrative levels, includ- Kreindler, Isabelle. (ed.) 1985. Sociolinguistic perspectives on ing local-level administration. It had become the lingua Soviet national languages: their past, present and future. franca of the USSR. Berlin: Mouton de Gruyter. Neroznak, V.P. (ed.) 1994. Krasnaia kniga iazykov narodov Mikhail Gorbachev came to power in 1985, and his Rossii (Red book of the languages of the peoples of Russia). restructuring program, perestroika, focused govern- Moscow: Academiia. mental attention on economic and political problems. Smith, Graham (ed.) 1990. The nationalities question in the Until 1989, both language and nationality policies Soviet Union. London: Longman. remained essentially unchanged from previous years; LENORE A. GRENOBLE Soviet Union: Successor States Geographically, the former Soviet Union occupied one the lands of Kazan and Astrakhan Khaganats (Turkic- sixth of the Earth’s land surface stretched from the speaking people), defeating the Mongol Horde. They Baltic Sea in the West to the Pacific Ocean in the East, also conquered Finns in the northwest (Finno-Ugric from the Arctic Ocean in the North to the Black Sea language family), and proceeded further to the east, and Tian Shan mountains in the South. It accommo- adding Siberia, the Far East, and Alaska to the Empire. dated a number of ethnic groups with more than 100 languages, which can be combined into six major lan- Russian Empire. By the end of the nineteenth centu- guage families. There were 15 Union Republics with a ry, the Russian Empire celebrated its last acquisition, total population of more than 240 million people, who the territory of Caucasus and Turkistan (present Central shared common Soviet political experience for about Asia). To run such a huge multiethnic state would have 70 years. been impossible without promoting the use of a single The Soviet multilingual state originated from the language in all spheres of life including administration, state of Kievan Rus (ninth–twelfth centuries CE). By legislation, education, etc.; hence, the Tsarist govern- the fourteenth century, Moscow accumulated enough ment began implementing its policy of russification. power to become a new political center of Slavic- The policy implied a ban or restrictions on the use of speaking people. The expansion of Russia to the east, languages other than Russian in government, educa- north, and south to the areas populated with non-Slavic tion, the press, and books. In order to enforce this pol- people started in the sixteenth century. Russia annexed icy, the government encouraged the migration of ethnic 998  SOVIET UNION: SUCCESSOR STATES Russian population into other parts of the Empire and unifying the scripts into Cyrillic, in order to strength- elevated the literacy rate through the study of Russian. en the Soviet identity and begin the creation of the The policy of russification was implemented by the Soviet nation. The Latin script created for different Tsarist government in all of the provinces of the ethnic groups in the 1920s and 1930s was replaced Russian Empire since the 1860s. However, in different with a modified Cyrillic. Extra letters were added to parts of the Empire, the implementation of this policy the alphabet to reflect those sounds in non-Slavic lan- was not uniform. On the one hand, in Ukraine any pub- guages that were not present in Russian. lications in Ukrainian language were prohibited, and At the same time, the Soviet government encour- the only language of instruction in Ukrainian schools aged development of the literary language of each eth- and universities was Russian. Moreover, the Ukrainian nic group, and each ethnic group could proudly name and Belarusian languages were considered to be varia- national writers who wrote in their language. Each tions of the Russian language. On the other hand, peo- Union or autonomous republic had radio and later TV ple in Turkistan were not restricted in the use of their channels broadcasting in their national language. languages. Also, the Tsarist government did not Books, newspapers, and magazines were published in attempt to change the Arabic-based alphabet of the national languages as well. Nevertheless, the aim of Turkistanese into Cyrillic script. Turkistan enjoyed this the Soviet government to strengthen the role of the freedom as a result of the Tsarist government’s policy Russian language was never abandoned. Official prop- of avoiding unrest in newly acquired territory populat- aganda put the Russian language as the language of ed by people of different faith. However, the govern- Lenin, Stalin and the revolution, the language of the ment encouraged Russians to move to Turkistan, thus largest and greatest nationality, and the language of providing a base for cultural and linguistic changes. By great Russian writers. 1917, most of the major languages of the former At primary and secondary education levels, people Russian Empire had their own alphabets: for example, generally had a choice of the language of instruction Central Asians and Azeri people used Arabic script, since there were classes taught in Russian or in the lan- Baltic peoples used the Latin alphabet, and Armenians guage of the eponymous nationality, Kyrgyz in the and Georgians had alphabets of their own. Kirgiz SSR, for example, or Moldovan in the Moldavian SSR. At tertiary institutions, Russian was Soviet Union. During the time of the Union of Soviet the major language of instruction, with some excep- Socialist Republics or USSR (1922–1991), the lan- tions. For example, Ukrainian was used as the language guage policy was a complex issue, revised with every of instruction in social sciences, art, and literature in change of the Soviet leadership. The Russian language Kiev, or Tajik was used in social sciences, art, and lit- preserved its dominating role and continued to serve erature in Tajikistan. Similar arrangements existed in as the lingua franca, enabling communication across each of the 15 Union Republics. All official govern- the country’s numerous ethnic groups. However, the ment business was conducted in Russian, with the Soviet government imposed several changes in lan- exception of local governments in the Union Republics guage policy. In the mid-1920s, the Soviet government that used local languages as well as Russian. However, began a campaign to develop alphabets based on Latin military officers gave their orders in Russian. The pol- script for most non-Slavic groups of the country. The icy of russification that was continued by the Soviet Arabic script of the Muslim population was changed government created a single literary language, that was into Latin script. However, the strong and established used by most of the people in the USSR. traditions of written language in Georgia and Armenia The government encouraged ethnic Russians to withstood the government’s campaign, and the ancient move to other parts of the country in order to promote alphabets were preserved. The underlying rationale in cultural convergence. However, the first wave of the change of alphabets was the attempt of the Soviet Russian migrants did not go willingly. During the col- government to cut off all the ethnic groups from their lectivization campaign of the 1920s and 1930s, mil- cultural, ethnic, religious, and ideological roots, and to lions of wealthy peasants had their property begin implementation of the totally new culture and confiscated and were banished to Siberia, the Far East, ideology of the Soviet state. Also, alphabets were cre- or Central Asia, thus diluting the local population. ated for those small ethnic groups who had never had Another wave of migration took place during and after a written language before. World War II, when plants and factories were moved, In the late 1930s, there was yet another sudden together with their workers, from potential war zones change in the Soviet language policy, which was only in the European part of the USSR to Siberia and possible due to tough political measures used by the Central Asia. The last major wave of migration Soviet government. The change, initiated under the occurred in the 1950s during the virgin land campaign, rule of Stalin, was aimed at achieving the huge task of when young people from all over the country came to 999 SOVIET UNION: SUCCESSOR STATES the Kazakh SSR with the aim of growing wheat in its who arrived or were born in those countries after the vast wild steppe. After the completion of the campaign 1940s were not legally considered citizens unless they and establishment of kolkhozy (collective farms) in passed a language examination. The Laws on Kazakhstan, some of these people remained in Citizenship, adopted in the Baltic countries after inde- Kazakhstan or moved to the neighboring republics. pendence, made most of the representatives of the During World War II, there was also a relocation of Russian-speaking minority (Belarusians, Russians, whole ethnic groups away from the war zones. and Ukrainians) ‘noncitizens’. This policy created eth- Germans, Crimean Tatars, and Koreans were forcibly nic tensions between the titular nations and Russian- relocated to Central Asian Republics, as the Soviet speaking minorities. For example, in Estonia the authorities feared their collaboration with the occupa- population of Russian-dominated Narva demanded tion army. secession or at least regional autonomy. This reshuffling of nations within the country led to A unique language policy is observed in Kyrgyzstan. a multinational composition of each republic of the Although the Kyrgyz language was declared the state Soviet Union. Often, children of at least five different language, the Russian language was given official status ethnic groups could be sitting in one class. By 1991, in 1996 and became the state language along with the year of dissolution of the USSR, around 25 million Kyrgyz in 2000. The reasons were continuing ties with of Russians lived outside Russia. Russia, financial considerations, and the disappearance of the threat of russification, since the majority of the Post-Soviet Era. After the dissolution of the Soviet Russian-speaking population had left the republic. Union in 1991, 15 new countries appeared on the Another change in language policy was demon- world map. This event was indicative of many dramat- strated in the attempts of some of the newly independ- ic changes in the lives of the peoples. Along with ent countries to change the Cyrillic alphabet to either major economic, political, and social transformations, Latin or Arabic. The changes of the scripts were aimed there were changes in the language policies in each of at highlighting the idea of independence from Russia, the new countries. of national state-building, and as a symbol of the One of the prominent changes was that the status of revival of the nation’s culture, traditions, and lan- the Russian language in all of the new countries, with guage. Financial considerations and the enormous the exception of the Russian Federation, was reversed. complexity of this task prolonged the change of the From its dominating position, it now became the lan- script in some of the republics. Other republics post- guage of Russian-speaking minorities. The languages poned it indefinitely. In Turkmenistan, a Latin-based of the titular nations of each new post-Soviet country Turkish alphabet replaced Cyrillic. Moldova switched were declared the state languages, and they are to a Latin-based script. Tajikistan is gradually imple- increasingly replacing Russian in all spheres of life, menting a change to Arabic. In Uzbekistan, the script including education, mass media, and administration. has been changed to Latin. In Azerbaijan, the alphabet Under the Soviet policy of russification, the Russians reform of 2001 implies transition to the Latin alpha- did not have to learn languages of titular nations of bet. Kyrgyzstan retains the Cyrillic script. other Soviet republics. In Estonia, for example, In the Russian Federation, the language situation according to the 1989 census, around 80% of ethnic also changed after 1991. The Russian language is still Russians did not speak Estonian at all. Dramatic the lingua franca of this large multinational state. changes in the language policies in the newly inde- However, the uncontrollable inflow of the foreign, pendent republics forced a sizeable proportion of the mainly English-based, words in mass media and col- Russian-speaking population (i.e. representatives of loquial language since the cultural isolation of Russia nontitular ethnic groups) to leave the places they were from the West ended together with the Soviet Union. born in or lived in for decades and move to Russia, This trend changed the Russian language dramatically. although a considerable number of Russians still live The literary language created during the Soviet time is outside the Russian Federation. Here are some figures deteriorating, and new language standards are being for 1998: 34.7% of the total population of Kazakhstan created. were Russians; 30.4% in Latvia; 28.1% in Estonia; 22% in Ukraine, 18% in Kyrgyzstan, 13.2% in Language Families Belarus; 13% in Moldova; 8.7% in Lithuania; 6.7% in Turkmenistan, 6.3% in Georgia; 5.5% in Uzbekistan; The 15 former Soviet Union republics bring together 3.5% (this number is declining due to emigration) in six language families: Indo-European, Altaic, Uralic, Tajikistan, 3% in Armenia, and 2.5% in Azerbaijan. North-Western, North-Eastern, and South-Caucasian. An outstanding example is that of the Baltic coun- There are also some limited language groups and sep- tries, Estonia, Latvia, and Lithuania. The Russians arate languages in the far eastern Siberia, which can 1000  SOVIET UNION: SUCCESSOR STATES FAMILY Group Main Languages Indo-European East-Slavic Russian, Belarusian, Ukrainian Baltic languages Latvian, Lithuanian Romance languages Moldavian/Romanian Iranian language Tajik, Kurd, Osetin languages Armenian language Altaic Turkic Azeri, Kazakh, Kyrgyz, Turkmen, Tajik, Tatar, Uzbek and other languages Manchu-Tungus Evenk, Even languages Mongolian Buryat, Kalmyk languages Uralic Finno-Ugric Estonian, Finnish, Khanty, Mansi, Mari, Mordvin, Karelian, Udmurt, Komi, Komi-Permyak, Veps languages Samoyedic Nenets, Selkap languages North-Western Adyghey, Kabardyn, Abkhaz, Caucasian Circassian languages South-Eestern Chechen, Lezginian languages Caucasian South Caucasian Georgian language Paleoasiatic Chukcha, Koryak, Kamchadal languages and others Figure 1. Language Families and Groups. be grouped as a Paleoasiatic group. The languages of republics of the former Soviet Union as well as in some this group are not related and are placed together only countries of Eastern Europe (former socialist block). geographically. Three dialects of Russian, i.e. northern, southern, and central, are only distinct from each other by minor vari- ations in the pronunciation of some consonants and The Indo-European Family vowels, unlike the significant differences in dialects of European languages. Literary Russian is based on the East Slavic Languages: The Russian Federation, Moscow (central) dialect and is written in Cyrillic. The Ukraine, Belarus Russian. The largest successor to the Soviet Union is Ukrainian. The second largest Slavic state of the for- the Russian Federation. Its total population in 1998 mer Soviet Union is the Ukraine. In 1998, its popula- was 146,001,176, which was represented ethnically as tion was 49,153,027 with the following ethnic follows: 81.5% of Russians, 3.8% of Tatars, 3% of composition: 73% Ukrainians, 22% Russians, 1% Ukrainians, 1.2% of Chuvashs, 0.9% of Bashkirs, Jewish, and 4% other ethnic groups. The Ukrainian 0.8% of Belarusians, 0.7% of Mordvins, and 8.1% of language is the state language; Russian and Polish are others. The Russian Federation is home to more than also spoken. The Ukrainian language is also spoken in 70 distinct ethnic groups, some of which are extreme- Ukrainian communities in Belarus, Russia, Poland, ly small and the six mentioned above are more than a and Slovakia. million of people in each. There are 21 autonomous Ukrainian is the direct descendant of the language republics in the Russian Federation as well as ten spoken in Kievan Rus and is written in the Cyrillic autonomous districts and one autonomous region. alphabet. Distinct differences between Russian, In spite of enormous ethnic diversity, the official lan- Belarusian, and Ukrainian languages emerged some- guage of the Russian Federation is Russian. The lan- where between the twelfth and thirteenth centuries; guage is also spoken as a secondary language in other however, development of the standard Ukrainian 1001 SOVIET UNION: SUCCESSOR STATES language was postponed due to long political subordi- rise of separatist tendencies from Slav-populated areas nation to Russia and Imperial Russia’s policy of of Dnestr basin. The Cyrillic alphabet was replaced restrictions in usage of the Ukrainian language. with Latin. Modern literary Ukrainian language began its devel- opment only at the end of the eighteenth century. Armenian: Armenia Belarusian. The third branch of east Slavic lan- Armenian is spoken by a wide Armenian population guages is spoken in Belarus (or White Russia). In 1998, worldwide. It is the language of Armenians in Armenia, there were 10,366,719 people and 77.9% of them were Turkey, and other parts of the former Soviet Union, Belarusians. There were also 13.2% Russians, 4.1% Middle East, America, Romania, Poland, France, and Polish, 2.9% Ukrainians, and 1.9% others. The others. In 1989, in Armenia itself, there were 3,344,336 Belarusian language used to be the official language of people and 93% of them were Armenians, 3% were Lithuania in the thirteenth–sixteenth centuries when Azeri (nearly all Azeri emigrated in 1993), 2% were Belarus was a part of Lithuania. After the merging of Russians, and 2% were others. It is traditionally Poland and Lithuania in the sixteenth century, Belarus believed that the Armenian alphabet was created by the found itself under Polish jurisdiction until the eigh- scholar and bishop St. Mesrop on the basis of Greek teenth century. There are a large number of borrowings and Aramaic scripts sometime around 405–406 CE. In from the Polish language. Belarusian became the offi- the twelfth century, there were two more letters added cial language of Belarus from 1990. Russian language to the alphabet. Armenian literature appeared by the became the second official language of Belarus in fifth century CE and the modern literary Armenian 1996. Belarusian is written in the Cyrillic alphabet. started developing in the nineteenth century. Baltic Languages: Latvia, Lithuania Iranian Languages: Tajikistan, The Russian Federation, Georgia, Armenia, and Azerbaijan Latvian. Is the official language of Latvia. Out of the total population of Latvia in 1998 (2,404,926), there Tajik. Is spoken mainly in Tajikistan. There were were 56.5% Latvians, 30.4% Russians, 4.3% 6,440,732 people in Tajikistan in 1998 with the follow- Belarusians, 2.8% Ukrainians, 2.6% Polish, and 3.4% ing ethnic composition: 64.9% Tajiks, 25% Uzbeks, others. In 1989, Latvian language replaced Russian in 3.5% Russians (the number is declining due to emigra- administration, education, media, etc., although tion), and 6.6% others. Tajik language has been the Russian and some other languages are still used in official language of the country since 1989, although everyday life. Latvian is written in the Latin alphabet. Russian is used in government and business. There are a large number of borrowings from Arab, Uzbek, and Lithuanian. Is the official language of Lithuania. In Russian languages. Before 1930, it was written in 1998, there were 80.6% Lithuanians out of the total Arabic, when the script was changed to Latin; in 1940, population of the country of 3,620,756 people, 8.7% it was changed to Cyrillic. The postindependence gov- Russians, 7% Polish, 1.6% Belarusians, and 2.1% oth- ernment is gradually implementing a change to Arabic. ers. Along with the official language, Polish and Russian are spoken as well. The first written docu- Osetin and Kurd. There are also other small groups ments in Lithuanian are recorded back to the sixteenth of Iranian-speaking people. Osetins are dispersed century. The Lithuanian language is written in the throughout Caucasus (Georgia, the Russian Latin alphabet. Federation). The first written document in the Osetin language is dated back to 941 CE on the basis of the Greek alphabet. In 1844, Shegren created the alphabet Romance Languages: Moldova on a Cyrillic base, which was changed to Latin in the Moldavian/Romanian. The only country of the for- 1920s. In 1938, the Northern Osetian alphabet was mer Soviet Union where people speak one of the changed to Cyrillic and the South Osetian alphabet Romance languages is Moldova. The population of was changed to Georgian. The latter was changed to Moldova consisted of 4,430,654 people in 1989 and Cyrillic in 1954. constituted 64.5% Moldovans, 13.8% Ukrainians, Kurds populate areas in Turkmenistan, Armenia, 13% Russians, 3.5% Gagauz, 2% Bulgarians, 1.5% Georgia, and Azerbaijan, Turkey, Afghanistan, and Jewish, and 1.7% others. In 1989, the Moldavian lan- Lebanon. The Kurds of the Russian Empire used the guage (virtually the same as Romanian) was declared Arabic alphabet; however, in 1921, it the Soviet gov- the official language; however, in 1994 the enforce- ernment changed it to Armenian, in 1929 to Latin, and ment of the Law on Languages was relaxed due to the in 1946 to Cyrillic. 1002  SOVIET UNION: SUCCESSOR STATES Altaic Group Kyrgyz. Is the state language of Kyrgyzstan with a population of 4,685,230 in 1996. A little more than half Turkic Languages: Azerbaijan, Kazakhstan, the population were Kyrgyzs (52.4%), 18% Russians, Kyrgyzstan, Moldova, Turkmenistan, The Russian 12.9% Uzbeks, 2.5% Ukrainians, 2.4% Germans, and Federation, Uzbekistan 11.8% other ethnic groups. The Kyrgyz and Russian Turkic-speaking peoples live mainly in the Central languages are the state languages. Kyrgyz language Asian Republics, although there are some Turkic used Arabic alphabet up to 1926 when it was changed groups in the middle Volga region of Russia (Bashkirs, to Latin; it was changed to Cyrillic in 1940. Tatars, Chuvash) and in the Caucasus (Azeri, Balkars, Karachays, Kumyks, and Nogays). In Siberia, there Turkmen. Turkmenistan is a country of 4,518,268 are numerous groups of Turkic-speaking population (July 2000 est.) people. In 1995, there were 77% who live between the Ural and Lake Baikal (Altais, Turkmens, 9.2% Uzbeks, 6.7% Russians, 2% Khakass, Tofalars, Shors, Tuvans) and also Yakuts Kazakhs, and 5.1% others. The languages spoken are who mainly live in the middle Lena basin and Dolgans Turkmen, Russian, and Uzbek. The Turkmen language in the Arctic. became the official language of Turkmenistan in 1990. A Latin-based Turkish alphabet was introduced in Azeri. Azeri is the state language of Azerbaijan. The 1993, replacing Cyrillic. total population of the country in 1998 was 7,748,163 people, 90% of whom were Azeri, 3.2% Dagestani, Manchu-Tungus Languages: The Russian Federation 2.5% Russians, 2% Armenians, and 2.3% were repre- Another branch of Altaic languages, Manchu-Tungus sented by other ethnic groups. The Russian language is languages, is spoken by the Evenks, Evens, and other still used in the country. small groups dispersed throughout eastern Siberia. The literary Azeri language began its development These languages had never been written before the from the eleventh century. Modern Azeri is based on Russian revolution; presently, they are written in Baku and Shemakhan dialects. The original Arabic Cyrillic. script used by Azerbaijani people was replaced by Latin in 1929 and by Cyrillic in 1939. After 1991, the government introduced a language reform, part of Mongolian Languages: The Russian Federation which was the change to Latin script. The Mongolian languages are spoken in the Lake Baikal region (Buryats) and to the west of the lower Kazakh. Is mainly spoken in Kazakhstan, which has Volga (Kalmyks). These languages got their first the second largest population, after Uzbekistan, of the alphabet, Cyrillic, after the Russian revolution. former Soviet Central Asia (16,733,227 people (1996)). The Kazakh language is also spoken in The Uralic Group Sinkiang Uygur autonomous region in China and in Uzbekistan, Mongolia, and Afghanistan. In 1996 there Finno-Ugric Languages: Estonia, The Russian were 46% Kazakhs, 34.7% Russians, 4.9% Federation Ukrainians, 3.1% Germans, 2.3% Uzbeks, 1.9% The Finno-Ugric languages are spoken in Estonia and Tatars, and 7.1% others in Kazakhstan. Kazakh people the European part of the Russian Federation from used the Arabic alphabet, which was changed into northwest (Karelians, Finns, and Veps) to the upper Latin and later to Cyrillic. The Kazakh language is the Volga river and the Ural mountains (the Mari, state language of Kazakhstan. Mordvins, Udmurt, Komi, and Komi-Permyak) to lower Ob’ river basin in Siberia (the Mansi and Uzbek. Uzbek is spoken primarily in Uzbekistan and Khanty). The areas of Finno-Ugric-speaking popula- in some parts of Kazakhstan, Tajikistan, Turkmenistan, tion are intertwined with areas of other language fam- Afghanistan, and China. In 1996, there were ilies such as Slavic, Turkic, Baltic, and others. Close 49,153,027 people in the country; 80% of them were contacts with other people led to a large number of Uzbeks, 5.5% Russians, 5% Tajiks, 3% Kazakhs, 2.5% borrowings from Russian, Turkic, Romance, Iranian, Karakalpaks, 1.5% Tatars, and 2.5% others. The Uzbek German, and Baltic languages. Practically all of the language is the official language of the country; how- Finno-Ugric languages, except Estonian, use the ever, Russian and Tajik are used as well. Although the Cyrillic script. Arabic script was used for written language before the Bolshevik Revolution, it was changed first to Latin in Estonian. The total population of Estonia was 1927 and to Cyrillic in 1940. The postindependence 1,431,471 people in 1998. Out of them, Estonians con- tendency was to change Cyrillic into Latin. stituted 65.1%, Russians 28.1%, Ukrainians 2.5%, 1003 SOVIET UNION: SUCCESSOR STATES Belarusians 1.5%, Finns 1%, and others 1.8%. 1.8% Abkhaz, and 5% others. Georgian is the official Estonians are the largest group of Finno-Ugric people language, although Russian, Armenian, and Azeri are in the former Soviet Union. Out of two main dialects, used as well. northern and southern, the former (or Tallinn) dialect Georgian is written in its own alphabet, which is served as a base for the Estonian literary language. believed to have been derived by King Parnavaz in the The language is most closely related to Finnish, third century BCE on the basis of the Aramaic script. Livonian, Ingrian, Karelian, and Veps languages. It is The script was subjected to revision under the influ- written in Latin. ence of Greek alphabet when Georgians embraced Christianity in the fourth century CE. In 1926, the Soviet government attempted to replace the alphabet The Samoyedic Languages: The Russian Federation with Latin with no success. Georgian is the only south The Samoyedic group is also quite dispersed in the Caucasian language with old literary traditions. Other Siberian tundra and taiga from Kola Peninsula to the south Caucasian languages, such as Mingrelian, Swan, Yenisey river (spoken by Nenets), around the middle and Laz, acquired written scripts and developed their Ob’ river (spoken by Selkup) and in Taymyr Peninsula literary languages mainly after the Russian (spoken by Nhanasan). Samoyedic languages are Revolution. agglutinative in structure and are spoken by a small number of people. Paleoasiatic Languages The Caucasian Family At the outer edge of far eastern Siberia, there are numerous ethnic groups that speak different languages. This group is made up of more than 40 languages, that One group is the Chukcha (Luorawetlan), Koryak are spoken by about six million people. There are three (Nymylan), and Kamchadal (Itelmen), which are spo- main branches: North-Western (Abkhazo-Adyghian), ken from Chukotka to Kamchatka. Another group is North-Eastern (Nakho-Dagestanian), and South the Eskimo-Aleut group. There are also completely (Kartvelian). isolated languages of people inhabiting Sakhalin Island, lower Amur river, Kolyma lowland, and middle North-Western: The Russian Federation, Georgia Yenisey river. Some of these languages were influ- (Abkhazia) enced by neighboring Yakuts, who speak a Turkic lan- Adyghey, Kabardyn, Abkhaz. It is characteristic of guage, and all of them borrowed considerably from these languages that they obtained alphabets after the Russian. The written scripts of paleoasiatic languages Russian revolution. Adyghey language was first writ- were developed after the Russian revolution. ten in Arabic script, which was later changed to Latin (1927) and then to Cyrillic (in 1938). The Abkhazian References language was written in Latin since 1928; the script was changed to Georgian in 1938, and then to Cyrillic Baskakov, Nikolai A. 1953. The Turkic language of Central in 1954. Kabardin language was also first written in Asia: problems of planned culture contact. London: Central Asian Research Centre. Latin (from the mid-1920s), and from 1936 in Cyrillic. Baskakov, Nikolai A. (ed.) 1982. Razvitie natsionalnykh iazykov v sviazi s ikh funktsionirovaniem v sfere vysshego North-Eastern Languages: The Russian Federation, obrazovaniia (Development of national languages in relation to their functioning in higher education). Moskva: ‘Nauka’. Dagestan, Azerbaijan Bol’shaya sovetskaya entsyklopedia, edited by Prokhorov, A.M. Chechen is spoken in Chechnya and Dagestan (the Moscow: Sovetskaya Entsyklopedia, 3rd edition, 1973 (Great autonomous republics in the Russian Federation). The Soviet encyclopedia, New York: Macmillan, Inc., 1983). original Arabic script was changed to Latin in 1927 Chinn, Jeff, and Robert Kaiser. 1996. Russians as the New and to Cyrillic in 1938. Minority: ethnicity and nationalism in the Soviet Successor States. Boulder, CO: Westview Press. Lezginian language is spoken in Dagestan and Comrie, Bernard, Gerald Stone, and Maria Polinsky. 1977. The Azerbaijan. The Arabic script was used for writing Russian language since the revolution. Oxford: Clarendon before the Russian revolution; in 1928, it was changed Press; as The Russian language in the twentieth century, 2nd into Latin and in 1938 to Cyrillic. edition, Oxford: Clarendon Press, 1996. Huttenbach, Henry R. 1990. Soviet nationalities policies: ruling ethnic groups in the USSR. London, New York: Mansell. South Caucasian Languages: Georgia Kirkwood, Michael (ed.) 1989. Language planning in the Soviet Union. London: Macmillan. Georgian. Georgia’s population in 1998 reached Laitin, David D. 1998. Identity in formation: the Russian- 5,019,538 people. There were 70.1% Georgians, 8.1% speaking populations in the near abroad. Ithaca and London: Armenians, 6.3% Russians, 5.7% Azeri, 3% Osetians, Cornell University Press. 1004  SPAIN Lewis, Glyn E. 1972. Multilingualism in the Soviet Union: Menges, K.H. 1995. The Turkic languages and peoples: an aspects of language policy and its implementation. The introduction to Turkic studies, 2nd, revised edition. Hague: Mouton. Wiesbaden: Harrassowitz. Magner, Thomas F., and William R. Schmalstieg (eds.) 1970. Smith, Graham (ed.) 1996. The Baltic states: the national self- Baltic linguistics. University Park: Pennsylvania State determination of Estonia, Latvia, and Lithuania. University Press. Basingstoke: Macmillan. Matthews, William K. 1951. Languages of the U.S.S.R. ALFIA ABAZOVA Cambridge: The University Press. Spain Spain, a country on the Iberian Peninsula in Western them their language, Latin. This new language gradual- Europe, occupies 504,782 square kilometers (194,897 ly absorbed all of the languages, with the exception of square miles) and has a population of about 40 million Basque, spoken on the peninsula. As a result, all lan- people. It is home to Castilian Spanish, a language that guages currently spoken on the Iberian Peninsula, except has left its linguistic mark in former colonies through- Basque, are Romance languages. It is assumed that the out Latin America and the Philippines, as well as large original languages of Spain (Iberian, Celtiberian, and Spanish-speaking communities in a number of other Basque) influenced the Latin spoken on the peninsula, countries. There are, however, many other languages contributing to the characteristics of the various vernac- that also claim Spain as their home; these include his- ulars. A dialect continuum developed on the peninsula; torical languages (Iberian, Celtiberian, and over time, the Latin of the various parts began to differ Mozarabic), other Romance languages (Catalan, more and more from one another. By the eighth century Galician, Aranese, Aragonese, and Asturian), and one CE, five distinct language groups appeared: Galician- language isolate (Basque). Portuguese, Asturian-Leonese, Castilian, Aragonese, and Catalan. Three particular areas in the north gradual- Historical (Extinct) Languages ly gained certain social prestige, thus establishing their eventual domination: Galician-Portuguese, based around Several inscriptions dating from the fifth to the first Santiago de Compostela in the northwest; Castilian, centuries BCE have been found in eastern Spain in a around Burgos in the north; and Catalan, around language known as Iberian. Most of these inscriptions Barcelona in the northeast. are written in the Iberian script, but a small number A now-extinct Arabic-influenced Ibero-Romance were written in the Greek script. Scholars have unsuc- language, Mozarabic, developed in the south during cessfully attempted to find a relationship between this the Middle Ages and was spoken by Mozarabs unknown non-Indo-European language and Basque, a (Christians under Moorish rule), Muslims, and Jews. language isolate spoken in northern Spain. No standard language developed because Mozarabic Celtiberian is an extinct continental Celtic language was rarely written; when it was written, it was written (as opposed to the more familiar insular Celtic lan- in either the Arabic or Hebrew script. Information on guages such as Gaelic, Irish, Welsh, etc.) that was spo- this language comes down to us in the following ken in northeastern Spain. The Celtiberians may have forms: poems written in the Arabic or Hebrew script inhabited Spain as early as eight BCE. Celtiberian that form part of longer compositions in Arabic or inscriptions and texts date between early two BCE to Hebrew; words and word lists written in the Arabic around one CE. There was no Celtiberian script; earli- script; place names; some medieval and early modern er inscriptions are written in the Iberian script, and Arabic-Latin and Arabic-Spanish dictionaries; and some of the later texts are written in the Latin alpha- certain post-Reconquest legal documents. Slowly, bet. These texts and inscriptions provide important Portuguese, Castilian, and Catalan speakers recon- grammatical information but little by way of vocabu- quered the territories where Mozarabic was spoken. lary as they tend to be short. Mozarabic was probably mutually comprehensible with these varieties of Romance but was gradually Introduction of Latin absorbed by the new dominant language until its dis- At the beginning of the second century BCE, the appearance as a separately identifiable language by 13 Romans entered the Iberian Peninsula and brought with CE. Some have argued that Mozarabic influenced the 1005 SPAIN other languages as it was being absorbed, but this is ple in the autonomous communities of Catalonia, only apparent in vocabulary. Valencia, the Balearic Islands, and Aragon. From the eleventh to the fourteenth centuries, Catalan was con- sidered a prestigious language comparable to French, Modern Languages Italian, Castilian, and Portuguese. Literature and philo- Because of its instrumental position during the sophical works thrived as Catalan replaced Latin and Reconquest, Castile became a dominant force on the Provençal as the language of cultural and literary pro- peninsula. Castile became politically superior to the duction. From the sixteenth century, however, Catalan other regions, and Castilian, the language of the court, declined in prominence as Castile’s power and control government, and the expanding empire, gained pres- increased. Catalan lost its prestige during this time and tige as a result; by the fifteenth century CE, its use as largely became a spoken language. This persecution and the language of culture and administration spread to Castilianization continued until the Renaixença non-Castilian areas. In some areas like Galicia, ‘Renaissance’ of the second half of the nineteenth cen- Castilian became the language of the elite, giving rise tury. Catalan has perhaps had the most successful to diglossia—a situation wherein a speech community revival of any of the other minority languages; this may uses two languages (or two very distinct varieties of be due to the fact that Catalan had an extensive literary one language) for different social contexts or for per- heritage that was written, in contrast to Galician which forming different functions. had more of an oral literary heritage and Basque which Attempts to use language policies to strengthen did not have much of a literary heritage. In 1914, the Spanish national identity began to be considered in the Institut d’Estudis Catalans ‘Institute of Catalan Studies’ seventeenth century CE. In addition, three important was set up by the Mancomunitat, the predecessor of the institutions furthered the position of Castilian through- current Generalitat, the governing body of Catalonia. out Spain: the Catholic Church, the education system, Galician (galego) is spoken by approximately 2.3 and the military. Charles III declared in 1768 that million people in northwestern Spain in the Castilian was to be used as the language of administra- autonomous region of Galicia, where it is one of the tion and education throughout Spain. The official languages. Galician was an important literary Castilianization of Spain continued until the latter half language during the Middle Ages and gained prestige of the nineteenth century when a resurgence of cultur- as a result, especially in lyric poetry; speakers of al activities, including codification and elaboration, in Castilian and Leonese frequently composed poetry in non-Castilian languages occurred; the Romantic Galician because of this prestige. Over the course of Movement motivated these ‘renaissances’. In 1932, the the fourteenth century, however, Galician lost ground king abdicated, and the four-year Second Republic was to Castilian due to the large influx of Castilian nobles proclaimed. Greater powers in the form of statutes of and Castilian-speaking priests. This relegated Galician local autonomy were given to Catalonia, Galicia, and to the low position of a popular spoken form used by the Basque Country during this time, although the Civil peasants. Galician was maintained by the large num- War meant that, with the exception of Catalonia, they ber of Galician speakers who immigrated to other were not able to experience this autonomy properly. parts of the Spanish empire, particularly the Americas. With the end of the Civil War came a dictatorship Galician did not revive until its Rexordimento of the (Franco’s regime), which repressed regional languages second half of the nineteenth century; the Academy of to centralize the country. Speaking non-Castilian lan- the Galician Language was founded in 1906. guages was viewed as antipatriotic and was considered Basque (euskara) is currently spoken by about rebellious; punishments included fines and imprison- 565,000 people in the Autonomous Basque Community ment. In 1966, the Freedom of Expression Law was and the Autonomous Community of Navarre. Not only passed, which removed strict censorship and allowed is it the oldest language still spoken on the Iberian instruction in non-Castilian mother-tongue languages Peninsula but it is also the only non-Indo-European lan- and publishing in these languages. guage; it is a language isolate—its relationship with Franco died in 1975, and three years later, the 1978 other languages has not yet been (nor, perhaps, will ever Constitution was passed. This turned Spain into a be) established. Basque speakers accepted Castilian Western-style democracy and resulted in the legal dominance more readily than the Catalan speakers recognition of the other languages. Castilian was still because, in exchange, they were able to retain their the official language of the country, but each of the rights. Unlike Catalan and Galician speakers, Basque autonomous communities could declare their own offi- speakers did not have a large body of written literature cial language alongside Castilian. to unite their language. The Basque language divided Catalan (català) (called Valencian (valencià) in into several dialects, some of which are no longer mutu- Valencia) is spoken by approximately 6.5 million peo- ally intelligible. 1006  SPANISH AND IBEROROMANCE LANGUAGES Aranese (aranès) is spoken by approximately 6,000 some schools as an optional extracurricular subject. It people in the Val d’Aran, a Pyrenean valley on the is not used in television; however, it can be heard on French–Spanish border forming a part of the Catalonia the radio, and some infrequent magazines are pub- Autonomous Community. It is a dialect of Gascon, a lished in Aragonese. There is a newspaper in North variant of Occitan, spoken in southwestern France and Aragon that includes a weekly Aragonese supplement. is, along with Catalan and Castilian, an official language Asturian is spoken in the Asturian Autonomous in the Val d’Aran. In April 1983, Catalonia enacted the Community and the north and west parts of Castile- Linguistic Normalization Law, as a result of which five León, and although it is not recognized as an official steps were taken in support of maintaining Aranese: language, its protection and promotion is stated in the Aranese was declared an official language of the Val Statute of Autonomy of Asturias. Asturian is present in d’Aran, and the right to know and use this language in some preprimary and primary schools as the language relations and in public ceremonies within the territory of instruction and is offered as an elective subject in was recognized; the Generalitat of Catalonia (the gov- primary and secondary schools. At the University of erning body of Catalonia), along with institutions in the Oviedo and in teacher-training college, Asturian is Val d’Aran, became obligated to do what was necessary offered as an option. Asturian television programs to guarantee the knowledge and normal use of Aranese practically do not exist, although the parliament has and to encourage its normalization; the Aranese form of approved the procedural stage for a project to create a place names in the Val d’Aran was declared the official public television and radio broadcasting institution in form; the Generalitat became obligated to supply the Asturias. There is also one weekly newspaper, Les means necessary to guarantee the teaching and use of Noticies, extracts in other newspapers, and a few mag- Aranese in the schools of the Val d’Aran; and the azines published in Asturian. However, the presence of Generalitat was required to take the necessary steps to Asturian in government is minimal; only a few depart- ensure that Aranese is used in the mass media of the Val ments accept documents written in Asturian, and it is d’Aran. The immediate consequences of this law include rarely used in courts, legislative texts, or public signs. the creation of a commission to standardize Aranese, the publication of textbooks in Aranese for their use in References schools, and a sociolinguistic survey to provide a basis for possible future actions. One of the most important Campbell, George L. 1991. Compendium of the world’s lan- steps, however, was the introduction of Aranese into guages. London and New York: Routledge, 2nd edition, 2000. schools. Aranese is the language of instruction during Lacombe, Georges. 1952. Langue Basque. Les Langues du the first few years of education, although Castilian and Monde, ed. by A. Meillet and Marcel Cohen. Paris: Centre Catalan take leading roles later. The Parliament of national de la recherche scientifique. Catalonia reintroduced the General Council (Conselh Mar-Molinero, Clare. 2000. The Iberian Peninsula: conflicting Generau) into the Val d’Aran, which regularly uses linguistic nationalisms. Language and nationalism in Europe, ed. by Stephen Barbour and Cathie Carmichael. Aranese, and certain government powers were attributed Oxford and New York: Oxford University Press. thereto. In addition, there is one program a week on the Price, Glanville. 1998. Encyclopedia of the languages of Catalan television service in the Aranese language and a Europe. Oxford and Malden, MA: Blackwell Publishers. short daily program in the radio schedule; one Catalan Siguán, Miguel. 1992. España Plurilingüe. Madrid: Alianza newspaper has a weekly Aranese supplement. Editorial; as Multilingual Spain, translated by Miguel Siguán. Amsterdam: Swets & Zeitlinger, 1993. Aragonese is spoken in a few Pyrenean valleys by ROBERT A. CLOUTIER about 30,000 people in the Aragon Autonomous Community, where it is, along with Castilian and See also Basque; Diglossia; Indo-European 4: Catalan, an official language. Aragonese appears in Romance; Spanish and Iberoromance Languages Spanish and Iberoromance Languages Spanish and the Iberoromance languages derive from previously spoken Celtic and non-Indo-European Latin, which was brought into the Iberic peninsula in languages. At the time of the Germanic invasion, in the third century BCE, and superseded most of the the fifth century CE, the Iberic peninsula was not yet 1007 SPANISH AND IBEROROMANCE LANGUAGES fully romanized, as shown by the survival of the stress in both closed and open syllables, the change /f/ Basque language in the northern region, near the gulf  /h/ in word-initial position (/f/ is retained as such of Guascogne, a language presumably spoken by a before /ue/), further palatalization of /kt/ ( /it/ in all much larger part of the population in pre-Roman Western Romance and  /tʃ/ only in Spanish), and leni- times. The Germanic people who poured into the area tion of voiced stops in the internal position. The vowel of the Roman empire left deep linguistic traces in the system includes five phonemes, /a/, /e/, /o/, /i/, and /u/, Iberic languages, much in the same way as in the with no distinction between long and short vowels, and other Romance languages. Peculiar of the Iberic area there are no cases of vowel reduction. The most impor- is the influence of Arabic, following the period of tant phonological feature that builds an inner isogloss domination from the eighth century BCE to the within Spanish varieties is the existence of a distinction ‘Reconquista’, fully accomplished only in 1492. between /s/ and /θ/, typical of peninsular standard, but Drawing the linguistic limits of the Iberoromance not found in Andalusian and American varieties (see area is not without problems. Castilian Spanish and below). Morphologically, vernacular varieties still spo- related vernaculars, Portuguese, and Galician certainly ken today are Leonés, in the province of Leon, which belong to the Iberic branch of Romance, but the position shares a number of features of Gallego and Portuguese, of Catalan remains more controversial: both geographi- Argonés, in the province of Huesca, and Andalusian, cally and linguistically, Catalan was closer in origin to much closer to Castilian, in Andalucia. This last variety the Galloromance branch, especially to Occitan; only has among one of its features the so-called seseo, that is, later cultural and historical developments created more convergence of /θ/ and /s/ in /s/ (the opposite phenome- solid links with the languages of the Iberic peninsula. non, called ceceo, i.e. convergence of both phonemes in The Iberoromance languages share with the rest of /θ/, is also sporadically attested). Since seseo is also Western Romance the palatalization of Latin /kt/, which found in all Latin American varieties, it was formerly became /it/, and the retention of -s plural. Other features assumed that the latter were based on Andalusian; how- of Iberoromance include the use of verbs derived from ever, there is evidence for the existence of two spirants Latin tenere (in Portuguese) and stare as auxiliaries (the in early Mexican Spanish. Most likely, the two spirants latter also found in the Southern Italian varieties were articulatorily closer to each other than /s/ and /θ/, excluding Sicilian) and the prepositional accusative, and eventually merged in both Latin America and with the extension of the preposition a to animate defi- Andalucia, but became more distinct in Castilian. So /θ/ nite direct objects, which, however, is nowadays fully is a Castilian innovation, which dates after the conquest standardized only in Spanish (note that the preposition- of Mexico. al accusative is also found in most Southern Italian In the field of morphology, the largest dialectal varieties, in Sardinian and in Rhetoromance; a similar variation is found in the pronominal system and the development in Rumanian involves the preposition pe, use of verb forms for second person plural. Standard from Latin per). peninsular Spanish has the following set of (accented) personal pronouns: Spanish (Castilian) Singular Plural With its more than 330 million speakers, Spanish is 1. yo nosotros/nosotras one of the most spoken languages in the world, second 2. tú vosotros/vosotras only to Chinese (data from the 1996 edn. of 3. él/ella ellos/ellas Ethnologue). It is the official language of Spain as well Polite form usted ustedes as of Argentina, Bolivia, Columbia, Costa Rica, Cuba, Chile, Ecuador, El Salvador, Guatemala, Honduras, The polite form usted/usetedes, which, in the stan- Mexico, Nicaragua, Panamá, Paraguay, Peru, Puerto dard, agrees with third-person singular/plural verb Rico, Dominican Republic, Uruguay, and Venezuela, forms, was created in the course of the Middle Ages; as and it is spoken as mother tongue by more than 20 mil- in some other Romance varieties, most notably Italian, lion speakers in the United States. there has been a long oscillation between third-person Given its diffusion, Spanish has different standards. forms and second-person plural. The form vos (an ear- Usually, a distinction is made between the peninsular lier second person plural) is still found in literary penin- standard, based on Castilian, and the American stan- sular Spanish of the nineteenth century and lies at the dard, whose prestige center is Mexico City; the most basis of voseo, i.e. the substitution of second-person sin- relevant deviations from this standard in Latin gular tú with vos (see below). The form vosotros/voso- America are found in Argentina. tras for familiar second-person plural only occurs in Special features of Spanish as a Romance language Castilian; the other varieties have ustedes, either fol- include diphthongization of Latin short /e/ and /o/ under lowed by second-person plural verb forms (Andalusian, 1008  SPANISH AND IBEROROMANCE LANGUAGES Canarian) or by third-person plural verb forms (Latin be set up for stressed and unstressed syllables. In America). In Argentina, where voseo is accepted at the stressed syllables, eight oral vowels and five nasal vow- level of the written standard, vos is used for familiar els are distinguished: /i/, /e/, /ε/, /ə/, /a/, /ɔ/, /o/, /u/ and second-person singular, followed by verb forms derived / ˜/,  /e˜/, /a ˜ /, /o ˜/, /u ˜ /. In unstressed syllables, the number from second-person plural, which no longer correspond of distinctions is considerably reduced: in particular, to the forms used in the peninsular standard (vos hablás, final unstressed vowels are reduced to /ə/, /ɐ/, /u/ and ‘you (sg.) speak’; the second-person plural of hablar in non-final vowels to /i/, /u/, /ə/, /ɐ/. In Brazilian peninsular standard Spanish is hablaís), and ustedes is Portuguese, unstressed vowels are reduced to a lesser used for second-person plural, both familiar and polite, extent: final unaccented vowels are /i/, /u/, /a/ (realized followed by third-person verb forms. At the substandard as [ɐ]) and nonfinal unaccented vowels are /i/, /e/, /a/, level, voseo is found in all Latin America, with the /o/, /u/; furthermore, there is no phonemic accented /ə/. exception of Mexico and Peru, the countries that main- Peculiar to Portuguese are the inflected infinitive tained closer links with Spain after the conquest, due to and constraints on clitic placement. The inflected the existence of the Viceroyalties. infinitive is the outcome of the Latin imperfect sub- Another peculiarity of Castilian in the field of pro- junctive. Since by regular sound change its forms nouns is so-called leísmo. In the literary standard, leís- became identical to the forms of the bare infinitive in mo is confined to the use of the clitic le instead of lo the first- and third-person singular, the whole para- for masculine animate definite objects, thus matching digm was reanalyzed as an infinitive with personal (but only for the masculine singular) the distribution of endings. Nowadays, it displays the following forms: prepositional accusative in the case of noun phrases Singular Plural (or accented pronouns), as in Maria vi a Juan/Maria le vi vs. Maria vi el libro/Maria lo vi (some speakers, 1. eu falar falarmos especially in the Madrid and Valladolid areas, also use 2. falares falarem le for inanimate objects). 3. ele/ela falar falarem A variety of Spanish is Judeo-Spanish, or Sefardí, (from the verb falar, ‘to speak’). which derived from the language spoken in Spain at the time of the expulsion of Sephardic Jews, in 1492. European Portuguese has retained a trace of the The expelled Jews moved to different countries, main- once pan-Romance Tobler – Mussafia Law, according ly on the shores of the Mediterranean, where small to which no pronominal clitics could precede the verb communities thrived until the seventeenth century. when the latter was placed in sentence-initial position. Their language, which retains a number of archaic fea- Hence, we find in European Portuguese groups like tures (for example, retention of initial /f/), is still spo- disse-mo, ‘he told me’, to be compared with Spanish ken today by small groups, mostly in Israel. me dijo, or Italian mi disse (in Brazilian Portuguese The first written attestations of Spanish date to the this constraint does not hold; in general, clitics are pre- tenth century CE. Early poetic compositions in the posed to a much larger extent than in European Iberic peninsula were written in a southern vernacular, Portuguese; for example, they are preposed with Mozarabic, the language spoken during the Arabic infinitives: European Portuguese buscar-me, ‘to look domination. The first literary text in Medieval Castilian for me’, Brazilian Portuguese me buscar). is the epic Cantar del mio Cid, written around 1140, The first attestations of Portuguese and Galician which narrates the deeds of the hero of the Reconquista. (which in ancient times were not yet distinct) date back to the twelfth century CE; Galician-Portuguese lyrics flourished especially at the court of King Denis Portuguese (1279–1323). Portuguese is the official language of Portugal, where it is spoken by about 10 million speakers, and of Galician (Gallego) Brazil; according to the 1995 edition of the World almanac, the total speakers of Portuguese were 170 Galician is spoken by about four million people, most million. European and Brazilian Portuguese differ of them living in the region of Galicia, in Spain, considerably from each other in various respects. where it has the status of official language together Portuguese vocalism is rather complicated, if com- with Spanish. It is intimately related to Portuguese, pared with the system of neighboring Spanish, and and the reasons why it is considered a separate lan- more similar to the French system, on account of the guage, rather than a vernacular variety, are ultimately existence of a phonemic opposition between oral and political. In the archaic period, until the sixteenth cen- nasal vowels, and of extensive vowel reduction in tury CE, there was only one literary variety, referred unstressed syllables. Two different vowel systems can to as Galician-Portuguese; only in the course of the 1009 SPANISH AND IBEROROMANCE LANGUAGES sixteenth century did Galician separate from the auxiliary use of anar, ‘to go’, with the infinitive, Portuguese, although some distinctions emerged, as a which yields a periphrastic preterit: va cantar, ‘he consequence of political developments, in the period sang’, equivalent to Spanish cantó. The prepositional from 1350 to 1500. The standardization of Galician accusative did not, in origin, belong to Catalan; today, it was a slow and uncertain process, with conflicting has spread from Spanish to some extent, and has tendencies either toward adherence to the Portuguese become the standard in the case when the object is a standard, or to a deeper distinction; only after the strong personal pronoun, or the indefinite tots, ‘all’, death of Franco, with the new Spanish constitution ‘everybody’, or in cases where word order could cause that recognized official status to minority languages, confusion between the subject and the object. has Galician achieved full recognition. The first attestations of Catalan are somewhat later than those of Spanish and Portuguese, because for a long period Catalan poets used Provençal, which had a Catalan (Catalá) higher prestige. The first written document is a liturgi- Catalan is spoken by about five million people, for the cal book dating to the beginning of the thirteenth cen- most part living in Catalunya, in the province of Valencia tury. In the course of the same century, the Catalan and in the Balearics, where Catalan is the official lan- kings extended their reign to the Balearics, Valencia, guage together with Spanish; small Catalan-speaking and Murcia; consequently, Catalan acquired official communities are located in the South of France and in status. Catalan was also the official language of the Northern Sardinia. Catalan is also the official language kingdom of Aragona, until unification with Castilla in for about 31,000 speakers in the Republic of Andorra. 1479. Spanish then superseded Catalan as the official As already remarked, Catalan was in origin closer language; only in the nineteenth century did separatist to Galloromance, and in particular to Occitan; also movements give an important contribution to the geographically, Catalan was closer to the area of the maintenance and the standardization of the language. Provence: Catalan-speaking communities in the early Middle Ages were located in the South of France and References in present-day Catalunya, and expanded to Valencia Badia i Margarit, Antoni M. 1951. Gramática histórica catalana. during the Reconquista; Mallorca and the Balearics Barcelona: Enciclopèdia Catalana. were conquered by Catalan speakers in 1229. As a Badia i Margarit, Antoni M. 1962. Gramática catalana, 2 vols. result of historical developments, Catalan nowadays Madrid: Gredos. builds a linguistic bridge between the Iberic and the Harris, Martin, and Nigel Vincent (eds.) 1988. The romance languages. London–Sydney: Croom Helm. Galloromance areas. Meyer Lübke, W. 1890–1902. Grammatik der romanischen Features peculiar to Catalan are the vowel system, Sprachen. Leipzig: Fues. consisting of eight vowels— /i/, /e/, /ε/, /ə/, /a/, /ɔ/, /o/, Posner, Rebecca. 1996. The romance languages. Cambridge: and /u/— and the loss of diphthongization of Latin Cambridge University Press. short /e/ and /o/ (but in there is evidence for early diph- Sala, M. 1982. El español de America. Bogotá: Instituto Caro Y Cuervo. thongization in Medieval Catalan). Palatalization of Sephiha, H.V. 1986. Le Judéo-espagnol. Paris: Entente. initial /l/ is a feature of modern Catalan, having start- Thomas, E.W. 1969. The syntax of spoken Brazilian ed in the sixteenth century (in e.g. lleit, ‘milk’, [ʎeit], Portuguese. Nashville: Vanderbilt University Press. cf. Spanish leche, French lait). Williams, E.B. 1962. From Latin to Portuguese, 2 edition. Catalan shares with standard Spanish the distinction Philadelphia: University of Pennsylvania Press. Wright, R. 1982. Late Latin and early romance in Spain and between the auxiliaries ser and estar, although the Carolingian France. Liverpool: Francis Cairns. extent to which estar is not as wide as in Spanish (ser Zamora Vicente, A. 1967. Dialectología española, 2nd edition. can also occur when it means ‘to be in a certain place’, Madrid: Gredos. where Spanish has estar). An innovation of Catalan is SILVIA LURAGHI Spectral Analysis Speech production and speech acoustics are two main in relation to the linguistic message the speaker intends areas of study in experimental phonetics. Speech pro- to deliver. Speech acoustics is concerned with the rela- duction is concerned with how the vocal organs move tionship between the moving vocal organs and the 1010  SPECTRAL ANALYSIS acoustic signal that they create and also with the way in constricted in the speaker’s mouth (place of articula- which the listener is able to recover the linguistic mes- tion). These phonetic distinctions generally only sage from the acoustic signal. ‘Spectral analysis’ refers emerge acoustically by transforming a waveform into to any technique that transforms an acoustic signal into another kind of display, known as a ‘spectrum’, which its ‘frequency components’. Spectral analysis is an displays the amplitude of air pressure variation as a indispensable part of speech acoustics, and it has appli- function of frequency. It is this transformation that is cations in areas such as clinical speech pathology and at the core of spectral analysis. the development of systems for automatically synthe- What follows is the description of a type of signal sizing and recognizing speech. known as a ‘sinusoid’, which is the building block of spectral analysis, and then a brief introduction to Fourier’s theorem, which shows that any signal can be Nature of an Acoustic Signal represented by the sum of sinusoids—thereby allow- An acoustic signal is created whenever an object ing the spectrum to be obtained for any waveform. moves, causing disturbances to the air pressure in its vicinity. These air pressure changes are propagated Sinusoids outward in all directions from the object and, if they are sufficiently large, a sound is heard when they reach Imagine a wheel with a number of scoops fixed to a a listener’s ears. The loudness of the sound is related water mill that is turned at a constant speed by a flow- to the amount or ‘amplitude’ of air pressure change ing river. Suppose a footpath runs alongside it so that relative to the atmospheric pressure. Different speech half the wheel is below the footpath and the other half sounds are heard because the moving vocal organs can is above it (Figure 2), and that the radius of the wheel be configured in a variety of ways and create different is one meter and exactly one revolution is completed patterns of disturbance to the air pressure. every second. If at several points in time during the A ‘speech pressure waveform’ or simply ‘wave- wheel’s cycle, measurements are made of the height of form’ is a record of the amplitude of air pressure one of the scoops above or below the path, the result- changes as a function of time that result from speech ing plot is a sinusoid. Since (in this example) the production. A waveform provides much information wheel completes one revolution of the cycle every about the differences between broad groups of speech second, the ‘frequency’ of the sinusoid is said to be sounds: in Figure 1, the pattern of air pressure varia- one cycle per second or 1 Hertz (abbreviated to 1 Hz). tions is very different for the oral stop /t/ compared The sinusoid has an axis of amplitude (of height in the with the fricative /s/ or the vowel /æ/. But it is gener- case) and an axis of time, but it is possible to represent ally much less useful for showing the acoustic differ- the same information in a spectrum that has axes of ences within any of these classes (i.e. stop, fricative, amplitude and frequency. For this example, the spec- vowel), e.g. acoustic differences stemming from the trum of the sinusoid is a single vertical line, of ampli- different positions where the airstream is blocked or tude 1 (meter), at a frequency of 1 Hz (Figure 3a) s t æ m p Amplitude Figure 1. Waveform of the word ‘stamp’ produced by an adult male speaker of Australian English. The horizontal line corresponds to atmospheric pressure. The vertical lines show the approximate 0 100 200 300 400 500 location of the boundaries between the Time (ms) speech sounds. 1011 SPECTRAL ANALYSIS 1 Amplitude (m) 0 ¼ ¾ −1 ½ 0 ¼ ½ ¾ Time (s) Figure 2. The circle represents a revolving wheel and the dots are one of the scoops at different time points on the wheel as it revolves. The horizontal dotted line is the path. The values around the circle are the times at which the scoop reaches that point in the wheel’s cycle. A plot of the height of the scoop above or below the path as a function of time results in the sinusoid on the right. has the same size as the one that was originally consid- 2 −1 0 1 (a) ered but that revolves at twice the speed. In this case, 1 the amplitude is still 1 unit, but 2 cycles are completed 0 1 2 every second—hence, the resulting frequency is 2 Hz (Figure 3c). The third type of change is one of ‘phase’, −2 −1 0 1 2 2 which refers to where in the cycle the measurements are first taken. For example, if the first measurement in 1 (b) Figure 2 is taken a quarter of a cycle earlier, when the 0 1 2 scoop is level with the path and rising, the sinusoid is Amplitude 2 itself shifted a quarter of a cycle back in time (Figure Amplitude −1 0 1 3d). Notice that the spectrum is unchanged, because 1 (c) neither the amplitude nor the frequency is affected by a 0 1 2 change in phase. It turns out that as far as distinguish- ing acoustically between speech sounds is concerned, 2 phase changes are more or less irrelevant; hence, they −1 0 1 1 (d) are considered no further in this discussion. 0 1 2 0 ½ 1 Frequency (Hz) Periodic Waveforms and Fourier Analysis Time (s) The waveform at the top of Figure 4 is said to be ‘peri- Figure 3. Waveform and spectrum of a sinusoid of amplitude odic’ because it has a basic pattern, or ‘cycle’, that 1 and frequency 1 Hz (a); amplitude 2 and frequency 1 Hz (b); repeats itself in time. The rate at which the pattern in amplitude 1 and frequency 2 Hz (c); and amplitude 1 and a periodic waveform is repeated is known as its ‘fun- frequency 1 Hz and phase-shifted 14 cycle earlier (d). damental frequency’. This waveform repeats itself 2.5 times a second, and hence its fundamental fre- corresponding to the wheel of radius 1 unit that com- quency (which is abbreviated F0 or f0 and pronounced pletes one revolution every second. ‘F-zero’) is 2.5 Hz. Now consider the effect on the resulting sinusoid of A listener hears a periodic waveform as having changing the configuration of the wheel, path, river, ‘pitch’. The pitch is perceived to increase in approxi- and scoop in three different ways. First, there might be mately equal steps on the frequency scale up to about a wheel of twice the size (in a different river) that hap- 1,000 Hz, and thereafter it increases approximately pens to revolve at exactly the same speed as the one logarithmically. A note played on a musical instru- that has been considered so far. The resulting sinusoid ment has a periodic waveform and an identifiable and spectrum in Figure 3b show that the amplitude has pitch, as do most vowels that are produced in speech. doubled but the frequency is the same—therefore, the On the other hand, the sound produced from rustling spectrum shows a vertical line of 2 units at a frequency leaves or the slamming of a door has an ‘aperiodic’ of 1 Hz. Second, there might be yet another wheel that waveform with no discernible repeating pattern. 1012  SPECTRAL ANALYSIS Accordingly, listeners do not hear them as having and 7.5 Hz for sinusoids c and d, respectively. The pitch in the same sense. spectrum of a periodic waveform always consists of a Fourier analysis, named after the French mathe- set of discrete lines on the frequency axis and is known matician Joseph Fourier (1768–1830), is a technique as a ‘line spectrum’. that allows a waveform to be decomposed into a set of The examples so far have been of mathematically sinusoids. The meaning of ‘decomposed’ is that if the pure waveforms. As far as speech production is con- sinusoids that result from Fourier analysis are summed cerned, most vowels and many voiced consonants are (i.e. played back at the same time), then the original produced with the vocal folds (vocal cords) opening waveform is exactly reconstructed. In Figure 4, the and closing at a rapid rate (typically around 100 times sum of the simple sinusoids b, c, d yields the complex per second or 100 Hz in adult male speakers and just waveform in a. below 200 Hz in adult female speakers), and this peri- There is a second important property of Fourier odic vocal fold vibration results acoustically in a analysis: the sinusoids that result from the decomposi- waveform that is sufficiently periodic that the harmon- tion of the type of periodic waveform shown in Figure ics are generally detectable in the spectrum. In Figure 4a have frequencies at multiples of the periodic wave- 5, there are peaks in the spectrum at intervals of about form’s fundamental frequency. This is why the lowest 90 Hz that are the result of the harmonics, giving rise frequency sinusoid that is produced from Fourier to a sawtooth effect in the spectrum. This 90 Hz inter- analysis in Figure 4, namely sinusoid b (known as the val is predictable from the waveform in Figure 5 in the ‘fundamental’), also has a frequency of 2.5 Hz, while following way. The waveform shows that there are 4 the next two sinusoids higher in frequency (known as cycles in 0.043 seconds. The fundamental frequency is the ‘2nd and 3rd harmonics’) have frequencies of 5 given by the number of cycles per second, which in this case is 4/0.043 Hz or just over 90 Hz. Voiceless sounds are produced without vocal fold vibration, and the airstream becomes turbulent as it is channeled at high speed through a narrow gap at the consonant’s place of articulation. This turbulent airstream, which is audible as the hissing sound in the [s] of ‘see’ or the [ʃ] of ‘she’, results in an ‘aperiodic’ signal that has no cycle or repetitive pattern. In contrast Figure 5. Top: Waveform of four pitch periods of the [æ] of Figure 4. A periodic waveform and the three sinusoids that ‘stamp’ from Figure 1 and their durations. Bottom: The spec- are the result of its Fourier analysis. a, b, c, d denote ampli- trum of this waveform showing the first three harmonics (ver- tude values on the waveforms at the same time point and tical dotted lines) and the frequency location of the first three their relationship is a  b  c  d. formants (F1, F2, F3). 1013 SPECTRAL ANALYSIS to a periodic signal, the spectrum of an aperiodic signal Source–Filter Theory of Speech Production does not have equally spaced harmonic components Speech production is often characterized as the com- but is ‘continuous’: that is, there are spectral compo- position of a ‘source’ and a ‘filter’. The source causes nents at all frequencies (Figure 6). the air in the vocal tract to vibrate: for voiced sounds, the source is the vibrating vocal folds, and for voice- Formants less sounds it is a turbulent airstream. In speech pro- duction, the shape of the vocal tract acts as a filter Consider the effect of tapping two glass jars, one long because it modifies the effects of the source. The dif- and quite wide and the other short and thin. These taps ferent contributions to the spectrum of the source and result in two quite different sounds (one sounds much filter can be seen in Figures 5 and 6: the source results lower in pitch than the other). The reason for these dif- in short-term variation that rides up and down on a ferences is that each jar has its own set of ‘resonance longer-term trend line (whose peaks are the formants), frequencies’ —these are the frequencies at which the which is the acoustic effect of the filter. air inside the jar is likely to vibrate at highest ampli- In speech production, the source and filter are large- tude, once the air is set in motion by the tap against its ly independent of each other because they arise from side. The frequencies at which the jar resonates with processes that can themselves be independently var- highest amplitude depend predominantly on the jar’s ied. In voiced speech, for example, a speaker can pro- shape. So too in speech production. If you set yourself duce the same vowel but can vary the pitch (as when to produce an [æ] vowel (analogous to one of the jars) singing an [æ] vowel on a rising pitch). The acoustic and flick a finger against your throat, you will hear a consequence of this independence of the source and muted sound that decays quickly. If you do it again but filter is that the formant frequencies are negligibly set yourself to produce an [i] vowel, you will hear a affected by changes in the harmonic structure. If the quite different sound. Since [æ] and [i] are produced same [æ] vowel as in Figure 5 were produced at twice with very differently shaped vocal tracts, the set of res- the pitch, the frequency interval between the harmon- onance frequencies, which in the case of the acoustic ics (short-term variation) would double, but the shape analysis of voiced sounds are known as ‘formant fre- of the long-term trend line due to the filter (vocal tract quencies,’ are also quite different. shape) would be largely unaffected. In the spectrum of a speech sound, the formant or resonance frequencies show up as amplitude peaks and are labeled F1 (the first formant frequency), F2, Spectrograms F3, …, Fn (Figures 5 and 6). The first two formant fre- A spectrogram is a three-dimensional representation quencies provide listeners with the most important that shows how spectral information changes in time. information for distinguishing between phonetically It is calculated by dividing the waveform into equal different vowels. They are proportional, respectively, time sections or ‘windows’ and applying a Fourier to how widely open the lips are and the position analysis to the waveform within each window. Each between the lips and vocal folds of greatest narrowing analysis produces a spectrum of exactly the kind that in the vocal tract. has been considered so far but in which the amplitude is represented by a gray scale (white is zero amplitude; black is maximum amplitude). In Figure 7, which is a spectrogram of the waveform in Figure 1, the first R three formants in the vowel are clearly visible at about R R 700, 1,600, and 2,300 Hz. Log. Amplitude Amplitude Figure 7 is an example of what is often known as a ‘wideband spectrogram’, and it has a ‘bandwidth’ of 300 Hz. Any two components whose frequency differ- ence is less than the bandwidth are indistinguishable on the frequency axis. Since for this speaker the funda- mental frequency for the [æ] vowel was estimated at 90 50 ms 3,000 5,000 7,000 Hz (Figure 5), the separate harmonics, which occur at Frequency (Hz) Time 90 Hz intervals, are indistinguishable on this wideband spectrogram, because their frequency separation is less Figure 6. Waveform of the section of [s] from ‘stamp’ in than the 300 Hz bandwidth. The ‘temporal resolution’ Figure 1 and its spectrum (right). A trend line is superim- posed on the spectrum. Three of the resonance frequencies of a spectrogram, which can be used to determine for [s] between 3,000 and 7,000 Hz are marked as R above whether two events are separately visible on the time the spectrum. axis, can be estimated from the reciprocal of the band- 1014  SPECTRAL ANALYSIS s t æ m p 8,000 s t æ m p 3,000 2,000 4,000 6,000 Frequency (Hz) 2,000 Frequency (Hz) 0 200 ms Time 1,000 Figure 7. Wideband spectrogram and waveform of ‘stamp’ in Figure 1. The vertical dotted lines mark the approximate loca- tion of boundaries between the speech sounds. The white lines on the spectrogram are the first three formant frequencies. width. For this wideband spectrogram, the temporal res- olution is 1/300 second or just over 3 milliseconds. Accordingly, the individual vocal fold closures for the [æ] vowel are separately visible as vertical lines or ‘stri- 200ms Time ations’ (Figure 7), since the interval between them, which is approximately 1/90 second or 11 milliseconds, Figure 8. Narrowband spectrogram of ‘stamp’ in Figure 1. is greater than the spectrogram’s temporal resolution. The white lines are the first three formant frequencies. Figure 8 is a ‘narrowband spectrogram’ that was calculated with a 45 Hz bandwidth. In this case, the for example). This separation could be achieved by vertical striations for [æ] are no longer visible in the applying Fourier analysis, resulting in a spectrum with vowel, because the temporal resolution, which is given two separate lines at these frequencies. Therefore, if by 1/45 second (just over 22 milliseconds) is greater Fourier analysis were applied to a spectrum of speech, than the interval between vocal fold closures. But the then the result, known as a cepstrum, should be a harmonics are visible because the frequency differ- separation of the rapidly and slowly varying compo- ence between them (90 Hz) is greater than the 45 Hz nents, i.e. of the source from the filter. Since they bandwidth. This type of spectrogram has often been appear in different parts of the cepstrum, the source used for measuring the extent and direction of changes can be quite easily removed, leaving only the contri- in pitch. From Figure 8, it is immediately clear that the bution from the filter. word was produced on a falling pitch. If Fourier analysis is applied once more, but this time to the cepstrum with the source removed, the Cepstrally and LPC Smoothed Spectra result is another spectrum: it very much resembles the Speech scientists often need to be able to estimate the original spectrum of the speech signal, but it is con- separate contributions to a speech signal of the source siderably less bumpy, precisely because the contribu- and the filter. Two techniques for doing this, which can tion from the source component has been mostly only be applied to signals that have been digitized on removed. One of the main advantages of this spec- a computer, are ‘cepstral analysis’ and ‘linear predic- trum, which is known as a ‘cepstrally smoothed spec- tive coding’ (LPC). trum’, is that it is much easier to see the formant The starting point for cepstral analysis is the obser- frequencies, since they are not confounded with the vation that was made earlier: the spectrum of a speech short-term effects of the source (Figure 9). The part of sound is made up of a rapidly changing part that is due the cepstrum that includes information about the filter, to the source and a slowly varying trend line that is the known as ‘cepstral coefficients’, provides consider- result of the filter. Consider in the light of this the able information about the acoustic differences problem of separating two sinusoids that are merged between phonetically distinct sounds. For this reason, together in a waveform, one of which varies rapidly cepstral coefficients are often used in systems for and the other slowly (at frequencies of 100 and 5 Hz, automatic speech recognition. 1015 SPECTRAL ANALYSIS given rise to an acoustic speech signal can be recon- F1 structed by estimating the different scale factors on successively delayed versions of that signal. This is F2 what LPC tries to do. The scale factors are known as ‘LPC coefficients’. F3 The Fourier analysis of the LPC coefficients results in an ‘LPC-smoothed spectrum’. As with cepstral Log. Amplitude F4 analysis, the spectrum is smooth because the contribu- tion from the source is minimal. LPC analysis also allows the formant frequencies to be estimated, because there is a direct correspondence between them and the LPC coefficients. The formant frequencies in the spectrograms in Figures 7 and 8 were automatical- ly tracked using LPC analysis. References 0 1,000 2,000 3,000 4,000 Fant, G. 1960. The acoustic theory of speech production. The Frequency (Hz) Hague: Mouton. ———. 1973. Speech sounds and features. Cambridge, MA: Figure 9. Cepstrally smoothed spectrum of the waveform in MIT Press. Figure 5 showing the frequencies of the first four formants. Flanagan, J. 1972. Speech synthesis, analysis, and perception. New York: Springer-Verlag. Harrington, J., and S. Cassidy. 1999. Techniques in speech The basis for the LPC analysis of speech, which in acoustics. Dordrecht: Kluwer. other fields is sometimes known as ‘autoregressive Johnson, K. 1997. Acoustic and auditory phonetics. Cambridge, MA: Basil Blackwell. analysis’, is as follows. If a speech signal is the result Kent, R., and C. Reid. 1992. The acoustic analysis of speech. of passing a source (vibrating vocal folds or a turbulent San Diego: Singular. airstream) through a filter (representing the shape of Ladefoged, P. 1962. Elements of acoustic phonetics. Edinburgh: the vocal tract), then it can be shown that the same sig- Oliver & Boyd. nal is equal to the sum of several ‘scaled’ and ‘delayed’ Lehiste, I. 1967. Readings in acoustic phonetics. Cambridge, MA: MIT Press. versions of itself added to the source. ‘Scaled’ means Lindblom, B., and S. Sundberg. 1971. Acoustical consequences that the speech signal’s amplitude is linearly changed: of lip, tongue, jaw and larynx movement. Journal of the the amplitude of the entire signal might be halved, for Acoustical Society of America 42. 830–43. example. ‘Delayed’ means that the entire speech signal Olive, J., A. Greenwood, and J. Coleman. 1993. Acoustics of is shifted back in time: if the signal is delayed by a frac- American English speech: a dynamic approach. New York: Springer-Verlag. tion of time, then all its values start a fraction of a time Pickett, J. 1998. The acoustics of speech communication. earlier. The factor by which each delayed signal is Boston: Allyn & Bacon. scaled is dependent on the characteristics of the filter Stevens, K. 1998. Acoustic phonetics. Cambridge, MA: MIT and therefore on the shape of the vocal tract. Hence, a Press. model of the vocal tract shape (filter) that could have JONATHAN HARRINGTON Speech Acts In distinguishing between constative and performa- Constative language reports on reality. We eval- tive language, the twentieth-century English philoso- uate constative language as true or false. ‘The snow pher J.L. Austin initiated ‘speech act theory’, which is white’ is an example, as is the following: ‘All has influenced linguistics and literary criticism. men are mortal; Socrates is a man, so Socrates is Constative language refers to things; performative mortal’. The first example describes a visual phe- language does things—it is speech in action. nomenon. The second details a logical relation’s 1016  SPEECH ACTS reality and describes Socrates as a man and all men as describe two people as being husband and wife: it mortal. makes them husband and wife, but only in the context Both examples are statements with a determinable of certain felicity conditions. If the person uttering the sense and reference, usually called ‘meaning’. In addi- performative lacks authority to perform marriages, the tion to simply conveying their meaning, statements man and the woman are not made husband and wife. may be uttered to bring about effects in listeners or Austin lists general felicity conditions for perfor- readers. The statement ‘There is a hungry lion in the matives: that recognized conventions specify who room’ produces fear in the room’s human occupants, should pronounce the performative, by what proce- in addition to informing them about the lion. Although dures, and to what effect; that the participants enact clearly a constative sentence, this sentence accom- these conventions correctly and completely; and that plishes a nonlinguistic speech act: it frightens. the person uttering the performative intends it seri- Sentences may be uttered simply to say meaningful ously. If, after the jury has returned a ‘not guilty’ ver- words, but in saying these words, a certain effect may be dict, the judge says to the accused, ‘I sentence you to achieved. Both cases are speech acts, but whereas the death,’ we would have an infelicitous performative first aims to achieve a linguistic effect—namely, that the because of conventions being violated. If John Doe listener understands what is said—the second may aim says, ‘I promise to pay’ without intending to keep the at nonlinguistic effects—for example, evoking a certain promise, the performative is infelicitous. Austin does emotional response. Austin refers to constative sen- not consider promises or death sentences uttered by tences as ‘locutions’ and to utterances intended to an actor playing a character on stage to be performa- achieve a different effect as ‘illocutions’, and the partic- tives because they are not intended seriously: they are ular intent of illocutions is labeled ‘illocutionary force’. fictions. A sentence that mentions a person allows us to under- Speech acts’ relations to intention are matters of stand to which person the sentence refers. If, in addi- ongoing debate. In Limited Inc (1988), the philosopher tion, the sentence commands that we despise or admire Jacques Derrida claims that Austin privileges intention that person, the sentence is an illocution. An illocution as the anchoring center of the context that grants a per- that not only aims to achieve an effect but is also suc- formative felicity. Linguistic signs, i.e. words, sen- cessful at achieving it is called a ‘perlocution’. All com- tences, and so on, must be able to function in their mands are thus illocutions, but only commands speakers’ or writers’ absence. Listeners can understand resulting in obedience are perlocutions. a voice echoing in a canyon even if the speaker died Illocutionary force is distinct from locutionary during enunciation. The voice repeats its message sense, reference, and meaning, even though illocutions despite the speaker’s absence. Signs can be iterated, or necessarily involve locutions. Consider the following repeated, beyond the context of their production. This illocution: ‘Impeach George W. Bush, President of the inherent repeatability, argues Derrida, guarantees that United States: he lied about Iraq’s weapons of mass speech acts cannot be fully explained by any given destruction!’ This example has illocutionary force context and cannot wholly depend on intention. A whether the words ‘he lied about Iraq’s weapons of speech act always has a context, but iteration prevents mass destruction’ are or are not legally meaningful (is that context from ever completely determining the Bush’s lying an impeachable offense?) or do or do not speech act’s meaning. As written or verbal signs, per- have a referent (did Bush lie?). If the example’s illo- formatives function in the radical absence of their cutionary force results in Bush’s impeachment, it has writers or speakers and, thus, their intentions. achieved a perlocutionary act. Michelle Rosaldo’s essay in Language in society What Austin calls ‘performatives’ constitute a par- (1982), ‘The things we do with words: Ilongot speech ticular class of illocutions. The typical performative acts and speech act theory in philosophy’, argues that includes a singular first-person pronoun as the subject a focus on intention, especially in relation to promises, (‘I’) and an active verb in the present tense indicative. biases speech act theory toward modern individualist ‘I promise to pay’ and ‘I declare the meeting over’ are ideology by emphasizing autonomous psychological examples. ‘I declared the meeting over’, a constative states. Miller’s Speech acts in literature (2001) locution, reports, but does not perform, a past act and extends Derrida’s arguments to claim that Austin’s lec- can be judged true or false. tures at once display and subvert sexist and imperialist Performatives are ‘felicitous’ or ‘infelicitous’ rather ideologies. Miller details how Austin’s examples tie than true or false. Felicitous performatives conform to intention to a white male speaker’s sovereignty. particular conditions, which are often institutional However, Miller argues, Austin unintentionally dis- conventions, whereas infelicitous performatives do mantles the constative/performative opposition, opens not. A classic example comes from the marriage cere- the possibility of performatives preceding and estab- mony. ‘I pronounce you husband and wife’ does not lishing their own felicity conditions, and discloses 1017 SPEECH ACTS other features of speech acts that militate against the Butler, Judith. 1997. Excitable speech: a politics of the perfor- very ideologies that Austin’s arguments harbor. mative. New York and London: Routledge. Derrida, Jacques. 1988. Limited Inc. Evanston: Northwestern Countering arguments such as Derrida’s, the University Press. philosopher John Searle maintains that intentionality Miller, J. Hillis. 2001. Speech acts in literature. Stanford: is a legitimate foundation of speech act theory. Searle Stanford University Press. differentiates between intentionality as a prelinguistic Petrey, Sandy. 1990. Speech acts and literary theory. New York feature of human and animal life and intentionality in and London: Routledge. Rosaldo, Michelle Z. 1982. The things we do with words: language, which is ultimately derivative of intentional- Ilongot speech acts and speech act theory in philosophy. ity as a biological endowment. Language and speech Language in Society 11(2). 203–37. acts are to be explained in terms of intentionality, Searle, John R. 1969. Speech acts: an essay in the philosophy of rather than intentionality being explained in terms of language. Cambridge and New York: Cambridge University language. Reasserting context as a controlling factor, Press. ———. 1977. Reiterating the differences: a reply to Derrida. Searle extends Austin to define types of illocutions (to Glyph 1. 198–208. request, to congratulate, and so on) and their condi- ———. 1979. Expression and meaning: studies in the theory of tions of felicity. speech acts. Cambridge and New York: Cambridge University Press. ———. 1983. Intentionality: an essay in the philosophy of mind. Cambridge and New York: Cambridge University Press. References Tsohatzidis, Savas L. (ed.) 1994. Foundations of speech act the- Austin, J.L. 1961. Philosophical papers, Oxford and New York: ory: philosophical and linguistic. London and New York: Oxford University Press, 2nd edition, 1970, 3rd edition, Routledge. 1979. ROBERT S. OVENTILE ———. 1980. How to do things with words: the William James lectures delivered at Harvard University in 1955. Oxford See also Austin, John Langshaw; Ideology and Lan- and New York: Oxford University Press. guage; Searle, John Speech Perception The speech stream is a continuously varying signal known as anticipatory coarticulation. As a result, the that, contrary to the listener’s impression of silence acoustic properties of [s] preceding [u] will be different between words, does not contain any pauses. The pri- from those of [s] preceding [i]. Thus, the pronunciation mary goal of research in speech perception is to illu- of the initial sound in the above words provides minate the way in which the listener converts this acoustic cues about the initial sound itself and about continuous signal into a sequence of discrete, mean- the immediately following sound at the same time. This ingful units. This process draws on perceptual, lin- parallel transmission of information can lead to a seg- guistic, and cognitive factors. mentation problem in that it is difficult or impossible to At least three major issues need to be taken into chop up the speech signal in chunks that correspond to account when considering the mapping between only a single speech sound. In other words, it is diffi- acoustic parameters and linguistic units. The first con- cult—if not impossible—to tell exactly where the con- cerns the segmentation of the speech stream. Speech is sonant ends and the vowel begins. a highly efficient means of communication in which A second related issue concerns the lack of lineari- multiple layers of information are transmitted in paral- ty. Although the listener perceives the speech signal as lel. Compare, for example, the words ‘sea’ and ‘Sue’. a linear sequence of units (the size of which is still The former consists of an [s] consonant followed by a under debate), the acoustic cues to these units do not vowel [i], which is pronounced with unrounded lips, necessarily occur in a corresponding left-to-right order whereas the latter consists of the same initial consonant in the speech stream. Consequently, acoustic proper- followed by a rounded vowel [u]. During production of ties appearing later in the speech stream may carry the initial consonant in ‘Sue’, lip rounding will start in information that is crucial for the identification of an anticipation of the upcoming rounded vowel, a process earlier occurring speech sound. 1018  SPEECH PERCEPTION The third issue concerns variability or the lack of stop consonants such as [b] or [d]. Discrimination of invariance. The properties of the acoustic signal that these sounds was not any better than their identifica- are thought to elicit perception of speech sounds (the tion. For example, in a typical experiment with syn- acoustic ‘cues’) are never exactly the same. Sources of thetic speech, an important formant frequency of the variability are, for example, differences in vocal tract consonant was manipulated. Typically, [b] has a for- size, speaking rate, phonetic context, emphasis, and mant at 1100 Hz, whereas the equivalent formant of [d] intonation; these can significantly affect acoustic lies at 1,800 Hz. With the help of a speech synthesizer, parameters of the speech signal. For example, the most researchers were able to produce sounds with formants salient frequencies (‘formants’) of the vowel in the that lie somewhere in between. When listeners were word heed are approximately 300 and 2,300 Hz for an asked to identify such intermediate consonants, it adult male and 500 and 3,100 Hz for a child. However, turned out that listeners seem to use a particular fre- despite these very different formant frequencies quency as the break-off point: all consonants whose caused by differences in vocal tract size, listeners per- formant exceeded this threshold frequency were identi- ceive both utterances as containing the vowel [i]. This fied as [d], and all others were identified as [b]. In illustrates what is known as the invariance problem: another experiment, listeners were presented with a acoustic cues to a particular speech sound may not be pair of consonant stimuli and asked to tell whether the constant but may instead vary according to the cir- stimuli were the same or different. When the formants cumstances under which they occur. Thus, listeners of consonants both exceeded or both undercut the have to compensate for such differences, or ‘normalize threshold frequency, the listeners seemed to answer at the input’. One of the primary issues in speech per- random. If one consonant fell into the [d] range and the ception, then, is how the listener achieves an invariant other into the [b] range, however, listeners were con- percept despite great variability in the acoustic input. sistently able to categorize them properly, even if the Speech perception experiments usually involve formant frequencies were close together. This pattern manipulated natural speech or synthetic computer- of results is known as categorical perception. generated speech. Systematic manipulation of individ- It was originally thought that categorical perception ual attributes of speech enables researchers to occurred only with speech sounds and not with non- determine which acoustic properties are necessary and speech sounds, and this would suggest that the percep- sufficient cues for a particular percept. In applying this tion of speech engaged specialized mechanisms. methodology to consonants, early research in the However, later experiments with carefully controlled 1950s revealed that English listeners use two primary nonspeech materials have shown patterns of categori- acoustic cues for determining where in the mouth cal perception as well. In addition, animals including exactly a consonant is articulated: the frequency of the the chinchilla, macaque, and Japanese quail have also release burst and the formant transitions from the con- been shown to have human-like categorical percep- sonant into the following vowel. When the researchers tion. These data concerning categorical perception do exposed listeners to consonant–vowel sequences, not readily support postulation of perceptual mecha- keeping the burst frequency constant, the listeners per- nisms that were specially evolved or adapted for ceived a [p] when an [i] was following, but a [k] when speech. Instead, they seem to favor an interpretation an [ɑ] was following. Thus, the exact same cue can based on general auditory mechanisms and psychoa- participate in different perceptions. In highlighting coustic sensitivity. that individual acoustic attributes are highly context The finding of categorical perception suggested that dependent, these experiments also exposed the invari- listeners did not perceive any differences between ance problem. stimuli belonging to the same category. This, however, Debate continues concerning the extent to which the may have been based on the particular response cate- perception of speech involves the use of biological gories that were typically used in identification and mechanisms evolved especially for speech. Evidence discrimination experiments (e.g. /b/ or /d/, ‘same’ or that speech sounds are perceived differently from their ‘different’, respectively). The use of more sensitive nonspeech analogs was first presented in the early response measures reveals that even though listeners 1960s. These studies specifically examined the way in do assign the same label or fail to distinguish between which these two types of sounds are identified and dis- stimuli of the same category, they are in fact aware of criminated. Most types of stimuli (e.g. musical tones, subtle differences. These findings indicate that colors) are much better discriminated than they are although categories play an important role in the per- identified. The greater the physical difference between ception of speech, they are not monolithic but have an two stimuli, the better their discrimination. This was internal structure to which listeners are sensitive. shown for nonspeech sounds as well. However, this The importance of speech categories leads to was not true of certain speech sounds, most notably the question of how they are established. This debate 1019 SPEECH PERCEPTION centers around the issue of whether speech categories typically represent frequency along a linear scale, it is are innate or result from exposure to the ambient lan- known that the auditory system warps the signal such guage. Crucial evidence in this debate is typically that its ultimate representation is more nearly logarith- drawn from perception experiments with infants. mic in nature. A thorough understanding of these Findings from discrimination experiments with infants transformations at the auditory periphery and higher as young as one month old suggest that they divide a levels along the auditory pathway may well have sig- speech continuum in a way very similar to adults, with nificant implications for the current view of acoustic two clearly defined categories and a sharp boundary at cues and their variability. the adult location. Additional research has shown that A second area concerns the relation between the infants up to approximately six months of age can not speech signal and higher levels of organization of the only discriminate speech categories from their native grammar. The first few decades of research on speech language but also from just about any other language, perception emphasized the cataloguing of acoustic as well. However, in the second half of the first year of cues, that is, the acoustic information that listeners life, infants seem to lose their sensitivity to nonnative need to extract from the speech signal to recognize distinctions. Presumably, the decrease in sensitivity to individual speech sounds. Since the 1980s, however, contrasts that do not play a role in the native language increased interest in the way words are recognized has allows for an increase in attention to other aspects of led to research on what is often referred to as spoken or the speech signal that play a role in word learning, auditory word recognition. A central issue for this such as sentence structure and intonation. Acquisition research area is if and to what extent ‘higher-level’ of speech categories can thus be understood as the linguistic and cognitive information that is not present result of the interaction between initial psychoacousti- in the speech signal contributes to word recognition. cally based sensitivities and an increasing awareness Research has shown that when presented with a of the structure of the language to be learned. sequence of speech sounds containing an ambiguous Because the perception of speech draws on many initial consonant, listeners will classify that consonant sources of knowledge, theories of speech perception such that the entire string will result in an existing word often account for only a few of its components. Three instead of a nonword. For example, listeners will report general classes of models may be distinguished. The hearing ‘beef’ rather than ‘peef’ when presented with motor theory of speech perception deals with acoustic the string ‘eef’ preceded by a sound that is ambiguous variability by claiming that the listener has specialized between /b/ and /p/. Conversely, listeners will classify neural mechanisms to convert the speech signal into the same ambiguous initial consonant as /p/ rather than invariant representations of articulatory gestures. /b/ when it is followed by ‘eace’. Findings such as These articulatory gestures are the object of speech these are often considered as evidence that lexical perception; i.e. this theory assumes that the listener information (knowledge about what constitutes a word) attempts to faithfully reconstruct how the perceived affects the listener’s interpretation of acoustic–phonet- speech sounds were articulated by the speaker. The ic information. The extent to which there is feedback theory of acoustic invariance claims that invariant from higher levels of linguistic representation, such as acoustic properties do reside in the speech signal. By the lexicon, is still very much under debate. using specialized neural mechanisms, the listener directly extracts these invariants from the speech sig- References nal and maps them onto phonetic features. Finally, pat- tern recognition models claim that speech perception Gernsbacher, Morton Ann (ed.) 1994. Handbook of psycholin- is much like statistical pattern classification. No spe- guistics. San Diego, CA: Academic Press. Goodman, Judith C., and Howard C. Nusbaum (eds.) 1994. cialized mechanisms are required. Instead, the unit of The development of speech perception: the transition from recognition and the structure of categories (e.g. based speech sounds to spoken words. Cambridge, MA: MIT on prototypes or exemplars) are determined by the Press. nature of the speech signal and general properties of Hardcastle, William, and John Laver (eds.) 1997. The handbook the mammalian auditory system. of phonetic sciences. Oxford and Cambridge, MA: Blackwell. There is a growing recognition that a detailed Harnad, Steven (ed.) 1987. Categorical perception: the ground- analysis of the speech signal alone will not be suffi- work of cognition. Cambridge: Cambridge University Press. cient to obtain a genuine understanding of the way in Jusczyk, Peter. 1997. The discovery of spoken language. which speech is perceived. Consequently, researchers Cambridge, MA: MIT Press. have started to incorporate findings from additional Kuhl, Patricia. 1989. On babies, birds, modules, and mechanisms: a comparative approach to the acquisition of vocal communi- areas. One such area concerns the way in which the cation. The comparative psychology of audition: perceiving speech signal is transformed by the auditory system. complex sounds, ed. by Robert J. Dooling and Stewart H. Although the measurements used in phonetic analysis Hulse. Hillsdale, NJ: Lawrence Erlbaum Associates. 1020  SPEECH PROCESSING Liberman, Alvin M. 1996. Speech: a special code. Cambridge, Nygaard, Lynne C., and David B. Pisoni. 1995. Speech percep- MA: MIT Press. tion: new directions in research and theory. Speech, lan- Marslen-Wilson, William (ed.) 1989. Lexical representation guage, and communication, ed. by Joanne L. Miller and and process. Cambridge, MA: MIT Press. Peter D. Eimas. San Diego, CA: Academic Press. Miller, Joanne L., Raymond D. Kent, and Bishnu S. Atal (eds.) Schouten, Martin E.H. (ed.) 1987. The psychophysics of speech 1991. Papers in speech communication: speech perception. perception. Dordrecht: Kluwer. Woodbury, NY: Acoustical Society of America. Warren, Richard M. 1999. Auditory perception: a new analysis Moore, Brian C.J. (ed.) 1995. Hearing. San Diego, CA: and synthesis. Cambridge: Cambridge University Press. Academic Press. ALLARD JONGMAN Speech Processing Speech processing (SP) is used to facilitate the com- Speech Recognition munication between humans and computers. A system that can converse with a human in a natural and unre- Jill, the user of the computer system, can communicate stricted way has long been seen as the quintessential with Sprach using either a microphone or a telephone challenge in computing. interface. In both cases, a speech recognizer is needed As a field, SP encompasses a wide variety of disci- to convert the acoustic signal into a sequence of words. plines and technologies. From a scientific perspective, Given that a telephone samples its input signal at a it draws on research from Signal Processing, lower frequency, the quality of speech recognition is Acoustics, Linguistics (Phonology, Syntax, Pragma- lower than with a microphone. tics, etc.), and other areas. Speech recognizers are typically designed to solve To present the main concepts of SP, we will be two distinct problems: isolated word recognition and introducing a hypothetical computer information sys- continuous speech recognition. Recognizing continu- tem, Sprach (the reader may think of an automated air ous speech is more difficult than isolated speech travel reservation agent or a museum information sys- recognition due to the presence of disfluencies (e.g. tem). The architecture of Sprach is shown in Figure 1. false starts or interjections such as ‘uh’ or ‘hmm’) and We will describe how a user can engage in a con- prosody effects (e.g. intonation and rhythm). The versation with the information system. The sample accuracy of speech recognition also depends on the transcript of the dialogue is shown in Figure 2. size of the vocabulary used. It is much easier to tell Text-to-speech Interaction module Speech recognizer User model Dialogue module dialogue history Database interface Transaction module Transaction processor IR interface Figure 1. Sprach: a generic speech- based information system. 1021 SPEECH PROCESSING Jill: I would like to fly to Seattle on Friday. Back End Sprach: I am looking for flights from Detroit to Seattle on Friday, March 2. What time of the day do you want to leave? The back end of the system is where transactions are Jill: I would rather fly early in the morning. processed. Our sample system contains three back Sprach: American flight 143 leaves Detroit at 7:05 AM and arrives in Seattle at 9:28 AM. A ticket costs $590.00. Should I book ends — an information retrieval (IR) interface, a data- one first-class ticket using your Discover card? base system interface, and an interface to a transaction Jill: Yes, please. processing component. The IR interface is used to Figure 2. A sample dialogue between Jill and Sprach. retrieve documents (e.g. e-mail messages) from a server. The transaction processing and database inter- faces can be used for a large variety of applications apart individual digits from 0 to 9 than to recognize the such as hotel or air travel reservation systems. names of all employees in a given company. Given the variability in speech characteristics among speakers, Text-to-Speech Generation most practical speech recognizers also need to be trained for a particular user. When Sprach has decided what to say, it still has to fig- One mathematical formalism underlying speech ure out how exactly to utter it. A text-to-speech (TTS) recognition is the noisy channel model. It is used to component is used for that purpose. The input to TTS guess the most likely utterance, given an acoustic signal. is most often a grammatical sentence, either retrieved The most likely utterance W ˆ among the space Λ of all from a sentence dictionary or produced by text genera- utterances W ∈Λ is identified according to the formula: tion from a logical form. The input sentence is decom- posed graphemes (textual units) which are then ˆ = arg max P(W|A) = arg max P(A|W)P(W) W converted into phonemes (phonetic units). The process W ∈Λ W ∈Λ of grapheme-to-phoneme translation is often ambigu- ous, especially in languages with inconsistent pronun- In the formula, P(A|W) is called the acoustic model ciation such as English. In English, the same phoneme and is used to estimate the most likely mappings /f/ can be generated from multiple graphemes (f, ph, between the acoustic signal A and the input utterance. gh, and ff), while the same grapheme can be translated The term P(W) is known as the language model and is to more than one phoneme (e.g. gh can be /g/, /f/, or used as a measure of the grammaticality of a given silent, among other pronunciations). The two main candidate utterance W. Other techniques used in speech approaches used in grapheme-to-phoneme translation processing are Hidden Markov Models, Dynamic Time are rule-based (using linguistically tested translation Warping, and Gaussian Mixture Models. For addition- pairs) and corpus-based (derived from large corpora). al reading, consult the book by Huang et al. To make the output sound natural, TTS systems make use of intonational (prosody) information. Prosody fea- Dialogue Management tures are expressed by controlling input parameters of the speech synthesizer. Some prosody features such as Sprach’s dialogue module shown in Figure 1 requires pauses can be inferred within a certain margin of error two locations to store the history of the interaction from the syntactic structure of the sentence (e.g. the with the user. One of them is long-term memory (or presence of semicolons, commas, and other punctua- user model) in which user preferences and information tion). Features such as timber, dialect, and speech rate extracted from past dialogues is stored. For example, are also often included in the generation process, mostly long-term memory may contain the city where Jill to produce voices that seem to be of different characters. lives or her favorite airline. On the other hand, the short-term memory is used to represent choices that Evaluation can be derived from the current dialogue, e.g. the day and time of the desired flight. The speech recognition research community has In practical speech-based dialogue systems, dia- adopted a set of evaluation metrics that are used to logue interactions between users and the system are gauge progress over the years. One of these metrics is often preprogrammed. Recently, VoiceXML has the error rate, which is simply the percentage of word emerged as a standard for encoding information for insertions, deletions, or substitutions that are needed to such applications. VoiceXML is used to annotate com- convert the output of the speech recognizer to the cor- puter–human dialogues. Some sample tags are: rect text. State-of-the-art recognition systems achieve <MENU> : a list of options, <FORM> : a mechanism around 10% word error rates on unconstrained sponta- for user input, and <EMP> : for emphasized speech. neous speech with a vocabulary of more than 10,000 VoiceXML and equivalent markup language facilitate words. In other words, in a given sentence, one in ten the development of practical speech-based systems. words is incorrectly identified. However, when the text 1022  SPEECH PROCESSING (NEUROBIOLOGY) is constrained, word error rate declines dramatically to Huang, Xuedong, Alex Acero, and Hsiao-Wuen Hon. Spoken approximately 4% for continuous speech with a 1,000- Language Processing: A Guide to Theory, Algorithm and System Development. Upper Saddle River, NJ: Prentice-Hall. word vocabulary and less than 1% for isolated word Jelinek, Frederick. 1998. Statistical methods for speech recog- recognition. nition. Cambridge, MA: MIT Press. Evaluating speech generation is significantly more Jurafsky, Daniel, and James Martin. 2000. Speech and language difficult and remains a research challenge. processing. Englewood Cliffs, NJ: Prentice-Hall. Pallett, David S., Jonathan G. Fiscus, John S. Garofolo, Alvin Martin, and Mark A. Przybocki. 1998 Broadcast News References Benchmark Test Results. Proceedings of the DARPA Cole, Ronald, Joseph Mariani, Hans Uszkoreit, Annie Zaenen, and Broadcast News Workshop, 1999. Victor Zue (eds.) 1997. Survey of the state of the art in human www.voiceml.com. VoiceXML markup language (2001). language technology. Cambridge: Cambridge University Press. DRAGOMIR R. RADEV Speech Processing (Neurobiology) From a physics point of view, the speech signal is based fied by the ear, which is divided into three parts: (1) on a change in pressure propagated through the elastic The external ear, which comprises the cartilaginous medium ‘air’, which leads to a specific particle move- pinna and a canal (external auditory meatus) leading to ment. The speech sound is encoded as a complex parti- a vibratory motion-sensitive membrane (tympanum). cle movement in the frequency and intensity domain As a consequence of the resonance properties of the and is superimposed on nonspeech-related particle ear canal, the pinna, and the head, the frequency range movement (noise). Listening to speech-embedded back- of 1,500 to 7,000 Hz (the main speech range) is pre- ground noise means (1) detecting the particle movement ferred while other frequencies are reduced (passive fil- with a sensory system, (2) filtering out nonspeech-relat- tered). (2) The middle ear, which comprises the very ed particle movement, and (3) analyzing the remaining small bones of the ossicle chain (malleus, incus, speech-related part of the signal and decoding the stapes), transmits the vibratory motions of the tym- meaning. Therefore, hearing sounds, especially listen- panic membrane to the much smaller membrane of the ing and comprehending language, is a very complex ‘oval window’ of the inner ear. This part amplifies the ability, the neurophysiological basis of which is not yet signal mechanically by increasing the pressure (gain) completely understood. This is even more important due to the ratio of the effective areas of the tympanum when the speech signal being focused upon is disturbed to the small oval window and by exerting a lever effect by the noise of other speakers, e.g. in the case of listen- of the oscillating ossicles. (3) The inner ear or cochlea, ing to a single speech signal during a noisy party. The where the motion of the ossicle chain is coupled to the ability to extract only the relevant parts of the air parti- fluids of the cochlear canals, where the sensory cells cle movement related to the speech signal of a certain (hair cells) are located in the organ of Corti of the scala speaker (so-called ‘cocktail party effect’) is an enor- media. Here, transduction from a vibratory motion to mous achievement of the neuronal structures involved. an electrical signal takes place. The following stations However, this powerful discrimination of an acoustic of the so-called auditory pathway are shown in Figure signal is not limited to human language processing. For 2 of Brain Organization and Auditory Pathways. example, seabirds returning to their colony are able to The process of language analysis takes place at an find their straying offspring by differentiating their spe- incredible speed. A trained listener can repeat sen- cific call (‘identificators’) among the calls of up to some tences (‘shadowing’) with a latency of only 220 ms. 10,000 other chicks. Although the physiological basis of This is evidence for the assumption of parallel analysis analyzing complex sounds is phylogenetically old and processes during language comprehension. Experi- quite common in animals, this does not diminish the ments show that words are almost comprehended after enormous achievement of the human brain to compre- only 100 ms, a time period that comprises only one hend spoken language. to two phonemes. The sound signals are mostly dis- The speech signal (encoded as the complex move- criminated by phylogenetically old neuronal feature ment of air particles) is gathered, filtered, and ampli- detectors. This is why most mammals (e.g. tested in 1023 SPEECH PROCESSING (NEUROBIOLOGY) chinchillas) are able to discriminate different plosives about 400 ms, whereas the final word in the sentence in a choice reaction experiment. After words have been ‘The pizza was too hot to eat’ does not. However, a detected and recognized, the cognitive part of language sentence with an intermediate semantic anomaly such comprehension sets in: analyzing the meaning of the as ‘The pizza was too hot to drink’ would elicit an word, analyzing the meaning of phrases and sentences, N400 with a smaller amplitude, since the peak ampli- and analyzing connotative meaning delivered by con- tude corresponds to the extent of the semantic anom- textual information, the speaker’s prosody, or nonver- aly. In contrast, a sentence with a syntactic anomaly bal communication signals. More than milliseconds are like ‘Turtles will spit out things they does not like to required to understand language in that sense; it may eat’ would elicit a positive component (P600), peaking take up to several seconds before a listener ‘gets the about 600 ms after the onset of the word ‘does’. In point’ of a certain utterance (see time scale in Figure 2 recent years, the N400 has been shown for many lan- of Brain Organization and Auditory Pathways). In guages, including American Sign Language (ASL). An addition, such complex analyses and judgments of additional language-related ERP component is the complex utterances are performed by integrative N280, which can be seen as an indicator for process- processes of many different brain areas. During lan- ing closed-class words (function words such as arti- guage processing, participating brain processes can be cles, conjunctions, or prepositions) in contrast to investigated with electrophysiological or brain-imag- open-class words (content words such as verbs, nouns, ing techniques. At the moment, the minority of experi- and adjectives). ments are conducted with invasive electrical recording of single or few neurons. Most electrophysiological References results are based on noninvasive observation of the syn- chronous activity of huge groups of neurons (from Binder, Jeffrey R., J.A. Frost, Thomas A. Hammeke, R.W Cox, some 100,000 up to a few million cells). and Thomas E. Prieto. 1997. Human brain language areas identified by functional magnetic resonance imaging. The Modern electrophysiological and brain-imaging Journal of Neuroscience 17. 353–62. techniques such as electroencephalography (EEG), Brown, Colin M., and Peter Hagoort (eds.) 1999. The neu- functional magnetic resonance imaging (fMRI), or rocognition of language. Oxford: Oxford University Press. positron emission tomography (PET) have led to a Fitch, R. Holly, Steve Miller, and Paula Tallal. 1997. new empirical basis for the investigation of cognition Neurobiology of speech perception. Annual Review of Neuroscience 20. 331–53. (see Neurolinguistics). While event-related fMRI was Gazzaniga, Michael S. (ed.) 2000. The new cognitive neuro- introduced only a few years ago, other techniques had sciences, 2nd edition. Cambridge, MA: MIT Press. already provided a substantial number of results. For Katz, Jack, Nancy A. Stecker, and Donald Henderson (eds.) example, for more than two decades the use of event- 1992. Central auditory processing: a transdisciplinary view. related potential (ERP) techniques has yielded new St. Louis: Mosby. Kutas, Marta. 1997. Views on how the electrical activity that the insights into the physiological processes underlying brain generates reflects the functions of different language language. The language-related wave form of the scalp structures. Psychophysiology 34. 383–98. EEG first described by Marta Kutas and Steven A. Müller, Horst M., Jonathan W. King, and Marta Kutas. 1997. Hillyard in 1980 was the so-called N400, a large neg- Event related potentials elicited by spoken relative clauses. ative ERP component related to semantic analysis of Cognitive Brain Research 5. 193–203. Ojemann, George A. 1991. Cortical organization of language. sentence processing. For example, the final word in The Journal of Neuroscience 11. 2281–7. ‘The pizza was too hot to cry’ elicits a huge bilateral Pulvermüller, Friedemann. 2000. Words in the brain’s language. negativity at posterior electrode positions between 250 Behavioral and Brain Sciences 22. 253–336. and 600 ms after word onset, with a peak latency of HORST M. MÜLLER Speech Production Research on speech production studies the processes to the motor movements of the articulatory organs. and representations that are involved in producing an Although language production has historically been the utterance, from generating the concept to be expressed, subject of less research than comprehension, recent 1024  SPEECH PRODUCTION years have seen increased interest in the question of expressed, the features expressed in the pre-linguistic how speakers are able to produce complex and well- message that is the output of conceptualization vary formed speech in real time. This increased interest has across languages. been accompanied by important methodological devel- During the next stage, formulation, the speaker opments. Until relatively recently, language production converts the prelinguistic message into linguistic form. was mainly studied through observational methods, by Formulation is subdivided into stages of grammatical which researchers collected corpora of naturally occur- encoding and morpho-phonological encoding. Each ring spontaneous speech, particularly speech errors, stage involves building abstract structure, and retriev- and used these to develop theories of the processes that ing content to fill that structure. Grammatical encod- underlie speech production. Observational methods ing is concerned with developing syntactic structure. obviously present methodological difficulties, such as The speaker must select a syntactic structure that can observer bias, infrequency of some phenomena, and express the intended meaning, and congruent lexical the extent to which any corpus is representative of the content. Research suggests that grammatical encoding language as a whole. Some of these problems have is a two-step process of functional processing fol- been solved with the advent of powerful computers that lowed by positional processing. In functional process- can collect and process large quantities of natural lan- ing, the speaker retrieves lexical concepts (the guage. However, methodological developments have semantic content of a lexical entry) and assigns func- also made it possible to study the production of both tional relations (subject, direct object, indirect object, well- and ill-formed utterances under rigorously con- etc.) to them. This is the stage at which at least some trolled laboratory conditions. Experimental manipula- agreement relations are computed. Functional level tions now allow such diverse possibilities as inducing representations are unspecified for linear order. Next, the production of particular syntactic structures or in positional processing, the speaker generates a syn- speech errors of interest, as well as investigating the tactic frame and associates it with the lemma, or syn- link between how we visually perceive the world and tactic content, of each relevant lexical entry. This how we describe it. syntactic frame specifies in which linear order the There is perhaps surprising consensus concerning individual elements are to appear, i.e. each phrase the overall architecture of the language production appears in the position that it will occupy in the final system. The model has been articulated in more detail utterance. by psycholinguists such as Willem Levelt on the basis The output of grammatical encoding forms the of recent experimental research; but in essential input to morpho-phonological encoding, where the respects the currently accepted architecture of the lan- speaker translates the syntactic frame and its content guage production system remains largely unchanged into appropriate articulatory gestures. This involves a from that proposed by Merrill Garrett in the 1970s on number of intermediate stages. First, each word’s the basis of speech error data. According to Willem morphological makeup (involving the relevant word Levelt’s account, production comprises three main forms) must be retrieved. Next, the individual sounds, stages: conceptualization (generating the prelinguistic syllables, and intonational properties of these mor- message), formulation (translating the message into phemes must be specified. Finally, a process of pho- linguistic form), and articulation (executing the motor netic encoding maps from phonological words to programs that realize the linguistic representation). gestural scores, i.e. a pronunciation plan is generated. Each has its own characteristic input and output repre- The last stage in speech production is articulation, sentations. Another component, the self-monitor, is where the speaker executes the abstract gestural assumed to check for errors or other problems at each scores. This involves the storage of phonetic plans in level of processing. an articulatory buffer, their retrieval as needed, During conceptualization, the speaker conceives an unpacking the hierarchy of motor programs, and issu- intention to express, and selects relevant information. ing instructions to the neuromotor circuits that control The representations implicated here are assumed to be the various articulatory muscles. propositional structures, i.e. prelinguistic representa- This overall architecture is widely accepted, but tions that involve expressions (or constituent parts of uncertainty remains concerning the details of the pro- those expressions) that can have truth values. They are duction processes. One question of particular interest also assumed to specify the speaker’s perspective, the is how activation passes between different elements of information structure of the sentence, and any features the production system. Some researchers believe that that are encoded by the grammar (e.g. tense informa- activation can only flow forward, with later levels of tion) of that language. Because languages differ in processing (e.g. morpho-phonological processing) terms of which semantic features are grammatically being unable to affect earlier levels (e.g. functional 1025 SPEECH PRODUCTION processing). Others believe that activation can flow References backward as well as forward. Similarly, it is not clear Bock, Kathryn. 1996. Language production: methods and whether processing is discrete, so that processing methodologies. Psychonomic Bulletin & Review 3. 395–421. must be completed at each level before processing can Bock, Kathryn, and Willem Levelt. 1994. Language production. begin at the next level, or cascaded, so that processing Grammatical encoding. Handbook of psycholinguistics, ed. of later levels can begin while processing is still under by M.A. Gernsbacher. London: Academic Press. Dell, Gary. 1986. A spreading activation theory of retrieval in way at an earlier level. It is also unknown as to how sentence production. Psychological Review 93. 283–321. far the incrementality of speech production extends: Garrett, Merrill. 1980. Levels of processing in sentence pro- elements appear to be processed as they become duction. Language production: Vol. 1. Speech and talk, ed. available, but the minimal unit of processing at each by B. Butterworth. London: Academic Press. level remains unclear. More glaring gaps in our Levelt, Willem. 1989. Speaking: from intention to articulation. Cambridge, MA: Bradfor Books. understanding of speech production concern the Vigliocco, Gabriella, and Robert Hartsuiker. 2002. The inter- nature of conceptualization, and the coordination of play of meaning, sound, and syntax in sentence production. syntactic structure-building, on the one hand, and lex- Psychological Bulletin 128. 442–72. ical retrieval on the other. HOLLY BRANIGAN AND ROBERT J. HARTSUIKER Speech Production: Neurobiology It takes only about two seconds for speakers to com- have proved to be more vulnerable to damage in and pute the correct meaning, grammatical, and sound around the Sylvian fissure of the left hemisphere, in structures to encode their ideas into speech. How does the regions indicated in Figure 1. the brain accomplish this encoding so fast? Which Articulation suffers when there is damage to parts of the brain participate in each level of linguis- Broca’s area. Damage in this part of the brain, locat- tic processing? Neurolinguists have attempted to ed just forward of the oral motor cortex, impairs the answer these questions by studying the speech of ability to sequence the movements of articulators to stroke and cerebral trauma victims who have lan- produce the sounds of the language. In Broca’s apha- guage impairments called ‘aphasias.’ These sia, speech becomes nonfluent and effortful, as if the researchers tried to localize, or identify, the regions of speaker must consciously control the movements of the brain responsible for different linguistic process- the articulators to produce spoken words. In fact, the es, such as grammatical encoding and word compre- utterances of speakers with Broca’s aphasia frequent- hension. However, continuing research has shown ly consist of only one word at a time, with longer that even if a part of the brain is necessary for a cer- utterances being very rare with little grammatical tain aspect of linguistic processing, it is unlikely that complexity. this is the only part of the brain involved in that par- Although all aphasics make some errors in pronun- ticular linguistic function. Furthermore, studies have ciation, the use of sound-related information is most shown that brain damage in different parts of the brain impaired in conduction aphasia, which is caused by can lead to similar language impairments, whereas the damage near the auditory association area in the tem- inverse is also true: the same locus of brain damage poro-parietal cortex. These aphasics produce words may lead to different language impairments in differ- that are off-target by one or two sounds, then repeat- ent people. Knowledge about the contribution of var- edly correct themselves as they attempt to produce the ious brain regions to language use is increasing daily target form, e.g. ‘I bought a bus, but, boot, book’. because of the development of noninvasive technolo- Thus, they are aware that they have produced an incor- gy such as event-related potentials and neuroimaging rect word form, but are not sure, until they hear them- techniques such as positron emission tomography, selves say it, what the correct sound combination is. single-photon emission computed tomography, and Importantly, subjects do not consistently make errors functional magnetic resonance imaging. These tech- on the same word across trials, indicating that the niques confirm that many parts of the brain contribute impairment affects the pronunciation of word forms to language processing in all linguistic tasks. but that the underlying knowledge of the sound struc- Nevertheless, various aspects of speech production ture of words is preserved. Conduction aphasics also 1026  SPEECH PRODUCTION: NEUROBIOLOGY Oral motor area Broca’s area Parietal lobe Frontal lobe Occipital lobe Temporal lobe Sylvian fissure Auditory Wernicke’s area Figure 1. Speech areas in the left association area hemisphere. have impaired auditory short-term memory, which Wernicke’s aphasics also produce nonwords, called interacts with their deficit in pronunciation and results ‘neologisms’, as substitutions for real words. Their in severe impairments of oral repetition. speech is fluent but full of pronouns and ‘empty’ Linguistic impairments found in agrammatic apha- words such as thing, stuff, do, make, and get. Because sia primarily affect words or word parts carrying of unrelated word substitutions, neologisms, and grammatical information, such as grammatical inflec- empty speech, it is frequently difficult to understand tions of verbs and nouns and function words, such as the speech of Wernicke’s aphasics. Consequently, articles and prepositions. Most people with Broca’s although they produce complex sentences, it is usual- aphasia are agrammatic, but not all agrammatics have ly impossible to determine whether sentence structures damage in Broca’s area. However, all people with are being used correctly. agrammatism have damage in or near the Slyvian fis- The meaning level of language is selectively impaired sure. In agrammatism, speakers often omit free-stand- in Alzheimer’s disease and semantic dementia. In these ing grammatical words, such as prepositions and syndromes, word finding is moderately to severely articles, but they substitute incorrect inflected forms impaired, leading to the production of related words for correct ones, for example, ‘Yesterday dog bark from the same meaning category or more general terms me’. Note that in English it is not clear whether the (e.g. saying ‘horse’ or ‘animal’ in response to a picture of word form ‘bark’ has been substituted for ‘barked’ or a deer). These patients score poorly on tasks requiring whether the past tense inflection (-ed) has been omit- full knowledge of word meaning, and their spontaneous ted. In other words, it is hard to determine from speech is often empty of content, as in Wernicke’s apha- English data whether grammatical inflection is com- sia. These syndromes are characterized by diffuse loss of pletely unavailable to the aphasic or whether a simpler, neurons in the temporal lobe and often throughout the uninflected form of a word has been chosen. However, cerebral cortex. Because all aphasics, regardless of the research on aphasic speakers of highly-inflected lan- location of cortical damage, produce these same types of guages like Russian has shown that the process semantic errors, it is unlikely that semantic processing is involves the substitution of one inflected form for localized to a specific area of the brain. another. This means that grammatical information is, It is interesting that researchers have not found a in principle, still available to the aphasic, even though particular type of damage that is associated with errors the wrong form may be selected. affecting word order or sentence structure, although The connections between the sound form of a word many researchers see agrammatism as a syntactic and its meaning are impaired by damage in or near impairment. In contrast, all aphasics show varying Wernicke’s area. This impairment affects both com- degrees of impairments of pronunciation, word find- prehension and production: single-word comprehen- ing, and understanding of complex sentences. These sion is impaired, unrelated words are substituted for findings, in conjunction with those of neuroimag- appropriate ones (e.g. ‘My boss is an envelope’), and ing studies, support the assertion that many areas of patients have no awareness of their own errors. Some cerebral cortex contribute to each level of linguistic 1027 SPEECH PRODUCTION: NEUROBIOLOGY processing, although the areas immediately adjacent Garrett, Merrill. 1992. Disorders of lexical selection. Lexical to the Sylvian fissure seem to hold particular impor- access in speech production, ed. by W.M.J. Levelt, 143–180. Cambridge, MA: Blackwell. tance for language use. Levelt, W.M.J. 1989. Speaking: from intention to articulation. Cambridge: Cambridge University Press. References MacWhinney, Brian, and Elizabeth Bates (eds.) 1989. The crosslinguistic study of sentence processing. Cambridge: Blumstein, Sheila E. 1995. The neurobiology of language. Cambridge University Press. Speech, language and communication, ed. by J.L. Miller and Obler, L. K., and K. Gjerlow. 1999. Language and the brain. P.D. Eimas. California: Academic Press. Cambridge: Cambridge University Press. Caplan, David. 1992. Language: structure, processing and dis- orders. Cambridge, MA: MIT Press. LORI J.P. ALTMANN Speech Synthesis Speech synthesis is the creation of speech sound by Parametric Coding means of a machine or a computer. The development This type of synthesizer makes use of the source–filter of speech synthesis dates back to the end of the eigh- theory of speech production and strings together sever- teenth century when Wolfgang von Kempelen built a al acoustical properties of speech sounds (by means of speaking machine in Vienna. His apparatus was based a set of rules). The sound source is characterized by on a model of the human vocal tract, with bellows for voiced pulses or noise. Subsequently, filters are used to lungs, a conical resonator that could produce vowel modify the source, i.e. to produce different resonances sounds, and constricted portions to create consonantal or formants of vowel sounds or to pass high-or low-fre- sounds. Another pioneer was Alexander Graham Bell quency information of consonants. In some systems, who, together with his brother, built a talking head several parameters can be altered to convey gender and with the pharynx, tongue, velum, lips, and teeth made age. For example, the word ‘seek’ is generated by first from wood, rubber, and cotton. In the twentieth centu- passing noise through a high-pass filter (at about 5,000 ry, with the invention of electronic circuitry, the vocal Hz), followed by a low first formant (350 Hz) together tract transfer functions were simulated by (combina- with a high second formant (2,200 Hz) and high third tions of) sound sources and filters. formant (3,000 Hz) for the vowel /i/, a short silence to simulate pressure buildup before unvoiced plosives and Stored Speech a sharp cutoff sound for /k/. Rules are used to simulate The task of the synthesizer is to create and produce natural timing, stress, and prosody and to create natu- sounds that are recognized as speech. One way to ral connections between the speech sounds. accomplish this is to make use of stored utterances. In The advantage of formant-based synthesis is that it general, the quality of prerecorded speech is high, but requires neither a large storage capacity nor high com- the vocabulary is fixed and restricted, and is only use- putational demands. As the acoustic differences ful for reproduction (e.g. in cars and aeroplanes). between speech sounds can be varied very precisely, Moreover, despite coding techniques, this synthesis this type of synthesis is commonly used for research method requires considerable memory capacity for purposes, for instance, to examine which acoustic storage. More flexible systems require the conversion properties of speech sounds determine the identity of of words to a string of phonemes, make use of pronun- the speech sound. The disadvantage of this type of syn- ciation rules (between phonemes and sounds) stored in thesis is that it proves to be very difficult to connect the dictionaries, and add durational and prosodic informa- different parameter tracks between adjacent sounds in tion. Several text-to-speech synthesis techniques exist, a natural way. The same speech sound produced in a none of which generate speech that sounds completely different context sounds different (e.g. aspirated, natural. At present, the two most common (computer) unaspirated). Ideally, each different realization of the techniques are parametric coding (also called formant same speech sound (phoneme) should be represented synthesis by rule) and concatenative synthesis. by the synthesizer. Moreover, speech sounds do not 1028  SPRACHBUND occur neatly one after the other during production due information retrieval (e.g. banking), for warnings (e.g. to coarticulation. Another problem involves the assign- in cockpits), in reading machines for the blind, for ment of the correct timing and intonation (cf. ‘unit’ and speech-impaired people, for foreign language learn- ‘undo’), and stress differences (cf. ‘project’ and ‘to ing, in toys, and in speech research. The improvement project’). Although different sets of rules exist to dis- of speech synthesis lies primarily in a more natural tinguish pronunciations, and to assign syntactic struc- generation of the prosody of speech. Although syn- ture (to specify the pronunciation of sound sequences), thetic speech is highly intelligible, it usually takes half it remains very difficult to determine the right settings a sentence to know that it is not human. This is often at the right time. because the listener is distracted by small details, such as timing, voice quality, overarticulation, or unnatural Concatenative Synthesis intonation. Whatever the use of synthesized speech, the issues of intelligibility and naturalness will remain Some of the above-mentioned problems can be solved challenging in the forthcoming years. by concatenating pieces of natural recorded speech. For this purpose, diphones and demisyllables are usu- ally used. A diphone is a piece of speech that stretch- References es from the middle of one speech sound to the middle Allen, J.A., M.S. Hunnicutt, and D.H. Klatt. 1987. From text to of the following one. They are cut from several record- speech: the MITalk system. Cambridge, England: Cambridge ed utterances, stored, and recombined during synthesis University Press. Bailly, G., and C. Benoit. 1992. Talking machines — theories, of an utterance. A demisyllable is half a syllable, with models and designs. Amsterdam: North-Holland. the cut in the middle of the vowel. Synthesis of unre- Klatt, D.H. 1980. Software for a cascade/parallel formant syn- stricted English requires approximately 1,000 thesizer. Journal of the Acoustical Society of America 67. diphones or demisyllables. Concatenative synthesis is 971–95. usually highly intelligible and sounds very natural, ———. 1987. Review of text-to-speech conversion for English. Journal of the Acoustical Society of America 82. 737–93. because the articulatory effects from one phoneme to Klatt, D.H., and L.C. Klatt. 1990. Analysis, synthesis and the the next are included. Although the advantage of this perception of voice quality variations among female and type of synthesis is clear, discontinuities can still arise male talkers. Journal of the Acoustical Society of America at the boundaries of the diphones or the demisyllables. 87. 820–57. Moreover, it is not possible to record all existing Kohler, K.J. 1991. Prosody in speech synthesis: the interplay between basic research and TTS application. Journal of words spoken with all possible pronunciations. As this Phonetics 19. 121–38. type of synthesis also requires considerable storage Liberman, A.M., F. Ingemann, L. Lisker, P. Delattre, and F.S. capacity, voice coding techniques are used to com- Cooper. 1959. Minimal rules for synthesizing speech. press the speech sounds. Journal of the Acoustical Society of America 31. 1490–9. Pierrehumbert, J. 1981. Synthezing intonation. Journal of the Acoustical Society of America 70. 985–94. Application Pisoni, D.B. 1997. Perception of synthetic speech. Progress in The best synthesis technique depends on the require- speech synthesis, ed. by J. Van Santen et al., 541–60. New York: Springer-Verlag Inc. ment of the intended application. There is always a Van Heuven, V., and L.C.W. Pols. 1993. Analysis and synthesis trade-off between maximizing speech quality, mini- of speech-strategic research towards high-quality text-to- mizing memory capacity, and computation time. speech generation. Berlin: Mouton de Gruyter. Speech synthesis is used, among other things, for ASTRID VAN WIERINGEN Sprachbund The notion of Sprachbund (or linguistic area) has been high degree of similarity in syntactic form, some sim- introduced in linguistic theory by Nikolaj Trubetzkoj ilarity in the principles that govern morphology, and a in 1928. In his famous proposition XVI, put forward great number of common culture words—sometimes on the occasion of the First International Congress of also an exterior similarity in the inventory of sounds. Linguists at The Hague, Trubetzkoj defines In modern linguistic theory, Sprachbund usually des- Sprachbünde as made up of languages displaying a ignates an area characterized by a number of linguistic 1029 SPRACHBUND features shared by a number of languages either unre- This perspective has been recently applied to the lated or from different subgroups of the same family. study of Circum-Baltic languages (Koptjevskaja- The following linguistic areas have been identified so Tamm and Wälchli 2002) and Mediterranean lan- far: (i) The Balkans (Greek, Albanian, Serbo-Croatian, guages (Ramat and Stolz 2002). In both the Bulgarian, Macedonian, and Rumenian); (ii) South Circum-Baltic and the Mediterranean area, there are Asia (Indian subcontinent); (iii) Meso-America; (iv) several convergence/contact phenomena on the Amazonia; (v) Ethiopia; and (vi) Europe. microlevel, that interact with phenomena on the Each of the aforementioned linguistic areas displays macrolevel. This distinction between microlevel and a number of distinctive features. For instance, the macrolevel phenomena nicely accounts for the internal Balkan linguistic area, which is by and large the most heterogeneity of linguistic areas. The claim underlying famous linguistic area identified so far, is characterized such studies is that the notion of Sprachbund tends to by the following features: (a) formation of future and overemphasize the overall macrocontact, which is of perfect tense using an auxiliary; (b) replacement of course justified in certain areas (as the well-estab- infinitives by finite verb forms; (c) postponed articles; lished Balkan linguistic area). In other areas, however, and (d) merging of dative and genitive case forms. intensive microcontacts superimposed on one another Areal linguistics (i.e. the subfield of linguistics may create the impression of an overall macrocontact which aims at uncovering linguistic areas) has been among the languages, which has not necessarily being facing the problem of the number of features necessary there. The obvious consequence of this assumption is to distinguish between proper Sprachbünde and simple that areal phenomena are worth looking at in their own contact phenomena among a group of languages. right, even if we are not faced with linguistic areas. According to a widely held view, there should be genealogical diversity of the languages belonging to a References linguistic area, although it cannot be excluded that Bisang, Walter. 1998. Areal typology and grammaticalization: shared features may be due to language contact even in processes of grammaticalization based on nouns and verbs the case of genetically related languages. Moreover, the in East and mainland South East Asian languages. Studies in similarities that characterize a linguistic area should Language 20. not be trivial, i.e. they should not be the result of typo- Campbell, Lyle, Terrence Kaufman, and Thomas C. Smith-Stark. 1986. Meso-America as a linguistic area. Language 62. logical developments that take place, under the same Dahl, Östen. 2001. Principles of areal typology. Language circumstances, in all or most of the world’s languages. typology and language universals, ed. by Martin Language typology, being traditionally concerned with Haspelmath, Ekkehard König, Wulf Österreicher, and the classification of languages based on grammatical Wolfgang Raible. Berlin–New York: Mouton de Gruyter. features independent of genetic relationship, may pro- Ferguson, Charles A. 1976. The Ethiopian language area. Languages in Ethiopia, ed. by M.L. Bender, J.D. Bowen, vide areal linguistics with some evaluation criteria. and Charles A. Ferguson. London: Oxford University Press. Similarly, historical linguistics may help areal linguists Haspelmath, Martin. 1998. How young is standard average to exclude that some postulated areal features are due European. Language Sciences 20. to genetic inheritance. Koptjevskaja-Tamm, Maria, and Bernhard Wälchli. 2002. An In some recent developments of areal linguistics, the areal typological approach to the Circum-Baltic languages. The Circum Baltic languages. Their typology and contacts, notions of areality are considered to be of interest irre- ed. by Östen Dahl and Maria Koptjevskaja-Tamm. spective of whether the phenomena examined can be Amsterdam–Philadelphia: Benjamins. described in terms of linguistic areas in the traditional Masica, Colin P. 1976. Defining a linguistic area. South Asia. sense. This change in scientific perspective has led to a Chicago: University of Chicago Press. rather composite approach to linguistic diversity that Nichols, Johanna. 1992. Linguistic diversity in space and time. Chicago: University of Chicago Press. could be labeled as ‘areal typology’ (Östen Dahl 2001): Ramat, Paolo. 1998. Typological comparison and linguistic to the extent that typologically interesting features are areas: some introductory remarks. Language Sciences 20. not evenly spread geographically but tend to cluster in Ramat, Paolo, and Thomas Stolz (eds.) 2002. Mediterranean certain areas, areal typology can be considered as the languages. Papers from the MEDTYP Workshop, Tirrenia, study of the geographical distribution of linguistic fea- June 2000. Bochum: Universitätsverlag Dr. N. Brockmeyer. Sandfeld, K. 1930. Linguistique Balcanique. Problèmes et tures, rather than the characteristics of individual areas. Résultats. Paris: Klincksieck. In this vein, Johanna Nichols’ important monograph Trubetzkoj, Nikolaj S. 1930. Proposition 16. Actes du Premier (1992) has provided a complex model of linguistic Congrès International de Linguistes à la Haye, du 10–15 diversity, which involves two main areal patterns, Avril 1928. Leiden: A.W. Sijthoff. namely spread and residual zones, the former being ANDREA SANSÒ areas in which languages tend to spread and mix quick- See also Balkans; Historical Linguistics; Lan- ly and repeatedly and the latter being zones where lan- guage: Contact—Overview; Trubetzkoy, Nikolai guages tend to remain undisturbed over long periods. Sergeyevich; Typology 1030  SRANAN Sranan Sranan (or Sranan Tongo, lit. ‘language of Suriname’) Sranan is somewhat unusual among creoles in gen- is the name of the English-lexicon creole language that eral in that it is relatively well documented in its early has been used in Suriname (in the northeast of South stages of development. Many written documents— America) since the late seventeenth century. In the either in printed or manuscript form—are available past, it has also been referred to as Negro-English, from the early eighteenth century onward. This is Nengre, or, in a rather derogatory manner, Taki-Taki. largely due to the work of the Moravian Brethren, a It is spoken by some 350,000 people in Suriname, German missionary organization that was very active French Guyana, and the Netherlands, either as a first in Suriname and that produced a large number of reli- or as a second language. It is the native language of gious texts in Sranan as well as a number of invaluable most Surinamese people of African descent, while it descriptive works such as dictionaries and grammars. serves as an interethnic lingua franca between the Apart from this, there are several printed language other ethnic groups, which include Amerindians, primers and dictionaries that were authored by mem- Indians, Javanese, and Chinese. Although it does not bers of the colonial elite. The availability of these early have any official status—the only official language in documents has enabled linguists to carry out detailed Suriname is Dutch—it is being used more and more in diachronic research on Sranan, especially on the devel- formal contexts, such as education, the media, politics, opment of its phonology and syntax (Arends 2002). and public information. Apart from this, there is also a As a creole language, Sranan is the product of a flowering literature in Sranan, especially poetry. process of language contact involving a number of dif- The reason why an English-lexicon creole is spoken ferent languages from different language families. in a country that was a Dutch colony throughout most Historical research has shown that while the African of its existence is purely historical. Before it became a population consisted of many different ethnolinguistic Dutch possession in 1667, Suriname had been an groups, the majority among them belonged to one of English colony for 17 years, and it is generally the three larger groups, who spoke either Gbe, assumed that the basis of the Sranan lexicon stems Kikongo, or Akan (Arends 1995). This means that the from that period. However, although most of the major linguistic input into the creolization process, as English had left the colony by 1680, this did not put a far as the African side is concerned, came from these definitive stop to the presence of the English in languages. The predominant role of Gbe, Kikongo, Suriname. Apart from the fact that some of the ‘old’ and Akan is confirmed by the fact that the African ele- English planters returned to Suriname later, new ment of the Sranan lexicon is overwhelmingly derived English planters settled there as well. This means that from these three languages. This is further supported the window of opportunity for Sranan to acquire an by research into other domains such as phonology, lex- English-based lexicon remained open for a longer time ical semantics, and morphosyntax. than the 30-year period between 1651 and 1680. As far as the European population is concerned, it Apart from the English element, the Sranan lexicon is important to realize that although Suriname was a contains several other layers as well. First, due to the Dutch colony from 1667 onwards the Dutch were fact that many of the planters in the early period were never a majority among the European population. Portuguese-speaking Jews, a number of Portuguese- From the late seventeenth to the early nineteenth cen- derived words have been incorporated into the lan- tury, it was the Portuguese Jews who were the numer- guage. Second, many items from local domains, such as ically most important group among the European flora and fauna, have Amerindian names, mainly from population, which, apart from the Dutch, included Arawak and Carib. Third, a fair number of words have Germans, French, Scandinavians, and others. As a been adopted from some of the African languages spo- result, Dutch did not become a majority language ken by the slaves, especially Gbe (a Kwa language, spo- among Suriname’s Europeans until well into the nine- ken in Ghana, Togo, and Benin), Akan (another Kwa teenth century. This may explain why Sranan was language, spoken in Ghana), and Kikongo (a Bantu lan- widely used by the Europeans among themselves, guage, spoken in Congo-Brazaville and Congo- even though ‘officially’ they held it in low esteem. Kinshasa). Finally, ever since the beginning of Dutch Below, the major features of each of the linguistic rule, a large number of Dutch-derived words have been subsystems will be briefly discussed (largely based on borrowed, a process that continues to the present day. Bruyn 2002; see also Adamson and Smith 1995). It 1031 SRANAN will be followed by a few remarks on the ‘verbal arts’, Word-final nasals are velarized, something that is an important activity in traditionally oral languages reflected in the spelling, as in Sranan, which is such as Sranan. pronounced [Sranang]. Lexicon. About three quarters of the basic vocabulary Morphology. The four morphological processes items are derived from English, while most of the operative in Sranan are conversion, compounding, suf- remainder is from Dutch. As for nonbasic vocabulary, fixation, and reduplication. Conversion (also known as this consists mainly of words derived from Dutch, multifunctionality or zero derivation) refers to the der- although some items have been taken from other ivation of a word, e.g. a verb, from another word, e.g. sources, such as Portuguese, Amerindian languages a noun, without any overt change in form. For exam- (Arawak, Carib), and African languages. Some exam- ple, from the adjective ebi ‘heavy’, both a noun ebi ples are katibo ‘slave’ ( Portuguese), kruyara ‘dug ‘weight’ and a verb ebi ‘to weigh’ have been derived in out canoe’ ( Arawak), awara ‘palm species’ ( this way. Compounding is a rather common process in Carib), agama ‘lizard species’ ( Gbe), pinda ‘peanut’ Sranan, especially when both elements are nouns, as in ( Kikongo), gongosa ‘gossip’ ( Akan), and nyan man-pikin ‘son’ (lit. ‘man child’) and uman-piki ‘eat’ ( Wolof). Although most English function words ‘daughter’ (lit. ‘woman child’). One of the few cases were not retained in Sranan, many functional elements of inflection is the use of the noun man ‘man’ as an are still expressed by English-derived words. This is a agentive suffix as in siki-man ‘sick person’ (lit. ‘sick result of the fact that English-derived content words man’) and bere-man ‘pregnant woman’ (lit. ‘belly from the Sranan lexicon were recruited to fulfill that man’). Finally, reduplication, the creation of a new role. Such a process of grammaticalization results in a word by (partially) doubling an already existing word, content word becoming more grammatical, often being is quite common. It can be used to create new words reduced both in its semantic content as well as its pho- belonging to a category different from the base word, netic substance. For example, while the English article as appears from sisibi ‘broom’ ( sibi ‘to sweep’). ‘the’ was not adopted in Sranan, two definite articles a (singular) and den (plural) developed out of the demon- Syntax. Sranan is a strict S(ubject)-V(erb)-O(bject) strative dati ‘that’ ( English ‘that’) and the personal language with a strong tendency toward an isolating pronoun den ‘they’ ( English ‘them’)respectively. morphology. The latter appears from the fact that tense Similarly, while the English indefinite article ‘a’ was (T), mood (M), and aspect (A) are expressed through not retained in Sranan, a new indefinite article wan ‘a’ independent particles, which are preposed to the verb, developed, based on the numeral wan ‘one.’ As for the rather than through inflection. The TMA system being words that were retained from English, their meaning far too complex to be discussed in any detail, suffice it is not necessarily the same as that of its English ety- to say that the distinction between stative verbs mon. This appears, for example, from anu ( ‘hand’), (‘love’) and nonstative verbs (‘eat’) is of paramount which means both ‘hand’ and ‘arm,’ and futu ( importance for the functioning of the system. For ‘foot’), which means both ‘foot’ and ‘leg.’ Semantic example, while the bare form of a nonstative verb indi- shifts such as these often have their origin in one or cates past tense, a bare stative verb indicates present; more of the African languages spoken by the slaves, to indicate present tense, a nonstative verb has to be many of which have only one word for ‘handarm’ preceded by the particle e. Compare the following and for ‘footleg,’ respectively. African origins are examples: mi nyan fisi ‘I ate fish’; mi lobi fisi ‘I love also responsible for the existence of a word class fish’; and mi e nyan fisi ‘I’m eating fish’. While the known as ideophones, words whose only function is to distinction between stative and nonstative verbs also intensify or specify the meaning of another word with plays a role in the use of the particle ben, other factors, which they occur in a fixed combination. For example, such as discourse structure, come into play here as the ideophone fáán, used to intensify the meaning of well, rendering this area of Sranan syntax too complex the adjective weti ‘white,’ is probably from Gbe. An to be treated any further here. Like many other creoles, example is a weti so fáán ‘he is so very white’ (lit. he Sranan has two copula forms, one (de) for location, white so IDEOPHONE). possession, and existence, and the other (a) for nomi- nal predication (although de is sometimes used here as Phonology. In phonology, there is a clear tendency well). Adjectival predicates are treated on par with ver- toward an open syllable structure, which becomes bal predicates, i.e. they normally follow the subject clear from the tendency to add an extra vowel to without an overt copula being inserted in between, as English-derived words ending in a consonant, such as in yu futu bigi ‘your feet are big’ (lit. ‘your feet big’). udu ‘wood’ ( ‘wood’) and waka ‘walk’ ( ‘walk’), To express intensity or contrast, both verbal and adjec- or to insert a vowel into some consonant clusters. tival predicates may be clefted, with a copy of the 1032  SRI LANKA PORTUGUESE predicate left behind, as in na bigi yu futu bigi ‘Your ating the story with remarks, songs, or even entire ‘sub- feet are really big’ (lit. ‘is big your feet big’). Finally, stories’ of their own. The importance of songs, e.g. as a syntactic phenomenon known in many creoles is the an emotional outlet for the slaves, is already apparent serial verb construction, where one subject is connect- from early sources, where reference is made to a social ed with two or more main verbs, which together form activity known as pree ‘play’ in which dance played an one semantic unit, as in Rudy ben tyari den buku kon important role. A similar role was played by various na ini a oso ‘Rudy has brought the books into the kinds of drama, whose origin similarly lies in the plan- house’ (lit. ‘Rudy has carried the books come at in the tation period and which continue to be performed to the house’). In this sentence, the meaning of what is present day (for splendid collections of Sranan oral lit- expressed by the preposition ‘to’ in English is erature, see Herskovits and Herskovits (1936) and expressed by the verb kon ‘come,’ which forms a series Voorhoeve and Lichtveld (1975)). with the verb tyari ‘carry’. Sranan has a wide variety of different types of serial verb constructions, for the References expression of direction, location, instrumental, dative, Adamson, Lilian, and Norval Smith. 1995. Sranan. Pidgins and benefactive, causative, comparative, completion, and Creoles: an introduction, ed. by Jacques Arends, Pieter complementation. Since both predicate clefting and Muysken, and Norval Smith, 219–32. Amsterdam and serial verbs are common features of many West Philadelphia, PA: Benjamins. African languages, it seems justified to interpret the Arends, Jacques. 1995. Demographic factors in the formation occurrence of these constructions in Sranan as reten- of Sranan, The early stages of creolization, ed. by Jacques Arends, 233–85. Amsterdam and Philadelphia, PA: tions from the African languages spoken by the slaves. Benjamins. ———. 2002. Young languages, old texts: early documents in Verbal arts. The domain of language use known as the Suriname Creoles. In Carlin and Arends, 183–205. the ‘verbal arts’ includes such activities as story telling Bruyn, Adrienne. 2002. The structure of the Surinamese creoles. and the performance of song and drama. Probably the In Carlin and Arends. 153–82. Carlin, Eithne, and Jacques Arends (eds.) 2002. The languages best known genre is the so-called Anansi tori, a type of of Suriname. Leiden: KITLV Press. story named after the trickster-spider Anansi, but Herskovits, Melville, and Jean Herskovits. 1936. Suriname including other types of folk tales as well. Although the folk-lore. New York: Columbia University Press. canonical context for telling Anansi tori is at funeral Voorhoeve, Jan, and Ursy Lichtveld. (eds.) 1975. Creole drum: wakes, they may be told on other occasions as well. an anthology of Creole literature in Surinam. New Haven and London: Yale University Press. Both the content and the performative structure of these JACQUES ARENDS tales have their roots in West Africa. The basic pattern is the call-and-response structure known from many See also Aspect; Function Words; Grammaticaliza- African-American oral genres, with the story-teller tion; Mood; Pidgins and Creoles; Saramaccan; being interrupted by members of the audience punctu- Serial Verb Constructions Sri Lanka Portuguese On May 18, 1498, the Portuguese explorer Vasco da their word stock from Portuguese, but developed inno- Gama, having discovered a sea route around the Cape of vative grammars that combined structural features from Good Hope, arrived off the coast of Calicut in south- Portuguese and local languages with new elements. One western India. This event marked the beginning of the of the most important of these creoles is the one still first large-scale European presence in South Asia since spoken in Sri Lanka (formerly Ceylon). Sri Lanka the expedition of Alexander the Great. As a result of Portuguese had a longer reign as a lingua franca (lan- Portuguese contacts, creole languages with Portuguese- guage of intergroup communication) than its sister based lexicons developed in various coastal communi- creoles in India and possessed a more copious litera- ties on the subcontinent, such as Daman, Chaul, ture. In addition, it is rivaled for current vitality only by Nagapatnam, and Cochin. Collectively known as Indo- the northern dialects spoken in Korlai, near Bombay Portuguese, these languages all adapted the majority of (see Clements 1996) and Daman. Sri Lanka Portuguese 1033 SRI LANKA PORTUGUESE is also the most fully described variety of Indo- The Dutch in their turn were supplanted in 1898 by Portuguese, thanks to the work of Wesleyan missionar- the British, who, with their victory over Kandy in 1815, ies in the early 1800s and to the description published in completed their dominion over the entire island and held 1900 by a Goan jesuit priest, Sebastião Rodolfo it until Sri Lanka regained its independence in 1948. The Dalgado (see Further Reading below). Because of its British did not copy the Dutch in establishing local fam- longevity and voluminous literature, the language is ilies and thus never adopted the creole themselves, but important not only for the study of Indo-Portuguese but continued to use it as an expedient lingua franca until its also for investigation into the general processes functions could be taken over by English. This replace- involved in language contact. ment appears to have occurred around the middle of the Generally, only one language other than Portuguese nineteenth century. Certainly by 1875, Sri Lanka was spoken in the settings in which the various vari- Portuguese is reported by the Portuguese creolist, eties of Indo-Portuguese developed. This fact sets Adolpho Coelho, to be ‘almost exclusively the language Indo-Portuguese apart from many other creoles (par- of the descendants of the Portuguese and Dutch’. ticularly those labeled ‘plantation creoles’), which Following Dalgado’s description of Sri Lanka developed in multilingual settings. In Sri Lanka, two Portuguese, published in 1900, we heard little further major indigenous languages are spoken: Sinhala, an until it was ‘rediscovered’ in the 1960s by the late Dr. Indo-Aryan language, and Tamil, a language of the D.E. Hettiaratchi of the University of Sri Lanka, Dravidian family. The initial development of Sri Peradeniya, and editor of the Sinhala Encyclopedia. Lanka Portuguese took place among Sinhala speakers Currently, the creole is known to be spoken only by along the southwestern coast of the island. ‘Portuguese Burghers’ in the east-coast towns of Contacts between Portugal and Sri Lanka began in Trincomalee and Batticaloa in the predominantly the early sixteenth century, and these eventually led to Tamil-speaking area of the island. Due to the modern the launch of an expedition from Goa in 1517 to estab- civil strife in Sri Lanka, recent figures on the number lish a fortified trading post at Colombo. Over the fol- of speakers are not available, but clearly the language lowing 100 years, the Portuguese extended their is moribund. All Sri Lanka Portuguese speakers are control around the entire coast, leaving the independ- bilingual in Tamil, with some also speaking Sinhala ent Kingdom of Kandy in the central highlands. and/or English. In many families, Tamil has supplant- Throughout the area under Portuguese dominion, ed the creole as a home language. This is partly due to communities that used Sri Lanka Portuguese as their marriage with non-creole speakers and partly due to first language arose as a result of unions between the fact that men must often seek work elsewhere on Portuguese soldiers and local women. the island, where they are in a purely Tamil or Sinhala- Beginning with the union of the Spanish and speaking milieu. Portuguese crowns in 1580, events in Europe weak- The sounds of sixteenth-century Middle Portuguese ened Portugal’s ability to control its vast overseas pos- are generally found in Sri Lanka Portuguese, although sessions, which then began to fall prey to the rising their pronunciation has taken on a local flavor. For power of Holland. The Dutch began cultivating rela- example, Sri Lanka Portuguese /v/ is pronounced [υ], a tions with Kandy in 1602, and in 1632 the newly sound found in languages throughout the region, and enthroned Raja Sinha II asked for their assistance in that can be heard in the typical South Asian pronuncia- expelling the Portuguese. After a long period of con- tion of vine or wine. On the other hand, Sri Lanka flict, the last Portuguese strongholds capitulated in Portuguese sometimes preserves a Middle Portuguese 1658. From this point on, Sri Lanka Portuguese was pronunciation now lost in standard varieties of isolated from Standard Portuguese. Thus, unlike many Portuguese; thus, the first consonants of cegaa creoles, it was not later influenced by the standard lan- (‘arrive;’ Standard Portuguese chegar) and juustu (‘cor- guage from which it derived its word stock. rect;’ Standard Portuguese justo) are pronounced [tʃ] Interestingly, because Sri Lanka Portuguese was and [d], while in both Portugal and Brazil they have already established as a lingua franca for communicat- changed to [ʃ] and []. Some Middle Portuguese ing with local people, the Dutch continued to use it sounds have also disappeared; for example, x (pro- throughout their rule, and it even became the home nounced [ʃ]) has been replaced by /s/, as seen in basu language of the families that the Dutch started locally. (‘below;’ Standard Portuguese baixo). More important- Dutch and Sri Lanka Portuguese thus came to stand in ly, the entire set of nasalized vowels has disappeared, a relationship known as diglossia, with the former although an m usually appears as evidence of former being used as a ‘high’ (i.e. socially prestigious) lan- nasalization. This can be seen in examples such as guage in administration, the courts, and the Dutch paam (‘bread’; Standard Portuguese pão) and amiyaam Reformed Church, and the latter as a ‘low’ language (‘tomorrow’; Standard Portuguese amanhã). Finally, for informal face-to-face communication. Sri Lanka Portuguese has developed some new sounds. 1034  SRI LANKA PORTUGUESE For example, as in both Sinhala and Tamil, the creole Portuguese quer vir); falaa tiña (‘had told;’ Standard now has a contrast between long and short vowels, Portuguese. tinha falado); naa-pooy botaa or pa-botaa which can be seen in pairs such as: ooy (‘eye’; Standard naa-poy (‘can’t put;’ Standard Portuguese não pode Portuguese olho) vs. oy (‘today’; Standard Portuguese botar). The categories that these verbal structures rep- hoje), and triisti (‘sad’; Standard Portuguese triste) vs. resent are generally those of Tamil and Sinhala rather isti (‘this’; Standard Portuguese este). than Standard Portuguese; the completion category is In its grammar, even more than in its phonology, Sri one example, but many others arise. Moreover, Lanka Portuguese has a predominantly South Asian Standard Portuguese verbal categories that are not rather than European character. This can be seen in found in Sinhala and Tamil, such as the subjunctive, such features as word order, noun and verb categories, are also not found in the creole. and complex sentence structures involving relative Complex sentence structures are also generally clauses and the quotative construction (explained modeled on those of Sinhala and Tamil. Relative claus- below). Moreover, Standard Portuguese characteristics es, for example, always come before the noun they such as noun gender and subject–verb agreement are modify and do not use relative pronouns. An example absent from the creole. is shown below with Sinhala and Tamil equivalents: The order of the main elements in a Sri Lanka SLP botus diñeeru ja-daa pesaam Portuguese sentence is subject–object–verb; for exam- Sinhala mahattea salli diy-e miniha ple, eev eli-pa diñeeru jaa-daa (‘I him money gave’ or Tamil niinka calli-ya ku/u-tt-a aa4 ‘I gave him the money’; Standard Portuguese Dei o Gloss you money gave person dinheiro para/a ele or Lhe dei o dinheiro). Adjectives and other noun modifiers precede their nouns: nosa (‘The person to whom you gave the money’; Standard noov ravkiiña (‘our new violin’; Standard Portuguese Portuguese a pessoa a quem o Senhor dou o dinheiro). nossa rauquinha nova). Postpositions rather than Another construction typical of Dravidian lan- prepositions are found: %lm%%ra riiva (‘wardrobe on’ guages and shared by Sinhala is the so-called ‘quota- or ‘on top of the wardrobe’; Standard Portuguese tive construction’, which marks the object of an overt acima do armário). In all these word-order characteris- or implied verb of thinking, naming, speaking, etc. tics, Sri Lanka Portuguese follows the model of Sinhala The quoted material is followed by the quotative par- and Tamil rather than that of Standard Portuguese. ticle falaa or falaatu in Sri Lanka Portuguese, kiyl in The role that nouns play within a sentence structure Sinhala, ent u in Tamil. The quotative particle derives is indicated by case suffixes and postpositions, again, as in all three¯ languages from a verb meaning ‘say’. In in Sinhala and Tamil. The dialect spoken in Batticaloa the example below, both the verb ‘say’ and the quota- has four case suffixes; for example, ɔɔ mi ‘man’ has the tive particle appear: following forms: ɔɔ mi-pa (object of verb, recipient, ‘to SLP eli jaa-falaa eev ja-kaa- falaatu the man’), ɔɔ mi-su (possessor, ‘man’s’), ɔɔ mi-ntu falaa (location, ‘in the man’), and ɔɔ mi-ntaa (person spoken Sinhala eyaa kiwwe mam kiww-e kiyl to, person holding something, ‘with the man’). Tamil avan connaan naan coll-i-t-t-an entu The verbal system is interesting because, unlike ¯¯ ¯ gloss he said I have told QUOTE Sinhala and Tamil, Sri Lanka Portuguese uses prefixes to indicate tense; thus, for the verb kuziñaa ‘cook’, the (‘He said “I have told [him]”’ or ‘He said he had told following forms are found: ta-kuziñaa (present), jaa- [you]’; Standard Portuguese Ele disse que ‘eu [lhe] kuziñaa (past), and lo-kuziñaa (future). These struc- falei’. or Ele disse que [lhe] tinha falado.) tures are also quite unlike Standard Portuguese: The vast majority of Sri Lanka Portuguese words compare cosinha ‘(he or she) cooks’, cosinhou ‘(he or are Portuguese in origin, but many other items are bor- she) cooked’, cosinhará ‘(he or she) will cook’. Other rowed from Dutch, English, Sinhala, and Tamil, such prefixes indicate completion, obligation, negation, and as gɔ rgal ‘throat’ (Du. gorgel), d ɔ k t a ‘doctor’ (Eng.), conditionality, as in mes-kuziñaa ‘must cook’, and suura ‘toddy, a drink made from¯ fermented ¯¯ palm sap’ naa-kuziñaa ‘won’t cook’. These prefixes were devel- (Si. sur), neeli ‘paddy, unhusked rice’ (Ta. nellu). oped from a variety of Standard Portuguese sources: Occasionally, Sri Lanka Portuguese preserves a auxiliary verbs (ta- from está ‘be’), adverbs (jaa from Middle Portuguese word no longer current in Standard ja ‘already’, lo- from logo ‘soon’), adjectives (mes- Portuguese, such as izmaleeru (‘beggar’; Standard from mister ‘necessary’), etc. Some auxiliary verbs Portuguese esmoleiro, archaic). appear before the main verb, as in Standard Literary Sri Lanka Portuguese is largely a nine- Portuguese; others appear after the verb, as in Sinhala teenth-century phenomenon. Kenneth David Jackson’s and Tamil; some can appear in both positions. book Sing without shame demonstrates that an oral Examples are: kera vii (‘intends to come; Standard tradition existed, which, because of its ties with the 1035 SRI LANKA PORTUGUESE traditions of other Portuguese-speaking communities, Coelho, Adolpho F. 1967. Os dialectos românicos ou neolatinos probably originated during Portuguese rule on the na África, Ásia e América. Bolletim da Sociedade de Geografia de Lisboa, 2:3 (1880), 3:8 (1882), 6:12 (1886), island. The earliest extant printed records of the creole reprinted in Estudos linguisticos crioulos, ed. by Jorge are the work of Wesleyan missionaries, who made Morais-Barbosa. Lisbon: Academia Internacional de extensive use of Indo-Portuguese in both India and Sri Cultura Portuguesa. Lanka (see Fox 1819.) Either through ignorance or Dalgado, Sebastião Rodolfo. 1900. Dialecto indo-português de because of a desire to dignify the low-prestige creole, Ceylão, Lisbon: Imprensa Nacional; reissued with introduc- tion by Ian R. Smith, Lisbon: Comissão Nacional para as the missionaries dressed Sri Lanka Portuguese in the Comemorações dos Descobrimentos Portuguêses, 1998. trappings of the more prestigious standard language to Fox, William Buckley. 1819. A dictionary in the Ceylon- which it is related, producing a hybrid variety combin- Portuguese, Singhalese and English languages ... to which is ing Standard Portuguese spelling and generally prefixed a compendium of the Ceylon-Portuguese language. Standard Portuguese and English grammar with some Colombo: Wesleyan Mission Press. Jackson, Kenneth David. 1990. Sing without shame: oral tradi- aspects of creole word structure, particularly in the use tions in Indo-Portuguese verses with transcription and analy- of verb prefixes. sis of a nineteenth-century manuscript of Ceylon Portuguese. The fact that we have reason to be skeptical of nine- Amsterdam: John Benjamins and Macau: Cultural Institute teenth-century Sri Lanka Portuguese texts should of Macau. make us equally skeptical of nineteenth-century creole Smith, Ian R. 1978. Sri Lanka Portuguese phonology. Trivandrum: Dravidian Linguistics Association. texts in general, and of those purporting to represent ———. 1979. Convergence in South Asia: a creole example. other varieties of Indo-Portuguese in particular. Lingua 48. ———. 1979. Substrata vs. universals in the formation of Sri References Lanka Portuguese. Papers in pidgin and creole linguistics, No. 2. Canberra: Pacific Linguistics. Clements, J. Clancy. 1966. The genesis of a language: the for- IAN R. SMITH mation and development of Korlai Portuguese. Amsterdam, Philadelphia: John Benjamins. See also Pidgins and Creoles; Tamil Standard Language A standard language presupposes diglossia. Diglossia Lingua franca is the common language of social or is a situation where dialectal varieties of a language in commercial communication among polyglot societies. a polity function differently. The dialect adopted as the Johor Malay was the lingua franca in the Malayan official medium of communication in formal settings world. During the colonial years, Johor Malay was the becomes the standard language. Consequently, a stan- language of publication when the press came to Malaya dard language rises from a contention of several (preindependent Malaysia) from Singapore via Johor, a dialects. A standard language is the medium of southern Malay state, to Kuala Lumpur (see Asmah). In instruction in national education, official correspon- the contention of standard language, the victory nor- dence in written communication, and mass communi- mally goes to the dialect with historical advantage. The cation in mass media like newspapers, television, present modern Standard Malay, however, defies the films, books, and magazines. historical norm. Johor Malay, originating from the Standard Malay in Malaysia went through interlan- southern part of the Malay Peninsula, was the Standard guage and intralanguage contentions. At the brink of Malay until Kedah Malay replaced it. Since 1988, independence in 1957, Malay became the national lan- Kedah Malay was decreed as Standard Malay because guage of Malaysia and English would be phased out the writing system of this northern Malay dialect is gradually in ten years. To equip Malay with the role of phonemic, i.e. its spelling directly corresponds with the a modern language, a semigovernment body called pronunciation. The [a] sound in an open syllable in the Dewan Bahasa dan Pustaka was set up in 1956 to word-final position is pronounced as /a/ as in /baca/ standardize the Malay spelling, and to coin Malay sci- (read) and /cerita/ (story) in this dialect. In comparison, entific terminology. At this interlanguage stage, Malay /baca/ and /cerita/ are uttered as /bace/ and /cerite/ in rose above English, Mandarin, and Tamil, among oth- Johor Malay. Due to these user- and education-friend- ers, as the standard language in all formal settings. ly reasons, students and teachers were instructed to 1036  STANDARD LANGUAGE adopt the standard [a] pronunciation in oral examina- tic future. There is little research on the relation tions, school lessons, and public speeches. In short, between Singapore and China Mandarin. Singapore current Standard Malay in Malaysia is a result of both exemplifies another complex inter- and intralanguage interlanguage and intralanguage competitions. situation in the linguistic life span of standard language. At the end of 1993, however, the Prime Minister of Based on the language of mass media in the United Malaysia announced that English would be the lan- States, especially the English used in Hollywood guage of instruction for science and technology at the movies and American sitcoms on television, Standard varsity level. There was an outcry over the threat of American English is the standard language that rises English on Standard Malay, which has been the medi- above African American English, Cajun English, Latino um of instruction from primary to tertiary education. English, and Chicano English. American English is Internationalism was used to describe the new era of different from British English as in these contrasts: change dawning on young polities like Malaysia, which windshield–windscreen, hood–bonnet, truck–lorry, once preoccupied itself with nationalism. Although expressway–motorway, elevator–lift, mail–post, dia- Malay remains as Malaysia’s standard language, its lin- per–nappy, lawyer–barrister, mean– nasty, and guistic path transformed from the interlanguage stage to dumb–stupid. Mistaken as American English, the word the intralanguage phase and another modern interlan- talented along with other American English words like guage attrition in the face of prosperity. In fact, as early dutiable, presidential, lengthy, finalize, normalcy, and as the 1980s, English-educated Malay elites were irregardless were once abhorred as nonstandard speaking Malayish, a mixture of Malay and English. English. The term Americanism, first used by John To give another language situation, four languages, Witherspoon in 1781, refers to the use and construction namely English, Malay, Mandarin, and Tamil, are des- of English vocabulary and sentences in the United ignated as the official languages in Singapore following States that are different from those in Great Britain. a 1956 official report. Through the 1991 Improving pri- Americanism implies a new English dialect. When do mary school education report, English has eclipsed the two dialects become two different languages? By other three official languages as the medium of instruc- the rule of thumb, a situation of two languages arises tion in Singapore. While the three official languages are when two dialects become mutually unintelligible. taught as mother tongues in school, English dictates the Languages like Urdu and Hindi spoken in Pakistan academic future of the students because failing to mas- and India, however, are mutually intelligible; so are ter the English language is a disadvantage in an educa- Danish, Norwegian, and Swedish spoken by Danes, tion system that offers English as the first language. Norwegians, and Swedes, respectively. Interestingly, there are two varieties of English in Singapore, namely Singapore English or Singlish and References Standard Singapore English. Regarded as closely con- Asmah, Hj. Omar. 1992. The linguistic scenery in Malaysia. nected to Chinese in terms of ethnicity and language, Kuala Lumpur: Dewan Bahasa dan Pustaka. Singapore English lacks complicated grammatical rules Baugh, Albert C., and Thomas Cable. 1951, 2001. A history of and displays a fair deal of loanwords from Hokkien, a the English language. London: Routledge. Chinese dialect, and Malay. Typical phrases of Collins, James T. 1996. Devaluasi Bahasa dan Pemasaran Bahasa: Perkembangan di Negara-negara yang Berbahasa Melayu Singapore English are, ‘Don’t play-play’ (Do not fool (Language devaluation and language marketing: changes in around), ‘Catch no ball’ (Do not understand), and the Malay-speaking countries). Jurnal Dewan Bahasa 40. suffixation of the particle -lah as in ‘go-lah, eat-lah, Ferguson, Charles A. 1959. Diglossia. Word 15. please-lah’. In addition to providing emphasis and a Fromkin, Victoria, and Robert Rodman. 1974, 1998. An intro- sense of persuasion to the three words, the particle-lah duction to language. Fort Worth, TX: Harcourt Brace. Gopinathan, S. 1994. Language policy changes 1979–1992: pol- can incur a negating sense of disagreement to please. itics and pedagogy. Language, society and education in Despite the vibrant pragmatics, i.e. situational meaning Singapore: issues and trends, ed. by S. Gopinathan, Anne of words when in use, this variety of local English is Pakir, Ho Wah Kam, and Vanithamani Saravanan. Singapore: considered better than not knowing English at all and is Times Academic Press. also thought to be associated with the lower classes and PuruShotam, Nirmala Srirekam. 2000. Negotiating multicultural- ism: disciplining difference in Singapore. Berlin: de Gruyter. low-paid occupations. Students are required to adopt Sew, Jyh Wee. 1996. Hakisan Bahasa di Malaysia (Language Standard Singapore English as the medium of commu- erosion in Malaysia). Jurnal Dewan Bahasa 40. nication and examination. Lately, as China becomes the Sweeney, Amin. 1987. A full hearing: orality and literacy in the new world economy that offers plenty of business Malay world. Berkeley: University of California Press. opportunities, the go-north investment strategy gathers Thomason, Sarah. 2001. Language situation. Edinburgh: University of Edinburgh Press. more momentum among Singapore entrepreneurs. JYH WEE SEW Mandarin in Singapore, although has not made a sig- nificant comeback, seems to have a promising linguis- See also: Diglossia; Language Planning 1037 STORY GRAMMAR Story Grammar The rediscovery of early structuralist work—for exam- the outcome. A canonical episode includes these six ple, of Propp’s research on the structure of Russian constituents, where G, A, and O are obligatory in a folk tales—and the development of transformational- well-formed story. generative grammar led to the emergence of story Mandler used the following dog fable to illustrate grammars in the 1970s. Linguists such as Morton these story constituents: Prince, George Lakoff, and Teun van Dijk and cogni- tive psychologists such as Jean Mandler and Nancy ● (S) It happened that a dog had got a piece of Stein contributed to this effort. meat and was carrying it home in his mouth. The story grammar approach holds that (1) simple Now on his way home he had to cross a plank stories exhibit a typical structure and (2) there is gen- lying across a stream. eral agreement among listeners whether a given text ● (B) As he crossed he looked down and saw his represents a story. For example, the following set of own shadow reflected in the water beneath. phrases is considered a story, however trivial: ‘A man ● (R) Thinking it was another dog with another was very happy, then he married a vain and domineer- piece of meat, ing woman, then, as a result, he was very unhappy,’ but ● (G) he made up his mind to have that also. the following phrase is not: ‘Electrons are constituents ● (A) So he made a snap at the shadow, of atoms.’ ● (O) but as he opened his mouth the piece of meat Among psycholinguists and cognitive psycholo- fell out, dropped into the water, gists, story grammar was viewed as an instance of a ● (E) and was never seen again. schema, an idealized model of the constituents of a The rules reflect the property of a story that the sen- story, and the relationship among the constituents. The tences do not simply represent a sequence of unrelat- story schema has been expressed in terms of rewrite ed events. Rather, the events are causally linked as rules analogous to those of transformational grammar. specified by the rules. A story may include multiple Representative of other story grammars, the Mandler episodes that may be sequenced, recursively nested, or grammar includes the following rules: exhibit some other relation. For example, an episode ● Story → Setting and Episode(s) may be interrupted because of an obstacle that the pro- ● Episode → Beginning cause Complex Reaction tagonist must overcome. cause Goal Path cause Ending Story grammars have been criticized for various ● Complex Reaction → Reaction cause Goal reasons. Computational linguists in the tradition of ● Goal Path → Attempt cause Outcome Roger Schank and Robert Abelson took the story grammar approach to task for abstracting the structure This set of rules, as well as additional rules, gener- of stories from their content. Similar criticisms have ates a hierarchy, whose terminal elements are the sur- also been raised vis-à-vis the syntactic approach to face sentences of a story. sentences. In each case, the principal criticism is that The setting (S) describes the background that grammatical categories cannot be defined independent enables the events in the story. Usually, the principal of the semantic relations of the constituents to one characters of the story are introduced in the setting. An another. As a result, the grammars provide no source episode gives a story its narrative component. An for the inferences that readers routinely make. In addi- episode is defined from the perspective of a protago- tion, the story schema, like other schemas, is either too nist who is faced with a problem and tries to solve it. general or too simple to do justice to the wide variety The problem is triggered by events described in the of narratives. Other critics argued that story grammars beginning (B) of the episode. Then, the protagonist are technically not grammars at all because they do not exhibits a reaction (R) to the problem, and he or she specify terminal elements in the manner of transfor- sets a goal (G) and makes an attempt (A) to achieve mational-generative grammar. the goal. The result of the attempt is narrated in the In defense of story grammars, it should be noted outcome (O). The episode concludes with an ending that they are not mutually exclusive from other knowl- (E), which may report the protagonists’ reaction to the edge sources at different levels of complexity, whether outcome, state a general moral, or simply emphasize they are scripts, schemas, etc. And it is important to 1038  STRUCTURAL INTERFERENCE note that the story grammar approach has produced stories. Reading times adjusted for sentence an impressive body of empirical research validating length and other factors were longer at the begin- critical assumptions about the narrative structures ning of episodes than for the remaining con- specified by story grammar. Consider evidence from stituents. This pattern has been interpreted in three different paradigms: structural judgments, recall terms of readers’ effort at calling up the episode patterns, and reading times. schema into memory. ● Judgments of stories: Researchers had people The story grammar approach continues to survive read stories of diverse contents and divide the pas- in successor approaches, including causal chain theo- sages into parts. The structures produced by par- ry and other theories of narrative, as described in ticipants correlated well with the a priori structure Goldman et al.’s (1999) volume cited below. of stories provided by story grammar theory. ● Recall patterns: Research revealed that people remember stories better if the stories are well References formed. For other stories, people tend to gener- Goldman, Susan R., Arthur C. Graesser, and van den Broek ate ‘improved’ structures, much in the manner of Paul (eds.) 1999. Narrative comprehension, casuality, and subjects who ‘corrected’ a passage with an unfa- coherence: essays in honor of Tom Trabasso. Mahwah, NJ: Lawrence Erlbaum Associates. miliar sentence structure to fit with their expec- Just, Marcel, and Patricia Carpenter. 1987. Psychology of read- tations. Further empirical studies have shown ing and comprehension. Needham, MA: Allyn & Bacon. that people’s recall protocols were predicted by Mandler, Jean. 1987. Psychological reality of story structure. the importance of constituents in the story hier- Discourse Processes 10. 1–29. archy, with beginning, attempt, and outcome Prince, Morton. 1973. A grammar of stories. The Hague: Mouton. being best recalled. Propp, Vladimir. 1968. The morphology of the folktale. ● Reading times of story sentences in self-paced Bloomington: Folklore Society and Indiana University. reading experiments were found to be sensitive Van Dijk, Teun A. 1977. Text and context. London: Longman. to the episodic structure of simple two-episode KARL HABERLANDT Structural Interference As the name signifies, structural interference denotes 1.4. Syntactic: I am reading this book since morn- the intrusion of various linguistic structures of one lan- ing (transfer of L1, i.e. Hindi syntactic fea- guage in the use of another. These intrusions can be at ture—the use of present continuous tense for the level of phonology, morphology, syntax, discourse, present perfect continuous—into L2, i.e. pragmatics, and semantics. Some examples of struc- English) tural interference in terms of transfer are given below: 1.5. Semantic: In India, use of ‘auntie’ for all eld- erly women 1. Linguistic Transfer 1.6. Discoursal: I respectfully submit the following 1.1. Morphological: Doctorji, Masterbabu, [trans- few lines for your kind consideration (transfer fer of honorific like ‘ji’, ‘babu’, from Indian of politeness rules). languages (L1) into English (L2)] 1.2. Lexical: /computer eta kinib lagib / 2. Cultural Transfer Computer one buy should first person 2.1. Lexical: What is your good name? (transfer of I should buy a computer the Indian expression ‘shubh-naam’) [transfer of L2 lexical item, ‘computer’ into 2.2. Discoursal: My nose has been cut; I can no L1, in this case Assamese) longer show my face to anyone. (transfer of a 1.3. Word order: concerned authorities (transfer of Hindi idiomatic expression from L1 to L2). L1 i.e. Hindi word order, i.e. adjective  noun into L2, i.e. English) 3. Psycholinguistic Transfer 1039 STRUCTURAL INTERFERENCE A psycholinguistic transfer is one that is determined the contrastive analysts (linguists comparing the by the learner’s perception of the relationship structures of two/more languages) hypothesized that between L1 and L2 and certain characteristics of L1 language learning is habit formation and an old habit rules and items. Thus, when a German speaker is affects the formation of a new habit. Therefore, they termed impolite by an English speaker, she or he per- claimed that by carrying out an explicit analysis of ceives the two languages to be similar at the level of similar and dissimilar features of different languages, syntactic, discoursal, and pragmatic features. For we could arrive at a foolproof theory of structural example, when a German speaker says, ‘You should interference. When this theory was found to have close the window’, in a situation where a native loopholes, other theories emerged, stressing the speaker of English will prefer using ‘Can/could you processes of language acquisition. For example, the close the window?’, she or he is not being impolite creative constructionists assume that L1 (first lan- but is transferring the politeness strategy from her or guage/mother tongue) and L2 (second/foreign lan- his L1 to L2 (cf. House and Kasper 1981). Again, guage) acquisition proceed in a similar manner. when a Hindi speaker says, ‘You are lucky, isn’t it?’, Hence, L2 acquisition is largely unaffected by L1 she or he uses her or his knowledge of the Hindi ques- transfer, which in turn means structural interference is tion tag hein na and transfers the rule of the applica- not related to language acquisition. The interlanguage tion of hein na to English. hypothesis proposed by Selinker (1972), on the other Structural interference is closely context bound. hand, recognizes the major influence of the native Increasing research evidence shows that learners can language in the learning and use of a second lan- produce a significantly more fluent, grammatical, and guage, but it also mentions some four other processes transfer-free interlanguage (i.e. the separate linguistic like overgeneralization of target language rules, system evinced when adult L2 learners attempt to strategies of communication, strategies of second lan- express meaning in a language that they are in the guage learning, and transfer of training. Interlanguage process of learning) in some social contexts rather differs both from the native (NL/L1) and the target than in others. language (TL/L2), with fossilization (a stage where Structural interference can also be strategic: for L2/TL learning stops due to pressure of communica- example, Ervin-Tripp (1969) pointed out that L2 tion/use) being the most significant feature. This the- learners might protect themselves from the conse- ory of adult L2 learning is now extended to child quences of sociolinguistic errors by maintaining non- language acquisition as well. native features in their speech, i.e. they might Research on the various aspects of structural maintain structural interference to protect themselves interference has contributed significantly toward from being assessed on the basis of native-speaker effective learning and teaching of second lan- norms and expectations. Similarly, there can be delib- guages, particularly, teaching of English as a for- erate structural interference to exhibit solidarity, or to eign language. serve business purposes. Researchers have noted that salespersons use different dialects and intonations depending upon their judgment of the customers’ eco- References nomic, linguistic, and ethnic background. In social Cohen, A. 1990. Language learning. New York: Newbury research, some specialists suggest the use of the sub- House. jects’ dialect by the researchers to make the popula- Ervin-Tripp, S. 1969. Sociolinguistics. Advances in experimen- tal social psychology, Vol. 4, ed. by L. Berkowitz. New tion sample under observation feel at home. York: Academic Press. Moreover, there are different transferability criteria. Faerch, C., and G. Kasper (eds.) 1983. Strategies in interlan- The learner’s assessment of the interlocutor’s recep- guage communication. London: Longman. tive knowledge of different language codes as well as House, J., and G. Kasper. 1981. Politeness markers in English processing constraints such as the degree of automa- and German. Conversational routine: explorations in stan- dardized communication situations and prepatterned speech, tization are relevant factors in transfer as well as in ed. by F. Coulmas. The Hague: Mouton Publishers. production. James, C. 1980. Contrastive analysis. Harlow: Longman. There are various theories about what, how much, Krashen, S. 1981. Second language acquisition and second lan- why, and how a structural interference occurs. The guage learning. Oxford: Pergamon Press. notion of structural interference was primarily Parasher, S.V. 1999. Communicative styles in Indian English. Language matters, ed. by K.V. Tirumalesh. New Delhi: advanced by linguists interested in teaching English Allied Publishers. to learners of other languages during the 1950s. ———. 1977. Current trends in error analysis. JSL. Under the influence of the structural approach in lin- Selinker, L. 1972. Interlanguage. IRAL 10. 219–231. guistics and the behaviouristic theory in psychology, SHREE VALLABH PARASHER 1040  STRUCTURALISM Structuralism Structuralism is a mode of inquiry that consists in It is important to point out that the term ‘struc- interpreting the phenomena it looks at as made up of turalism’ came to acquire markedly different charac- relations among the various entities rather than as teristics in the United States, where it is used to refer those entities per se. The particular units are thus to the set of methodological principles brought to bear defined solely by virtue of the network of relations on linguistic analysis by the followers of Leonard into which they enter. They are, in other words, Bloomfield, whose 1933 book Language has been defined in negative terms rather than in terms of posi- referred to as the Bible of American structuralism. tive contents, so that any change in any one of the rela- Among the distinguishing traits of Bloomfieldian tionships will automatically affect the entire set of structuralism are its stubborn aversion to questions of relationships within the given structure. Another way meaning and its close ties with behaviorist psycholo- of putting this is to say that all structures constitute gy. Bloomfield was also keen on making linguistics a fully integrated systems, in which the elements are genuinely scientific discipline, which, under the terms fully dependent on one another and are sensitive to the of the philosophy of science then in vogue, meant rig- most minute of alterations taking place in any part of orously restricting evidence to empirically available any given system. Because each structure is fully and data. exhaustively defined by the relations among the ele- But, apart from these crucial differences, ments, it makes little sense to speak of universal struc- Bloomfield’s structuralism was at one with its tures; each structure is, as it were, a law unto itself. Yet European counterpart in insisting that language be another feature of structures in general is that they pro- viewed as a self-contained whole and that the entities vide us with a snapshot of the phenomena rather than be identified relationally rather than in terms of any the evolutionary stages through which those phenom- positive content. The key relations were those of con- ena pass; they are static and synchronic par excellence. trast and equivalence. The sound system, or phonolo- Furthermore, many structuralists have been at pains to gy, of a given language, for instance, was seen as point out that the structures they describe are posited being made up of units called ‘phonemes’, whose as such and not inherent in or latent to the phenomena existence was predicated on the identification in the themselves. language in question of a set of contrasts with other It is important to stress that structuralism did not phonemic units, each of which was to be likewise initially emerge as a school of thought or a philo- posited on the strength of the same principle. The lit- sophical tendency. Yet, toward the end of the nine- mus test of a contrast was a ‘minimal pair’, which is teenth century and in the early decades of the a pair of words identical in every respect except for twentieth century, structuralism had established itself the particular contrast as in cat and mat, in which the as a major force to reckon with, reaching its pinnacle only difference is marked by the contrast between the of glory by the 1930s. Structuralism swept across phonemes /k/ and /m/, or in the pair mat and mate, in almost all fields of inquiry, making significant contri- which the only difference consists in the change of butions to the humanities and the social sciences. vowels. Where two phonetically distinct items did not Among the most important names associated with the enter into a relationship of contrast but instead mani- movement are Ferdinand de Saussure, Claude Lévi- fested what is technically known as ‘complementary Strauss, Roland Barthes, and Louis Althusser. distribution’ (that is, where one occurs, the other can Ferdinand de Saussure, hailed as the Father of never occur), the two variants were to be regarded as Modern Linguistics and author of Cours de systematic variants of the same phoneme called ‘allo- Linguistique Générale (1916; A course in general lin- phones’. The same procedures were then applied to guistics), which was published posthumously, inaugu- other levels of analysis, notably word structure, where rated the structuralist revolution in linguistics. (It has, the analysis yielded analogous entities such as ‘mor- however, been argued that some of the key principles phemes’ and ‘allomorphs’. At the sentence level, the of structuralism may actually date back to thinkers structuralist method yielded what is referred to as such as Leibniz, Marx and Engels, Humboldt, Herder, ‘immediate constituent analysis’. Thus, a sentence and so on.) such as The dog barked all night was analyzed as con- 1041 STRUCTURALISM taining the immediate constituents ‘the dog’ and It should not be concluded from the foregoing ‘barked all night’. The resulting constituents were fur- remarks that structuralism in its classic sense is by ther analyzed into the immediate constituents ‘the’ now a spent force. Quite on the contrary, Saussure’s and ‘dog’ and ‘barked’ and ‘all night’, respectively, teachings continue to inspire generations of scholars and the constituent ‘all night’ further on into ‘all’ and and is still a point of reference for so-called post- ‘night’, with the process ending once the ultimate structuralism, whose advocates, far from rejecting constituents had all been identified. Finally, as what tout court the legacy of Saussure’s teachings, use would have come as an unpleasant surprise to them as a springboard for further advancing his Bloomfield himself had he lived long enough to wit- insights, albeit to logical consequences unimagined ness the development, the techniques of structural by the Swiss savant. As for Bloomfield’s legacy, it too analysis were carried over to the analysis of mean- arose from the ashes and gained a further lease on life ings, paving the way for a structurally oriented theory in the form of what is referred to as neo- of semantics. Bloomfieldeanism, notably in the work of Bernard Unlike its European counterpart, Bloomfieldean Bloch (1907–1965). structuralism invested a considerable amount of effort in identifying what came to be known as ‘discovery References procedures’, that is, a set of methodological principles Abraham, S., and F. Kiefer. 1966. A theory of structural seman- with guaranteed results that would automatically and tics. The Hague: Mouton. infallibly generate the right analysis from a given set Bartsch, R., and T. Vennemann (ed.) 1973. Linguistics and of empirical data. These discovery procedures came neighboring disciplines. Amsterdam: North-Holland. under heavy attack from Noam Chomsky and the fol- Bloomfield, Leonard. 1933. Language. New York: Holt, lowers of the new paradigm of transformational-gen- Rinehart & Winston. Cassirer, E. 1945. Structuralism in modern linguistics. Word 1. erative linguistics that he inaugurated in the 1950s. Chafe, Wallace L. 1971. Meaning and structure of language. Chomsky dismissed the whole idea of discovery pro- Chicago: University of Chicago Press. cedures and claimed that a linguistic theory should Chomsky, Noam A. 1957. Syntactic structures. The Hague: instead aim at attaining explanatory, rather than mere- Mouton. ly observational or descriptive, adequacy. Culler, Jonathan. 1975. Structuralist poetics: structuralism, lin- guistics, and the study of poetics. London: Routledge & Despite the major differences between Kegan Paul. Bloomfieldean linguistics and generative grammar, it Derrida, Jacques. Of grammatology, English translation by is nevertheless true to say that Chomsky’s approach G.C. Spivak. Baltimore, MD: Johns Hopkins University is structuralist in an extended sense of this Press. term—although it is also the case that, thanks to the Dinneen, Francis. P. 1967. An introduction to general linguis- tics. New York: Holt, Rinehart & Winston, Inc. tireless criticism of the early Bloomfieldean practices Francis, W. Nelson. The structure of American English. New by the early transformational-generative grammarians, York: Harcourt, Brace & World. the term ‘structuralism’ itself fell into disrepute and Hall Jr., Robert A. (ed.) 1987. Leonard Bloomfield: essays on acquired pejorative connotations that survive now. his life and work. Amsterdam and Philadelphia: John Chomsky retained the basic idea of language being Benjamins. Hockett, Charles. F. A course in modern linguistics. New York: structured but opted for an atomistic approach to the Macmillan. notion of structure instead of the holistic vision that Hymes, D., and Fought, J. American structuralism. The Hague: underwrote the European (mainly French) approach. Mouton. The elementary building blocks of the structure were Joseph, J.E. Bloomfield’s Saussureanism. Cahiers Ferdinand de no longer negatively defined but were invested with Saussure, 43. Lévi-Strauss, Claude. 1963. Structural anthropology. New positive attributes. Generative phonology adopted as York: Basic Books. its basic building blocks ‘distinctive features’ that Saussure, Fedinand de. 1916. Cours de Linguistique Générale. were universal and not language specific. The so- Paris: Payot, English translation by Wade Baskin, A course called phonemes were from now on seen as merely in general linguistics. New York: Philosophical Library, contingent and language-specific combinations of 1959. Wells, R. 1957. De Saussure’s system of linguistics. Readings these features. This important theoretical intervention in linguistics, ed. by M. Joos. Washington, DC: American permitted Chomsky and his followers to posit a uni- Council of Learned Societies. versal base for all languages, shifting the focus of KANAVILLIL RAJAGOPALAN attention to language as an innate species-specific attribute of man and redefining linguistics itself as a See also Bloomfield, Leonard; Chomsky, Noam; branch of cognitive psychology. Lévi-Strauss, Claude; Saussure, Ferdinand de 1042  STYLISTICS Stylistics Stylistics is the description and analysis of the vari- below a time). What actually constitutes the ‘norm’ is ability of linguistic forms in actual language use. The not always explicit in literary stylistics, since this concepts of ‘style’ and ‘stylistic variation’ in language would presuppose the analysis of a large collection of rest on the general assumption that within the lan- nonliterary texts. However, in the case of authorship guage system, the same content can be encoded in identification, statistical approaches were pursued at a more than one linguistic form. Operating at all lin- relatively early stage. For example, by counting specif- guistic levels (e.g. lexicology, syntax, text linguistics, ic lexical features in the political letters written by an and intonation), stylisticians analyze both the style of anonymous Junius in the 1770s and comparing them specific texts and stylistic variation across texts. These with a large collection of texts from the same period, texts can be literary or nonliterary in nature. Generally and with samples taken from other possible contempo- speaking, style may be regarded as a choice of lin- rary authors, the Swedish linguist Ellegård could iden- guistic means; as deviation from a norm; as recurrence tify, in the 1960s, the most likely author of those letters. of linguistic forms; and as comparison. The concept of style as recurrence of linguistic Considering style as choice, there are a multitude of forms is closely related to a probabilistic and statistical stylistic factors that lead the language user to prefer understanding of style, which implicitly underlies the certain linguistic forms to others. These factors can be deviation-from-a-norm perspective. It had already been grouped into two categories: user-bound factors and suggested in the 1960s that by focusing on actual lan- factors referring to the situation where the language is guage use, stylisticians cannot help describing only being used. User-bound factors include, among others, characteristic tendencies that are based on implicit the speaker’s or writer’s age; gender; idiosyncratic norms and undefined statistical experience in, say, preferences; and regional and social background. given situations and genres. In the last resort, stylistic Situation-bound stylistic factors depend on the given features remain flexible and do not follow rigid rules, communication situation, such as medium (spoken vs. since style is not a matter of grammaticality, but rather written); participation in discourse (monologue vs. of appropriateness. What is appropriate in a given con- dialogue); attitude (level of formality); and field of text can be deduced from the frequency of linguistic discourse (e.g. technical vs. nontechnical fields). With devices in this specific context. As for the analysis of the caveat that such stylistic factors work simultane- frequencies, corpus linguistic methods are becoming ously and influence each other, the effect of one, and increasingly important. With the advent of personal only one, stylistic factor on language use provides a computers, huge storage capacities, and relevant soft- hypothetical one-dimensional variety. Drawing on this ware, it is now possible to compile very large collec- methodological abstraction, stylistic research has tions of texts (corpus (singular), corpora (plural)), identified many correlations between specific stylistic which represent a sample of language use in general, factors and language use. For example, noun phrases and thus enable exhaustive searches for all kinds of lin- tend to be more complex in written than in spoken lan- guistic patterns within seconds. This methodology is guage in many speech communities, and passive voice based on the general approach of style as probability, occurs much more frequently in technical fields of dis- by allowing for large-scale statistical analyses of text. course than in nontechnical ones. For example, by using corpora, the notion of text- Style, as deviation from a norm, is a concept that is type—defined by co-occurrences of specific linguistic used traditionally in literary stylistics, regarding liter- features—has been introduced to complement the ary language as more deviant than nonliterary language extralinguistic concept of ‘genre’. The linguistically use. This not only pertains to formal structures such as defined text types contradict traditionally and nonem- metrics and rhyme in poems but to unusual linguistic pirically established genre distinctions to a consider- preferences in general, which an author’s poetic license able extent. In particular, many spoken and written allows. Dylan Thomas’s poetry, for example, is charac- genres resemble each other linguistically to a far greater terized by word combinations that are semantically extent in terms of text-types than previously assumed. incompatible at first sight and, thus, clearly deviate Style as comparison puts into perspective a central from what is perceived as normal (e.g. a grief ago, once aspect of the previous approaches. That is, stylistic 1043 STYLISTICS analysis always requires an implicit or explicit com- A different, although conceptually similar, tradi- parison of linguistic features between specific texts, or tion of linguistic stylistics was established by British between a collection of texts and a given norm. In linguists in the 1930s and came to be called British principle, stylistically relevant features such as style Contextualism. The most important proponents of markers may convey either a local stylistic effect (e.g. British Contextualism include John Rupert Firth, an isolated technical term in everyday communica- M.A.K. Halliday, and John Sinclair. Their work is char- tion) or, in the case of recurrence or co-occurrence, a acterized by a clear focus, firstly, on the social context global stylistic pattern (e.g. specialized vocabulary in which language is used and, secondly, on the in- and passive voice in scientific texts). depth observation of natural language use. From the From the multitude of linguistic approaches to point of view of British Contextualists, linguists need style, two linguistic schools of the twentieth century to describe authentic language use in context and have exerted the most decisive influence on the devel- should not confine themselves to invented and isolated opment, terminology, and the state of the art of stylis- sentences. Additionally, linguistics is not considered as tics: the Prague School and British Contextualism. an intuition-based study of abstract systems of form as, The central dictum of Prague School linguistics, for example, in the merely formal description of going back to the Bauhaus School of architecture, is autonomous syntactic rules (as in Chomsky’s approach ‘form follows function’. Firmly established since the to language), but as the observation-based and empiri- 1920s, some of this dictum’s B most important propo- cal analysis of meaning encoded by form. This nents are Lubomír Dolez el, Bohuslav B Havránek, approach allows for insights into the immense variation Roman Jakobson, and Jan Mukar ovský. These lin- within language. It is a fact that depending on the con- guists have paid particular attention to situation-bound text of situation, all speakers use different ‘registers’ stylistic variation. A standard language is supposed to (i.e. different styles of language, depending on the have a communicative and an esthetic function that topic, the addressee, and the medium in a given context result in two different ‘functional dialects’: prosaic of use). Note that there is, of course, a clear correspon- language and poetic language. More specific function- dence between the concept of register and the Prague al dialects may, of course, be identified; for example, School’s notion of functional dialect. Although largely the scientific dialect as a subclass of prosaic language, abandoned by mainstream linguists in the 1960s and which is characterized by what is called the ‘intellec- 1970s due to the prevailing Chomskyan school of tualization of language’—lexicon, syntax, and refer- thought, it had already been suggested by Firth in the ence conform to the overall communicative function 1950s that large collections of text were a prerequisite that requires exact and abstract statements. for an empirical approach to stylistic variation. Thus, it A very important notion is the distinction between does not come as a tremendous surprise that, among ‘automatization’ and ‘foregrounding’ in language. others, Sinclair set out to develop computerized corpo- Automatization refers to the common use of linguistic ra that could be used as empirical databases. devices which does not attract particular attention by the With corpus linguistics now a standard methodolo- language decoder, for example, the use of discourse gy, stylistic analyses have reached an unprecedented markers (e.g. well, you know, sort of, kind of) in sponta- degree of explanatory adequacy and empirical accura- neous spoken conversations. Automatization thus corre- cy. For example, stylistic features that are beyond most lates with the usual background pattern, or the norm, in linguists’ scope of intuition, such as the nonstandard language use—it encompasses those forms and struc- use of question tags in English-speaking teenagers’ tures that competent language users expect to be used in talk, are now feasible in quantitative terms. More a given context of situation. Foregrounded linguistic importantly, there is no longer a bias toward fore- devices, on the other hand, are usually not expected to grounded phenomena that tend to catch the linguist’s be used in a specific context and are thus considered attention. A computer, in contrast, does not distinguish conspicuous—they catch the language decoder’s atten- between conspicuous and common phenomena and tion (e.g. the use of old-fashioned and/or very formal provides an exhaustive array of all kinds of patterns, words such as epicure, improvident, and whither in depending solely on the search query. Thus, the fuzzy spontaneous spoken conversations). Foregrounding thus concept of ‘norm’ is about to be put on an empirical captures deviations from the norm. It is obvious that footing since the accessible corpus norm represents what is considered as automatized and foregrounded the norm of a language as a whole. language use depends on the communication situation at Stylistics is a linguistic branch that is immediately hand. In technical fields of discourse, for instance, spe- relevant to foreign language teaching. This applies to cialized vocabulary items tend to be automatized (e.g. both linguistic and literary stylistics. Language learn- lambda marker in molecular biology), but in everyday ers must know which linguistic devices are preferred communication become foregrounded devices. by native speakers in specific contexts. Without such a 1044  SUMERIAN linguostylistic competence, communication errors highly standardized, is also affected by stylistic varia- may be made in interacting with native speakers, such tion, as deliberate misspellings in the language of as using highly formal words in informal settings. advertising and popular culture (e.g. 2 for to/two/too, Also, learners must have command of text-typological lynx for links) reveal. On the other hand, words and knowledge, which is important, for example, in writ- syntax are linguistic devices that, in principle, are sub- ing essays. As for literary texts, language learners ject to transfer between media, although there are clear should acquire a firm understanding of those levels of medium-dependent preferences of lexical and syntac- description where stylistic variation may occur (e.g. tic choice that need to be investigated further. by analyzing Hemingway’s syntactic simplicity and, The objective and unbiased approach to stylistic moreover, its function). variation in authentic language use is a cornerstone of It should be noted that a specific style is sometimes modern descriptive linguistics. Unlike traditional ascribed to a language in its entirety. Although the grammar, it clearly rejects the normative prescription underlying norms remain largely unspecified, general of one specific style. tendencies of stylistic preference differ across lan- guages. This is particularly important for translators, but also for language learners. It is, for instance, com- References mon for German students of English to transfer the Biber, Douglas. 1989. Variation across speech and writing. German style of academic writing, which is character- Cambridge: Cambridge University Press. ized by heavy noun phrases, to their English essays. Coupland, Nikolas (ed.) 1988. Styles of discourse. London: Croom Helm. As with any other linguistic branch, stylistics is Enkvist, Nils Erik. 1973. Linguistic stylistics. The Hague: very much a work in progress. This is because the Mouton. object of inquiry constantly grows, evolving new and Esser, Jürgen. 1993. English linguistic stylistics. Tübingen: specialized fields of discourse (e.g. genetic engineer- Niemeyer. ing, computer sciences). Furthermore, new aspects of ———. 2000. Medium-transferability and presentation struc- ture in speech and writing. Journal of Pragmatics 32. stylistic variation come into existence, such as e-mails, Garvin, Paul L. (ed.) 1964. A Prague school reader on esthetics, a now widely used genre that seems to blur the tradi- literary structure and style. Washington: Georgetown tional distinction between spoken and written lan- University Press. guage. As for empirical approaches to style, new Halliday, M.A.K. 1978. Language as a social semiotic: the social corpora make it possible to address questions of style interpretation of language and meaning. London: Arnold. Joos, Martin. 1961. The five clocks: a linguistic excursion into not possible before. Also, recent theoretical develop- the five styles of English usage. New York: Harcourt, Brace ments will no doubt widen the scope of stylistics. and World. Drawing on British contextualists’ distinction between Leech, Geoffrey, and Michael Short. 1981. Style in fiction: a language substance (that is, sound waves in the phon- linguistic introduction to English fictional prose. London: ic medium and printed paper in the graphic medium) Longman. Semino, Elena, and Jonathan Culpeper (eds.) 2002. Cognitive and language form (that is, anything that can be trans- stylistics: language and cognition in text analysis. ferred from one medium into the other), it has been Amsterdam: Benjamins. suggested that stylistic analyses should clearly distin- Sinclair, John. 1991. Corpus, concordance, collocation. Oxford: guish between medium-dependent and medium-inde- Oxford University Press. pendent stylistic variation. Intonation, for example, is Weber, Jean Jacques (ed.) 1996. The stylistics reader: from Roman Jakobson to the present. London: Arnold. bound to the phonic medium and shows stylistic vari- JOYBRATO MUKHERJEE ation that cannot be mapped onto punctuation in a straightforward and monocausal way. With regard to See also Firth, John Rupert; Halliday, Michael the graphic substance, English orthography, albeit Alexander Kirkwood Sumerian Sumerian was a language spoken in the south of Nas.r (from the end of the fourth millennium to the mid- Ancient Mesopotamia (modern Iraq) and is most likely dle of the third). By the end of the third millennium, first attested in the archaic texts from Uruk and Jemdet Sumerian had died out for the most part as a spoken 1045 SUMERIAN language. However, it was still used in a wide variety sikil /kiskil/ (‘girl’), gir5-gir5-(r)e /gigre/ (‘to sink, to of literary, scholarly, and religious genres, and was pre- dive’), etc. Some have suggested that Sumerian may served in writing until the practical disappearance of have had lexical tones as a suprasegmental feature, the Mesopotamian civilization. Sumerian is an ISOLATE, which allegedly would explain the sometimes high i.e. it is not related to any other language or language number of homophonic terms. However, other factors family. Thus, our knowledge of Sumerian grammar and can explain this: different Auslaut consonants, conso- lexicon is mostly based on a large number of bilingual nantal clusters, etc. texts (in Sumerian and Akkadian), as well as a stream VOWEL HARMONY occurs with several verbal prefix- of scribal and scholastic traditions materialized in a es, and traces of it can be found in some nouns. The corpus of lexical lists and grammatical texts. explicit writing of vowel harmony developed through time in diverse ways. For instance, in the case of the verbal modal prefix /he-/ or /ha-/ (see below), the ortho- Phonology and the Writing Interface graphic evolution went from one allomorph (he2-), to Mesopotamian CUNEIFORM is the logosyllabic script two (he2- and ha-), and eventually to three (he2-, ha-, used for Sumerian. For instance, the sign AN (origi- and hu-). nally a star) could write the word an ‘sky, heaven’, as well as dingir ‘god’. As a consequence of this and other processes, cuneiform signs are polyphonic, i.e. Nominal Morphology they have several readings (e.g. the sign NE can be Sumerian is an AGGLUTINATIVE language. The gram- read {ne}, {de3}, {bi2}, etc.). On the other hand, the matical gender is based on an opposition between ani- writing interface shows HOMOPHONY, because differ- mate and inanimate nouns, but has no specific ent signs can have the same reading, such as {e}, morphological marker and only surfaces in mor- {e2}, {e3}, {e4}…, {e11} (numerical indexes are used phosyntactical relations between pronouns and their to distinguish between homophonous signs). antecedents. Grammatical number (plural vs. singu- Furthermore, scribes used determinatives preceding lar) does not need to be marked in writing (lugal or following some nouns and names, simply as read- ‘king’ or ‘kings’), but can be made explicit through ing aids: KI ‘place’ in lagaški; DINGIR ‘deity’ in d suffixation (lugal-e-ne /lugal-ene/ ‘kings’) or redupli- nanna, etc. cation (lugal-lugal ‘kings’). The absence of marker Our knowledge of Sumerian PHONOLOGY is limited and lexematic reduplication are probably simple by the nature of the writing interface. The vocalic sys- orthographic conventions to write the plural, as may tem is quite simple: /a/, /i/, /e/, /u/. The possible exis- be indicated by the construction of plurals with redu- tence of an /o/ vowel remains unsubstantiated. Vocalic plication of the adjective (dingir gal-gal  god- length does not seem to be phonological. The conso- great-great ‘great gods’) or with reduplication and an nantal system includes three series of stops (/b/ and additional suffix (dingir gal-gal-e-ne  god-great- /p/; /d/ and /t/; and /g/ and /k/), three sibilants (/s/, /z/, great-PLURAL ‘great gods’). and /š/  IPA ), two laterals (/l/ and /r/), three nasals In the nominal system, there are ten cases, which (/m/, /n/, and /g~/ or /"/), and a uvular or pharyngeal of are marked by attaching suffixes to Noun Phrases undefinable nature (/h/ or /¦/). Furthermore, the exis- (NPs). Moreover, cases can also be indicated by insert- tence of different ‘extra phonemes’ (segments that ing prefixes in verbal forms (all elements preceding would not be immediately obvious in the writing inter- the stem are called prefixes): face) has been proposed. Among these, the velar nasal /g~/ or /"/ is now commonly accepted by all Sumerologists—for instance, both {mu} and {gu10} Nominal Prefixes in r are written withB the same sign, MU—as well as /d / suffixes verbal chain (most likely /r /). To further complicate things, some ERGATIVE /-e/ final consonants (Auslaut) seem to drop (kala ABSOLUTIVE /-∅/ ‘mighty’  /kalag/), but the same signs have readings GENITIVE /-ak/ with and without Auslaut: kala  kalag; dug4  du11 DATIVE /-ra/ /-na-/ LOCATIVE /-a/ /-ni-/ ‘to speak’. COMITATIVE /-da/ /-da-/ Our knowledge of Sumerian PHONOTACTICS is limit- TERMINATIVE /-eše/ ({-še3}) /-ši-/ ed. Although the cuneiform system, as most syl- ABLATIVE– labaries, precludes the writing of initial and final INSTRUMENTAL /-ta/ /-ta-/ or /-ra-/ consonantal clusters, it is quite likely that Sumerian LOCATIVE– did have them. In compounds, medial consonantal TERMINATIVE /-e/ /-e-/ (or /-i-/) clusters were also sometimes difficult to indicate: ki- EQUATIVE /-gin/ ({-gin7} GIM) 1046  SUMERIAN NPs are usually called ‘nominal chains’ by The problem of the marking of the so-called marû Sumerologists, because all the suffixes are heaped at stem is still widely discussed. It is likely that all verbs the very end of the last syntagm: had two stems. Affixation verbs perhaps marked the marû stem with an affix /-e/; reduplication verbs (like dumu lugal kalam-ma-ka-ke4-ne-ra gar ‘to place’) marked it with partial reduplication /dumu lugal kalam-ak-ak-ene-ra/ (ga2-ga2), as opposed to complete reduplication in son-king-nation-GEN-GEN-PL-DAT ¦amtu¢ forms (gar-gar); alternating verbs (such as e3 ‘for the son of the kings of the nation (Sumer)’ ‘to go out’) with their ‘expanded form’ (/e3-d/); and Nonetheless, the syntagmatic structure is not so suppletive or complementary verbs (such as dug4) much that of a chain as it resembles a series of Chinese with completely different lexemes (e) (see Yoshikawa boxes (Zólyomi 1996): 1993: 1–56, 95–104, 114–26). However, it is also pos- sible that many verbs did not have two different stems, β[dumu α[lugal kalam-ak]α-ak-ene]β-ra and that the only way to distinguish ¦amtu¢ from marû in those verbs was through concord (i.e. through Sumerian is an ERGATIVE language: the subject of an pronominal affixes). According to this, /-e/ would not intransitive verb presents the same marker as the mark marû but agreement for the 3rd sg. subjects of object of a transitive verb (the absolutive case), while transitive marû forms (intransitive marû forms have the subject of a transitive verb presents a marker (the /-∅/ in the 3rd sg.). ergative case) that is different from that of the intran- Verbal forms can present both prefixes and suffixes. sitive verb. Thus, Sumerian has /-e/ as the ergative suf- Although almost no actual verbal form exhibits all the fix, and /-∅/ as the marker of absolutive case: possible affixes, an ideal table of slots would be as in lugal-e e2 mu-un-du3 ‘the king built the temple’ Table 1. lugal i3-tuš ‘the king sat down’ Whereas some MODAL PREFIXES always indicate nin-e in-tud-en ‘the queen bore me’ the same mood (ga- → cohortative; nu- →negative; ša- nin i3-tuš ‘the queen sat down’ → affirmative), others can mark either deontic or epis- temic modality, depending on their interaction with the In fact, Sumerian exhibits split ergativity in its mor- other elements within the TAM system (Civil 2000): phology. The ergative alignment is strictly followed only in the nominal system. Independent personal pro- /ha-/  ¦amtu¢ → affirmative ha-na-sum nouns, imperatives, cohortative verbal forms, and a ‘I have indeed few nonfinite verbal constructions exhibit an accusa- given’ tive alignment. The system of verbal agreement shows /ha-/  marû → precative hu-mu-hul2-e a similar split: the ¦amtu¢ forms (perfective) are erga- ‘may you rejoice’ tive, whereas the marû forms (imperfective) show an /na-/  ¦amtu¢ → affirmative nam-mi-gub accusative pattern. ‘he set indeed’ /na-/  marû → prohibitive na-ab-pad3-de3 The Verbal System ‘do not tell’ Verbal stems are usually divided into two major cate- /bara-/  ¦amtu¢ → negative ba-ra-ra-dug4 gories: ¦amtu¢ (‘quick, sudden’ → perfective) and affirmative ‘I have never said marû (‘slow, fat’ → imperfective). However, these to you’ two labels, rather than native Sumerian grammatical /bara-/  marû → vetitive ba-ra-pad-re6 categories, reflect the understanding of the Sumerian ‘he shall not verb by Akkadian-speaking scribes. In an early gram- destroy’ matical text, lugud (‘short’) occurs instead of ¦amtu¢ , and gid2 (‘long’) instead of marû (Civil 2002)—the Other prefixes in this slot do not really indicate same two terms were used by ancient scribes in nam- modality: u3- marks anteriority and /inga-/ is actually ing signs according to their shapes. Thus, the same a proclitic connective particle. labels, when used to designate verbal stems, refer to The number, function, rank, compatibility, and the shape of the stems, which would be regarded as morphophonemic shape of the so-called CONJUGATION long because of reduplication and other possible PREFIXES are still a matter of discussion. For instance, changes. These two stems marked tense and aspect, /-m-/ is probably the noninitial allomorph of /mu-/, perfective and imperfective, as well as mood (deontic which explains why both prefixes cannot occur togeth- versus epistemic) when occurring with certain modal er. It is possible to argue for a rather simple system of prefixes (see below). conjugation prefixes with only four morphemes: /ba-/; 1047 1048 SUMERIAN TABLE 1 Structure of the Sumerian Verbal Chain  SUMERIAN /imma-/; /i-/ or /V-/; and /mu-/. The prefix /bi-/ would a verbal prefix. Moreover, in order to agree with plu- be a combination of the prefix /ba-/ followed by the rals, the pronominal prefixes can occur together with locative-terminative infix, and /imma-/ would be a plural pronominal suffixes. reduplication of /mu-/ (Karahashi 2000). The PRONOMINAL SUFFIXES are identical for all ver- All verbal forms seem to start with an obligatory bal forms with the exception of the 3rd person singu- conjugation prefix (/mu-/, /ba-/, or /i-/). The choice of lar and plural. A 3rd person in the absolutive case prefix seems governed by FOCUS (Vanstiphout 1985). shows concord with /-∅/ in the singular and with /-eš/ The prefixes /mu-/ and /ba-/ are mutually exclusive in the plural. An ergative with a marû form shows con- and complementary: /mu-/ is focused for a person but cord with /-e/ in the singular and /-ene/ in the plural. not for a locus, while /ba-/ is focused for a locus but The absolutive case of a transitive marû construction not for a person; and /i-/ is not focused; and hence it is (i.e. the accusative case with marû) does not agree indifferent to this opposition. In discourse, the /i-/ pre- with the pronominal suffixes but with the pronominal fix is preferred for supportive, nonsubstantial material prefixes (/-b-/ and /-n-/). The distribution of pronomi- (background), but the foreground of regular narrative nal elements in verbal forms is as follows: discourse is marked by /mu-/ or /ba-/ according to With regard to the choice of prefixes vs. suffixes, focus: the ¦amtu¢ forms follow an ergative pattern, whereas the marû ones show accusativity. Nonetheless, the pronominal suffixes used with the marû forms do point to an opposition between ergative subject and absolutive subject for the 3rd singular and plural (i.e. they ultimately have an ergative alignment). The SUFFIX /-ED/ can immediately follow the verbal stem and precede the pronominal suffix. Some con- sider the /-e-/ in /-ed/ the marker of marû (see above). The conjugation prefix /a-/ seemingly occurs in This suffix is much more frequent in nonfinite than in nonagentive passive constructions (an-na-sum ‘it was finite verbal constructions (in which it indicates given to him’). This prefix defocalizes the agent of the future in diverse modalities): e2-mu lu2 i3-bur3-de3 sentence, shifting the focus to any other argument ‘someone could break (/i-bur-(e)d-e/) into my house’; (dative, comitative, ablative, and so forth); see e2 du3-de3 igi-zu u3 dug3-ga nu-ši-ku4-ku4 ‘in order Yoshikawa (1995). The occurrence of /a-/, however, to build (/du3-(e)d-e/) the house you will not let sweet may also correspond to local or diachronic dialects (or sleep enter your eyes’; and ur-sag e2-a-na ku4-ku4- even to scribal idiolects), since in some texts it is da-ni ud me3-še3 gu3 ga2-gar-am3 ‘at entering (/ku4- extremely rare or completely unattested, while in oth- ku4-(e)d-a-ani/  enter-enter-ED-NOMINALIZER-his) ers it is quite frequent. his house, the warrior was a storm roaring towards The DIMENSIONAL INFIXES mark case functional rela- battle’. The suffix /-ed/ is never written as such: /-e-/ tions between the verb and NPs that may be explicit or is written almost exclusively after a consonant, and merely implicit in the sentence. The PRONOMINAL PRE- /d/ is written only when followed by a vowel. FIXES normally agree with the subject of transitive The IMPERATIVE exhibits a reverse order of verbal ¦amtu¢ forms (ergative) and the subject of both transi- constituents: it begins with the stem, which is followed tive and intransitive marû forms—the latter showing by all the prefixes; e.g. sum-ma-ab /sum-mu-a-b/ an accusative alignment. Furthermore, they can also ‘give (sg.) it to me’, sum-ma-ab-ze2-en /sum-mu-a-b- specify that a dimensional prefix (terminative, comita- enzen/ ‘give (pl.) it to me’. This phenomenon is simi- tive, etc.) refers to a 2nd or 3rd person, as well as the lar to the switch from proclisis to enclisis in the gender (/-n-/ for animate and /-b-/ for inanimate) of the imperative in other languages; e.g. Spanish me lo das verbal object or any syntactical argument marked with (‘you give it to me’) vs. dámelo (‘give it to me’). 1049 SUMERIAN Morphosyntax a syntactic and semantic unit: : igi — bar (‘to look at’ ← bar ‘to open’  igi ‘eye’), ki — ag2 (‘to love’ ← ag2 The NOMINALIZER suffix /-a/ can be attached to both ‘to measure’  ki ‘place’), etc. The second object of a nonfinite and finite verbal forms, and can be followed compound verb is very frequently in the locative–ter- by case endings and pronouns. When the nominalized minative case: nig2-dug3-ge al na-an-ga-am3-mi-in- verbal form agrees with an explicit (such as lu2 dug4 ‘sweet things (/nig2-dug3-e/) she has indeed also ‘man’→ who/that) or implicit noun that has an wished (/na-inga-bi-n-dug4/, al — dug4 ‘to desire, antecedent in another sentence, it constitutes the wish’)’. There are some double compound verbs, in equivalent to an English relative clause: ensi2 lu2 e2- which a whole compound verb becomes the nominal ninnu in-du3-a e2-uru-gir2-suki-ka-ni mu-na-du3 part of a compound verb, whose verbal member is an ‘the ensi-ruler that built the Eninnu (lu2... in-du3-a ← ‘auxiliary’ verb, such as ak (‘to do’) and dug4 (‘to /i-du3-a/), built her temple of the city of Girsu’. Due say’): šu tag — dug4 ← šu — tag ‘to cover, decorate’ to the high number of grammatical functions explicit- (← šu ‘hand’  tag ‘to touch’). ly marked, there is no obvious syntactical opposition between parataxis and hypotaxis (the latter would cor- A Sumerian Genderlect or Genrelect? respond entirely to nominalized verbal forms). For instance, the word order tends to be almost always Sumerian is called eme-gir15 (perhaps ‘native tongue’) Subject–Object–Verb in all sentences. in native Sumerian sources. In some Mesopotamian The verb TO BE (me) has a basic essential meaning scholarly texts, a few lexical items and grammatical and does not indicate existence, for which gal2 (‘to be forms are identified as eme-sal (perhaps ‘fine lan- there, to exist’) is used: pi-lu5-da ud-bi-ta e-me-a guage’). It has been argued that eme-sal was a ‘these were (/i-me-a(m)/) the conventions of old women’s language (Frauensprache) or genderlect, times’. It occurs more frequently in the form of an especially because the sign SAL can also be read enclitic copula, with the following suffixes: munus ‘woman’. The fact is that eme-sal is attested in compositions of very specific genres: cultic songs per- 1st sg. -me-en 1st pl. -me-enden formed by gala-priests; diverse texts containing 2nd sg. -me-en 2nd pl. -me-enzen Inanna’s speech; some laments over the destruction of 3rd sg. -(a)-m 3rd pl. -me-eš cities; a lullaby; about 30 proverbs or short sayings from rhetorical collections; an unpublished composi- E.g., digir-ra-ni dšul-utul12-am6 ‘his god is Šulutul’, tion (‘The song of the millstone’); and the ‘Dialogues ama-mu ze2-me ‘you are (/ze2-me-e(n)/) my mother.’ between two women’. No text is entirely written in The PRONOMINAL SYSTEM follows an accusative eme-sal, and there is no true consistency in its use; alignment. The pronominal subjects of both transitive hence, an otherwise ‘main-dialect’ text may present and intransitive verbs present the same marker, /-e/: some scattered eme-sal words. In most cases, the 1st sg. 2nd sg. 3rd sg. 3rd pl. SUBJECT ga2-e (me-e) za-e (ze) e-ne e-ne-ne DATIVE ga2-a-ra (ma-a-ra) za-a-ra e-ne-ra e-ne-ne-ra TERMINATIVE ga2(-a/e)-še3 za(-a/e)-še3 e-ne-še3 e-ne-ne-še3 COMITATIVE ga2(-a/e)-da za(-a/e)-da e-ne-da e-ne-ne-da EQUATIVE ga2(-a/e)-gin7 za(-a/e)-gin7 e-ne-gin7 e-ne-ne-gin7 The POSSESSIVE SUFFIXES can be attached to NPs as occurrence of eme-sal forms seems determined most- well as to nominalized verbal forms: ly by the genre of the text. 1st sg. -mu (-gu10) ‘my’ 1st pl. -me ‘our’ References 2nd sg. -zu ‘your’ 2nd pl. -zu-ne-ne, -zu- e-ne-ne, -zu-ne Attinger, Pascal. 1993. Eléments de linguistique sumérienne: ‘your’ La construction de du11/e/di «dire». Fribourg/Göttingen: Editions Universitaires/Vandenhoeck&Ruprecht. 3rd sg. an.-a-ni ‘his, her’ 3rd pl. -a-ne-ne ‘their’ Civil, Miguel. 2000. Modal prefixes. Acta Sumerologica 22, in 3rd sg. in. -bi ‘its’ -bi ‘their’ (prob- press. ably collective) ———.2002. The forerunners of marû and ¦amt u¢ in Old Babylonian. Riches hidden in secret places: ancient Near Sumerian has a number of COMPOUND VERBS, i.e. Eastern studies in memory of Thorkild Jacobsen, ed. by combinations of a verb and a direct object that become Tzvi Abusch, 63–71. Winona Lake, IN: Eisenbrauns. 1050  SWADESH, MORRIS Edzard, Dietz Otto. 2003. Sumerian grammar. Leiden: Brill. Vanstiphout, Herman L.J. 1985. On the verbal prefix /i/ in stan- Karahashi, Fumi. 2000. The locative-terminative verbal infix in dard Sumerian. Revue d’Assyriologie 79. 1–15. Sumerian. Acta Sumerologica 22, in press. Yoshikawa, Mamoru. 1993. Studies in the Sumerian verbal sys- MichaBowski, Piotr. 1980. Sumerian as an ergative language. tem. Hiroshima: The Middle Eastern Culture Center in Journal of Cuneiform Studies 32. 86–103. Japan. ———.2004. Sumerian. The Cambridge encyclopedia of ———.1995. The Sumerian verbal prefix /a-/. Acta ancient languages, ed. by Roger D. Woodard, 19–59. Sumerologica 17. 299–307. Cambridge: Cambridge University Press. Zólyomi, Gábor. 1996. Genitive constructions in Sumerian. Thomsen, Marie-Louise. 1984. The Sumerian language. Journal of Cuneiform Studies 48. 31–47. Copenhagen: Academic Press. GONZALO RUBIO Swadesh, Morris Like his teacher Edward Sapir, Morris Swadesh was a apparent lexical and structural similarities in different prolific data-gatherer and an avid student of lan- languages, and his interest in comparative historical guages. He learned Yiddish from his Russian parents. linguistics grew. Since his study of nearly extinct As an undergraduate at the University of Chicago, he languages was conducted with limited resources, he studied German and French, and as a graduate and felt the need for a standardized procedure for quickly postgraduate at Yale University he concentrated on collecting crucial data yielding clues about language Nootka, a Canadian indigenous language. On his reg- relationships. ular field trips through Canada, the United States, and To determine whether two given languages are Mexico, he collected data on more than 20 native lan- related, historical linguists usually use the ‘compara- guages. In Mexico, he developed programs for indige- tive method’. This means that they attempt to recon- nous people to attain literacy in their own languages. struct an ancestor language on the basis of cognates, Later, during World War II, he worked for the war i.e. arguably related words from different languages department, editing dictionaries, providing linguistic (English hound is a cognate of German Hund ‘dog’). analyses of foreign languages, and developing teach- Since cultural development is always accompanied by ing materials for Spanish, Russian, Burmese, and lexical innovations, Swadesh—and many linguists Chinese. before him—felt that presumably more stable ‘basic Swadesh did his first theoretical work on phonemic vocabulary’ would be the best place to start looking for analysis, i.e. the analysis of the sound structure or cognates. The basic vocabulary of a language phonology of languages. Sapir, Leonard Bloomfield, describes body parts and functions, such as skin, Nikolai Trubetzkoy, and others had already advanced blood, drink, natural phenomena like water, sky, bird, the concept of the ‘phoneme’, an abstract representa- smoke, immediate sense experiences, and physical tion of sound types. Swadesh’s contribution was to dimensions, such as long, red, cold. While the concept develop a set of principles to help the phonologist dis- of basic vocabulary had been informally used before, cover phonemes on the basis of the distribution of Swadesh made it explicit by drafting a list of 100 word sounds in a given language. If, for example, one sound meanings that a fieldworker investigating any lan- always occurs at the beginning of words, while anoth- guage could use for identifying the basic vocabulary of er always occurs at the end, Swadesh suggested that that language. The use of this list, now generally these particular sounds are in ‘complementary distri- known as the ‘Swadesh list’, has drawn criticism from bution’ and thus could potentially be instances of the its inception. Many linguists believe that it is impossi- same sound type or phoneme. These principles were ble to enumerate universal meanings, and that the later applied to word and sentence elements by Zellig identification of semantically equivalent words in dif- Harris. Distributional analysis thus became a general ferent languages is often highly problematic. ‘discovery procedure’ for the basic elements of lin- Nevertheless, the list has become a widely used tool in guistic structure and has remained an integral part of comparative linguistics. linguistic methodology to this day. Even though the notion of basic vocabulary was In his extensive investigations of numerous lan- already contentious, Swadesh pressed forward and guages, Swadesh gained an increasing appreciation of used his list for lexicostatistics, a quantitative method 1051 SWADESH, MORRIS for measuring the similarity of languages. If the basic deliberations tend to be overshadowed by the bolder vocabulary of one language matches that of another aspects of his thought. by over 90%, Swadesh argued, these languages must be closely related. Most linguists believe that the Biography reconstruction of ancestor languages provides more reliable evidence for the relatedness of languages Morris Swadesh was born in Holyoke, Massachusetts than statistical analyses; hence, lexicostatistics con- on January 22, 1909. He received his B.A. (1930) and tinues to be viewed with suspicion. Yet, moderate lin- M.A. (1931) for his dissertation on Nootka aspect, guists today concede that the Swadesh list and tutored by Edward Sapir, University of Chicago. He lexicostatistics may be useful for rough initial inves- followed Sapir to Yale University, New Haven, tigations or for situations where complete data are Connecticut, for Ph.D. work on Nootka semantics, simply unavailable—which is, in fact, close to what 1933, and then worked at Yale on synchronic phono- Swadesh had in mind. logical theory and on American English grammar in Even more controversially, Swadesh claimed that 1933–1937. He was Assistant Professor, University of the ‘decay’ of basic vocabulary could be used for Wisconsin, Madison in 1937–1939. He moved to ‘glottochronology’, the dating of ancestor languages Mexico City, where he became Director, Consejo de analogous to determining the age of fossils on the Lenguas Indígenas, and Director of Linguistics, basis of radioactive decay. Swadesh came to believe Departamento de Asuntos Indígenas, in 1939; he was that basic vocabulary decays at a rate of 14% over Professor, Instituto Politécnico Nacional de México, 1,000 years; hence, languages would retain on average Escuela de Antropología, and Departamento de about 86% of their basic vocabulary over this time Asuntos Indigenas, in 1939–1941. Swadesh was a lin- span. Thus, if the basic vocabularies of the two related guist for the War Department in New York City during languages are found to match by 70%, they can be World War II; he was Associate Professor, City assumed to have developed from a single language University of New York in 1948, but lost his teaching that existed approximately 12 centuries ago. The appointment and had his passport revoked because of assumption that basic vocabulary decay is generally ‘leftist’ views and activities. He worked as a librarian uniform has been seriously challenged. If one allows at the Boas Collection, Library of the American that languages, just like societies, may develop at dif- Philosophical Society, Philadelphia, PA, 1949–1953, ferent rates at different times, the assumption of steady and carried out independent fieldwork in 1953–1956. vocabulary decay in particular, and the glottochrono- He moved again to Mexico City, where he became logical method in general, is seriously undermined. Professor at the Instituto de Historia, Universidad Swadesh’s name has remained symbolic for lexico- Nacional Autónoma de México, and the Escuela statistics and glottochronology, but his central place in Nacional de Antropología e Historia, 1956–1967. He the continuing and highly ideological debate about was a member of the Linguistic Society of America these and related issues seems to rest on misunder- (LSA) in 1931, Life Member in 1937; member of standings and/or polemically slanted readings. He is, two special interest groups of the LSA in 1939; for example, accused of introducing lexicostatistics as President of the Linguistic Section of the 29th a shortcut for investigation, attempting to avoid the International Congress of Americanists, 1939; and edi- hard work of reconstruction. However, Swadesh stated tor of Word, 1946–1949. Swadesh died in Mexico City repeatedly that a detailed knowledge of the languages on July 20, 1967. under investigation is crucial, and that other data must be considered. Likewise, Swadesh is misleadingly cited as a supporter of the theory that all languages References have developed from a single ancestor (the ‘monogen- Anttila and Raimo. 1972. An introduction to historical and esis’ theory). He certainly conceded that the instinctive comparative linguistics. New York: Macmillan; 2nd edition, vocalizations of early humanoids may be called a as Historical and comparative linguistics, Amsterdam and species-specific ‘language’, but he also surmised that Philadelphia: Benjamins, 1989. Harris, Zellig. 1951. Methods in structural linguistics. as soon as arbitrary signs—i.e. true words and com- Chicago: University of Chicago Press; as Structural plex linguistic structures—entered the human reper- Linguistics, 1960. toire, diversification was the instant result. Hymes, Dell H. (ed.) 1964. Language in culture and society. These ideas were the focus of his major book, The New York: Harper & Row. origin and diversification of language, but he died of ———. 1971. Morris Swadesh: from the first Yale School to world prehistory. The origin and diversification of language, a heart attack before he was able to complete it. His by Morris Swadesh. Chicago: Aldine Atherton. arguments thus remained somewhat sketchy, which Lamb, Sidney M., and E. Douglas Mitchell. (eds.) 1991. may explain why Swadesh’s even-handed and careful Sprung from some common source: investigations into the 1052  SWEDISH AND SCANDINAVIAN LANGUAGES prehistory of languages. Stanford, CA: Stanford University ———. 1937. A method for phonetic accuracy and speed. Press. American Anthropologist 39(4). 728–32. Newman, Stanley. 1967. Morris Swadesh (1909–1967). ———. 1959. Linguistics as an instrument of prehistory. Language 43. 948–57. Southwestern Journal of Anthropology 15. 20–35. Sapir, Edward and Swadesh, Morris. 1932. The expression of ———. 1971. The origin and diversification of language. the ending-point relation in English, French, and German, Chicago: Aldine Atherton. edited by Alice V. Morris, Linguistic Society of America, PHILIPP STRAZNY Language Monographs, Vol 10. ———. 1934. The phonemic principle. Language 10. 117–29. See also Haas, Mary Rosamond; Sapir, Edward Swedish and Scandinavian Languages Scandinavian languages belong to the Indo-European syllabic words simplified the Ancient Scandinavian North-Germanic languages. North-Germanic lan- syllable structure. For example, the inscription in guages are spoken in the Nordic countries (Denmark the fifth century golden horn from Danish Gallehus, (including The Faroes and Greenland), Finland, ek hlewagastiR holtijaR horna tawido, ‘I, HlewagastiR Iceland, Norway, and Sweden). The concept of Nordic from Holt, made the horn’, results in the reconstructed languages (Swe. nordiska språk) covers all North- forms: *ek hlégestr hyltir horn táð a, in Old Norse. Germanic languages, and Scandinavian languages Plurisyllabic words like Hlégestr ‘dropped’ syllables. comprise the three languages Swedish, Danish, and Deletion of Ancient Scandinavian final –n caused a Norwegian. The prehistoric period in Scandinavia is changed infinitive form: fullian  Swe. fylla ‘fill’, lau- not well known, but by the second century CE, Ancient sian  Da. løysa ‘loosen’. The Umlaut (common Scandinavian language(s) were spoken in Denmark, Germanic vowel shift) complicated the morphology Southern and Central Sweden, and in southern Norway. structure from Ancient Scandinavian to Old Norse. Based on older language history the North-Germanic Particularly, /i/ in suffices influenced /a/, /o/, and /u/ in languages have been divided into West-Nordic word roots: ansiR  æsir ‘pagan gods’, fo¯tiR  fø´tr, (Norwegian, Icelandic, and Faroese) and East-Nordic ‘feet’. In words like mann
menn ‘man
men’ and (Danish and Swedish). A more recent division separates mús
my´ss ‘mouse
mice’, the vowel shift received North Scandinavian (Norwegian and Swedish) and phonemic weight: Swe. singular man vs. plural män. South Scandinavian (Danish) from Insular Nordic The three Scandinavian languages (and German) share (Icelandic and Faroese; cf. Torp 1982; Vikør 1993). The the development of /æ/, /ø/, and /y/, which are repre- latter view is central here. The division into West and sented in writing (Swe. ä and No./Da. æ for /æ/, Swe. East Scandinavian was based on isoglosses, for exam- ö and No./Da. ø for /ø/ and y for /y/). Swedish also has ple, the deletion of initial /j/ in West Scandinavian å, for /o/, which was recently adopted by Norwegian (Old Norse): Nynorsk ek, and Faroese eg, Swedish and Danish for the digraph aa. Diphthongs changed to jag and Danish jeg, ‘I’. Changes during the Middle ages monophthongs: lausian  Swe. lösa ‘loosen’. The in Norwegian/Swedish vs. Danish are considered more change of diphthongs to monophthongs did not affect significant than the West – East split (in the sixth – sev- Old Norse or some Norwegian dialects: Nynorsk has enth centuries CE). For example, the four-case system stein, ‘stone’, laus ‘loose’, and øy ‘island’, compared to in Icelandic (with Nominative, Genitive, Dative, and Swedish/Danish sten, lös/løs, and ö/ø. Accusative) developed into a two-case system in the Scandinavian languages (Nom. and Gen.). Another Danish vs. North Scandinavian Languages change was the loss of the synthetic verb conjugation, which in the Old Norse system expressed person, num- Three changes initiated a unique development for ber, and mood. Suffixed definite articles developed from Danish. The neutralization of unstressed vowels in the demonstratives, and the passive voice, which inflectional endings in Danish developed a central evolved from the reflexive pronoun sik ‘self’, for exam- vowel/schwaa, which is still spelled with an –e: Da. ple, Swe. kalla ‘call’ vs. kalla-s ‘be called’, are charac- holde, Swe. hålla ‘to hold’. This initiated a reduction teristic of mainland Scandinavian. of inflectional morphology and its functions. Within phonology, the syncope (deletion between Danish unvoiced stops after long vowels became consonants) of short, unstressed vowels in pluri- weaker pronounced (lenition). The older forms rep 1053 SWEDISH AND SCANDINAVIAN LANGUAGES ‘rope’, bita ‘bite’, høk ‘hawk’, and leka ‘play’ devel- Western Sweden), Central Swedish, North Swedish, oped into reb, bide, høg, and lege. This has not been Gutnish (Gotland), and East Swedish dialects (in consequently reflected in later spelling conventions in Finland, influenced by Finnish phonology). The differ- Danish. In Danish, the voiced stops shifted to frica- ences between the peripheral dialects are considerable. tives. The adjacent dialect areas in Southern Sweden Personal pronouns are compulsory to indicate sub- and Norway partly followed suit, but they did not ject person (only tense is expressed in the verbs): jag introduce the full set of fricatives. talar ‘I speak’, du talar ‘you speak’ (second-person Danish developed a glottal stop, which was paral- singular), han talar ‘he speaks’, vi talar ‘we speak’. leled by the development of a tonal system in Regular (titta–tittade–tittat ‘look, looked, looked’) Norwegian and Swedish, with two tonemes. The tones and irregular forms (springa–sprang–sprungit ‘run, make a difference between words such as Swe. ´anden ran, run’) of verbs exist. The definite–indefinite article ‘the duck’ (Swe. akut accent) and xanden ‘the spirit’ system is complex, since congruence of definiteness, (grav accent). The glottal stop is not reflected in gender, and number is expressed. Swedish (and Danish spelling; it is understood as a prosodic, not a Danish) only has Neuter (det-genus) and Uter (den- segmental feature. Insular Nordic has neither the glot- genus) gender. For example, det stora huset ‘the big tal stop nor the toneme system. These three phonetical house’, ett stort hus ‘a big house’, and, den gamla changes added to the differences between Danish and staden ‘the old town’, en gammal stad ‘an old town’. Swedish/Norwegian; Danes and Swedes today have Word order (basically SVO) has syntactic func- difficulties understanding each other. tions, for example, indicating the type of clause (main–subordinate; declarative–question): Han kom- mer inte (main clause) ‘He will not come’, vs. Han Swedish säger, att han inte kommer ‘He says, that he will not The separate Swedish development can be divided into come’; Han kommer snart ‘he will come soon’, vs. five periods: Runic Swedish 800–1225 CE, Classical Kommer han snart? ‘Does he come soon?’. Old Swedish 1225–1375, Late Old Swedish 1375– Swedish is the de facto official language of 1526, Early Modern Swedish 1526–1732, and Modern Sweden. About 8.2–8.3 million speakers of the nine Swedish from 1732–. These periods are based on fea- million inhabitants speak it as a first language; the tures of the written language. The Runic inscriptions remainder speak it as a second language. In Finland, (Germanic by origin) were possibly used as early as Swedish is an official national language, with about the second century CE in Denmark. Runes were used 300,000 speakers. Emigration to North America start- in carvings in hard material such as wood, bone, and ed already in the seventeenth century, but had its peak stone; they have been well retained in runic stones. in the late 1800s. The Swedish targets were mainly the Runes were used for almost a millennium, even for farming areas around the Great Lakes. About one mil- everyday purposes in the late Middle Ages. lion people left Sweden for North America; about The Scandinavian languages were stabilized by the 70,000 speak Swedish today (the United States of translations of the New Testament (in 1526 to America). Modern trans-Atlantic migration has target- Swedish) and the Bible (in 1541, Gustav Vasa’s Bible). ed the West and East coasts. Printing had been introduced in 1483. Some of the major typological changes of Swedish from a synthet- Danish ic to a more analytic language were completed by the end of the fifteenth century. In 1732, the newpaper Danish and Swedish have experienced similar devel- Then Swenska Argus started appearing (similar to the opments for their written languages. Old Danish is set English The Spectator), which had a modernizing to 800–1100 CE, Early Middle Danish 1100–1350, impact on the writing of Swedish. Late Middle Danish 1350–1525, Early Modern Danish The administration of the Swedish kingdom was 1525–1700, and Modern Danish 1700–. Christianity centralized early and efficiently by the Vasa dynasty had a clear impact on this process, both during the (1523–1654). It used a Stockholm-based chancery eleventh century, and with the Lutheran Reformation style, which facilitated the standardization of Swedish in the sixteenth century. Low German language con- and the evolution of a spoken standard Swedish. The tact effects were extensive during the Late Middle cultivation of Swedish was supported by the founda- Danish period. tion of the Swedish Academy in 1786 by King Gustav Danish Vikings created a strong position for Danish III. During this period, French also had a great influ- in Northern Europe. It has retained a dominant posi- ence on the court and its language. tion vis-á-vis Norwegian, Icelandic, and Faroese (and The Swedish dialects are divided into six major Greenlandic). Danish was an urban high-status variety groups: South Swedish (Scania and adjoining in Norway for seven to eight centuries. This has provinces), Guthnic dialects (Northwest of Scania; brought the Norwegian vocabulary close to Danish. 1054  SWEDISH AND SCANDINAVIAN LANGUAGES Danish has itself been under continuous pressure from After the Danish secession (1814) of Norway, German and Latin. Danish dominance was challenged. It remained a Similar to Swedish, the oldest written sources, in highly central issue during the whole period of the addition to runic inscriptions, are regional law texts Norwegian–Swedish Union, 1814–1905. One lan- from the thirteenth century. Beginning in the four- guage political direction was headed by the self-taught teenth century, legends, chronicles, and religious liter- linguist Ivar Aasen (1813–1896), who undertook a ature were written in and translated into Danish. The study of south and west-Norwegian dialects. He influential Bible of King Christian III in 1550 was to a founded modern Norwegian dialectology, and through large extent based on Danish dialects. From that peri- his grammar (1864) and dictionary (1873) based on od onward, the Zealand dialects, including Roskilde these dialects, he created a standard written and Copenhagen, have been important. The standardi- Norwegian, Nynorsk. This was in line with the ideas of zation process was largely adapted to Copenhagen the nationalistic Romantic era. Nynorsk was given upper-class people’s style. The German Oldenburg official recognition in Parliament in 1885. The support monarchy, which ruled the country for four centuries for Nynorsk has decreased steadily since the 1940s. up to the mid-nineteenth century, preferred German. Another approach was more pragmatic. The teacher Also, Latin and later French were prestigious lan- Knud Knudsen (1812–1895) Norwegianized the estab- guages, whereas Danish remained the ‘people’s’ lan- lished, written, Danish-based language according to guage until the eighteenth century. With the writer the middle-class townspeople’s speech habits. These Ludvig Holberg (1684–1754), a stable literary devel- were influenced by southeastern dialects and Danish. opment started, which pointed out the route to present- This created the second Norwegian standard, Bokmål day standard Danish. (or Dano-Norwegian). Through two spelling reforms Three main Danish dialect areas are recognized: (1907 and 1917), Bokmål became more adapted to Jutish in Jutland, Insular Danish in Funen and Norwegian. After the first translation from Danish to Zealand, and ‘East-Danish’, in Bornholm and the lost Bokmål in 1919, Bokmål has been considered a sepa- provinces in Southern Sweden. Similar to Swedish, an rate language from Danish. About 1880, the two forms extensive dialect leveling has taken place recently. of Norwegian were connected to the political struggle Danish is the first language of about 4.8 million of between on the one hand a more radical, parliamentary the 5.1 million inhabitants of Denmark. It is the offi- wing—supporting Nynorsk—and a more urban, Oslo- cial language of Denmark, and has a secondary official based conservative wing—supporting Bokmål. status in the Faroes and Greenland. About 45,000 Since the linguistic differences were small it was speak Danish in North Schleswig, where it can be suggested already in the 1880s, that the two standards studied at all levels in school. Danish emigration to be amalgamated to a common Norwegian (Samnorsk). North America was extensive in the nineteenth and The idea slowly gained ground, but after fierce oppo- twentieth centuries. Today, 35,000 descendants speak sition during the 1960s, the official support for this the language (United States of America), many of view has been downgraded. A natural leveling has par- whom live on the West Coast. tially closed the linguistic gap, but attitudinal resist- ance is still strong on both sides. Bokmål is used by more than eight out of ten Norwegians as their school Norwegian language and written code. Norwegian stems from Old Norse (West Compared to Danish and Swedish language poli- Scandinavian), 700–1350 CE. The Middle Norwegian cies, Norwegian official policy has been both more lib- period is set to 1350–1525, and Modern Norwegian eral in relation to dialects, and more puristic regarding 1525–. Old Norse represents a more synthetic and language contact effects (from English). Norwegian Modern Norwegian represents an analytic stage of the purism is less protective than the Icelandic attempts to language. Written forms of Old Norse include the fully replace the influx of international loanwords. Icelandic Sagas. The historian Snorri Sturlasson Norwegian dialects are divided into five groups: (1179–1241) termed the Old Norse of his day Danish. Western, Central (covering the inland valleys), The idea of different languages (Danish, Icelandic, Eastern, Trønder, and Northern dialects. Nynorsk and Norwegian) has thus been promoted later. its base dialects (Western and Central), have greater Middle Norwegian was under heavy pressure from similarities with the Insular Nordic languages. Danish. Written Norwegian was practically replaced About 95% of the Norwegians speak the official by Danish writing by 1525. Further, the Danish Bible Nynorsk or Bokmål as their mother tongues. About of Christian III was used in Norway. This made the 800,000 Norwegians emigrated to North America. attempt to create a stable, written Norwegian unsuc- Today, 80,000 know Norwegian in the United States. cessful. A Danish-based high variety developed in the Many of them migrated to the shores of the Great towns. Lakes, as did the Swedes. 1055 SWEDISH AND SCANDINAVIAN LANGUAGES Linguistic Examples of Swedish, Danish, and References Norwegian Barð dal, Jóhanna, Nils Jörgensen, Gorm Larsen, and Bente Swedish Danish Norwegian Martinussen. 1997. Nordiska - Våra språk förr och nu (Nordic—our languages in the past and present). Lund: Bokmål Nynorsk Studentlitteratur. ‘one’ ett et et eit Haugen, Einar. 1976. The Scandinavian languages: an intro- duction to their history. London: Faber & Faber Ltd. ‘eye’ öga øje øye auga Klintborg, Staffan. 1999. The transience of American Swedish, ‘weak’ svag blød bløt blaut Lund Studies in English, 98. Lund: Lund University Press. ‘week’ vecka uge uke veke Molde, Bertil, and Allan Karker (eds.) 1983. Språkene i Norden ‘I’ jag jeg jeg eg (The languages in the Nordic countries). Stockholm: ‘she’ hon hun hun ho Esselte, Copenhagen: Gyldendal and Oslo: Cappelen. ‘we’ vi vi vi vi, me Torp, Arne. 1982. Norsk og nordisk før og nå (Norwegian and ‘who’ vem hvem hvem kven Nordic languages in the past and present). Oslo: ‘behind’ bakom bagved bakom attom Universitetsforlaget. ‘not’ inte ikke ikke ikkje Vikør, Lars. 1993. The Nordic languages: their status and inter- ‘fifty’ femtio halvtres femti femti relations. Oslo: Novus Press. ‘swear’ svära bande banne banne ‘count’ räkna tælle telle telje JARMO LAINIO ‘window’ fönster vindue vindu vindu, See also Celtic Languages; Finnish and Finnic Lan- vindauge guages; German; Old English; Scandinavia Sweet, Henry There were three significant phonetician-linguists in He read Classics (‘Greats’) at Oxford, but he was Britain in the nineteenth century: Alexander Melville hardly assiduous, spending most of his time working, Bell (1819–1905), Alexander John Ellis (1814– virtually alone, in the field of Germanic languages. At 1890), and Henry Sweet (1845–1912). Bell is the time of his graduation, he had published not only remembered primarily for his ‘Visible Speech’ alpha- the Pastoral Care, but had critically reviewed a num- bet and the theory of general phonetics that underlies ber of scholarly works in the academic press, and had it; Ellis for his studies of English pronunciation, both read three papers to the Philological Society. Within a synchronic and diachronic; and Sweet for his erudite year of graduating, he published his History of English presentation, in a consistently structured and intelli- sounds (1874). gible manner, of material on phonetics and various Throughout his life, he maintained a wide spectrum forms of medieval English. Of the three, it is Sweet of academic interests, including phonetics, spelling who has exerted the strongest influence over the reform, shorthand, grammar, the teaching and learning direction that the subjects have taken over the past of languages, general linguistics, the history of English 130 or so years. and other Germanic languages, and literature. Since he Sweet’s interests in the older Germanic languages had to fulfill two roles, that of the scholar and, second- began relatively early. One of his schoolteachers (and ly, the private teacher whose books for undergraduates an Anglo-Saxonist), Thomas Oswald Cockayne and other learners would generate a certain amount of (1807–1873), may have been instrumental in introduc- income, he adopted the general strategy of publishing ing him to the two languages. On leaving school, he advanced material first, then simplifying some of it in undertook a year’s study of Comparative and works written intentionally for the student learner. Germanic Philology at Heidelberg University. A con- Meetings of the Philological Society in London sequence was that he began to collect materials during were the locus for the exposition of many of his views his teenage years for what was later to become his on language and languages, and he read papers there Student’s dictionary of Anglo-Saxon (1897). While he on a variety of subjects, especially the contemporary was still an undergraduate at Oxford, his edition of forms of a number of languages, including Danish, King Alfred’s Pastoral Care was published. German, Icelandic, Irish Gaelic, and Swedish. In his 1056  SWEET, HENRY later years, his field of interest widened to take comparative philology, with particular reference to the account of aspects of the linguistic structure of Arabic, Neo-Grammarian thesis on language change. Chinese, Norwegian, Portuguese, Sanskrit, and Welsh. His Handbook of phonetics (1877) inspired a whole Biography tradition of phonetic studies, especially in Britain and Europe. It expounds a general phonetic theory, illus- Henry Sweet was born in London on September 15, trated by phonetic transcriptions of various languages, 1845. He was educated at private schools, but primari- and concludes with an excursus on the nature of pho- ly at King’s College School, London. He received his netic notational systems, including a pivotal discus- B.A. in 1873 from Oxford University, Ph.D. (Hon sion of Broad and Narrow Romic notations— Causa) from Heidelberg University in 1875; and equivalent in most respects to phonemic and allophon- LL.D. (Hon Causa) from Glasgow University in 1890. ic notations. Two works on phonetics for the student He studied Comparative and Germanic Philology at learner, the Primer of spoken English (1890) and the Heidelberg University, 1863–1864. He was an under- Primer of phonetics (1890), present a less detailed graduate student at Balliol College, Oxford, account of phonetic theory. The text of the Primer of 1869–1873. He obtained his only full-time teaching spoken English is essentially a translation of another post, Reader in Phonetics, at Oxford University in of Sweet’s influential works, his Elementarbuch des 1901. He was President of the International Phonetic gesprochenen Englisch (1885), written for German Association from 1887 until his death. He was learners of English at a time when radical changes Corresponding Member of the Munich Academy of were under way in European schools and universities Sciences, and of the Royal Prussian Academy of in the teaching and learning of modern languages. A Sciences, and an Ordinary Member of the Royal further work on English phonetics, the Sounds of Danish Academy. Sweet died in Oxford on April 30, English, appeared in 1908. The practical study of lan- 1912. The character of Henry Higgins in George guages (1899) discusses in detail the theories and Bernard Shaw’s (1856–1950) Pygmalion is based to practices of teaching and learning languages. some extent on that of Sweet. Sweet’s ideas about grammar and grammatical analysis developed relatively slowly. His New English References grammar (1892, 1898) deals with the contemporary Henderson, E.J.A. (ed.) 1971. The indispensable foundation. A and historical forms of the language. Two simplified selection from the writings of Henry Sweet. London: Oxford accounts, based on the work, are his Short historical University Press. English grammar (1892) and the Primer of historical MacMahon, M.K.C. 2001. Modern language instruction and grammar (1902). phonetics in the later nineteenth century. Geschichte der Sprachwissenschaften, ed. by S. Auroux et al. Berlin and The edition of the Epinal glossary (1883), together New York: Walter de Gruyter. with a companion volume, The oldest English texts Sweet, H. 1871. King Alfred’s West-Saxon version of Gregory’s (1886) and King Alfred’s Orosius (1883), provided pastoral care. With an English translation, the Latin text, notes, scholars with the necessary tools for advancing the and an introduction. London: Trübner & Co. [published 1872]. critical study of the earliest periods of Old English. ———. 1874. A history of English sounds from the Earliest Period, including an investigation of the general laws of The series of works on earlier forms of English, writ- sound change, and full word-lists. London: Trübner & Co. ten expressly for the novice learner rather than the ———. 1876. An Anglo-Saxon reader in prose and verse, with scholar, began with the Anglo-Saxon reader of 1876, grammatical introduction, notes, and glossary. Oxford: followed by a less demanding account, the Anglo-Saxon Clarendon Press. primer in 1882. Three further intermediate textbooks ———. 1877. A handbook of phonetics, including a popular exposition of the principles of spelling reform. Oxford: for students of Old English were Ælfric’s Homilies Clarendon Press. (1885), Extracts from Alfred’s Orosius (1885), and the ———. 1882. An Anglo-Saxon primer with grammar, notes Second Anglo-Saxon reader (1887). For students and glossary. Oxford: Clarendon Press. embarking on the study of Middle English, there were ———. 1883. The Epinal glossary, Latin and Old-English of two Middle English primers (1884, 1886). the eighth century. Photo-lithographed from the original MS … with transliteration, introduction, and notes. London: His Icelandic Primer (1886) is a classic example of Trübner & Co. his ability to present the essential facts of a medieval ———. 1885. Elementarbuch des gesprochenen Englisch language in a form that is immediately attractive to the Grammatik, Texte und Glossar. Oxford: Clarendon Press. learner. The essential linguistic information is ———. 1886. An Icelandic primer, with grammar, notes and described synchronically, with no diachronic or cog- glossary. Oxford: Clarendon Press. ———. 1886. The oldest English texts, edited with introduc- nate Germanic materials to distract the learner. Sweet’s tions and a glossary. London: Trübner & Co. History of language (1900), written for the general ———. 1888. A history of English sounds from the Earliest reader, is an exposition of the principles that underlie Period, with full word-lists. Oxford: Clarendon Press. 1057 SWEET, HENRY ———. 1890. A primer of phonetics. Oxford: Clarendon Press. ———. 1899. The practical study of languages. A guide for ———. 1890. A primer of spoken English, Oxford: Clarendon teachers and learners. London: J.M. Dent & Co. Press. ———. 1900. The history of language. An introduction to the ———. 1892/1898. A new English grammar, logical and his- principles of comparative philology. London: J.M. Dent & Co. torical. Part I: Introduction, phonology, and accidence ———. 1908. The sounds of English. An introduction to pho- (1892); Part II: Syntax (1898). Oxford: Clarendon Press. netics. Oxford: Clarendon Press. ———. 1897. The student’s dictionary of Anglo-Saxon. Wrenn, C.L. 1946. Henry Sweet. Trans Philological Society. Oxford: Clarendon Press. MICHAEL K.C. MACMAHON Switch-Reference Switch-reference (first described by William Jacobsen (Chickasaw data are from Munro, P. ‘When ‘Same’ is in ‘Switch-reference in Hokan-Coahuiltecan’, 1967) is not ‘Not Different’, 1983: 223) one of several subject-linked reference tracking mech- In (1a) and (2a), the underlined morphemes labeled anisms that apply outside the domain of a single ‘SS’ show that the subject of the marking clause has the clause. Switch-reference markers indicate whether the same referent as the subject of the reference clause; in referent of an argument (typically the subject) of one (1b) and (2b), the morphemes labeled ‘DS’ show that clause is the same as or different from the referent of the subjects of the marking clause and the reference an argument (typically the subject) of another clause. clause have different referents. Switch-reference systems are found in a wide range of Switch-reference marking obligatorily tracks the languages in North America, Australia, Papua New reference of subjects across clauses within a sentence, Guinea, North Asia, and Africa. even when other morphology makes the switch-refer- ence marking redundant. For example, the agreement Typical Switch-Reference Systems prefixes on the verbs in (1b) show that the marking clause has a second-person subject while the reference In a typical switch-reference system, the verb of one clause has a first-person subject; but a different-sub- clause, the marking clause, is marked to indicate ject suffix is still required on the verb of the marking whether or not its subject has the same referent as the clause. Similarly, the agreement prefixes on the verbs subject of another clause, the reference clause. The in (1a) show that both clauses have first-person sub- marking clause can be subordinate to the reference jects, but the same-subject suffix is still required. clause, as in the Maricopa examples (1ab) or coordi- More complex examples show that the switch-ref- nate with the reference clause as in the Chickasaw erence marking is sensitive to the syntactic subject, examples (2ab). Languages limit what syntactic rela- rather than to a logical or underlying subject or agent. tionships hold between marking and reference clauses: When one clause is subordinate to the other, the in some languages, switch-reference is used across a marking clause is always the subordinate one, wide range of coordination and subordination relation- regardless of linear order. For example, in Maricopa, ships, while in others it is used in more limited struc- a subordinate marking clause can be initial, final, or tures. Switch-reference applies only between clauses embedded in the middle of its reference clause. that combine to form a single syntactic sentence. (1a) paan nya-m-chew-k m-maa-uum Typical Variation in Switch-Reference bread when-you-make-ss you-eat-INCOMPLETE Systems ‘When you make the bread, you’ll eat it’ (1b) paan nya-m-chew-m ‘-maa-uum Individual languages deal differently with a number bread when-you-make-ds I-eat-INCOMPLETE of predictable structures, including (1) the assignment ‘When you make the bread, I’ll eat it’ of switch-reference marking when the referents of (2a) hi’lha-cha talowa the subjects of the relevant clauses are overlapping, dance-COORDINATE  ss sing rather than entirely different or completely identical; ‘He danced and sang’ (2) the assignment of switch-reference marking when (2b) hi’lha-na talowa the subject of one or both of the clauses does not have dance-COORDINATE  DS sing a referent; (3) the combination of switch-reference ‘He danced and he (someone else) sang’ markers with other functional markers (for example, 1058  SWITZERLAND in Chickasaw (2a), -cha marks both coordinate and Analyses same subject); and (4) the use of switch-reference- There are two basic approaches to switch-reference: marked verbs that appear to have a primarily dis- syntactic/semantic and discourse/pragmatic. course function as transitions between sentences or paragraphs. Syntactic/semantic approaches, like Daniel Finer’s 1985 The formal grammar of switch-reference and Uncommon Variants in Switch-Reference George Broadwell’s 1997 article in Atomism and bind- Systems ing, analyze the assignment or interpretation of switch- reference markers as being controlled by the syntactic In some languages (e.g. Pima), switch-reference mark- structure of the sentence. These accounts take the ers are not verbal affixes, but independent particles. In assignment or interpretation of reference of subjects as other languages, the system is extended from subjects basic to the structure, with other functions derivative to include other arguments. For example, Peter Austin from this primary use. (in ‘Switch-reference in Australia’ in Language 57 (1981):325) describes Warlpiri as having a three-way Discourse/pragmatic approaches, like David Rising’s contrast: if the subordinate clause subject is coreferen- 1992 Switch reference in Koasati discourse and Yan tial with the subject of the main clause, the subordinate Huang’s 2000 Anaphora: A cross-linguistic study, clause verb is marked with-karra; if it is coreferential assume that the primary use of switch-reference mark- with the object or dative of the main clause, the mark- ers is to mark topic or other discourse continuity and er is -kurra; otherwise the marker is -ngkarni/-rlarni. the marking of subject identity or lack of identity is In still other languages, the switch-reference system is derivative. restricted to a subset of persons, or interacts with a person hierarchy as in Quechua. References Austin, Peter. 1981. Switch-reference in Australia. Language 57. Other Reference Tracking Systems Broadwell, George A. 1997. Binding theory and switch-refer- ence. Atomism and binding, ed. by Hans Bennis, Pierre Switch-reference must be distinguished from a num- Pica, and Johan Rooryck. Dordrecht: Foris Publications. ber of other systems used by languages to indicate Finer, Daniel. 1985. The formal grammar of switch-reference. identity or lack of identity of reference of arguments New York and London: Garland Publishing, Inc. across clauses. For example, a number of African lan- Haiman, John, and Pamela Munro (eds.) 1983. Switch-refer- ence and universal grammar. Amsterdam: John Benjamins. guages have a ‘logophoric’ pronoun or verbal affix, Huang, Yan. 2000. Anaphora: a cross-linguistic study, Oxford which appears in clauses that report someone’s ideas, Studies in Typology and Linguistic Theory. Oxford: Oxford words, or feelings. The use of the logophoric pronoun University Press. in the reported clause indicates that its referent is the Jacobsen Jr., William. 1967. Switch-reference in Hokan- same as the individual whose ideas, words, or feelings Coahuiltecan. Studies in Southwestern ethnolinguistics: meaning and history in the languages of the American are being reported. The logophoric form and the argu- Southwest, ed. by Dell H. Hymes and William E. Bittle. The ment it refers back to are not necessarily limited to the Hague and Paris: Mouton & Co. subject position and it is limited to reported contexts. Munro, Pamela. 1983. When ‘same’ is not ‘not different’. In Switch-reference, in contrast, seems to be much more Haiman and Munro. strictly linked to syntactic structure and much less Rising, David P. 1992. Switch reference in Koasati discourse. Dallas: Summer Institute of Linguistics and the University linked to particular semantic or discourse contexts; of Texas at Arlington. unlike logophoric forms, switch-reference markers are Stirling, Leslie. 1993. Switch-reference and discourse represen- never pronouns and typically are not part of the ordi- tation theory. Cambridge: Cambridge University Press. nary verb-agreement system. LYNN GORDON Switzerland The Swiss Confederation consists of 26 cantons small country has a population of 6,873,867 speaking (states, provinces) within an area of only 41,284 four languages: German, French, and Italian, which square kilometers. According to the 1990 census, this are shared with its neighbouring countries, and 1059 SWITZERLAND Romansh, a distinct Romance language, that is, one Territories of the four languages of Switzerland (1: descended from Latin and related to French, Italian, French; 2: German; 3. Italian; and 4: Romansh): Spanish, and others. This linguistic mix is the result of the historical processes leading to the composition of modern Switzerland. The original confederation was an oath of eternal allegiance taken between four German-speaking cantons in the mountainous centre 2 of the country, occurring in 1291. The German lan- guage had been brought to this area by the Alemannic 1 invaders, part of the expansion of Germanic peoples 4 into the old domains of the Roman Empire from the 3 fifth century onward. After 1291, the confederation expanded, adding more cantons, reaching its present composition only at the beginning of the nineteenth century. While some of these cantons were, like the The patterns of language use in communication original four, German speaking, those in the west among the language communities of Switzerland vary spoke French, while the incorporation of the southern widely, with the Italian and Romansh speakers far areas brought in Italian and Romansh. The linguistic more likely to adapt to their compatriots from the larg- boundaries shown on the map below are in fact older er communities. Indeed, English—a popular choice as than the state boundary of Switzerland. Even the the third school language—is now widely used as a increasing Germanisation of the Romansh area is a lingua franca. The situation is exacerbated by the per- process that began well before the area’s incorporation ceived dominant position of the German-speaking into the Swiss Confederation. community, and by the existence of diglossia in the All four of these languages are defined as ‘nation- German-speaking area. al languages’ in the federal constitution, and since 1996, when Romansh was added to the other three— Diglossia, as originally defined, is the existence of as ‘official languages’ of the confederation. Multiple two distinct forms of the same language within the same national languages are unusual, especially for Europe, community, each with its own functions. Ferguson’s and Switzerland has been cited as a possible model (1959) original article on this phenomenon uses for the European Union with respect to the accommo- German-speaking Switzerland as one of the canonical dation of several official languages. Although examples. In this area, Swiss Standard German, which Switzerland is at a national level quadrilingual, it is is very similar but not identical to Standard German, is not the case that the four languages are evenly dis- used for most written functions and such highly formal tributed through the population, or that all areas of spoken genres as university lectures, while local Switzerland are multilingual. In practice, despite dialects, known collectively as Schwyzertüütsch, are much internal migration and the large foreign popula- used for most spoken purposes. These dialects are so tion in Switzerland, most areas of the country are different from Standard German as to be practically essentially monolingual. Each of the three major lan- mutually unintelligible with it. The differences are pri- guages has its territory in which it functions as the marily due to varied histories of their sound and gram- single language of public life. The Swiss school sys- mar systems, but there are also lexical items unique to tem aims to provide all Swiss with some knowledge the Swiss dialects. The table below presents a few of a second national language; however, this goal is examples of these characteristic phonological and lexi- by no means fully achieved. The following statistics cal distinctions: and map show the numerical and geographic distribu- tion of the country’s four languages. English Zurich German Standard German Numbers and percentages of Swiss population who newspaper Zytig Zeitung nominated one of the national languages as their main house Huus Haus to take out uusenee herausnehmen language in the 1990 census: let’s go home Gömer häi Gehen wir nach German: 4,374,694 63.6% Hause listen lose zuhören French: 1,321,695 19.2% one, two, three, äis, zwoo, drüü, eins, zwei, drei, Italian: 524,116 7.6% four, five vier, föif vier, fünf Romansh: 39,632 0.57% I was there I bi dëët gsy ich war dort 1060  SWITZERLAND There is widespread reluctance among German- census, held every ten years, and Romansh’s territo- speaking Swiss, and in many cases inability, to use ry has been shrinking under pressure from German Standard German easily for spoken purposes, espe- for many centuries. Virtually all speakers of cially for informal ones; yet, Standard German is what Romansh are fluent in German, often in addition to other Swiss learn at school. other languages as well. There have been and contin- Historically, the French and Italian areas of ue to be, however, strenuous efforts to ensure the sur- Switzerland had local dialects, but the situation in vival of Romansh. both these areas differs from that found in German- The precarious state of the language is not helped speaking Switzerland. The local dialects of French by the existence of the five written versions. In 1982, belong to the Franco-Provençal group, a family of a unified written form of the language was introduced dialects sharing language attributes with both north- greatly facilitating the publication of official materials ern French (to which Standard French belongs histor- in Romansh. Even works of literature are now being ically) and the Occitan languages of the south of written in this standard. Founded in 1919 as an France. However, to all intents and purposes, these umbrella organization for the growing number of soci- dialects are now lost in Switzerland. With the excep- eties and groups founded in the previous century to tion of a few very old people, everyone in the French- promote and protect Romansh, the Lia Rumantscha, speaking parts of Switzerland speaks Standard with its seat in Chur, is the main organisation coordi- French. Standard French is not identical to Parisian nating and driving the work of language maintenance French; rather, it incorporates a few older features that and language planning for Romansh. Although not a have been replaced in Parisian French, as demonstrat- statutory body, the Lia Rumantscha is financed by ed in the table below: very generous grants from both federal and cantonal governments. Its activities include translation services, English Swiss Standard Parisian terminology creation and definition, publication, and French French language promotion. In those municipalities with a he must be il faut lui aider il faut l’aider majority of Romansh speakers, Romansh is used as the helped language of instruction at junior levels of school, and seventy septante soixante-dix is available as a subject at secondary and even tertiary (sixty plus ten) levels of education. eighty huitante quatre-vingts The linguistic situation in Switzerland is complicat- (four twenties) ed, and is a matter of intense discussion within the ninety nonante quatre-vingt-dix country itself, not least because of the perception that (four twenties plus ten) the linguistic communities of Switzerland do not so much live together as past one another. Pride in this The local dialects of Italian are still spoken, but until ‘diversity in unity’, however, is an important feature of recently had low status. Standard Italian was regarded Swiss identity. as the language of prestige. In recent decades, howev- er, the Swiss dialects of Italian have gained in usage and prominence as a mark of local identity. References Romansh has been spoken continuously in the Clyne, Michael. 1995. The German language in a changing southeast portions of Switzerland since Roman Europe. Cambridge and New York: Cambridge University times. Linguistically, it shares features with both Press. northern Italian dialects and French. Like other lan- Dürmüller, Urs. 1997. Changing patterns of multilingualism: guages of Switzerland, particularly German, it exists from quadrilingual to multilingual Switzerland. Zurich: Pro Helvetia. as a chain of dialects that have been standardised in Duval-Valentin, Marianne. 1983. La situation linguistique en various forms since the time of the Reformation in Suisse. Language reform: history and future, Vol. 1, ed. by the sixteenth century. Of these five written forms, István Fodor and Claude Hagège, Hamburg: Buske. those of the upper Rhine valley and the lower Ferguson, C.A. 1959. Diglossia. Word 15. 325–44. Engadine (the left- and right-hand ends of area num- Watts, Richard J. 1991. Linguistic minorities and language con- flict in Europe: learning from the Swiss experience. A lan- ber 4 on the map) have enjoyed particular support. guage policy for the European community, ed. by Florian However, despite its long history and its official sta- Coulmas. Berlin and New York: Mouton. tus at both federal and cantonal levels, Romansh is The activities of the Lia rumantscha are described at <http:// endangered. A steadily decreasing proportion of www.liarumantscha.ch>. Swiss report Romansh as their first language in the RAY HARLOW 1061 SYLLABLE: STRUCTURE Syllable: Structure The syllable is defined as a unit of cognitive organiza- weight of a syllable (see, for example, Hayes 1980). tion, comprised of a ‘sonority peak’. Sonority is Thus, if stress is attracted to ‘heavy’ syllables in a lan- traditionally defined as intrinsic loudness (Chomsky guage only material from the rhyme can contribute to and Halle 1968). It is rather unclear as to what sonor- this weight. An example of such a language is Creek ity might translate into in verifiable phonetic terms, (Haas 1977). Accent falls on the final syllable or the but it is generally accepted that the major phonologi- second syllable from the right, whichever is an even cal classes can be ranked from high to low sonority as number of syllables from the left, except that a heavy follows: vowels, glides, liquids, nasals, and obstruents. syllable restarts the count. Following researchers such as Itô, (1989), the seg- Os'ana ‘otter’ ments of a word are exhaustively parsed into syllables. hitot'i ‘snow’ This is termed ‘syllabic licensing’. For example, 'icki ‘mother’ marking syllable boundaries with parentheses, a word like system is syllabified as follows: [(sIs)(təm)]. Evidence for the nucleus–coda division is a little hard- It is a question of some interest whether syllables er to come by, but has also been adduced from stress sys- are further divided into smaller constituents above the tems. There are a number of languages where stress is segment level. Many researchers assume that a sylla- attracted to syllables where only the nucleus contributes ble is subdivided into onset and rhyme constituents, to syllable weight. For example, in Khalkha Mongolian where the rhyme constituent is perhaps further subdi- (Street 1963), the leftmost long vowel is stressed. If there vided into a nucleus and coda. The onset is composed is no long vowel, the first syllable is stressed. of all the segments that precede the peak of the sylla- bos'uul ‘fugitive’ ble. The nucleus is composed of all the vocalic mate- bari'aad ‘after holdin’ rial of the peak; all the material to the right of the 'ali ‘which’ nucleus is in the coda. An English syllable like trounce xoyərdu'aar ‘second’ [trawns] would have the following structure: x'ötəlbərə ‘leadership’ σ ar'aasaa ‘from one’s own hand’ O R Closed syllables are irrelevant to the computation of stress in Khalkha. t r N C While terms like onset, rhyme, nucleus, and coda are still quite widely used, the world of subsyllabic constituency was thrown into question with the publi- a w n s cation of Hayes (1989). Hayes argued against these The subsyllabic constituents presented are abbreviated notions and for a moraic characterization of the sylla- in the obvious fashion and a Greek sigma ‘σ’ is used ble. The basic idea—according to Hayes—is that there for a syllable. are no subsyllabic constituents. Instead, there is only Evidence for the onset-rhyme division comes from the mora. The mora is a unit of syllabic quantity. Light two main sources: phonotactics and stress. First, in syllables have one mora; heavy syllables have two (or many languages, while there are a number of restric- more) moras. If we assume that English syllables have tions on what consonants can occur together in an onset, a maximum of two moras, we might represent the there are very few restrictions on what segments can co- English word trounce as follows: occur together across the onset-rhyme division. In σ English, for example, the only known restriction that plausibly spans the onset-rhyme divide holds that (in many dialects) coronal consonants cannot be followed t r µ µ by the diphthong [ju], e.g. beauty [bjuɾi] vs. duty [duɾi], *[djuɾi], etc. (see, for example, Hammond 1999). a w n s A second traditional argument for this division comes from stress. It is generally thought that only Onset consonants are direct dependents of the syllable material from the rhyme of a syllable contributes to the node. Elements that we would have formerly put in the 1062  SYNTACTIC CATEGORIES rhyme or nucleus are now docked to mora nodes The key generalization about Latin is that only a (abbreviated as Greek mu ‘µ’). The number of moras coda consonant can trigger compensatory lengthening. is a function of how weight is assigned in the language If an onset consonant deletes, there is no compensa- in question. The peak of the syllable is generally tory lengthening. If we had an onset constituent, this thought to be always moraic. Long vowels are always would make the wrong prediction. A deleted onset bimoraic. Coda consonants may or may not generate might leave some sort of slot for a neighboring vowel their own mora. (There are controversial aspects of the to lengthen with. This characterization of syllabic proposal. For example, onset consonants always attach weight accounts for several things. First, it directly to the σ, or can they dock to the first mora? Can there describes what can contribute to stress attraction. be moraless syllables? Do moraless coda consonants Second, it accounts for why compensatory lengthen- attach to the final mora or directly to the σ?) ing never results from loss of an onset consonant. The moraic theory of the syllable accounts directly With the advent of ‘Optimality Theory’, subsyllab- for quantity–sensitivity in stress systems. If stress is ic constituency has come back into vogue. In their attracted to closed syllables, it is because coda conso- presentation of an optimality–theoretic theory of the nants are moraic in the language in question. If stress syllable, Prince and Smolensky (1983) used the terms is attracted to long vowels, but not closed syllables, it onset, peak, and coda. This has led a number of is because coda consonants are not moraic in the lan- researchers to reintroduce subsyllabic constituency guage in question. into syllabification theory, although Prince and Where the moraic theory makes new predictions is its Smolensky acknowledge that they use those terms rejection of any sort of onset constituent. Hayes argues only as a convenience. It is an unresolved empirical that this is required to account for compensatory length- question as to whether Optimality Theory can be ening, the phenomenon whereby a vowel lengthens directly reformalized without these subsyllabic con- when a neighboring consonant is deleted. For example, stituents. Hayes cites Latin, where an [s] is deleted before a coro- nal sonorant, triggering lengthening just in case the [s] References was a coda. Thus *[kasnus] ‘gray’ is realized as [kanus] Chomsky, Noam, and Morris Halle. 1968. The sound pattern of while *[snurus] ‘daughter-in-law’ is realized as [nurus]. English. New York: Harper & Row. Under moraic theory, this has a straightforward Haas, Mary. 1977. Tonal accent in Creek. Studies in stress analysis. The [s] of *[ kasnus] is syllabified as a moraic and accent, Southern California Occasional Papers in coda of the first syllable. It is then deleted, leaving the Linguistics 4. Hammond, Michael. 1999. The phonology of English. Oxford: mora behind. The vowel then spreads to the second Oxford University Press. mora, resulting in a long vowel. Hayes, Bruce. 1985. A metrical theory of stress rules. Ph.D. dissertation, Massachusetts Institute of Technology, 1980. σ σ σ σ σ σ New York: Garland. µ µ µ µ µ µ µ µ µ µ µ µ Hayes, Bruce. 1989. Compensatory lengthening in moraic k a s n u s k a n u s k a n u s phonology. Linguistic Inquiry 20 Itô, Junko. 1989. A prosodic theory of epenthesis. Natural Contrast this with the case of *[snurus], where dele- Language and Linguistic Theory 7. tion of the [s] leaves no stranded mora. Prince, Alan, and Paul Smolensky. 1993. Optimality theory, ms., Brandeis University and University of Colorado. σ σ σ σ Street, J.C. 1963. Khalkha structure, Bloomington: Indiana µ µ µ µ µ µ University Publications. s n u r u s n u r u s MICHAEL HAMMOND Syntactic Categories The words and phrases that make up utterances can be term is also used with reference to classes of words classified as belonging to certain syntactic categories. and word roots as part of the vocabulary. These two The criteria for this classification are their function senses are often not distinguished. within the clause or phrase, their form, and their dis- A more traditional term that is often used synony- tribution, i.e. what other elements they occur with. The mously is ‘part of speech’. In traditional grammar, the 1063 SYNTACTIC CATEGORIES following parts of speech were distinguished: verbs, not possess this kind of form-class. If the term noun nouns, pronouns, adjectives, adverbs, determiners (arti- were used in the description of this language, it would cles), prepositions, and conjunctions. This classification mean something different from what it means in goes back to the ancient Greek and Latin grammarians English. For this reason, among others, the traditional and has been a cornerstone of grammatical description part of speech classification was never completely aban- for 2,000 years. It also plays an important role in most doned, and it was revived in generative grammar, which modern theories of grammatical structure, although otherwise strove for formal rigor like its predecessor, there have been many modifications on the status, the the structuralistic approach. number, and the definition of these categories. Both the traditional parts of speech and the form- In traditional grammar, the existence of the above- classes of structuralist descriptions are primarily con- mentioned categories was taken for granted. They cerned with the classification of words. They differ in were axioms in the model of description and served as what aspect of the words they were most interested in. the background for the definition of other categories The object of structuralist form-classes are words and for the analysis of sentences. It was assumed that as they appear in sentences, units that can be obtained all languages could be described by using the same set by segmenting utterances into parts and classifying of categories, although the actual forms and functions these parts. With this approach, the forms goes, gone, of nouns, verbs, adjectives, etc., could be different. A and going belong to different form-classes because similar view is taken within the framework of genera- they have different distributions and occupy different tive grammar, in which syntactic categories are used to slots within an utterance. Traditional grammar, on the define phrasal categories and remain undefined them- other hand, mainly dealt with lexemes, or abstract selves. It is often assumed that the distinction between units of the lexicon that can be realized by one or sev- nouns and verbs is universal. Furthermore, it is used to eral word-forms. Here, the lexeme GO, which is real- characterize other categories; i.e. the four major cate- ized by the word-forms go, goes, went, gone, and gories are distinguished with respect to how much they going is classified only once, namely, as a verb. function like a typical noun or verb. In this view, verbs The difference between classes of lexemes with are characterized as verbal and nonnominal, and nouns their potential forms and functions, and classes of are nominal and nonverbal. Adjectives are described word forms fulfilling a given function in a sentence, as having both verbal and nominal properties and has long been ignored. It was obscured by the habit of prepositions as having neither. This approach allows using the same label for both kinds of categories. For general statements about categories that include sever- example, the label ‘verb’ is used both for classes of al parts of speech. For example, in some languages, lexemes like GO and for word-forms used as predi- verbs and prepositions, the nonnominal categories, cates; in the clause don’t madam me! (meaning ‘don’t share characteristics that other categories do not have. call me madam’), the word madam is a verb in the lat- A different view was held by the early American ter sense, but not in the former. structuralists, who challenged traditional grammar in There are, of course, very obvious links between the second quarter of the twentieth century. Instead of the two kinds of classes, and these links were particu- assuming universally valid axiomatic categories, they larly strong in the Indo-European languages that had maintained that syntactic categories, or form-classes, been the model languages for linguistic description for should be established on language-particular grounds, a long time. However, the more linguists explored lan- that is, on the basis of the formal properties and the dis- guages of very different structure, the more it became tribution of elements of a particular language. A tool for clear that lexical classes (classes of lexemes) and syn- classifying linguistic elements was substitution frames, tactic word classes have to be established independ- e.g. ‘the ___ is good’ for nouns in English. All elements ently of each other. How the two are connected in that could fill the slot in such a frame belong to the same various languages has been an important topic in the distribution class. A careful analysis of the possible sub- field of language typology in the 1990s. The distinc- stitution frames of words leads to their classification tion between the category, or type, of a lexeme and the into form-classes. A grammatical description that strict- syntactic category of a word-form has also been inte- ly follows this view comes up with far more classes than grated into some recent formal theories of syntax. the traditional eight parts of speech. Also, these classes In a system with a limited number of classes (such were language particular and could not be used for lan- as the traditional part-of-speech system), there will guage comparison. For example, English count nouns always be items that display features of more than one could be described as a form-class of elements that may class or that do not fit clearly into one of the classes. appear together with an article and combine with a plu- This problem can be solved by assuming that classes ral marker. Obviously, another language, which lacks are not homogeneous but have an internal structure, articles or has other means of marking plurality, could such that there is a center containing the elements 1064  SYNTACTIC CATEGORIES that meet all criteria typical for the category and a word-forms that are possible and necessary in each periphery with elements that share only some of these function. Hence, in many languages, nouns cannot features. This conception of categorization became make e.g. tense distinctions, as verbs can, and to have popular in linguistics in the 1980s, when the concept a noun as predicate one often has to use a special of prototype was taken over from psychology. Within device, such as the copula verb in English (Jane was a this view, categories are also allowed to overlap; i.e. girl, not Jane girl-ed). words may belong to more than one class. A language can be said to lack a lexical distinction Another solution lies in categorizing lexemes more of parts of speech if any content word can be used in than once. Because lexemes have several characteris- any of the major syntactic constructions and may be tics that distinguish them and make them similar to combined with the whole range of grammatical mark- other lexemes, several parallel classifications can be ers. Whether such languages exist is still a matter of established, just as we classify languages according to debate. Certainly, languages differ in how sharp the their historical origin as well as their structural char- distinction between lexical classes is drawn, i.e. how acteristics or their sociolinguistic status. The idea of important it is for the grammar of the language and multiple classification of words has entered linguistic whether the elements of a class are more or less spe- theory only recently and will probably be further elab- cialized for syntactic functions. Latin is an example of orated in the future. a language with clear-cut lexical classes. In English, in The concept of prototype has also been used to solve contrast, there are many pairs like to look/a look or to the old dilemma of universal vs. language-particular move/a move. Some linguists regard these pairs as two categorization. For example, many linguists have been separate lexemes belonging to different classes, where- concerned with the question of whether all languages as others treat them as single multifunctional lexemes. can be said to distinguish nouns and verbs. To tackle Whereas the link between a lexical class and a syn- this question, a general definition of these categories is tactic function is only indirect and often rather weak, needed to ensure that the classes labeled ‘noun’ and matters are different for syntactically determined class- ‘verb’ are indeed categories of the same kind in differ- es of word-forms. Syntactic theories therefore have ent languages. One option is to define the center, or been more concerned with the latter, and the term syn- prototype, of the major lexical categories (nouns, tactic category is nowadays used in a sense that is only verbs, and adjectives) on the grounds of criteria related remotely connected to the traditional parts of speech. It to meaning and usage. Hence, prototypical nouns are covers phrases besides words and also more abstract words that designate objects, or time-stable entities, units that may lack a form, so-called empty categories. and are used for reference to discourse participants, i.e. A syntactic category in this sense can be described what or who is talked about. Prototypical verbs are by a set of features with values specifying the slot in words that designate actions and are used for the act of which items belonging to this category can occur. For predication, and prototypical adjectives designate prop- example, in the clause she is a linguist, the category of erties and are used for modification. Within a particular the word-form is will be described by the features ‘third language, one of these categories can be identified if person’, ‘singular’, and ‘present tense’. As mentioned words meeting the general definition are somehow dis- above, the features ‘nominal’ and ‘verbal’ had been tinguished from other words, i.e. are reasonably con- derived from the categories noun and verb. Developing sidered a lexical class by language-internal criteria this idea further, it becomes possible to characterize a such as grammatical inflections and syntactic function. syntactic category entirely by features. In the above Although the center of each category is determined by example, the syntactic category of linguist would thus the general definition, it is a matter of individual lan- be the set of the following features: {nominal, nonver- guages as to what further elements the class contains bal, singular}. Most contemporary syntactic theories and how classes are delimited. For example, although make use of such features but differ in the role attributed predicative action words (such as run, push, eat, sell) to them and their relation to syntactic categories. will be verbs in any language that distinguishes such a Another issue of current debate is the relationship lexical class, lexemes denoting states or emotions (like between syntactic categories and sentence structures. In love, fear, be happy, be hungry) may be verbs in one short, the question is which one comes first and should language and adjectives or nouns in another. be used to derive the other. This again is connected to The distinction may be complete, so that e.g. words the distinction between categories of the lexicon and that are identified as nouns cannot be used in the same categories of sentences and how these are related. When syntactic functions and environments as verbs or adjec- sentence structures are derived from word categories, it tives. However, we often find that lexemes are not that is supposed that lexemes or word roots are preclassified strictly specified but are used in several functions. In in the lexicon. For example, the lexicon of English will this case, they may be distinguished by the particular contain an item kiss (noun) and an item kiss (verb), the 1065 SYNTACTIC CATEGORIES former defining a noun phrase as in she gave him a big Huddleston, Rodney. 1988. English grammar: an outline. kiss, and the latter a verb phrase as in you may kiss the Cambridge: Cambridge University Press. Lefebvre, Claire, and Lisa Travis (eds.) 1999. A propos des caté- bride. According to another position, it is the respective gories syntaxiques.  Revue québécoise de linguistique 27-2. construction that makes kiss a noun in the first example Plank, Frans. 1997. Word classes in typology: recommended and a verb in the latter, whereas in the lexicon, the item reading. Linguistic Typology 1. kiss is not categorized. Radford, Andrew. 1997. Syntactic theory and the structure of English. A minimalist approach. Cambridge: Cambridge University Press. References Sag, Ivan A., and Thomas Wasow. 1999. Syntactic theory. A formal introduction. Stanford: CSLI Publications. Bhat, Shankara D.N. 1994. The adjectival category: criteria for Sasse, Hans-Jürgen. 1993. Syntactic categories and subcate- differentiation and identification. Amsterdam, Philadelphia: gories. Syntax, ed. by J. Jacobs, A. von Stechow, W. Benjamins. Sternefeld, and T. Vennemann. Berlin and New York: de Croft, William. 1991. Syntactic categories and grammatical Gruyter. relations: the cognitive organization of information. Vogel, Petra, and Bernhard Comrie (eds.) 2000. Approaches to Chicago: University of Chicago Press. the typology of word classes. Berlin: Mouton de Gruyter. Crystal, David. 1967. English. Lingua 17. NICOLE NAU Davis, Henry, and Lisa Matthewson. 1999. On the functional determination of lexical categories. Revue québécoise de See also Empty Categories; Function Words; Lexi- linguistique 27(2). con: Overview; Phrase Structure; Word Syntactic Typology Syntactic typology is one of the major branches of lin- stand for Subject, Verb, and Object, respectively), guistic typology, aiming to classify human languages there are logically six possible combinations: SVO, as into different types on the basis of their shared syntac- in English, French, Hausa, Vietnamese, and Thai; tic properties, which are not necessarily the result of SOV, as in Persian, Japanese, Korean, Amharic, common origin or geographical contact. Like other Tibetan, Turkish, and Georgian; VSO, as in Welsh, branches of linguistic typology, the ultimate goal of Tongan, Classical Hebrew, Classical Arabic, and such studies is centered on the provision of (syntactic) Tagalog; VOS, as in Malagasy, Tzotzil (Mexico), and universals: those qualities that tend to be common in Cakchiquel (Guatemala); OVS, as in Hixkaryana most, if not all, languages, and enable linguists to (Brazil) and Barasano (Colombia); and OSV, as in make hypotheses in their further studies of languages. Xavante (Brazil) and Apurina (Brazil). The development of syntactic typology owes a lot These orders are, of course, the ‘unmarked’ order of to the American linguist Joseph Greenberg, whose elements in languages, an order sometimes being called interest in typology and language universals led him to ‘canonical order’. Obviously, in languages with rich propose the theory of ‘implicational’ universals based inflectional systems and therefore ‘free word-order’ pat- on syntactic properties (see below). terns, the canonical order may be altered and changed. There are some criteria based on which languages However, these possible compositions are not equal can be typologized syntactically; some of the most with respect to their frequency. It has been proved that important ones are: over 95% of the world’s languages use one of the pat- terns SOV, SVO, and VSO as their basic word order, (1) Word order. This criterion is the widest known and object-initial languages are extremely rare and and the oldest way of classifying the languages syn- uncommon. tactically. It was so common in the 1960s that one Since not all languages express overt subjects, the could easily claim that syntactic typology is mainly relative order of verb and object is often considered to known to be represented through word-order studies, be the most important from the point of view of typol- i.e. how different lexical items are arranged in various ogy. The main interest in classifying languages in this languages, to produce syntactic universals. way lies in the ‘implicational’ relationships, in that we Studies have shown that, in a transitive clause believe certain word orders make it likely for other which mainly has the elements S, V, and O (which constructions to occur (more technically speaking: if 1066  SYNTACTIC TYPOLOGY X, then Y). For example, an SVO language, such as Agreement between subject and predicate, in terms English, is likely to have prepositions rather than post- of gender, number, or case; as in He is a waiter/She positions (such as up the tree rather than *the tree up), is a waitress. This may also show itself in structures and auxiliaries before main verbs (such as John may where a word refers back to something mentioned come rather than *John come may). Such a language is previously, as in I know Jane quite well, she was my also likely to have relative clauses (beginning with secretary for long. who/which, etc.) after the noun they refer to, such as In this case, a syntactic typologist tries to find out if The boy who broke the glass escaped rather than *The different verb affixes mark person, number, and gen- who broke the glass boy escaped. Similar implications der of subject, object, or other noun phrases, and if may also be provided for other word orders; for exam- such agreement is obligatory. ple in a VSO language, we have prepositions rather than postpositions, while in an SOV, postpositions are (3) Universals of negative position. Negation is a more likely to appear than prepositions. syntactic (or more technically, “morphosyntactic”) The general principle behind these observations process in which a lexical item denies or inverts the appears to be a preference for consistency in the posi- meaning of another lexical item or construction. The tion of the main word (or ‘head’) in any construction lexical item that expresses negation is called negator. with regard to the items attached to it (or the ‘modi- The position of a negator as a syntactic element is fiers’). Hence, an SVO language such as English is a therefore a criterion for syntactic typology; that is, if ‘head first’ language and an SOV language such as the negator comes before or after the verbs, or in pre- Persian is a ‘head last’ language. In this case, the lin- fixing verbs, if it is attached to the whole verb, or is guist is interested in knowing if, for example, genitives placed before the verbal part. appear before or after the noun heads, or whether arti- cles, adjectives, and numerals come before or after the (4) Case-marking patterns. Case is a way of show- noun heads in a language with a particular word-order. ing grammatical relationship between certain kinds of Although being a very informative source for lin- words by variations in word structure. Languages guists, word-order typology has its own flaws because make use of different means for marking their cases. (1) it is only relevant to languages with Subjects and In some languages, case may be reflected by the use of Objects, (2) it only deals with languages with a basic affixes. In these languages, we either have preposi- word order, while many languages have mixed word tions (like English and Arabic) or postpositions (like orders, and (3) it is only relevant to basic sentences Turkish) or both (like Estonian). There are, of course, within a language (declarative, nonpronominal, etc.). some languages that do not make use of adpositions at The recent developments in syntactic typology have all, and make use of variations in word endings, like provided use with some other criteria as well (see Nos. Latin, Estonian, Russian, and Finnish. 2–9), but basic word order is always standing out as On the other hand, in noninflectional languages the most fundamental feature of syntactic typology. (like French and English), where sentence structure or word order encode the syntactic functions of the (2) Agreement. Agreement refers to a formal rela- words, we may associate cases with specific syntactic tionship between elements in which a form of one positions. word requires a corresponding form of another. This correspondence usually takes place with respect to case, person, number, and gender. For example, in (5) Wh-movement. Syntactic typologists are also Arabic, noun and their adjectives must correspond to interested in the position of the question words in sen- one another in number , gender, and case. Generally, tence, and see if it is initial, final, or other. we can assume three important domains in grammat- ical agreement: (I) The agreement between subject (6) Relative clause structures. As an implicational and verb phrase in terms of person and number, as in result of word-order studies, the questions concerning I go and she goes. Some languages, however, repre- the use of relatives before or after noun heads is an sent object–verb agreement, like Swahili, Basque, issue within the domain of syntactic typology. and Caucasian languages (see Ergativity below). (II) Nominal agreement with elements related to noun, (7) Causative constructions. A causative is a gram- such as determiners, adjectives, and appositions. matical or lexical indication of the causal role of a ref- Like the German sentence Sie sucht einen Jungen, erent in relation to an event or state expressed by a ihren kleinsten Sohn (she is looking for a boy, her verb. A causative may be indicated by a verbal affix youngest son), where the words Jungen (boy) and (black-blacken, able-enable), or a component of Sohn (son) are both accusative and masculine. (III) meaning in the verb, or a special construction, like 1067 SYNTACTIC TYPOLOGY George had John killed, in comparison with The sun References solidified the mixture. Aissen, Judith. 1999. Markedness and subject choice in optimality theory. Natural Language and Linguistic Theory 17. 673–711. (8) Ergativity. In some languages, the form of the Austin, Peter and Joan Bresnan. 1996. Non-configurationality object in a transitive clause is the same as that of the in Australian aboriginal languages. Natural Language and Linguistic Theory 14. 215–68. subject of an intransitive clause. Such languages are Bussman, Hadumod. 1996. Routledge dictionary of language called ‘Ergative’ languages, as opposed to languages and linguistics. London and New York: Routledge. like English, which are typologically entitled ‘nomina- Comrie, Bernard. 1995. Language universals and linguistic tive-accusative’ languages. In studying the ergative lan- typology. Chicago: University of Chicago Press. guages, the linguist has to check if the subject of a Legendre, Géraldine 2001. An introduction to optimality theo- ry in syntax. Optimality-theoretic syntax, ed. by Géraldine transitive clause has a different form; i.e. in some lan- Legendre, Jane Grimshaw and Sten Vikner, 1–27. guages, in intransitive clauses some subjects are treat- Cambridge, MA: MIT Press. ed like transitive subjects, and others like transitive Lyovin, Anatole V. 1997. An introduction to the languages objects. A remarkable number of Caucasian (Georgian, of the world. New York and Oxford: Oxford University Ubykh, Avar), Austronesian (Tongan), Australian Press. Greenberg, Joseph (ed.) 1963. Some universals of grammar (Dyirbal), and Mayan (Tzeltal) languages are ergative. with particular reference to the order of meaningful ele- ments. Universals of human language, 73–113. Cambridge, (9) Pro-drop languages. Some languages can MA: MIT Press. Mithun, Marianne. 1987. Is basic order universal? Coherence optionally drop pronouns at the beginning of sen- and grounding in discourse, ed. by S. Tomlin Russell, tences, as in Persian man Irâni hastam/Irâni hastam (I 281–328. Amsterdam: John Benjamins. am Iranian/Am Iranian) and one that does not usually Steele, Susan. 1978. Word order variation: a typological survey. do so, such as English. Pro-drop languages seem to Universals of human language, Vol. 4, ed. by Joseph behave somewhat differently over a range of construc- Greenberg, 585–623. Stanford: Stanford University Press. tions from languages that do not drop their pronouns. BEHROOZ MAHMOODI-BAKHTIARI Syntax Human languages are productive systems for relating No comparable substitution is possible for the sound (or more generally gesture, as in signed lan- sequence man left in (1). Second, an adverb modifying guages) and meaning. The systematic properties on the the sentence can easily be inserted between man and sound side are the realm of phonology while those on left, as in (3), but not between the and man, as in (4): the meaning side belong to semantics. What is in (3) The man obviously left between is syntax. One of the two most fundamental aspects of syntax is that utterances in a human lan- (4) *The obviously man left [* indicates an guage are structured. Specifically, a sentence is not ungrammatical sentence, a sequence of words merely a string of sounds. Minimally, the sounds must that is not a sentence of the language under be grouped together into words. But a sentence is not investigation] just a string of words either. The words are grouped This pattern of data indicates that the sequence The together into units, called ‘constituents’. Consider a man constitutes a unit while the sequence man left does simple sentence such as: not. There is a division of sentence (1) into two con- (1) The man left. stituents. The first part, the subject, is based on a noun (so is often called a noun phrase (NP)), and the second This is not simply a sequence of three autonomous part, the predicate, on a verb; hence a verb phrase (VP). words. The and man are closely associated in a way in A common notation for representing this division into which man and left are not. The former pair of words units, and the ‘names’ of the units, is displayed in (5). constitute a unit based on the noun man, as the fol- lowing two tests indicate. First, a ‘pro-form’ (in this (5) [S [NP The man] [VP left]] case, a pronoun) can substitute for the man: The interpretation of (5) is that the sequence The man (2) He left is a unit of structure, and this constituent is called an 1068  SYNTAX NP; left is a unit of structure labeled VP; and the entire This process too can go on without limit: sequence of words is a unit of structure called S (11) The man and the woman and the child left (Sentence). A more complete structural representation (12) The man and the woman and the child and the of (1) would also indicate the categories of the indi- dog left vidual lexical items The, a determiner, man, a noun, Another phenomenon that falls in the domain of and left, a verb. At this point, the labeled bracketing syntax is a particular kind of ambiguity, ‘structural notation becomes rather cumbersome. A standard ambiguity’. Human languages are filled with ambigu- alternative more graphically represents the structure in ity. Some instances of ambiguity are lexical, involving the form of a phrase structure ‘tree’ as in (6). two different words that happen to be pronounced (6) S and/or spelled alike. An example is NP VP (13) Meet me at the bank where I could intend that we gather after depositing Det N V our paychecks or that we get together to go fishing. The man left But many other instances of ambiguity do not depend The second fundamental syntactic property of on homonymous words. Each of the following exam- human languages is their infinitude. In every human ples feels like it is two different sentences in one, even language, ever longer sentences can be created by though the words of each have constant meaning. embedding one clause inside another, a process called (14) Sue interviewed a singer on the stage recursion. For example, the sentence in (1) can [Was the interview on the stage, or just the singer?] become the ‘direct object’ of another verb, as in (7), with tree representation (8). (15) Susan said John resigned in order to protect his reputation (7) The woman said the man left [Was the resignation or the statement to protect John’s reputation?] (8) S These ambiguities can be captured by positing two dif- NP VP ferent syntactic structures for each two-way ambigu- ous sentence. For (14), the prepositional phrase (PP) Det N V S on the stage can be inside the direct object NP (where The woman said NP VP it will modify singer) or outside that NP, but still inside the VP (where it will modify interview). The Det N V two structures are in (16) and (17), respectively. the man left (16) S And (7) can, in turn, be inserted into a still larger NP VP sentence: (9) The child believes the woman said the man left N V NP And so on, with no limit. As a consequence of this Sue interviewed Det N PP recursion, there is no longest sentence, and, thus, there are an infinite number of sentences. a singer P NP The kind of recursion just seen involves ‘subordi- on Det N nation’, placing one unit of structure inside another. Just as general in the languages of the world is coordi- the stage nation — putting constituents together in parallel with a coordinating conjunction like and: (17) S (10) S NP VP NP VP N V NP PP NP and NP V Sue interviewed Det N P NP Det N Det N left a singer on Det N The man the woman the stage 1069 SYNTAX A further striking syntactic property of many In both (24) and (25), Mary is the understood subject of human languages is the existence of dependency rela- solve the problem. In (24), additionally, the property of tions between an overtly realized constituent (a ‘eagerness’ is attributed to Mary, while in (25), ‘likeli- ‘filler’) and a ‘gap’ (marked by __), as in interrogative ness’ is not. This contrast is made particularly clear in sentences like (18). sentences with pleonastic (meaningless) subjects: (18) Who will Mary hire __ (26) *There is eager [ __ to be a thunderstorm] Who, the filler, is pronounced at the beginning of the (27) There is likely [ __ to be a thunderstorm] sentence, but is nonetheless interpreted as the direct object, i.e. in the gap position following the verb hire, The kind of dependency in (24), where the overt sub- parallel to (19). ject is understood in both sentences, is often called ‘Control’. The contrasting kind in (25) is sometimes (19) Mary will hire someone called ‘Raising’ (a term coming from a movement One common way, although not the only way, to analysis: Mary originates in the lower sentence, where represent this kind of dependency is through a ‘move- its semantic role is determined, and moves into the ment transformation’ whereby the interrogative word higher sentence). Passive sentences display a similar originates in its understood position, then ‘moves’ to dependency, this time between object position and sub- its superficial position at the beginning of the sen- ject position: tence, leaving a gap in its original position. (28) John was chosen __ Such filler-gap dependencies can span a great dis- tance, as shown in (20), where the edges of each embed- John is the semantic object of the verb, but is phono- ded clause are marked with brackets, and __ marks the logically realized in subject position. A particularly ‘understood’ position of the interrogative word. interesting dependency of roughly this sort is illustrat- (20) Who did Mary think [that John said [that ed in (29). Susan visited __ ]] (29) Mary is easy [ __ to please __ ] Compare: Here, Mary is understood as the object of the embedded sentence. (The subject of the embedded sentence is (21) Mary thinks that John said that Susan visited understood as free in reference.) Syntacticians have someone long been fascinated by contrasts like these, where, item The properties of long-distance relations of this type by item, a set of sentences seem identical, but have very (‘WH-dependencies’, so-called because in English, different interpretive properties, as in the following most interrogative words begin with wh) have been a three examples, based on the ones just discussed. major concern of syntacticians. (30) Mary is eager to please [Mary is understood While WH-dependencies can occur across unlimited subject of both sentences] distances, under some circumstances they are heavily (31) Mary is likely to please [Mary is understood bounded. Constraints on these dependencies (some- subject of embedded sentence only] times called 3‘island constraints’) have been another (32) Mary is easy to please [Mary is understood major topic of investigation. One island constraint object of embedded sentence only] involves structures built with coordinating conjunctions. Consider the hypothetical question form of (22) in (23): In addition to filler-gap dependencies, languages (22) Mary will blame [Sue and someone else] also display dependencies between pairs of overtly (23) *Who will Mary blame [Sue and __] realized items. One such dependency involves reflex- ive expressions, as in (33). This is a perfectly coherent thought, but it cannot be expressed with a WH-dependency. A coordinate struc- (33) John criticized himself ture is an island for such dependencies. Typically, a reflexive requires an ‘antecedent’, and the Two other types of dependencies involve an overt antecedent must be ‘higher’ in the tree than the reflexive: item in the subject position and a gap in a ‘lower’ posi- tion. At first glance, the examples in (24) and (25) (34) *Himself laughed seem identical to each other, but they turn out to have (35) *Himself criticized John significantly different properties. Reflexive dependencies, like many other dependen- (24) Mary is eager [ __ to solve the problem] cies, have locality restrictions. In English, the (25) Mary is likely [ __ to solve the problem] antecedent must generally be in the same clause as the 1070  SYNTAX reflexive (although many other languages have long- In Japanese, along with many other languages, the distance reflexives): structures are exactly the reverse. Verbs follow their direct objects, and the language has postpositions (36) *John said that [Mary criticized himself] instead of prepositions: (37) *John said that [himself would do the job] This is a stricter requirement than we saw for inter- (43) VP rogative filler-gap dependencies (as in (20)). On the NP V other hand, in contrast to (23), coordinate structures are not islands for reflexives: N N (38) John criticized Mary and himself Ken-o yatot-ta One major area of syntactic investigation centers on (44) PP the characterization of divergent locality constraints for the various kinds of dependencies. NP P Syntacticians have long been concerned with sys- N kara tematic ways that languages differ in their syntactic patterns. For example, while many languages are like Tokyo English in having filler-gap WH-dependencies, many The lexical item that a phrase is based on (V for VP, P others, such as Japanese, realize their interrogative for PP, etc.) is commonly called the ‘head’ of the words in their ‘understood’ positions. In terms of phrase. The constituent that serves as ‘direct object’ of movement analyses, interrogative words in these lan- the head is the ‘complement’. In these terms, the two guages do not move. Another systematic difference language types are characterized as head-complement among languages involves basic word order. English, and complement-head. along with many other languages, has the verb preced- In sum, syntax is concerned with the structural prop- ing its direct object, and, parallel to this, English has erties of sentences in languages. Of equal interest are the prepositions (i.e. prepositions precede their objects): general, universal properties, such as phrase structure, (39) Naomi hired Ken recursion, dependencies; and the systematically diverg- ing properties, such as relative order of heads and their (40) complements and how dependency relations are realized. VP V NP References hired N Brown, Keith, and Jim Miller. 1991. Syntax: a linguistic introduc- tion to sentence structure. London and New York: Routledge. Ken Carnie, Andrew. 2002. Syntax: a generative introduction. Oxford: Blackwell Publishers. (41) A package arrived from Tokyo Chomsky, Noam. 1988. Language and problems of knowledge. Cambridge, MA: MIT Press. Green, Georgia, and Jerry Morgan. 2002. Practical guide to (42) PP syntactic analysis. Stanford: CSLI Publications. Tallerman, Maggie. 1998. Understanding syntax. New York: V NP Oxford University Press. HOWARD LASNIK from N See Also Anaphora; Long-Distance Dependency; Tokyo Phrase Structure 1071  T Taboo Discussions of ‘taboo’ words must begin with an because they are considered to be a violation of norms understanding of the relationship between language of etiquette. The use of these words is unacceptable to and culture. The term originally comes from certain people in certain situations. As a society shapes Polynesian cultures, in which certain objects, actions, the individual, it passes on a form of moral judgment and words were considered to exert harmful power on about particular words. In such a society, there is no people. Some words were thought to be dangerous. If supernatural judgment or appeal to spiritual powers, it someone mentioned the name of an ancestor, for is rather a societal pressure against the use of these instance, negative consequences could result. Thus, words and phrases. speakers of a language had to avoid using certain Linguists have taken a neutral and descriptive names or words, just as they had to avoid offending stance on taboo words. The role of linguistic studies cultural norms and religious beliefs. Which words has been to document which words are avoided in were considered ‘taboo’ and to what degree any words what situations. Linguistics makes no moral or reli- were harmful depended on cultural norms. gious judgment on the use of taboo forms. Thus, for Societies that recognize supernatural powers and example, the journal Maledicta: The International are concerned about offending nonhuman powers con- Journal of Verbal Aggression includes articles listing sider words to be very powerful. Cultural norms dis- taboo forms used by various groups and subgroups of courage people from using certain words because the a culture. very utterance will cause harm either to the speaker or In English, words related to certain bodily functions to other members of the society. Melanesian cultures (death, defecation, urination, sexual intercourse) and often have a prohibition on mentioning the name of religious beliefs (hell, damnation, curse, swear) are ancestors to avoid any negative repercussions from the sometimes considered to be ‘taboo’, at least in partic- spirit world. Since personal names are related to com- ular segments of the population. Those who consider mon words, the vocabulary of some languages has the words to be offensive will often use a synonymous undergone a remarkably rapid change. Speakers of euphemism in place of the taboo word: ‘go number such languages find alternative ways of referring to two’ instead of ‘defecate’, ‘pass on’ instead of ‘die’. ancestors and find other ways of talking about the On the other side of the politeness scale are words that objects that are included in the ancestor’s name. These are less neutral and more likely to be limited in public avoidance practices may also extend to how a person settings: ‘take a shit’ for ‘defecate’, ‘croak’ for ‘die’. can talk about or address in-laws. Words themselves are not ‘taboo’, ‘dirty’, or ‘pro- The discussion of taboo words in the Western world fane’. Many of the words currently considered inap- is also related to cultural norms, but not typically to propriate in public settings were the neutral, normal the supernatural world. Certain words may be avoided term for an object or action in earlier forms of English. 1073 TABOO The term ‘shit’ was not always deemed inappropriate Racial epithets are also considered to be within the or impolite. In a similar way, many languages of the topic of language taboo. The use of racial epithets can world still treat bodily functions in a less euphemistic be seen as a form of aggression against others. The use manner. of names and words that denigrate a racial group is Since the use of particular words are determined by subject to societal norms. In the United States, there are the social setting, people use or avoid these forms based even laws restricting and condemning such language. on their desire to follow or violate the social norms of Even though portions of society and even govern- the particular situation. Violating social norms may be ment regulations seek to limit the use of taboo forms, considered appropriate when those norms are thought to it is the individual speaker who determines his or her be outmoded or prudish. American humorists such as own use of these forms. In informal settings among Lennie Bruce, for example, used taboo words in their co-workers or in dangerous situations (war, natural routines. Young people might use taboo words to show catastrophe, etc.), people may use taboo forms as a their disregard for the norms of their elders. means of expressing their emotions. In more formal The topic of taboo words has been extended to situations, in settings that are less homogeneous or in ‘cursing’ and ‘swearing’. The word ‘cursing’ is tied to less stressful situations, the use of taboo language is a worldview in which bad or evil can be caused by often more limited. using words. Likewise, ‘swearing’ has changed from ‘taking an oath’. Both words are now used to refer to offensive language. The terms are used as synonyms References for ‘profanity’, which has its roots in the religious Hughes, Geoffrey. 1998. Swearing: a social history of foul lan- notion of the ‘profane’ vs. the ‘sacred’. Such words are guage, oaths and profanity in English. London: Penguin. also deemed to be ‘dirty’ as in ‘dirty language’, ‘filthy Montagu, Ashley. 2001. The anatomy of swearing. Philadelphia, words’, etc. Newspapers and other public media PA: University of Pennsylvania Press. Reinhold, Aman. 1987. The best of Maledicta. Philadelphia: sources typically restrict the use of particular words, Running Press. even in direct quotations. Likewise, government regu- Spears, Richard A. 1998. Forbidden American English: taboo lations of broadcast media require certain words to be American English. New York: McGraw Hill. censored (‘bleeped out’). PETER J. SILZER Tagalog and Philippine Languages Philippine Languages time of the first Austronesian habitation of the Over 150 languages are spoken by the more than Philippines, the original language has diversified into a 76,500,000 Filipinos who live in an archipelago of number of fairly clearly defined subgroups (Blust around 7,000 islands that stretches over 1,500 kilome- 1991). ters from north to south, and about 800 kilometers Of the 167 Philippine ethnolinguistic groups cited from the most western point of Palawan to the most in Grimes (2000), at least ten are presently spoken by easterly point of Mindanao. Most of the languages are more than a million population, and constitute what dialectally diverse, with a number constituting exten- have been referred to as the ‘major’ languages of the sive dialect chains. Philippines. These are Tagalog, Sebuano, Ilokano, All Philippine languages belong to the Western Hiligaynon, Bikol, Samar-Leyte, Kapampangan, Malayo-Polynesian group of the Austronesian lan- Pangasinan, Maranao, and Magindanao. The first four guage family. The archeological record suggests that of these languages have considerable importance in the earliest Austronesian speakers arrived in the north- the Philippines. Tagalog, with some 15,000,000 native ern Philippines, probably from what is now called speakers, is the native language of Manila and a num- Taiwan about 4,500 years ago, at the beginning of the ber of surrounding provinces. It is also the basis of the migrations that later took them to the Indo-Malaysian national language, now known as Filipino. archipelago, and ultimately into the Pacific. Since the Furthermore, it has become the main language of 1074  TAGALOG AND PHILIPPINE LANGUAGES movies and comics, and much of the Philippine mass Spanish occupation of the Philippines for over 400 media. It is required to be taught in all the schools in years that had the most impact on the languages of the the Philippines, and is rapidly becoming the main sec- country. Probably every Philippine language has a ond language that people speak throughout the coun- large number of words that are ultimately of Spanish try. Sebuano, Ilokano, and Hiligaynon are widely origin. There is at least one language in the Philippines spoken as regional trade languages. Ilokano is the that shows a far greater influence from Spanish than main language of trade and wider communication spo- does Tagalog. This language, called Chavacano or ken throughout northern Luzon. It is also spoken in Zamboangueño, is a creole language spoken mainly in some areas of southern Mindanao and is the main Zamboanga City and surrounding areas. Philippine language spoken in the United States and The other major influence on Philippine languages other countries to which Filipinos have migrated. has come from English as a result of the 50 years that Sebuano is used not only in the Visayan area of the America colonized the Philippines. English is still Central Philippines, but also in much of southern spoken widely throughout the Philippines, is one of Mindanao. Hiligaynon is also spoken in some parts of the languages used in education, and continues to southern Mindanao. influence Philippine languages. At the other end of the scale, there are scores of lan- Most Philippine languages have sound systems that guages spoken by relatively small groups of peoples. are relatively simple. Most have between three and six Many of these languages are spoken in highly isolated vowels. Tagalog originally had three distinctive vow- areas, such as the remote, mountainous areas of els, i, a, and u, but two more vowels, e and o, have Mindanao, Palawan, and Luzon, as well as on small, developed as a result of influence from Spanish bor- isolated islands. Some of the smallest languages, and rowings. Many languages still retain the same vowel the ones most in danger of dying out in the very near system that has been reconstructed for Proto- future are those spoken by some of the approximately Austronesian, with four vowels, i, a, u, and ə (Reid 30 surviving Negrito groups, descendants of the origi- 1973). Similarly, most Philippine languages have nal non-Austronesian inhabitants of the Philippines, fewer than 18 consonants. They are nontonal, but whose original languages were ultimately replaced by some have lexical stress and distinguish vowel length. those of the culturally more dominant Austronesians. A number of the languages of northern Luzon, such as These include Arta, the language of fewer than a dozen Inialoi, exhibit complex morphophonemic variation. remaining members of a group in Quirino Province, Most Philippine languages have only two or three the Atta and Agta groups of Kalinga-Apayao and different kinds of syllables. Words are commonly built Cagayan Provinces, some of the Dumagat and Alta using just two syllable types, CV and CVC, where C groups of Isabela and Quezon Provinces, the Sierra stands for any consonant and V for any vowel. The Madre mountain range, and the islands off the eastern disyllabic Tagalog word bahay ‘house’ is typical of the coast of Luzon, as well as the many Ayta groups of great majority of common nouns in Tagalog and other Bataan and Zambales Provinces. Many of these lan- Philippine languages. Verbs are commonly morpho- guages are spoken by fewer than 1,000 people. On a logically highly complex, with various prefixes, infix- slightly larger scale are the languages spoken in the es, and suffixes providing both syntactic and semantic Cordillera Central, the massive mountain range in the information. A wide variety of reduplicative processes center of northern Luzon. These include closely relat- is also found. ed varieties of Ifugao, Bontok, Balangaw, Kankanay, and Kalinga, some of which have fewer than 50,000 speakers each. Similarly, the number of speakers of Tagalog several of the Manobo languages of Mindanao is esti- Tagalog is a nonconfigurational language (Kroeger mated to be no more than 10,000 to 15,000. 1993). Its basic sentences are predicate initial, with Prior to European settlement in the Philippines nominal complements typically following the predicate. beginning with the Spanish in 1521, Philippine lan- Noun phrase word order is not rigid, except that actors guages had already been strongly influenced by con- typically precede other actants, and full NPs which tact with traders from outside the country. The Chinese carry nominative case tend to occur later in the sen- (primarily speakers of Hokkien) had established tence. Since there is no copula verb, Tagalog sentences enclaves in major port areas (beginning around the may have a noun, an adjectival form, or a prepositional twelfth century CE). Similarly, traders speaking a vari- phrase as the predicate of a sentence: e.g. Doktor ang ety of Malay probably used in Brunei, Malaysia had bisita. ‘The visitor is a doctor.’ Maganda ang babae. firmly established themselves in the Manila area at ‘The woman is beautiful.’ Nasa kusina ang regalo. ‘The least 100 years prior to the arrival of the Spanish. Their gift is in the kitchen.’ NPs are typically introduced by influence on Tagalog was considerable. But it was the one of a small number of short, unstressed words, often 1075 TAGALOG AND PHILIPPINE LANGUAGES referred to as determiners, that provide information Relative clauses follow their head noun and are about the case, plurality, and personhood of the follow- linked to it by a ‘ligature’ na (following consonants) or ing noun, as in Table 1. Most common nouns are not -ng (following vowels). Their structure follows the so- inflected for plurality, but may have a plural word mga called gap strategy, with the gap corresponding to the (/maŋa/) following the determiner to mark plurality: nominative NP of the relative clause, e.g. ang babaeng e.g. Tumakbo ang mga batà . ‘The children ran.’ pumasok ‘the woman who entered’, ang bahay na Nominative forms typically mark an NP as specific or bibilhin ko ‘the house that I will buy’. A ligature also definite. They may also mark topicalized NPs and defi- links a main verb and its complement, e.g. Huwag nite NPs functioning as predicates. kang umiyak. ‘Don’t cry.’ Tagalog distinguishes three case-marked sets of per- Tagalog is typical of other Philippine languages that sonal pronouns, as shown in Table 2. In addition to pro- have traditionally been considered to be typologically nouns that refer to the speaker (first person), directly different from other languages in that they have been address the hearer (second person), or refer to a third assumed to have a unique type of grammatical system, person, Tagalog like other Philippine languages has dis- known as the ‘focus system’. This is a system charac- tinct forms for first person plural exclusive (‘we, not terized by the use of verbal affixes to indicate the the- you’), and first person plural inclusive (‘we all’). matic role of the NP bearing the nominative case in a Pronouns with dual reference (‘we, two’), while com- sentence. The term ‘focus system’ was first introduced mon in many Philippine languages, are used only in to describe the languages of the Philippines. rural areas where Tagalog is spoken. An irregular com- Subsequently, other Austronesian languages exhibit- bined form of Genitive 1s and Nominative 2p, kita, ing a similar type of grammatical system (such as occurs, e.g. Iniibig kita. ‘I love you.’ Third person pro- many of the languages in Taiwan, Sabah, northern nouns do not distinguish gender. Nominative pronouns Sarawak, and northern Sulawesi, as well as Malagasy, also function as predicates: e.g. Ikaw ang anak ni Pedro. Palauan, and Chamorro) have been described as hav- ‘You are the child of Pedro.’ Genitive forms express not ing a ‘Philippine-type’ syntax. only possessors that follow their noun heads, but also Basic verbal clauses in Tagalog have one of two actor participants in a clause: e.g. Ibinigay nila ang basic structures. ‘Actor focus’ verbs carry one of a set libro sa titser nila. ‘They gave the book to their teacher.’ of affixes on the verb that indicate that the actor is Pronominal forms, as well as clitic adverbs such as na expressed by the Nominative case. These may be ‘already’, pa ‘yet’, and din ‘also’, occur as second posi- monadic, expecting only one NP, e.g. with infix -um-: tion clitics attaching to the first constituent within the Pumasok si Nila. ‘Nila entered.’, or dyadic, in which clause, such as a negative auxiliary, e.g. Hindi ko pa rin case an additional NP expressing a theme argument binili ang libro. ‘I still haven’t bought the book yet’, or occurs, expressed by the oblique case when indefinite, a fronted adverb, e.g. Bukas na siya papasok. ‘He will or a partitive, definite NP, e.g. Bumili sila ng mangga. enter tomorrow already.’ ‘They bought mangoes’; Uminom noon ang mga batà. ‘The children drank some of that.’ Such sentences are TABLE 1 Tagalog Determiners syntactically intransitive. ‘Nonactor focus’ verbs carry one of a different set of affixes on the verb that indi- Nominative Genitive Oblique Locative cates whether some participant other than the actor Common ang ng /naŋ/ ng /naŋ/ sa carries Nominative case. In these sentences, the actor Personal carries Genitive case. All are syntactically transitive: Singular si ni –– kay Plural sina nina –– kina e.g. Goal focus, with –in: Bibilhin nina Juan ang mga mangga bukas. ‘Juan (and companions) will buy the mangoes tomorrow.’ Location focus, with –an: TABLE 2 Tagalog Personal Pronouns Bibilhan nila ng mangga ang mga batà. ‘They will buy mangoes from the children.’ Beneficiary focus, Nominative Genitive Locative with i-: Ibibili nila ng mangga ang batà. ‘They will Singular buy the child a mango.’ Instrument focus, with ipang- 1st person ako ko akin : Ipambibili nila ng mangga ang pera nila. ‘They will 1st pers. dual kata nita kanita buy mangoes with their money.’ 2nd person ka / ikaw mo iyo 3rd person siya niya kaniya In addition to the affixes that mark focus or voice, Plural verbs also carry a wide range of affixes marking tem- 1st pers. excl. kami namin amin poral aspect (perfective, imperfective, and contemplat- 1st pers. incl. tayo natin atin ed), and mood (volitional vs. nonvolitional). They may 2nd person kayo ninyo inyo also be derived with a causative affix, introducing an 3rd person sila nila kanila additional actant (a causer) into the clause. 1076  TAMIL The grammatical system of Philippine-type lan- Himmelmann, Nikolaus P. 1991. The Philippine challenge guages has been a topic of considerable controversy in to universal grammar. Arbeitspapier Nr. 15 (Neue Folge). Köln: Institut für Sprachwissenschaft, Universität zu linguistic analysis. Hardly any of the statements made Köln. in this article have gone unchallenged in recent times. Kroeger, Paul. 1993. Phrase structure and grammatical relations Various issues discussed in the literature are the follow- in Tagalog. Dissertations in Linguistics. Stanford: Center for ing: Do Philippine languages have a ‘subject’, and if the Study of Language and Information. they do, which NP is it? Are Philippine languages accu- Liao, Hsiu-chuan. 2004. Transitivity and ergativity in Formosan and Philippine languages. Ph.D. Dissertation. Department of sative, ergative, split-ergative, or some other type? Do Linguistics, University of Hawai’i. Philippine languages have a true passive construction? Mithun, Marianne. 1994. The implications of ergativity for a Do they have an antipassive construction? Do the ‘focus Philippine voice system. Voice: form and function, ed. by affixes’ constitute inflectional voice morphology on the Barbara Fox and Paul J. Hopper. Amsterdam and verb, or are they derivational ‘applicative’ affixes? Philadelphia: John Benjamins Publishing Company. Quakenbush, J. Stephen. 2003. Philippine linguistics from an SIL perspective: trends and prospects. Philippine Journal of References Linguistics 34. 1–28. Blust, Robert A. 1991. The greater Central Philippine hypothe- Reid, Lawrence A. 1973. Diachronic typology of Philippine sis. Oceanic Linguistics 30. 73–129. vowel systems. Current trends in linguistics, Vol. 11: ––––––. 2002. ‘Notes on the history of focus’ in Austronesian diachronic, areal, and typological systems, ed. by Thomas languages. The history and typology of Western A. Sebeok. The Hague: Mouton and Co. Austronesian voice systems, ed. by Fay Wouk and Malcolm Reid, Lawrence A., and Hsiu-chuan Liao. 2004. A brief syntac- Ross. Canberra: Pacific Linguistics. tic typology of Philippine languages. Special issue of De Guzman, Videa P. 1998. Ergative analysis for Philippine lan- Language and Linguistics 5(2) (Issues on Austronesian guages: an analysis. Studies in Austronesian linguistics, ed. Linguistics), ed. by Elizabeth Zeitoun. Taipei: Academia by Richard McGinn. Athens, Ohio: Ohio University Center Sinica. for International Studies-Center for Southeast Asia Studies. Schachter, Paul, and Fe T. Otanes. 1972. Tagalog reference Grimes, Barbara F. (ed.) and Joseph E. Grimes (consulting edi- grammar. Berkeley, Los Angeles, London: University of tor) 2000. Ethnologue: languages of the world, 14th edition. California Press. Dallas, Texas: Summer Institute of Linguistics. LAWRENCE A. REID Tamil Tamil is a language of the Dravidian family that origi- figure that may include second-language speakers; nated in southern India and spread in prehistoric times United Kingdom, 100,000; France, 60,000; Germany, to Sri Lanka. In the nineteenth century, the British 50,000; Switzerland, 40,000; United States, 35,000; colonial empire enabled Tamils from India to seek Australia, 30,000; Italy, 25,000; Mauritius, 22,000 of a work as laborers in Malaysia, Singapore, Sri Lanka, total of 100,000 ethnic Tamils; Netherlands, 20,000; Burma, southern Africa, Fiji, Reunion, Mauritius, and Norway, 8,000; Denmark, 7,000; Fiji 6,600; New the West Indies. Many well-educated Sri Lanka Tamils Zealand, 3,000; Sweden, 2,000. In all communities found higher status jobs in Malaysia and Singapore. outside India and Sri Lanka, younger generations are More recently, Tamils have looked to the Middle East, no longer learning Tamil, so that in some cases these the United Kingdom, North America, and Australia for figures exaggerate the number of actual speakers. employment. In addition, thousands of Sri Lanka Tamil exhibits ‘diglossia’: a formal or ‘High’ variety Tamils seeking asylum from the troubles in their home- is used in public speaking and most writing, while col- land have fled to India, Western Europe, North loquial or ‘Low’ varieties are used in face-to-face inter- America, and Australia. The distribution of the over 63 actions and in writing intended to convey an informal million native speakers of Tamil in various countries is chatty or ‘hip’ style. Differences between formal and roughly as follows: India, 58.6 million, primarily in colloquial varieties are found in pronunciation, gram- Tamil Nadu and neighboring states; Sri Lanka, 3 mil- mar, and vocabulary. Colloquial Tamil has much more lion; Malaysia, 1.3 million of 1.5 million ethnic Tamils; variation according to region and social class than for- South Africa, 250,000; Canada, 160,000; Singapore, mal Tamil does: the greatest differences are between 120,000 of 191,000 ethnic Tamils; Reunion, 120,000, a Sri Lanka and India, with different subvarieties within 1077 TAMIL each of these. Differences in the usage of different script, which descends ultimately from a variant of the social groups are also found, as, for example, between Brahmi script used during the reign of Emperor Ashoka Brahmin and non-Brahmin varieties in India. Some (273–232 BCE). The writing system is well adapted to examples are given Table 1 comparing formal Tamil represent the pronunciation of formal Tamil, except that with the principal colloquial varieties of India and a few symbols are superfluous, such as those that repre- Sri Lanka. sent the combinations /aj/ and /a/ as though they were In vocabulary that is not borrowed from unrelated diphthongs, and the symbols for [n] and [ŋ], which are languages like Sanskrit or English, formal Tamil has the just variants of /n/. A second script, the Grantha script, consonants / p t t  c k m n 1  r 5 l )  j / and the vow- was developed in the fifth century CE to write Sanskrit els / i e a o u /. All consonants except /r/ and /*/ are in South India. Grantha symbols for /d/, /+/ /s/, /h/, and found single (short) or geminate (long), and all vowels /k+/ have been adopted into the Tamil script to represent may be short or long. The consonants and vowels are sounds occurring in loanwords. essentially the same as those believed to have been used A Tamil word consists of a root plus a series of suf- in Proto-Dravidian, the common ancestral language of fixes. Word order is strictly ‘head-final’: that is to say, the entire Dravidian family. Unlike the other Dravidian adjectives and relative clauses precede the nouns they languages, Tamil has not augmented the single series of modify; adverbs, direct objects, and indirect objects voiceless stops /p t t / c k / with voiced or aspirated precede verbs; auxiliaries follow main verbs; and ones, although the voiceless stops have variants that postpositions are used rather than prepositions. Formal include voiced stops such as [b, d] and continuants; and Tamil is more conservative and exhibits characteristics for many Indian speakers, a series of voiced stops of an earlier stage of the spoken language. Thus, some occurs in loanwords such as bassu ‘bus’ and daakam of the suffixes of colloquial Tamil have developed ‘thirst’. Other sounds that occur only in loanwords are / from earlier postpositions, auxiliaries, etc., which still h /, / ʃ /, and / % /. Colloquial Tamil has generally lost appear as separate words in formal Tamil. Examples /5 /, replacing it with / 4 / or / j /; Indian varieties have can be seen in the last three examples in the table also lost / tF /, but Sri Lanka Tamil preserves it as / r / or above. / tF /. Two characteristics of the pronunciation of collo- Case suffixes and postpositions are used to indicate quial Indian Tamil distinguish it from Sri Lankan vari- the grammatical functions of nouns and pronouns. In eties: the insertion of / j / before / i / or / e / and of /  / head-final languages like Tamil, the distinction before / u / or / o / when these vowels begin a word; and between postpositions and case endings is sometimes nasalized vowels that result from the loss of / m / or / n fuzzy; indeed, as mentioned above, the latter can / at the end of a word, as in [ pɐjjε ˜) ] from pajjan ‘boy.’ develop from the former. To illustrate this point fur- Sri Lanka Tamil, for its part, is characterized by the ther, traditional grammars of formal Tamil recognize retracted pronunciation of / i / and / e / before certain eight cases: nominative (subject), accusative (direct consonants, as in [kAA kkrɐn] from keekkran ‘I ask’, object), instrumental (by using …), associative and the fronted pronunciation of / a / before alveolar (together with …), dative (to …), ablative (from …), consonants, as in [p%ll] from pallu ‘tooth’. genitive (of …), and locative (at …), but the ablative is Tamil has been written since the third century BCE. not a case in formal Tamil: rather, it is a combination Like many South Asian literary languages, it has its own of the locative case and the postposition iruntu, as seen in the example ‘from the house’ above. In colloquial varieties, however, the postposition and locative have TABLE 1 Comparison of Formal Tamil with fused into a single unit and can be considered a sepa- Colloquial Forms of India and Sri Lanka rate case form. Common Sri Lanka Formal Pronoun categories include the expected first, sec- Indian Colloquial Tamil ond, and third person forms (‘I/you/he’). In the first- Colloquial (Jaffna) person plural (‘we’), there is an additional distinction [I] gave ku/utteen ku/uttanaan ko/utteen (lacking in colloquial Sri Lanka Tamil) between ‘inclu- [I] gave ku/utti//een ku/uttitFtFan ko/uttu i//een sive’ and ‘exclusive’ (naam ‘we, including you’ vs. (perfective) naanka ‘we, not including you’). The second person [you] ask keekkriinka keekkrija4 kee/kitFiirka4 distinguishes between the familiar singular nii and the very rompa miccam mikaum boy pajjan poijan pajjan plural or polite singular form niinka; a singular polite from the ii//ulerntu ii//ulerntu ii//il iruntu form, niir, is also less commonly found. The third-per- house son pronouns are the most numerous and show the must go pookaum pooka ooum / pooka ee/um greatest variation from one variety of Tamil to another. pooka eeum First, there are three parallel sets of ‘deictic’ (pointing) didn’t come arale areelle ara illai pronouns, which are distinguished by their initial 1078  TAMIL vowel; for example, itu ‘it, this one’, atu ‘it, that one’, TABLE 2 Selected Forms for the Verb ko/u ‘give’ etu ‘which one’. These three sets can be called ‘proxi- (Formal Tamil) mal’ (i-), ‘distal’ (a-), and ‘interrogative’ (e-). In formal Imperative Tamil, each of these sets has six members; illustrating ko/u imperative ‘give’ with the distal set, these are: aar ‘he or she, that Infinitive (human, respectful) person’, aan ‘he, that (human, ko/u-kk-a infinitive ‘to give’ nonrespectful, male) person,’ aa4 ‘she, that (human, Tense forms ko/u-kkitF-aan rational present ‘he gives/is giving’ non-respectful, female) person’, atu “it, that (non- ko/u-tt-aan past ‘he gave’ human) one’, aarka4 ‘they; those (human) people’, ko/u-pp-aan rational future (potential/habitual) ‘he and aajka4 ‘they, those (non-human) ones’. In addi- will give; he gives; he used to give’ tion to these deictic pronouns, two further interrogative Forms built on infinitive stem pronouns are found: enna ‘what’ and jaar ‘who’. Many ko/u-kk-alaam permissive/speculative ‘may/could give’ ko/u-kk-a//um Optative ‘let [him] give’ other deictic words also have proximal, distal, and ko/u-kk-atu nonrational present ‘it gives/is giving’ interrogative forms, for example ippootu ‘now’, ko/u-kk-um nonrational future (potential/habitual) appootu ‘then’, eppootu ‘when’; ittanaj ‘this many’, ‘it will give; it gives; it used to give’ attanaj ‘that many’, ettanaj ‘how many’; and ippai ko/u-kk-aatu nonrational future (potential/habitual) ‘thus, like this’, appai ‘thus, like that,’ eppai ‘how’. negative ‘it won’t give; it doesn’t give; it used not to give’ Finally, the Jaffna dialect of Sri Lanka Tamil has a Forms built on present stem fourth set of addressee-proximal deictics in u-. For ko/u-kkitF-a pres. verbal adjective ‘[who] gives’ example, Jaffna Tamil distinguishes between itu ‘it, ko/u-kkitF-atu pres. verbal noun ‘giving’ this one near me/us’, utu ‘it, that one near you’, atu ‘it, Forms built on past stem that one (not near you or me/us)’, and etu ‘which one’. ko/u-tt-a past verbal adjective ‘[who] gave’ ko/u-tt-aal conditional ‘if [he] gave/gives’ Verbs do much of the grammatical work in a Tamil ko/u-tt-u past participle ‘having given’ sentence, and as a result there is a rich set of verb Forms built on future stem forms; a selection of these is given in Table 2. For reg- ko/u-pp-atu future verbal noun ‘giving’ ular verbs, the verbal root serves as the base of the Infinitive + auxiliary imperative. The infinitive is also closely related to the ko/u-kk-a mu/ij-um habilitative ‘can give’ ko/u-kk-a υee1/-um debitive ‘should/must give’ root form of the verb. Three tenses (present, past, and ko/u-kk-a υillaj simple negative ‘didn’t give/isn’t giv- future) are found, and these take additional suffixes to ing/won’t give’ agree with the subject. The infinitive and tense stems ko/u-kk-a maa//-aan future rational neg. (potential/habitual; also serve as the base on which other forms are built; volitive) ‘he won’t give; he doesn’t finally the infinitive and the past participle combine give; he used not to give’ ko/u-kk-a ppa/-utu passive ‘it is given’ with a variety of auxiliary verbs. In colloquial Tamil, ko/u-kk-a υaj-kkitF Causative ‘he makes [someone] give’ some of these auxiliaries are reduced to suffixes, as -aan seen in the example ‘[I] gave (perfective)’ above. Past participle + auxiliary Colloquial Tamil has also developed new formations ko/u-tt-u kko- not found in formal Tamil; an example is the past 1/- iɾu-kkitF-aan pres. progressive ‘he is giving’ ko/u-tt- formation of Jaffna Tamil seen above in the example iru-kkitF-aan pres. perfect ‘he has given’ ‘[I] gave’. ko/u-tt-u It can be seen from the examples that the human vs. υi /u-kitF-aan pres. perfective ‘he (definitely/com- nonhuman distinction found in the pronouns has par- pletely) gives/is giving’ allels in the verbal system; for example, the human present and future, built on the corresponding tense stems, are structurally different from the nonhuman any structural similarity to the positive form it negates. present and future, built on the infinitive stem. Less Moreover, the human future positive is not completely apparent is the fact that the positive and negative sys- equivalent semantically to the human future negative, tems of Tamil are somewhat independent of one anoth- which carries connotations of intention. er. In most languages, each positive form has a Among the constructions typical of Tamil and corresponding negative form with a similar structure. other Dravidian languages are the conjunctive partici- But in Tamil, this is not the case: the distinctions of ple, the relative clause, and the quotative. In the con- tense (present, past, future) and aspect (progressive, junctive participle construction, found throughout perfect, perfective) found in the positive are replaced the South Asian linguistic area, all but the final verb in the negative by just the simple negative, the future in a sequence of verbs are represented as past human negative, and the future nonhuman negative; of participles (emphasized in the following examples): these three, only the future nonhuman negative has dejalalitaa
in ii//ukku ppooj aaraj cantittaar ‘He 1079 TAMIL went to Jeyalalitha’s house and met her’ (literally (literally ‘“that cow I will-buy” QUOTATIVE Mohan ‘Jeyalalitha’s house having-gone, her met’); having-asked-even, “cannot” QUOTATIVE Raman tankau/aja pa5aja danapatijaj alualaka ka//attukk refused’). u iruntu taևևi taratara-entFu i5 uttu ppooj na/ucaalajil tuukkil poo//aarka “They pushed their own former References president out from inside his office building, went dragging him along the ground and hanged him in the Andronov, Mikhail Sergeevich. 1989. A grammar of modern and classical Tamil. Madras: New Century Book House. middle of the road” (literally “their-own former presi- Annamalai, E., and Sanford B. Steever. 1998. Modern Tamil. dent office building inside from having-pushed The Dravidian languages, ed. by Sanford B. Steever. along-the-ground having-dragged having-gone in- London, New York: Routledge. the-middle-of-the-road in-a-noose put”). Asher, R.E. 1985. Tamil, London, Sydney, Dover, NH: Croom Relative clauses in Tamil and other Dravidian lan- Helm. Britto, Francis. 1986. Diglossia: a study of the theory with guages are formed without relative pronouns, by using application to Tamil. Washington, DC: Georgetown a verbal adjective (emphasized in the following exam- University Press. ples); the relative clause precedes the noun it modifies, Gair, James W., S. Suseendirarajah, and W.S. Karunatilaka. 1978. and the verbal adjective is the last item in the relative An introduction to spoken [Sri Lanka] Tamil. Kelaniya: clause: [naan ko֒utta] pputtakam… “The book that I External Services Agency, University of Sri Lanka. Hart, Kausalya. c. 1999. Tamil for beginners. Berkeley: gave…” (literally “I having-given book”); inta aarajc- International and Area Studies, University of California at cijin mu/iil [e4ijaana] i2ajam [nammaj ellaam nic- Berkeley. cajam atirccikk u44aakkukitFa] maatiri taan irukkitFatu Lehmann, Thomas. 1989. A grammar of modern Tamil. “The information which has been released in the results Pondicherry: Pondicherry Institute of Language and Culture. of this research is of the very sort that will certainly sub- Schiffman, Harold. 1999. A reference grammar of spoken Tamil. Cambridge: Cambridge University Press. ject us all to shock” (literally “this research’s results-in Suseendirarajah, S. 1993. Jaffna Tamil. Thirunelvely, Sri Lanka: having-been-released information us all certainly University of Jaffna. shock-to subjecting type EMPHATIC is”). Zvelebil, Kamil. 1973. The smile of Murugan—on Tamil liter- Finally, the quotative construction marks the com- ature of South India. Leiden: E.J. Brill. plement of an overt or understood verb of saying, Zvelebil, Kamil, Yurii Glasov, and Mikhail Andronov. 1967. Introduction to the historical grammar of the Tamil lan- thinking, hearing, etc. with a QUOTATIVE comple- guage, transl. by L. Navrozov. Moscow: Nauka. mentizer derived from the verb en ‘say’. As can be seen in the following examples, Tamil makes no dis- tinction between direct and indirect discourse. ‘atil Web Resources enna Giceeam entu’ aGacarappa/a Gee1/aam ‘Don’t Grimes, Barbara F. (ed.) 2000. Ethnologue: languages of the be in a hurry [to say] “What’s special about that?”’ (lit- World. SIL International,<http://www.sil.org/ethnologue/> erally ‘“that-in what special” QUOTATIVE hurry Tamil Electronic Library <http://www.geocities.com/Athens/ 5180/index.html> don’t’); ‘anta maaaj naan Gaanki-kkokiteen’ entu Web Assisted Learning and Teaching of Tamil, University of mookan keeu-kko-um kuua, ‘muijaatu’ entu raa- Pennsylvania Language Centre, <http://ccat.sas.upenn.edu/ man matutt-u Giaar ‘Even when Mohan offered to plc/tamilweb/> buy the cow, Raman refused [saying] “[you] cannot”’ IAN R. SMITH Tarascan Tarascan is the name of a language (and its indigenous shortened to P’oré or P’uré simply to denote things speakers) spoken in the mountainous north central Tarascan. Many consider Purépecha preferable to region of the state of Michoacan, Mexico. Tarascan Tarascan, an anglicized version of the Spanish word speakers call themselves and their language tarasco. The generally accepted etymology of ‘tarasco’ Purépecha, variously written as Purhépecha, derives from stories of the first contact with Europeans, P’urepecha, P’orepecha, Phurhépecha, and sometimes who through various means acquired for themselves 1080  TARASCAN daughters of the Tarascan nobles and princes. 100,000 and 120,000. The Mexican National Tarháskua means ‘son-in-law’, ‘daughter-in-law’, or Indigenous Institute (Instituto Nacional Indigenista, or ‘father-in-law’ in the indigenous language, all relation- INI) in its most recent count puts the number at just ships which, in effect, the Spaniards acquired with the under 100,000 speakers in the state of Michoacán Tarascans. The repeated, and uninformed, use of the itself. Linguistic surveys have also revealed differing word is said to have spread among the Spaniards as a degrees of Tarascan-Spanish bilingualism throughout reference to the Indians. The etymology of the word the region, the smaller, less hispanicized towns, of ‘purépecha’ has two basic strands. Some have traced its course, with higher rates of Tarascan use. meaning to refer to migration, travel, and alliances Approximately 7,217 (8% of the Tarascan-speaking while other scholars take the word to mean ‘common population of the state) are monolingual. Despite the people’ (plebeyos). historical decline in numbers of speakers, Mexican The geographic range of spoken Tarascan is today census data since 1970 have revealed an average annu- limited to the state of Michoacán. Historically, howev- al increase of 2.6% in the number of speakers, com- er, the Tarascan empire (known to the Spaniards as the pared to an average annual growth rate of 2.1% for the ‘Western Empire’) was much larger. Prior to European state’s population as a whole. This increase in the num- contact, the Tarascans are thought to have exerted their ber of speakers is owed in part to the work of the political and cultural influence as far north as the pres- Purhépecha Language Academy (La Academia de la ent-day state of Durango. They maintained a milita- Lengua Purhé) and the Center for Investigations of rized border against the Aztecs, established as far east P’urhépecha Culture (Centro de Investigaciones de la as Guanajuato and the Valley of Toluca, and to the Cultura P’urhépecha) (Argueta 2003). The spoken lan- west their control reached the Pacific Ocean. Tarascan guage is characterized by clear dialect differences, presence is thus estimated to have covered an area some regional (Sierra, Cañada, Lake, as mentioned nearly equal in size to the empire of the Aztecs, above), but others in much closer proximity, from vil- extending over virtually all of present-day Michoacán lage to village (Friedrich 1975). and with settlements or influence in portions of the Interest in the ethnography and language of the states of Guerrero, México, Jalisco, Guanajuato, Tarascan people dates to the first half of the sixteenth Querétaro, Colima, Nayarit, Sinaloa, and Durango. century, shortly after contact with the Europeans. The Through the centuries the extent of spoken Franciscans accompanying the Spanish showed imme- Tarascan has shrunk significantly, to the point where it diate interest in the ethnographic and linguistic char- is now found in a well-defined area comprising about acteristics of the Tarascans, documenting their 3,500 square miles in north central and northwestern findings in the Relación de Michoacán, probably Michoacán only. Researchers commonly describe four begun in 1539 (finished and presented to the Viceroy distinct zones of Tarascan speech: the Zacapu Marsh in 1541). The first systematic linguistic works on region, the Lake Pátzcuaro region, including the town Tarascan were undertaken by the Franciscans around bearing that name and the villages surrounding the the middle of the sixteenth century, and at least three lake, an area called La Cañada de Chilchota (Los Once dictionaries and two grammars of the language were Pueblos), once containing 11 towns, now nine, to the produced in the 1600s. The earliest of these works, north and west of the Lake, and the mountainous titled Vocabulario en la lengua de Michuacan, was region due west of Lake Pátzcuaro called the Tarascan brought to fruition in 1559 by Fr. Maturino Gilberti Sierra (also known as La Meseta Tarasca). The Sierra and subsequently reedited in 1898 by Nicolás León, in hosts the largest percentage of Tarascan speakers 1901 by Antonio Peñafiel, and yet again in 1962 by (an estimated 60% as opposed to 19% in the next Ernesto Ramos Meza. largest zone, the Lake Region). Tarascan speakers are Despite more than four centuries of linguistic distributed in 95 of Michoacán’s 113 municipios. analysis, the origins of both the Tarascan language Researchers have observed that it is commonly and its people remain puzzling. It is certain that the thought throughout the region that ‘real’ Tarascan is Tarascan language bears no linguistic relationship spoken in the Sierra. with any of the original languages spoken in Mexico. The number and location of Tarascan speakers has Scholars generally agree that its genetic relation- historically been difficult to assess, given the remote ships are difficult, if not impossible, to define for nature of many of the villages where the language is any other languages as well. Attempts have been spoken. Through much of the middle part of the twen- made on various fronts to trace the origins of the lan- tieth century, the numbers were variously pegged at guage and its speakers. Linguistic and material cul- 50,000 to 80,000 speakers. More recently, some lin- tural evidence is cited to support the idea of a guistic survey counts have put the numbers as high as historic connection between the Tarascan language 1081 TARASCAN and the Pacific Coast of South America, most The strength of the Tarascan language (as measured notably with Quechua of Peru, but observed phonet- by the language surveys of the Mexican Census data) ic similarities with indigenous languages of Peru is encouraging. Only speakers over the age of 5 are may be merely coincidental. Recent evidence is counted for the language census. Government efforts interpreted by some to indicate linguistic, material in bilingual teacher education, bilingual schools, and culture, and even physiological links to Central language research and training efforts through the INI American groups. Other scholars do not believe the have together effected a slowing in the numeric evidence points unequivocally to a southern prove- decline of Tarascan speakers. Work by scholars in the nance and remain open to a northern or northwestern Instituto Michoacano de Cultura del Gobiemo del origin. Swadesh (1969) determined Tarascan to be an Estado and in the Colegio de Michoacán have resulted autonomous linguistic family, a finding with which in growing interest, training, awareness, and standard- others concur. ization (of the alphabet, training materials, and of The phonological characteristics of the Tarascan teaching methodology) for the survival of this linguis- language are relatively clear-cut. tic isolate. 6 vowels: i u e a o ï (predictably following the References consonants s, ts, and ts’) Argueta, Arturo. 2003. Instituto Nacional Indigenista, Perfil de 16 Consonants: Hablantes Indigenas, Estado de Michoacán. Web version at http://www.ini.gob.mx/monografias/purepechas.html P t ts Ch k Boyd, Maurice. 1969. Tarascan myths and legends. Fort Worth: b d g S J Texas Christian University Press. ill n r X Chamorro, Arturo. 1994. Emotion and rivalry in Purhepecha rh (a retroflexed ‘r’) music. Paper presented at the 46th International Congress of Americanists, Stockholm, July 4–9. Five of the consonants take phonologically distinc- Corona Nunez, Jose. 1957. La religi6n de los Tarascos. tive aspiration: Morelia: Anales del Museo de Michoacán, No. 4. p’ t’ ts’ ch’ k’ Friedrich, Paul. 1975. A phonology of tarascan. Chicago: University of Chicago Press. Despite its straightforward appearance, this Gómez Bravo, Lucas, Benjamin Pérez González, and Ireneo phonological pattern is subject to a complex set of Rojas Hernández. 2001. Intrudcción al idioma P’urhepecha. morphophonemic rules, which include vowel length- Morelia, Mich., Mexico: Universidad Michoacana de San Nicolas de Hidalgo, Centro de Investigaci6n de la Cultura ening, stop epenthesis, nasal assimilation, stop P’urhepecha. assimilation, retroflexion, vowel deletion, and final González, Juan L. Jr. 1985. Mexican and Mexican-American devoicing. Lengthy words comprised of many suffix- farm workers. New York: Praeger Publishers. es and multiple roots are a regular feature of the lan- INI 2002 (source: http://207.248.180.194/bibdf/ini/perfiles/ guage. Verbs are inflected for person, number, tense, purhepecha). Jasso Espinosa, Rodolfo. 1969. Grandeza de Michoacan. as well as mode and interrogation. Formality is also Mexico City: B. Costa-Amico. marked on the verbs (as well as on pronouns). Katzner, Kenneth. 1992. Languages of the World. New York: Ordered suffixation characterizes the verbal struc- Routledge. ture, such that verbs may develop extensive forms. León, Nicolás. 1904. Los tarascos, notas hist6ricas, etnicas y For example: antropo16gicas, Pt. 1. Mexico City: Imprenta del Museo Nacional. Mendieta y Núñez, Lucio. 1940. Los Tarascos. Mexico City: Unandontskuarhinsïndiksï (from uandontskuarhini Imprenta Universitaria. ‘to converse, to chat’) Pollard, Helen. 1993. Tariacuri’s legacy: the prehispanic Syllable patterns follow eight types: Tarascan state. Norman: University of Oklahoma Press. V VV VC CV CVC Relación de la ceremonias y ritos y población y gobierno de los indios de la provincia de Michoacán (1541.) 1956. CVV CCV CCVV Madrid. Word order is generally Subject–Object–Verb. Rubio Ortiz, Pascual. 1920. Historia de Michoacán. Morelia. A sample of written Tarascan (taken from the pop- Ethnologue. 2003. Summer Institute of Linguistics (source: ular folk song Dalia Tsïtsïki) is as follows: http://www.ethnoiogue.com/). Velásquez Gallardo, Pablo. 1978. Diccionario de la lengua Ay dalia tsïtsïki emenda anapu, ka shankar t’u sesarhati Phorhépecha. Mexico City: Fondo de Cultura Económica tsïpajka; ka aimindu tumbiecha ishï jarashtia su sha aru West, Robert C. 1948. Cultural geography of the Tarascan area. komjarhani erontashïn. Isï sha jaueru iamindu Washington, DC: Smithsonian Institution, Institute of Social tumbiecha nochkan ji tiempushin isï tsïperan, no para ji Anthropology. trabajimbo ekarha askia. WARREN D. ANDERSON 1082  TEACHING: CURRICULA Teaching: Curricula The term ‘curriculum’ refers to the content and aims of what came to be known as the ‘functional-notional’ a course, the teaching procedures that will be fol- or the ‘communicative’ approach to language teach- lowed, and the ways in which student learning will be ing emerged. This view saw the aim of language assessed. For some, curriculum is synonymous with learning as involving more than just the acquisition ‘syllabus’, that is, the listing and sequencing of course of linguistic structures. The aim was the achieve- content. The terms ‘syllabus’ and ‘method’ are also ment of communicative proficiency and the ability sometimes used interchangeably. In its broadest sense, to convey meaning in a range of communicative sit- a curriculum includes all of these: syllabus, aims, uations. This shift in the aim of language learning method, and means of assessment. programs had a dramatic impact on language-learn- A helpful way to summarize key issues in language ing curricula. curriculum design is through Richards and Rodgers’ The basis of this new approach comprised the con- notions of approach, design, and procedure. cepts of ‘functions’ and ‘notions’, based on the work ‘Approach’ refers to the theory of language and lan- of Austin and Searle, and of the applied linguist guage learning that underlies the particular approach Wilkins, as well as the concept of communicative to course development. ‘Design’ includes the objec- competence, based on the work of Hymes. Functions tives, organization, and content of the particular syl- are the things that people ‘do’ with language, such as labus type, kinds of teaching and learning activities, asking for information, greeting, apologizing, and teacher and learner roles, and the role of instructional inviting. Examples of notions are frequency, duration, materials. ‘Procedure’ describes the classroom tech- location, and quantity. Communicative competence niques and practices that might be employed within includes knowledge about language as well as the particular approach. methods of using language in appropriate ways to For a long time, the overriding approach to lan- achieve particular goals. Language-learning programs guage curriculum design was based on the structural were changed to make communicative competence, syllabus; that is, the listing of grammatical items to be rather than (just) linguistic competence, the overall learned. This was the basis of the grammar-translation goal. As a result, lists of functions and notions and method, which dominated language teaching from the communicative appropriateness were added to the lan- 1840s to the 1940s. Structural syllabi are still used guage-learning curriculum. today, but primarily with a different approach to The communicative approach is still used today, but methodology and in combination with other approach- with a number of subsequent refinements and redevel- es to syllabus design. opments. These include programs that focus on skills Situational language teaching, or the oral approach, development (such as listening for ‘gist’, deciphering became the primary avenue toward language teaching the main idea, and guessing vocabulary from context); in Britain in the 1950s. The aim of this approach was programs that focus on tasks (such as exchanging to develop learners’ spoken language skills in every- information, solving problems, making decisions, day situations. Hence, lessons might be labeled ‘eating reading a job advertisement, making an appointment, out’ or ‘at the bank’. The view of language in this and writing a résumé); and programs organized around approach was still structural, however, but with a focus a particular content area (such as first year university on oral practice and the use of sentence patterns and science and technology) or themes (such as food, fam- drills. The audiolingual method, which emerged in the ilies, employment, and pollution). Other syllabus types United States at the same time, was similar in its view include the ‘lexical’ approach, which focuses on the of language and use of methodology. acquisition of particular vocabulary items, and text- or A major shift occurred, however, in the late genre-based approaches, which use whole texts as the 1960s, when applied linguists began to question the starting point for language program development and theoretical assumptions underlying situational lan- consider what it is that learners need to know and guage teaching and audiolingualism. Drawing on the accomplish in order to produce an example of a par- work of scholars such as M.A.K. Halliday, Dell ticular genre. A further development is the ‘negotiated Hymes, John Langshaw Austin, and John Searle, syllabus’, in which teachers and learners negotiate the 1083 TEACHING: CURRICULA aims of the program, course content, and sometimes References methodology. Often, however, language-learning Brown, James Dean. 1995. The elements of language curricu- curricula will be ‘mixed’ or ‘multistranded’, in that lum, a systematic approach to program development. Boston they will combine a number of these program types in and London: Heinle & Heinle. order to be most useful and comprehensive for a par- Dudley-Evans, Tony, and Maggie Jo St. John. 1998. ticular group of learners. Developments in English for specific purposes. A multi-dis- ciplinary approach. Cambridge and New York: Cambridge James Dean Brown’s The elements of language University Press. curriculum (1995) describes the main components of Feez, Susan. 1998. Text-based syllabus design. Sydney: National a language-learning program as needs analysis, goals Centre for English Language Teaching and Research, and objectives, testing, materials, teaching, and eval- Macquarie University. uation. ‘Needs analysis’ refers to the identification of Hedge, Tricia. 2000. Teaching and learning in the language classroom. Oxford and New York: Oxford University Press. the language that students need in order to achieve Johnson, Robert (ed.) 1989. The second language curriculum. particular goals, along with the identification of the Cambridge and New York: Cambridge University Press. learners’ linguistic competence in relation to these Nunan, David. 1988a. Syllabus design. Oxford, New York: goals. Needs analysis is an especially important com- Oxford University Press. ponent of programs that are designed for specific pur- ———. 1988b. The learner-centred curriculum. Cambridge and New York: Cambridge University. pose language use, such as academic English and Richards, Jack, and Theodore Rogers. 1986. Approaches and business English programs. ‘Goals and objectives’ methods in language teaching: a description and analysis. refer to the overall reason or purpose of the program, Cambridge and New York: Cambridge University Press. as well as what learners are expected to be able to do Richards, J.C. 2001. Curriculum development in language teach- with language by the end of the program. ‘Evaluation’ ing. Cambridge: Cambridge University Press. Ur, Penny. 1996. A course in language teaching: practice and the- includes the ways in which learners’ progress will ory. Cambridge and New York: Cambridge University Press. be assessed in relation to the program’s goals and White, Ron. 1988. The ELT curriculum. Oxford and Cambridge, objectives. Once needs have been established and MA: Basil Blackwell. goals and objectives have been set, the curriculum Wilkins, David. 1976. Notional syllabuses. Oxford and New planner is able to go about adopting ready-made York: Oxford University Press. Yalden, Janice. 1983. The communicative syllabus. Evolution, materials for the program, adapting materials as design and implementation. Oxford and Elmsford, NY: required, or developing new materials. Decisions then Pergamon. proceed regarding teaching approaches and program BRIAN PALTRIDGE evaluation; that is, how the program might best be taught and how it can best be assessed in terms of its See also Genre; Second Language: Teaching; Teach- overall effectiveness. ing: Methods Teaching: Methods A teaching method, sometimes referred to as a Since there is general debate and disagreement about methodology, is a coherent and theory-based charac- all of these broader issues, it is not surprising to find that terization of the language-teaching process, usually the search for the ‘best’ language-teaching method has prescribing one set of classroom practices, proce- been fruitless. Studies that have sought to compare dif- dures, strategies and techniques and often proscribing ferent methods to determine their effectiveness have others. remained inconclusive, partly because there are so A method is always informed by a broader many variables involved in the language teaching and approach or underlying philosophy. This would nor- learning process, partly because, as a number of empir- mally include a set of beliefs about language and how ical studies have discovered, the actual activities in it works, about how people learn or acquire language, observed classrooms often seem indistinguishable no about curriculum design and general educational matter what method is being used, and partly because it approaches, about general learning theories, and about is almost impossible to ascribe cause and effect in any- the purposes of language learning. thing other than trivial or short-lived learning. 1084  TEACHING: METHODS The history of language-teaching methods has severely eroded through the development of transfor- largely been one of fluctuation between form and mational grammar and Noam Chomsky’s 1959 attack function, between learning about the target language on habit formation as a basis for language learning. and learning through actually trying to use it, and Many activities in the language-teaching classroom between focus on the written language and focus on today still show allegiance to audiolingual and behav- the spoken language. While there is little doubt that iorist principles. the vast majority of those people who successfully and The dominant paradigm over the last 30 years in lan- fluently use more than one language––and there are guage-teaching worldwide has been communicative many more bilinguals and multilinguals in the world language teaching. The approach foregrounds the use of than monolinguals––learned their languages outside language as a means of communication, with the the classroom by using them for real purposes in the assumption that much learning of language takes place real world, the issue of formal instruction is more through language in use rather than language as an complex. For example, language instruction in formal object of study. Communicative language teaching was classroom settings has in many cases not included the reinforced by advances in error analysis and second lan- purpose of learning to communicate in an everyday guage acquisition studies in the 1970s, which strongly sense, but has instead focused on gaining access to the suggested that learners proceed developmentally literary canon of the target language or has been for through a predictable series of stages in which the pro- the specific purpose of understanding ancient texts for duction of errors plays an important part. It is also historical, cultural, or archeological reasons. closely linked with the notion of language in context In the study of languages that are no longer spoken and with those sociolinguistic theories, especially naturally, focus is necessarily on understanding the M.A.K. Halliday’s systemic-functional grammar and mechanics of the written language––the grammar and Dell Hymes’ notion of communicative competence, that the meaning of words. The widespread study of Latin take context into account in their investigations of how and Greek in western schools from the Middle Ages people use language. The key concept in communica- onward had a major influence on the teaching of lan- tive language teaching is that learners learn through the guages more generally and meant that the grammar- use of language in meaningful activities. It is not simply translation method was the dominant––although not a matter of contextualization of the language––it must unchallenged––paradigm for language teaching until be use of language that is meaningful for the learners. well into the twentieth century. The development of language for specific purposes, The major challenge came late in the nineteenth where language learning occurs while the ostensible century when thinkers such as Henry Sweet in England focus is on communicating for some other purpose, can and Wilhelm Viëtor in Germany argued strongly for the be seen as a special case of communicative language primacy of speech over writing, the inductive learning teaching. Similarly, Stephen Krashen and Tracy Tyrell’s of grammar, and the avoidance of translation, using natural approach (1983), originating in studies of only the target language as far as possible. This ulti- English-speaking Canadian students studying in mately led to the development of the direct method, French-medium schools (‘immersion’), can be classi- also influenced by studies of natural language learning, fied as communicative language teaching. Both of these especially in childhood. The direct method was readily have been very influential at a policy level, and there adopted by private language schools, where native has been considerable work on language across the cur- speakers of the target language were recruited as teach- riculum and on bilingual and immersion education. ers. It was much less influential in formal education An allegiance to the principles of communicative systems, where nonnative speakers often struggled language teaching has shown up in national curricula in with the fluency needed to implement the method, and recent times, for example, in countries such as China, where the emphasis on spoken language made formal Korea, Japan, and Thailand; however, there are obsta- assessment of students much more difficult. cles to implementation in practice, and communicative The audiolingual method, the development of language teaching has been criticized for conflicting which owed much to the wartime need for fluent and with local cultural values and for its unrealistic expec- convincing speakers of other languages, was a build- tations of the teacher’s own language proficiency. ing-block approach based on a thorough analysis of The very concept of a language-teaching ‘method’ language grounded in structuralist traditions and on has also come under attack recently, with several theo- the theories of behaviorist psychology, with its empha- rists and researchers finding that the concept has little sis on stimulus–response conditioning and learning justification in actual classroom practices. Jack through habit formation. It became the dominant lan- Richards (1990) has suggested that teachers would be guage-teaching method by the 1960s despite the fact better advised to focus ‘not on the search for the best that its theoretical foundations were already being method, but rather on the circumstances and conditions 1085 TEACHING: METHODS under which more effective teaching and learning are Howatt, A.P.R. 1984. A history of English language teaching. accomplished’. Oxford: Oxford University Press. Krashen, Stephen D., and Tracy Terrell. 1983. The natural approach: language acquisition in the classroom. Oxford: Pergamon. References Richards, Jack C. 1990. The language teaching matrix. Brown, H. Douglas. 1987. Principles of language learning and Cambridge: Cambridge University Press. teaching. Englewood Cliffs, NJ: Prentice-Hall. Richards, Jack C., and Theodore S. Rodgers. 1986. Approaches Candlin, Christopher N., and Neil Mercer (eds.) 2001. English and methods in language teaching. Cambridge and New language teaching in its social context. London and New York: Cambridge University Press. 2nd edition 2001. York: Routledge. DAVID R. HALL Chomsky, Noam. 1959. A review of B.F. Skinner’s verbal behavior. Language 35. 26–58. See also Acquisition; Halliday, Michael Alexander Holliday, Adrian. 1994. Appropriate methodology and social con- Kirkwood; Hymes, Dell H.; Sweet, Henry; Teach- text. Cambridge and New York: Cambridge University Press. ing: Curricula Telugu Telugu is a Dravidian language, one of the language (1) Coastal Dialect –– Districts of Andhra families found in South Asia. The Dravidian family is Pradesh divided into three subgroups: South Dravidian, Central (i) North Coastal –– Vijayanagaram Dravidian, and North Dravidian. Telugu belongs to the Vishakapatanam Central Dravidian subgroup. Some other languages of Srikakulam this subgroup––Gondi, Konda, Pengo, Manda, Kui, (ii) Central Coastal –– West Godavari Kuvi––have sometimes been treated as a separate East Godavari group called South-Central Dravidian. Kolami, Naiki, Krishna Parji, and Gadaba are the other languages in Central Guntur Dravidian. (iii) South Coastal — Prakasam Telugu is mainly spoken in the State of Andhra Nellore Pradesh in the south central part of India. Telugu is (2) Rayalaseema Dialect also spoken in the neighboring states––Tamil Nadu, (i) Southern –– Chittoor Karnataka, Maharashtra, and Orissa. It is spoken by Rayalaseema migrants in South Africa, Mauritius, Fiji Islands, (ii) Central –– Cuddapah Malaysia, Singapore, and the United States of Rayalaseema Kurnool America. The earliest written records of Telugu from (iii) Western –– Anantapur the second century BC consist mainly of records of Rayalaseema names. Numerous inscriptions written in Telugu are (3) Telangana Dialect available from the sixth century AD. From the twelfth (i) Central –– Medak century onward, literary and grammatical treatises Telangana started being composed. On the basis of the available Ranga Reddy materials, some scholars have divided the history of Hyderabad the Telugu language into six periods: (ii) North –– Adilabad Telangana (1) Early Old Telugu AD 200–AD 700 Karimnagar (2) Late Old Telugu AD 700–AD 1200 Nizamabad (3) Early Middle Telugu AD 1201–AD 1400 (iii) South –– Mahaboobnagar (4) Late Middle Telugu AD 1401–AD 1600 Telangana (5) Early Modern Telugu AD 1601–AD 1900 (iv) East –– Warangal (6) Modern Telugu AD 1901–onward Telangana Regional variations give us the following picture of Nalgonda dialects and subdialects: Khammam 1086  TELUGU Telugu shares the presence of cerebral consonants If we had treated d and c as belonging to two different (t., n., I.) with the other Dravidian languages. These con- classes, the deletion of the final /u/ as well as the sub- sonants are not present in any of the Indo-European sequent assimilatory change from /d/ to /c/ (to match groups (the tongue position differs slightly from that the following /c/) could not be treated as an instance of used to produce English [d, n, l]) excepting in the the broader Sandhi rule. Indo-Aryan subgroup, which perhaps is due to the Vowel harmony in Telugu affects noninitial syllables: influence of Dravidian languages. Similarly, Telugu has long and short varieties of enimidi + lu → enumudulu central vowels (e, e¯, o, o¯), which again is true in other ‘eight’ pl Dravidian languages, but not in Indo-Aryan languages (e.g. erugu (to know) , e¯ rugu (to shit), to°a (thigh), All vowels in the first word assimilate (are harmo- to¯°a (with)). nized) to the vowel in the suffix -lu. This kind of har- Telugu draws a distinction between voiced and monization does not cross a geminate (doubled)_ voiceless consonants (e.g. /d/ vs. /t/), which is not the consonant or a consonant cluster: case in e.g. Tamil (a South Dravidian language). In Tamil, b, d, g, d., j occur only in postnasal posi- engili + lu → engiI.I.u tions––i.e. after n or m––whereas p, t, k, t., c occur [partly [pl] (remnants of word initially or in clusters. In Telugu, the following eaten food] eaten food) word pairs show that the difference between k/g or t/d is crucial: kampa (thorny wood) vs. gampa (basket), Telugu, like other Indian languages, is a postposi- a¯t,a (play) vs. a¯d.a (feminine), kanta (hole) vs. kanda tional language, i.e. particles equivalent to English (an edible root). prepositions (such as with) follow the noun or noun Aspiration of consonants is also a contrastive fea- phrase: ture that Telugu has acquired from Sanskrit and built into its orthographic system (e.g. d vs. dh); its ortho- atanu → atani to¯ graphic system is derived from the Ashok Brahmi [that he his with (= with him) script. Speakers of nonstandard varieties of Telugu, nonpolite] however, tend to drop aspiration in their speech. ‘Friction’ (of voiceless labiodentals) is a feature Nouns and pronouns that do not end in ‘u’ do not that Telugu has recently borrowed from English but it have distinct oblique form: is not integrated into its script, e.g. coffee (ka¯f¯), sofa (so¯fa¯). It is represented in the script by bilabial voice- ranga°u → ranga°i to¯ less aspirate, i.e. the sound [f] is written as ph. Ranga Ranga with Sandhi and vowel harmony are reflected in the a¯me → a¯me to¯ orthography that is phonemic (letters stand for individ- [that she her with (with her) ual speech sounds) and syllabic (letters stand for sylla- polite] bles). Keeping in mind the behavior of sounds in red.d.i → red.d.ito¯ Sandhi, Telugu consonants can be classified as follows: Reddy Reddy with (with Reddy) (1) Dorsals k, kh, g, gh (2) Apicals t., t.h, d., d.h, n, n., l, I, r In Telugu, the finiteness is expressed by the GNP (3) Laminals t, th, d, dh, , h, , s, s´, ®s (Gender Number Person) marker, which comes at the (4) Labials p, ph, b, bh, m end of a verbal expression. If the GNP marker is t/d and / are regarded as belonging to one category absent, the verbal unit is nonfinite. (laminal), as examples of Sandhi (microjuncture) phe- nomena show: a phonological rule of Telugu states a¯me ninna vacci ¯ ro¯ju veI.I.indi that a short vowel gets deleted between consonants of She yesterday come-perf. today go-perf-she the same category. This juxtaposes two consonants She came yesterday and went (back) today and causes one of them to assimilate to the other. For example: Here the marker ‘-i-’ in vacci and veI.I.indi shows completion of action––hence ‘perfective’ suffix. ‘nin- ca¯du + cukka → c a¯ c c u k k a na’and ¯ ro¯ju’ refer to time. As the action is complete, [substance [dot] people consider this also to be past tense. used as dot As in the case of other languages, Telugu kinship on one’s forehead] terms reflect certain cultural phenomena like the 1087 TELUGU ‘marriage’ relationship. In English, an ‘uncle’ could Because Telugu words tend to end in a vowel, the be maternal or paternal. Telugu has different terms for language has been described by Charles Philip Brown, them. The maternal uncle, the father-in-law, and the a great lexicographer, as ‘the Italian of the East’. paternal aunt’s husband are all referred to as ‘ma¯ma’. Similarly, the paternal aunt, the mother-in-law, and References the maternal uncle’s wife are referred to as ‘atta’; both Krishnamurti, Bh., and J.P.L. Gwynn. 1985. A grammar of the atta’s and the ma¯ma’s daughter/son can become Modern Telugu. Delhi: Oxford University Press. one’s spouse. Prakasam, V. 1985. The linguistic spectrum. Patiala: Punjabi In North India, where Indo-Aryan languages like University. Hindi and Punjabi are spoken, the maternal uncle, the Ramakrishna Reddy, B. 1986. Localist studies in Telugu syntax. Hyderabad: The Author. father-in-law, and the paternal aunt’s husband are Ranganathacharyulu, K. 1987. A historical grammar of inscrip- given different terms because the maternal uncle’s and tional Telugu (1401 A.D. to 1900 A.D.). Hyderabad: Dept of the paternal aunt’s son/ daughter cannot be taken as Linguistics, Osmania University. one’s spouse. VENNELAKANTI PRAKASAM Tense and Aspect Marking From the point of view of language typology, the most Aspect is not relational like tense, but focuses the common grammatical means expressing tense and internal temporal structure of events. The progressive aspect are past, future, perfect (these three expressing (also called continuous) indicates that an event is still tense) and progressive, imperfective, perfective (these in progress, relative to some explicit or implicit refer- three expressing aspect); cf. Dahl (1985), Bybee ence time, cf. John was dancing (when Mary came in), (1985), and Bybee and Dahl (1989). In the majority of where the reference time is given through Mary’s com- cases, the present is morphologically unmarked, which ing in. The perfective is used when the event is viewed is the reason why it is usually not regarded as a sepa- as a bounded whole, whereas situations described in rate grammatical morpheme. The six tense–aspect the imperfective are not seen as bounded. In Slavic expressions mentioned are found alone or in combina- languages, the imperfective/perfective distinction is tion: e.g. in the case of the pluperfect (John had seen mostly expressed morphologically, cf. Russian pisal the woman, past plus perfect) or past progressive (John ‘to write’ (imperfective) vs. napisal ‘to finish writing’ was seeing the woman, progressive plus past). Tense (perfective). and aspect may be expressed either by periphrastic or Now let us examine some common tense–aspect by inflectional (or bound) expressions. Typologically, expressions in detail. As for the perfect, there are at perfect and progressive have a very strong tenden- least four different types of periphrastic constructions cy for periphrastic expressions, while past, imper- (cf. Dahl 1985): first, copula plus past participle (e.g. fective, and perfective prefer the bound variant. Future in Hindi, Bulgarian); second, auxiliary ‘have’ plus past does not favor either periphrastic or inflectional participle (originally possessive constructions, e.g. in expressions. most Germanic and Romance languages); third, main Tense is a grammatical category that specifies a verb plus particle ‘already’ (e.g. in Yoruba, Isekiri); time at which the corresponding tenseless sentence is fourth, constructions with auxiliaries historically true. Compare I was often in Venice, uttered at a cer- developed from verbs meaning ‘finish’ or ‘throw tain time t: past expresses that there is a period of time away’ (e.g. in Sango, Ewe). The present perfect in before the speech time t at which the tenseless sen- English has the following uses (cf. McCawley 1971): tence I often be in Venice is true. Past indicates that the (a) to indicate that a state of affairs prevailed through- event at issue precedes speech time; future indicates out some interval stretching from the past into the that it follows speech time. Perfect (also called anteri- present (universal perfect—I’ve known Max since or) is used for events preceding speech time or some 1960); (b) to indicate the existence of past events other reference time and still being relevant at that ref- (existential perfect—I have read Principia erence time. Mathematica five times); (c) to indicate that the direct 1088  TENSE: SYNTAX effect of a past event still continues (stative/resultative happen some weeks later, with no preparatory actions of perfect—I can’t come to your party tonight—I’ve it at the present moment. caught the flu); (d) to report hot news (hot news per- The most widespread type of bound or inflectional fect–Malcom X has just been assassinated). tense-aspect is what Bybee and Dahl (1989) call the The progressive usually combines with the present tripartite system. The past/nonpast distinction is added and the past. Like the perfect, the progressive tends to the perfective/imperfective distinction, with perfec- towards periphrasis. The periphrastic expressions are tive combining only with past, and the past/nonpast locative, meaning ‘to be located in or at an activity’; cf. distinction only occurring in the imperfective. About Bybee and Dahl (1989). Explicit variants (copula plus every second language seems to have this tripartite locative adposition plus nominalized verb) are most system (e.g. Mandarin Chinese, French, Kurdish). frequent, e.g. in Irish tá sé ag dúnadh an dorais (‘he is Interestingly, the two distinctions are independent, at shutting the door’). Locative meaning is part of because it is also possible to have the perfective/ almost all progressives, even of the English progres- imperfective distinction but lacking past (e.g. in sive, although not overtly. But the English progressive Arabic), or to have past but lacking the perfective/ like in John is jumping is historically derived from the imperfective distinction, as in English. The tense– construction John is a-jumping, where ‘a’ is actually aspect systems of many Slavic languages are not sim- the locative preposition ‘on’. Although the progressive ply tripartite, however. Russian, for example, allows usually describes events that are still in progress at ref- present-imperfective, past-imperfective, past-perfec- erence time, note that e.g. the English progressive has tive, future-perfective, and future-imperfective. The a wider use. In John is writing his thesis, the activity system of Ukrainian is even richer; in addition to the need not be in progress at reference time but rather at a Russian possibilities, we also find perfect-imperfec- larger time-interval properly including reference time. tive, perfect-perfective, pluperfect-imperfective and In addition, examples like John is working on his the- pluperfect-perfective. sis everyday show that the progressive is also possible with habitual activities, at least in English. The future does not favor either periphrastic or References inflectional expressions. Common lexical sources for Bybee, J.L. 1985. Morphology: a study of the relation between the future are the following: first, an auxiliary with the meaning and form. Amsterdam: Benjamins. former meaning of ‘want’ (e.g. in Mandarin and Bybee, J.L., and Ö. Dahl. 1989. the creation of tense and aspect Swahili, and also English ‘will’); second, a verb mean- systems. Studies in Language 13. Comrie, B. 1976. Aspect. Cambridge: Cambridge University ing ‘to owe’ (e.g. in the Western Romance languages, Press. Korean, and also English ‘shall’, derived from ‘sceal- ––––––. 1985. Tense. Cambridge: Cambridge University Press. lan’); third, like the English going-to-future, a colloca- Dahl, Ö. 1985. Tense and aspect systems. Oxford: Blackwell. tion meaning ‘movement toward a goal’. The difference Hopper, P.J. (ed.) 1982. Tense––aspect: between semantics and between the English will-future and the going-to-future pragmatics. Amsterdam: John Benjamins. . McCawley, J.D. 1971. Tense and time reference in English. is easily felt in minimal pairs like there is going to be a Studies in linguistic semantics, ed. by C.J. Fillmore and riot here vs. there will be a riot here. Only the going-to- D.T. Langendoen. New York: Holt, Rinehart and Winston. future expresses that the riot is already in its preparation Tedeschi, P., and A. Zaenen (eds.) 1981. Tense and aspect. New phase and on its way to happening. The will-future, in York: Academic Press (= Syntax and Semantics 14). contrast, just makes a prediction, but the riot can also MONIKA RATHERT Tense: Syntax Often, theories of tense try to assign a constant mean- (capitals). Russian, for example, expresses a ing to morphological tenses like future, preterite, or telic semantic FUTURE by a combination of morpho- present. But a closer look at the data shows that logical perfective plus morphological present. A one has to distinguish between morphological tense German semantic FUTURE may be a morphological (small letters in this article) and semantic TENSE present. 1089 TENSE: SYNTAX While it is more or less known what the morpholo- (2) John said that Mary was pregnant gy of a morphological tense is, the semantics of a John PAST say that Mary PAST be pregnant semantic TENSE is notoriously unclear. Some regard John PAST say that Mary ∅ be pregnant TENSES as operators (e.g. Ogihara 1996; Dowty Ogihara’s system was also developed in order to 1979); others view them as constraints on variables account for the crosslinguistic contrast between (e.g. Partee 1973). The nature of the temporal relations English and Japanese. In Japanese, we find the seman- is debated, too: most people regard TENSES as two- tically right TENSE configuration since simultaneity place operators, but there are also some (Nerbonne under a matrix PAST is expressed by the TENSEless 1985; Ehrich 1992) modeling the three Reichenbachian ‘present’ and there is no need for the deletion rule: parameters E, R, and S as a three-place relation. For most researchers, TENSES locate the time variable of (3) Bernhard-wa Junko-ga byookida to it-ta the verb; others view them as restrictions on the possi- B.-TOP Junko sick-∅ that say- ble denotations of this variable (Dowty 1982; -SUBJECT PAST Nerbonne 1985). ‘Bernhard said that Junko was sick’ An interesting issue of morphological tense vs. Moreover, Ogihara can account for the English– semantic TENSE is given by the sequence-of-tense Japanese contrast of TENSE interpretation in relative (SOT) phenomena. Now it is widely accepted that clauses. A PAST TENSE relative clause embedded tense in attitude-complements may be ignored in the under a matrix PAST TENSE can have a simultaneous semantics, i.e. that tense in attitude-complements may interpretation by the application of the deletion rule: be no TENSE, that there is TENSE-deletion or empty TENSE. (4) John saw a girl who was running The interpretation of temporal expressions in John PAST see a girl who PAST be running embedded contexts is a complex issue; there are con- John PAST see a girl who ∅ be running trasts between temporal expressions in the scope of In Japanese, the simultaneity under a matrix PAST attitude verbs and those in adjunct clauses, and there is is expressed by the TENSEless ‘present’ and there is, significant cross-linguistic variation (Kusumoto 1999; again, no need for the deletion rule: Ogihara 1996). The semantic treatment of embedded tenses began (5) Mariko-wa naiteiru otoko- hanasi with Enç (1986, 1987). According to her, tenses are to noko-ni kaketa be considered referential expressions denoting times; Mariko-TOP cry-teiru ∅ boy-to talk-PAST they fill temporal argument positions. Tenses are sub- ‘Mariko talked to the boy who was crying’ ject to anchoring principles that are sensitive to GB- Expressions other than ‘real’ tenses but associated style binding restrictions. This works fine for matrix with PAST-TENSE-like interpretations also trigger clause but, as pointed out by Ogihara (1996), it runs TENSE deletion in Ogihara’s system. TENSE deletion into trouble when applied to multiple embeddings. In can be triggered by the perfect and by certain noun these cases, a tense has to meet contradictory anchor- complements: ing requirements: (6) John believes Mary to have claimed that she (1) John decided a week ago that in ten days at was innocent breakfast he would say to his mother that they (7) John’s claim that he was innocent is well known were having their last meal together Ogihara accounts for the phenomena by assuming In (1), the time of their last meal is understood as that a noun like claim can be associated with the syn- simultaneous to John’s speech (thought there is a past tactic features [+PAST], [−PAST], the perfect with the tense) and, considering the presence of the adverbial feature [+PAST], the past tense with [+PAST], and the ‘in ten days’, this time is in the future with respect to present tense with [+PRES]. Under these assumptions, speech time. the SOT rule is simply sensitive to the syntactic fea- A semantically more sophisticated analysis is tures [+PAST] and [+PRES]: proposed by Ogihara (1996). He developed an analy- (8) SOT rule: If a tense feature B is the local tense sis in which TENSES are Priorean operators, predi- feature of a tense feature A at LF, and A and B cates take a time variable as their argument, and the are occurrences of the same feature, A and the simultaneous interpretation is accounted for by the tense associated with A are optionally deleted. application of a TENSE deletion rule. When a PAST TENSE is c-commanded by (embedded in) another Ogihara’s analysis is problematic for the following PAST TENSE, the former is deleted and the embed- example, where the embedded future tense can only ded sentence becomes TENSEless: have a simultaneous interpretation: 1090  TEWA AND THE KIOWA-TANOAN LANGUAGES (9) Jen said that would move to Polish. In these languages, the present tense has different Cecilia Amherst interpretations in complement and relative clauses when Jen PAST-say that woll-PAST move to embedded under a matrix PAST. Cecilia Amherst Jen PAST-say that woll-∅ move to References Cecilia Amherst Dowty, D. 1979. Word meaning and Montague grammar: According to Ogihara, would has to be analyzed Synthese Language Library. Dordrecht: Reidel. as composed of a TENSEless future operator woll ––––––. 1982. Tense, time adverbs and compositional semantic and the PAST TENSE. Since the PAST TENSE on theory. Linguistics and Philosophy 5. 23–55. Ehrich, V. 1992. Hier und Jetzt - Studien zur lokalen und tem- woll is c-commanded by the matrix PAST TENSE, it poralen Deixis im Deutschen. Tübingen: Niemeyer. can be deleted, giving rise to the correct future-in- Enç, M. 1986. Toward a referential analysis of temporal expres- past interpretation. But since the TENSE deletion sions. Linguistics and Philosophy 9(1). 405–26. rule is optional, it should be possible to retain the ––––––. 1987. Anchoring conditions for tense. Linguistic Inquiry embedded PAST, giving rise to a shifted future-in- 18. 633–57. Kusumoto, K. 1999. Tense in embedded contexts. Ph.D. past reading: there is a time t preceding Jen’s speak- Dissertation. Department of Linguistics, University of ing, and after this time (potentially before Jen’s Massachusetts. speaking), Cecilia moves to Amherst. This is clearly Nerbonne, J.A. 1985. German temporal semantics: three- not a possible reading. In order to avoid the problem, dimensional tense logic and a GPSG fragment. Ph.D. Ogihara has to say that the SOT rule is obligatory in Dissertation. The Ohio State University. Ogihara, T. 1996. Tense, attitudes, and scope. Dordrecht: the case of would. Kluwer. Since it is a syntactic configuration that triggers the Partee, B. 1973. Some structural analogies between tenses and application of the deletion rule, Ogihara’s system cannot pronouns in English. Journal of Philosophy 70. 601–9. account for non-SOT languages such as Russian or MONIKA RATHERT Tewa and the Kiowa-Tanoan Languages The Kiowa-Tanoan language family consists of four began a process of cultural transformation that made branches: Kiowa, Towa, Tiwa, and Tewa. Although them exemplars of the Southern Plains culture area. they are not so radically divergent that they suggest a Part of the linguistic evidence for such an interpretation long time-depth to account for internal diversity, the rests on the well-established fact that speech communi- speech communities associated with them and their ties associated with every other branch of the Kiowa- component languages reside in two distinct cultural Tanoan language family was associated with sedentary areas of Native North America. Kiowa has long been settlements in the Rio Grande River Valley of northern associated with the Southern Plains cultural area. This New Mexico at the time of Spanish contact in 1540. area is often associated with its cultural emphases on Within the Tanoan group, the Towa branch consists the tipi, the horse, the bison, and the sun dance ritual of only one existing Pueblo community, Jemez complex. In contrast, Towa, Tiwa, and Tewa––usually Pueblo, although the extinct Pueblo of Pecos also grouped together as the Tanoan languages––represent spoke this language prior to abandonment. the Pueblo cultural area of the southwest. These cul- The Tiwa branch is divided into Northern and tures were known for their large, sedentary agricultural Southern Tiwa languages, each having multiple villages featuring elaborate ceremonial calendars. dialects. In Northern Tiwa are the dialects of Taos and Anthropologists have located the Kiowas near the head Picuris while Southern Tiwa consists of Sandia, Isleta, of the Missouri River around 1700. While their earlier and Ysleta del Sur. The last of these dialects was for- history is more difficult to construct, it is likely given merly spoken in the Rio Grande area but transported both archeological and linguistic evidence that the by Isletans who migrated with the Spanish to El Paso Kiowas abandoned a more Pueblo-like adaptation in in 1682 in the wake of the Pueblo Revolt of 1680. The what is today Northern New Mexico well prior to Tewa branch consists of two languages––Rio Grande Spanish exploration in the mid-sixteenth century, and and Arizona Tewa. Rio Grande Tewa is or was spoken 1091 TEWA AND THE KIOWA-TANOAN LANGUAGES in the following Pueblos––San Juan, San Ildefonso, common basic word order in the language appears to Nambe, Santa Clara, Pojoaque, and Tesuque. In the be Subject–Object–Verb and all languages according- last two of these Pueblos, the Tewa language has not ly feature postpostional phrases (e.g. house-into) continued except through the in-migration of speakers rather than prepositional phrases. from other Tewa-speaking pueblos. Of the dialects currently maintained, Santa Clara is usually regarded Historical Linguistics—An Overview as the most divergent. Arizona Tewa, which is current- ly spoken on First Mesa of the Hopi Reservation in The relatively close relationship of the Tanoan languages northeastern Arizona, was formerly spoken by the was first recognized by John Wesley Powell in 1891 on Southern Tewa, also known as Tano, who occupied the the basis of a lexical comparison of linguistic data that Galisteo Basin of Northern New Mexico prior to their he had collected. J.P. Harrington was the first to link this exodus from New Mexico after the (Second) Pueblo group to Kiowa in 1910 when he published a list of like- Revolt of 1696. Also known as the Hopi-Tewa or the ly cognate words they shared. Further comparative evi- Hano Tewa, the Arizona Tewa have continued to main- dence was amassed until the relationship was tain their language for more than 300 years despite definitively established by Kenneth Hale in 1967 when their relocation and accommodation to their more he successfully reconstructed the phonology of Proto- numerous Hopi neighbors. Kiowa-Tanoan––the ‘protolanguage’ or antecedent ancestral language, which precedes the emergence of the Kiowa, Towa, Tiwa, and Tewa branches. Some Structural Features of Kiowa-Tanoan Although Kiowa is the most recently detected Languages member of the family, it is not necessarily the most Most Kiowa-Tanoan languages have a large inventory divergent. The apparent divergence is greatly magni- of consonants, which includes a four-way contrast fied by Kiowa’s relative removal from recent contact between voiced, voiceless, aspirated, and glottalized with other related languages and by linguistic adapta- stops (e.g. b, p, ph, p’––where the voiceless bilabial is tion to a rather different cultural area. Although the accompanied by a closure of the glottis as in the Tanoan languages clearly share many more vocabulary American English expression ‘oh-oh’). The vowel sys- items, formal resemblances in grammar crosscut the tem usually contains six vowel qualities but also adds family and make Kiowa appear more like just another vowel length and an oral/nasal contrast to expand the related language rather than like one that is marginal- inventory. Some of the Kiowa-Tanoan languages have ly related. well-attested tonal contrasts. In Arizona Tewa, which Researchers have posited a deeper historical linguis- has two tones, nouns pronounced identically in terms tic relationship that would relate the Kiowa-Tanoan lan- of consonants and vowels but contrasting in pitch may guage family to the Uto-Aztecan family of North and be different words––’shirt’ (low tone) and ’digging Meso-America. The linguist Edward Sapir, in his clas- stick’ (high tone). The languages all have an elaborate sifications of 1921 and 1929, was the first to posit such verbal morphology. Verbs obligatorily consist of a a relationship and this largely impressionistic assess- pronominal prefix that codes the subject (and some- ment of verbal morphology later led to a more system- times the object as well), the verb stem, and one or atic examination of possible cognate lexical items by more tense–aspect suffixes. But verbs can also include George Trager and Benjamin Whorf in 1937. Since that prefixes that encode negation and adverbial manner time, collective historical linguistic scholarship has information, incorporated objects (such as things undermined anything that might approximate a ‘proof’ owned or possessed), and suffixes indicating that the of relationship and suggested that existing research is at verb is part of a dependent clause rather than the main best inconclusive. At the present time, this posited rela- sentence. The intricate pronominal prefixes of these tionship must await new research before it can be defin- language encode the person (I, you, he/she/it) and itively confirmed or disconfirmed. number (singular, dual, and plural (>2)) of the subject and inflect for stative and possessive constructions. Contemporary Speech Communities Other prefix sets supply pronouns for transitive sen- tences which often distinguish between agentive, Like virtually all indigenous languages that do not reflexive-reciprocal, and inverse constructions. The enjoy state support, many of the existing Kiowa- last of these provides a means of making the recipients Tanoan languages are spoken by relatively small popu- of action rather than its agents the subject of the sen- lations in which speakers are not evenly distributed tence. As such, they resemble passive sentences in across all generations. For the Kiowa, who now live English, but unlike English passives, Tewa inverse primarily in southwestern Oklahoma in Caddo, Kiowa, constructions remain fully transitive. The most and Comanche counties, there are about 300 adult 1092  TEWA AND THE KIOWA-TANOAN LANGUAGES speakers exhibiting a wide range of fluency. Towa, spo- While linguistic and speech-making skills were an ken at Jemez, is reported as spoken throughout the important resource associated with tribal leadership community of about 2,000. The Northern Tiwa-speak- among the Kiowa, the role of cultural models of speak- ing community of Taos, according to surveys conduct- ing provided by the theocratic elite is especially pro- ed in 1980, has about 800 speakers across all nounced in Tanoan communities. In many of these generations and represents, after Jemez, the next most communities, the religious language of the kiva––the vital of Kiowa-Tanoan speech communities. Picuris ceremonial chambers of native religious activity––pro- has slightly more than 100 speakers. Of the Southern vided a powerful, ideal model for everyday speech. Tiwa dialects, Sandia has about 150 while Isleta has ten Since this ceremonial register embodied both religious times that number. The Ysleta del Sur dialect is not flu- authority and the political power of the religious elite, ently spoken anymore. Some members of the commu- it provided a natural symbol of good and effective nity remember some of the distinctive vocabulary of speech, which accounts for the usual characterization their ancestral language. As for Tewa, estimates based of Pueblo cultures as ‘linguistically conservative’. on surveys of various kinds count 1,200 speakers of Such a cultural standard provides a model analogous to Rio Grande Tewa and about 300 for Arizona Tewa. In the standardized languages of state societies, which both communities, there are children still learning the enjoy a high prestige, in part, because they are the offi- language, although the percentage still learning the lan- cial languages of state-owned and controlled institu- guage in their homes may be slightly less than 50%. tions like public education and the broadcast media. Language renewal programs of various types exist in Some of the ideals embodied by kiva speech, which Southern Tiwa, San Juan Tewa, and some other com- have had a demonstrable effect on speakers of Tanoan munities as resources used by these native communi- languages, are its emphasis on indigenous purism and ties in their ongoing efforts to control and maintain the valorization of traditional forms. In the ceremonial their heritage languages. speech of the kiva, no mixing of language is permitted; such mixing is in fact considered a punishable offense. Given this cultural ideal of linguistic purity, it is not sur- Language and Culture prising to find well-attested patterns in all the Tanoan In all Kiowa-Tanoan speech communities, as in Native languages of rejecting possible loanwords in favor of American communities more generally, speech is extending or combining existing terms. For example, the highly regarded as a powerful form of action and not first word for automobiles among the Arizona Tewa was reduced, as in some Western folk views, to labeling not a loanword from English ‘car’ or ‘auto’ but rather the functions that linguists call ‘denotation’. In all the word wa-tege (wind-wagon). For both Rio Grande and Kiowa-Tanoan-speaking communities, traditional sto- Arizona Tewa, patterns of linguistic purism have been rytelling was viewed as an essential aspect of language well documented by Edward P. Dozier (Santa Clara socialization. In these stories children would experi- Tewa) and Paul V. Kroskrity. Although bilingualism in ence the early moral training of learning from the pos- Spanish for the Rio Grande Pueblos is traceable to the itive and negative models provided by the story colonial occupation of the seventeenth century, less than characters. In all communities, lively and engaging 5% of the Rio Grande Tewa vocabulary is from Spanish, performances––typically by parents and grandpar- and for the Arizona Tewa who actively removed them- ents––called attention to the dramatic power of lan- selves from Spanish domination around 1700, less than guage to entertain and instruct. All narrators use 20 Spanish loanwords are attested. For the Arizona ‘hearsay’ grammatical devices to traditionalize and Tewa, this pattern of resistance to loanwords has also create a proper sense of authenticity for their narra- extended to their Hopi neighbors. Despite bilingualism tives. Storytellers in the Kiowa tradition tell their sto- in Hopi, pervasive intermarriage, and other forms of cul- ries collaboratively with their audiences. They expect tural integration, they only have two Hopi loanwords. In considerable audience involvement as both narrators a similar fashion, the ceremonial emphasis on reproduc- and audience members openly comment on the story ing traditional forms as precisely as possible extends to as it progresses. In the Pueblo cultures of the Tanoan a cultural value on traditionalizing narrative and other communities, audience involvement is more limited to speech genres so that speakers lend their voices to the stylized interjections that encourage the storyteller to tradition of ‘speaking the past’. continue that the narrative and that demonstrate the involvement of the audience. Kiowa narrators espe- References cially value humor and use many puns toward that end, Davis, Irvine. 1979. The Kiowa-Tanoan, Keresan, and Zuni lan- whereas the verbal artistry of Tanoan storytellers is guages. The languages of native America: historical and more centered on unfolding their stories through the comparative assessment, ed. by Lyle Campbell and Marianne dialogue of their characters. Mithun, 390–443. Austin: University of Texas Press. 1093 TEWA AND THE KIOWA-TANOAN LANGUAGES Kroskrity, Paul V. 1993. Language, history, and identity: ethno- Trager, George L. 1946. An outline of Taos grammar. linguistic studies of the Arizona Tewa. Tucson: University of Linguistic structures of Native North America, ed. by Harry Arizona Press. Hoijer et. al., 184–221. Viking Fund Publications, No. 6. Kroskrity, Paul V. 1998. Arizona Tewa Kiva speech as a mani- New York: Viking Fund. festation of a dominant language ideology. Language ide- Trager, George L. 1967. The Tanoan settlement of the Rio ologies, practice and theory, ed. by Bambi B. Schieffelin, Grande area: a possible chronology, ed. by Dell H. Hymes Kathryn A. Woolard, and Paul V. Kroskrity, 103–23. New and William E. Bittle, 335–49. The Hague: Mouton. York: Oxford University Press. Watkins, Laurel J. 1984. A grammar of Kiowa. Lincoln, Palmer, Gus. 2003. Telling stories the Kiowa way. Tucson: Nebraska: University of Nebraska Press. University of Arizona Press. PAUL V. KROSKRITY Text Linguistics In the 1960s, widespread dissatisfaction with prevail- syntactic rule of English is the categorical ing models of linguistics, particularly in continental subject–verb–object order as in The baby (S) loves (V) Europe and the United Kingdom, led to the rise of her new toy (O). Violating that order by switching the ‘text linguistics’ as a new linguistic subdiscipline. At constituents around as in Loves (V) her new toy (O) that time, phonetics/phonology, morphology, and syn- the baby (S) produces an ungrammatical sentence. tax were considered to be the core fields of linguistic However, textual rules may easily be flouted in certain inquiry. Furthermore, most linguists restricted their contexts or for stylistic reasons and the text will still attention to the structure of the language system remain acceptable. For instance, a textual rule would (langue) at the expense of language in use (parole). be that pronouns normally appear after the element ‘Langue’ is the idealized knowledge of a language that they refer to, as in The baby loves her new toy. It makes a competent speaker has. ‘Parole’ refers to the less such a funny sound, where ‘it’ refers back to ‘her new than perfect utterances this knowledge results in, char- toy’. However, it is easy to image a context in which it acterized, for instance, by errors, repetitions, and hes- would be perfectly acceptable to change the order itations due to processing constraints or not paying around: It makes such a funny sound. The baby loves attention. The language system was thought to provide her new toy. a window into the human mind, while ‘language in At that point, text linguistics began to examine the use’ was deemed too unsystematic and limited to be of langue–parole distinction more critically. This was not real theoretical interest. a completely novel undertaking, as functional linguists Consequently, the two central concerns of early text had always been wary of the distinction. Functional linguists were (a) linguistic modeling above the level linguists of the Prague School (founded as the ‘Cercle of the sentence, and (b) a focus on language in use. Linguistique de Prague’ in 1926 by Villem Mathesius, The relationship between these two concerns has Joseph Vachek, and others), for instance, had always changed over time, however. Initially, most text lin- maintained that language can only be described on the guists wanted to show only that structural properties basis of authentic speech samples. As a result, they extended beyond the level of the sentence. In the had also dismissed the strict distinction between 1960s, Peter Hartmann, a German linguist, felt that the langue and parole. In addition to the Prague School of emphasis on phonemes, morphemes, and sentences functional linguistics, an important British group of was misplaced, as none of these elements alone con- functional linguists centered around John Firth, stituted communication. Communication only occurs Michael Halliday, and John Sinclair, a recent offshoot when these units are joined together in a text and, con- of which is the Sydney School. As a consequence of sequently, he viewed the text as the primary linguistic the developing interest in texts, functional sign (‘das originäre sprachliche Zeichen’). The search linguistics—a marginalized type of linguistic for a ‘text grammar’ or ‘text syntax’, therefore, inquiry—gained in importance. It became increasing- abounds in this early work. However, it soon became ly apparent to text linguists that the only way to inves- apparent that textual rules do not have the same status tigate textual structures was by turning to corpora of as syntactic rules. The flouting of a syntactic rule naturally occurring language. Text linguists also results in an ungrammatical sentence. For instance, a became increasingly interested in the patterns of 1094  TEXT LINGUISTICS speech (initially, ‘texts’ had been thought of mainly as instance, if a text about text linguistics is written for an written language). encyclopedia entry, the language used will differ sig- In 1981, Robert de Beaugrande and Wolfgang nificantly from a text about the same topic that is pro- Dressler developed a set of criteria in an effort to duced as so-called ‘cocktail party talk’. ‘Tenor’ refers define the characteristics of text as both (a) a unit of to the role structure that pertains between the interac- language above the sentence level and (b) language in tants, i.e. a text that is produced as a student–teacher use. These seven criteria are cohesion, coherence, exchange will differ from one that is produced as a intentionality, acceptability, informativity, situationali- marital communication. Finally, ‘mode’ refers to the ty, and intertextuality. symbolic organization of the text (e.g. written vs. spo- First and foremost, texts are characterized by inter- ken language). nal links. The word ‘text’ derives from Latin ‘textus’, The increasing focus on ‘language in use’ has which means ‘(inter)woven’. Grammatical links are blurred the boundaries between text linguistics and known as ‘cohesion’ and semantic links as ‘coherence’. other fields of linguistics that share this concern, such Cohesion may be achieved through repetition, pro- as conversation analysis, discourse analysis, the nouns, ellipsis, conjunctions, and similar devices. ethnography of communication, interactional sociolin- However, textual linkage does not suffice to make a guistics, or pragmatics. Even if it seems likely that dis- text. In the following example, all the links are present: course analysis will eventually fully absorb text ‘I went to the library. Libraries contain books. Big vol- linguistics, text linguistics has to be credited with umes are heavy’. Despite the fact that ‘library’ and opening up new horizons for linguistics––by looking ‘libraries’ are linked through repetition, and ‘books’ beyond the sentence, by trying to theorize context as and ‘volumes’ through synonymy, it is difficult to part of grammar, and by opening the doors for inter- image a context where this would be an appropriate disciplinary cooperation with education, literary text. The textual characteristics of ‘intentionality’ and studies, and other fields. ‘acceptability’ refer to the fact that it is essential that a sender (the speaker or writer) intends to produce a well-formed text, and that addressees (hearers or read- References ers) are prepared to accept the text as a communication Brinker, Klaus, Gerd Antos, Wolfgang Heinemann, and Sven F. that makes sense to them. Both senders and addressees Sager (eds.) 2000. Linguistics of text and conversation: an international handbook of contemporary research, Vol. 1. usually rely on their encyclopedic knowledge (knowl- Berlin and New York: Mouton de Gruyter. edge they have about the world that is not encoded in a de Beaugrande, Robert. 1997. New foundations for a science of particular text), in order to encode and decode texts. text and discourse: cognition, communication, and freedom ‘Informativity’ refers to the relationship between old of access to knowledge and society. Norwood, CT: Ablex. and new information that a text contains. A text con- de Beaugrande, Robert. 1980. Text, discourse and process. London: Longman. taining only old information is dull; a text containing de Beaugrande, Robert, and Wolfgang Dressler. 1981. only new information becomes impossible to process. Introduction to text linguistics. London: Longman. Therefore, senders produce texts that are in line with Eggins, Suzanne. 1994. An introduction to systemic functional what they believe their addressees will already know, linguistics. London: Pinter. or not yet know. ‘Situationality’ refers to the factors Halliday, M.A.K., and Ruqaiya Hasan. 1976. Cohesion in English. London: Longman. that make a text appropriate to a particular situation. Harweg, Roland. 1968. Pronomina und Textkonstitution. Finally, texts are ‘intertextual’ because they depend Heidelberg: Fink. upon a knowledge of other texts to be understood. Such Martin, James R. 1992. English text: system and structure. intertextual knowledge may refer to text types (we have Amsterdam and Philadelphia, PA: Benjamins. certain expectations about what a letter looks like, or an Petöfi, Janos S. (ed.) 1988. Text and discourse constitution: empirical aspects, theoretical approaches. Berlin and New encyclopedia entry) or to allusions (the slogan ‘Just say York: Mouton de Gruyter. know’ as used in a students’ union campaign alludes to van Dijk, Teun A. (ed.) 1977. Text and context. London: the antidrug campaign ‘Just say no’, and the intertextu- Longman. ality forms an additional layer of meaning). ––––––. 1978. Textwissenschaft. Tübingen: Niemeyer. As the two understandings of text (as linguistic unit Wilson, John, and Srikant Sarangi (eds.) 2000. Text linguistics at the millennium: corpus data and missing links. Special edi- above the level of the sentence, and as language in use) tion of Text, 20:2. Berlin and New York: Mouton de Gruyter. became increasingly integrated, the need for a gram- INGRID PILLER matical model able to incorporate the context has risen in importance. Halliday and other systemic functional See also Coherence in Discourse; Discourse Analy- linguists suggest that textual variation forms patterns sis; Firth, John Rupert; Genre; Halliday, Michael along three dimensions of the context: field, tenor, and Alexander Kirkwood; Relevance in Discourse; mode. ‘Field’ refers to context as social action. For Stylistics 1095 TEXT UNDERSTANDING Text Understanding Text understanding (Grosz et al. 1989; Scha et al. 1987; offer a densely packed interpretation context with tax- Singer 1990) is a cognitive process in which a text, i.e. onomic concept hierarchies, as well as default proper- a written sequence of ‘coherent’ sentences, is read and ties and predictive expectations that guide the the human reader is able to decode all or most of the recognition and incremental interpretation of objects propositions encoded in that source. (In case the read- and events. Text understanding processes operate on er is a computer, we refer to this process as automatic facts and basic propositions comprised of this a priori text understanding (Schank and Christopher 1981)). knowledge, either by finding a direct match with given Determining these propositions is, first of all, a chal- knowledge structures, or by inferring a proper linkage, lenging semantic/conceptual recognition problem. i.e. by employing reasoning mechanisms in order to Because texts such as newspaper articles, cooking link single propositions from a text to that background recipes, or business letters contain a large number of knowledge (Schank and Christopher 1981). Most of propositions, two additional layers help to structure the these inferences are merely plausible (e.g. inductive, textual information flow. One is constituted by struc- abductive, uncertain, or default-based) rather than tural organization patterns of texts themselves, which strictly deductive. As a result of incrementally assimi- reflect several mechanisms concerning how sets of text lating knowledge from texts, one acquires entirely new propositions are properly ‘packaged’. Conditions of propositions, modifies already known propositions, or textual well-formedness can then be stated in a ‘text’ even retracts propositions completely. grammar much in the same way as a standard sentence Text propositions are considered to be basic con- grammar reflects sentential and phrasal well-formed- ceptual units such as predicate-argument pairs. For ness criteria. The second important issue concerns the instance, in ‘Peter met John on the beach. They spent pragmatic factors underlying text understanding, i.e. three hours together.’, we distinguish a MEETING-1 assumptions both writers and readers rely on (such that event whose actors are PETER and JOHN, whose loca- certain beliefs are shared or implicit goals can be rec- tion is ON-BEACH and whose duration is 3-HOURS. ognized) in order to discern the intended (not only the From this, we may construct the following predicate- literal) interpretation of a text. Therefore, we will con- argument representations as the main propositions: sider text understanding from these three different MEET-ACTORS (MEETING-1, {PETER, JOHN}) angles, viz. the semantic/conceptual, structural, and MEET-LOCATION (MEETING-1, ON-BEACH) pragmatic level (Grosz and Candace 1986). MEET-DURATION (MEETING-1, 3-HOURS) In order to decode the contents encoded in a text, linguistic knowledge from various sources must be Even this small text fragment already features a pat- combined—knowledge of the morphological structure tern of text structure that guides our interpretation of of the words and their organization in the lexicon, as both sentences. We may only derive the third proposi- well as grammatical knowledge in terms of syntax and tion (specifying the duration of the meeting) when we semantics. Rather than just extending truth-functional interpret the pronoun ‘they’ at the beginning of the sec- semantic sentence analysis up to the text level (Kamp ond sentence as denoting ‘Peter and John’ as they and Uwe 1993), text understanding—and hence the meet on the beach. This is an example of a reference computation of text propositions—is deeply rooted in due to anaphora: an anaphoric expression (‘they’) the knowledge of the textual domain. This background referring to some referents previously introduced in knowledge has to be made explicit in terms of a the text (the antecedent)—in this example, the ad hoc domain knowledge base, a repository that contains group of people is composed of ‘Peter and John’. common-sense knowledge, as well as fairly special- Anaphoric reference is one of the basic text cohesion ized domain knowledge, e.g. from science, sports, or mechanisms that create local connections between economy. adjacent sentences at the semantic/conceptual level. As far as the organization of that knowledge is con- In order to account for such cohesion phenomena, cerned, research in text comprehension indicates that it special representational devices for the history of a seems beneficial to have structured knowledge repre- discourse (such as focus stacks, focus spaces, or cen- sentation formats (such as frames, scripts, or semantic tering lists (Walker et al. 1988)) must be available. networks) available. Such representation structures They help to keep track of the attentional structure of 1096  TEXT UNDERSTANDING a text, e.g. by recording the kinds of specific discourse has not understood a text, several proposals have been entities mentioned in the text, the order and syntactic made concerning how to assess the depth and accura- embeddings in which they were introduced, etc. Such cy of text understanding in a more or less indirect way. a device makes explicit the context created by the var- The following tasks have been considered useful in ious utterances in a text in terms of the salience status reaching that goal: of discourse entities and, thus, helps to minimize the • In the paraphrasing task, subjects are asked to number of alternative interpretations (ambiguities) rewrite the contents of a source text in their own that are likely to otherwise arise. words, while keeping in mind any piece of infor- Text cohesion mechanisms are supplemented by mation they have acquired from that source text. text coherence patterns that relate to the global organi- By way of comparing the propositions of the zational structure of a text. Examples of this are vari- source text with those from the newly generated ous forms of coherence relations, e.g. causality, paraphrase text, the level of detail of text under- contrast, or elaboration (Mann and Sandra, 1988), standing can be made explicit. hierarchically linking simple, as well as more and • In the question answering task, subjects are more complex proposition sets from the text. asked particular questions whose answer is con- Moreover, text macroorganization patterns such as tained only in the source text. The answer to a genre-specific superstructures or story grammars con- question may either be explicitly stated in the strain the overall structure of the entire text (e.g. the underlying text, or it may have to be inferential- composition rules underlying a weather report or a ly derived when it is merely implicit in the hospital discharge summary) (van Dijk 1977). proposition set. Correctly answering a question Discourse structures of this type can be signaled by is then taken as an indication of profound under- cue phrases (such as ‘but’, ‘for example’, ‘however’) standing of the source text. and are, therefore, easily recognized, but they may • In the summarizing task, subjects are asked to also require elaborate inferencing on rich knowledge compress the contents of a source text with structures. respect to the most relevant propositions con- At the pragmatic level (Cohen et al. 1990), text tained in it. Psycholinguists have a clear under- understanding relies on discourse constraints that hold standing of how one moves from a large set of irrespective of particular utterances in a text and char- basic propositions (micropropositions) to a gener- acterize a generally valid framework of human com- alized set of so-called macropropositions, which munication. They require the representation and is the equivalent of a summary at the representa- recognition of various presuppositions (knowledge tion level. Transformations operating on micro- assumed to be mutually shared between the writer and propositions generate macropropositions by way the reader), goals, plans, and intentions of the writer, of generalizing single statements along taxonom- and adherence to truly general communication princi- ic concept hierarchies, in turn eliminating irrele- ples based on plausibility and rationality (such as sev- vant details and abstracting single actions in more eral conversation maxims). These underpinnings of general event patterns (Kintsch and Teun 1978). text understanding are intended to keep it focused by Producing a reasonable summary is then consid- relying on a set of reasonable, socially accepted con- ered to be evidence of an adequate understanding ventions for felicitous and compact textual informa- of the source text. tion transfer. The pragmatic level of text understanding implies that utterances in a text contain more than they literally indicate in terms of semantic/conceptual information (a phenomenon called conversational References implicature). The reader’s construal of the implicit Cohen, Philip R. Jerry Morgan, and Martha E. Pollack (eds) intentions, plans, and beliefs that the writer had for 1990. Intentions in communications. Cambridge, MA; producing the text constitutes the key for a successful London, UK.: Bradford Book and MIT Press. Grosz, Barbara J., Martha E. Pollack, and Candace L. Sidner textual information transfer (Morgan and Georgia 1989. Discourse. editor, Foundations of cognitive science, 1980). ed. by Michael I. Posner, 437–68. Cambridge, MA; London, A problem with defining text understanding in UK: Bradford Book and MIT Press. terms of the recognition of text propositions is that it Grosz, Barbara J., and Candace L. Sidner. 1986. Attention, is almost impossible to reach a consensus among dif- intentions, and the structure of discourse. Computational Linguistics 12(3), 175–204. ferent readers as to what exactly constitutes the prop- Kamp, Hans, and Uwe Reyle. 1993. From discourse to logic. er set of propositions of an underlying text. Because Introduction to modeltheoretic semantics of natural lan- this makes it difficult to judge whether a reader has or guage, formal logic and discourse representation theory, Vol. 1097 TEXT UNDERSTANDING 42 of Studies in linguisitics and philosophy, 2 vols. gence. Vol. 1, ed. by Stuart C. Shapiro, 233–45. New York: Dordrect: Kluwer. John Wiley. Kintsch, Walter, and Teun A. van Dijk. 1978. Toward a model Schank, Roger C., and Christopher K. Riesbeck. 1981. Inside of text comprehension and production. Psychological computer understanding: five programs plus miniatures. Review 85(5), 363–94. Hillsdale, NJ: Lawrence Erlbaum. Mann, William C., and Sandra A. Thompson. 1988. Rhetorical Singer, Murray, 1990. Psychology of language. An introduction structure theory: toward a functional theory of text organi- to sentence and discourse processes. Hillsdale, NJ: zation. Text 8(3). 243–81. Lawrence Erlbaum. Morgan, Jerry L. and Georgia M. Green. 1980. Pragmatics and van Dijk, Teun A. 1977. Semantic macro-structures and knowl- reading comprehension. Theoretical issues in reading com- edge frames in discourse comprehension. Cognitive process- prehension. Perspectives from cognitive psychology, lin- es in comprehension, ed. by Marcel A. Just and Patricia A. guistics, artificial intelligence, and education, ed. by Rand J. Carpenter, 3–32. Hillsdale, NJ: Lawrence Erlbaum. Spiro, Bertram C. Bruce, and William F. Brewer, 113–40. Walker, Marilyn A., Aravind K. Joshi, and Ellen F. Prince (eds) Hillsdale, NJ: Lawrence Erlbaum. 1998. Centering theory in discourse. Oxford, UK: Oxford Scha, Remko J. H., Bertram C. Bruce, and Livia Polanyi. 1987. University Press. Discourse understanding. Encyclopedia of artificial intelli- UDO HAHN Thai and Tai Languages Thai (Bangkok Thai, Standard Thai, Central Thai, or Tai Ya and Tai Lue in Chiang Rai, and Phu Thai in Siamese) is the national language of Thailand. It is Nakhon Phanom. spoken by approximately 60 million speakers, at least The word Tai (without the h) has two major uses: half of whom speak other languages and dialects as (1) it refers to a group of languages, including Thai their mother tongues. The standard language, based on and Lao, which are spoken in a vast geographical area the speech of the elite and the upper middle class in of Southeast Asia ranging from southwestern China Bangkok, is used as the medium of instruction in and northern Vietnam to the Indian state of Assam; (2) schools throughout the country. Regional dialects out- it is the name of a language family, the Tai family, a side Bangkok and the Central Plains include Northern subgroup of Tai-Kadai or Kam-Tai. The term Kadai Thai (Kam Muang or Lanna), Northeastern Thai (Isan was coined by Paul K. Benedict to refer to a group of or Lao), and Southern Thai (Paktai). Although these languages closely related to Tai such as Li (Hlai), varieties are generally considered dialects of Bangkok Lakkia (Lajia), and Gelao, which are spoken mainly in Thai, they are more closely related to other languages China. The relationships between Tai and Kadai or Tai of the same family spoken in neighboring countries. and Kam (Dong) languages as well as between Tai- For example, Northern Thai is more similar to Shan or Kadai and other language families in Southeast Asia Tai Yai in northern Burma, and Northeastern Thai to are still controversial. According to Benedict, the Tai Lao. Strong regional identities exist in these regions, family is a branch of Austro-Tai, because it shows particularly among speakers of Southern Thai (most of strong affinities with Austronesian, another major whom are Muslim Malays) and northeasterners who family in Southeast Asia comprising more than 1,000 call themselves and their language Lao. The name languages. Scholars in China, on the other hand, view Thai-Isan is a newly coined term introduced by the Tai (so-called Daic) as a branch of Sino-Tibetan, a Thai kingdom to promote nationalism among north- superfamily that consists of Chinese (Sinitic), Tibeto- eastern speakers. Apart from Thai and its dialects, Burman, and Hmong-Mien (Miao-Yao). A view gen- other languages spoken in Thailand include Chinese erally held by linguists today is that Tai-Kadai or Daic (mainly in Bangkok and the central region), Malay in is a separate language family, which has no proved the south, and Mon-Khmer languages in the northeast. connection with either Austronesian or Sino-Tibetan. In addition, there are a number of hill tribe languages Hypotheses about Proto-Tai have been debated since such as Karen, Hmong, Yao, Lisu, Lahu, and Akha in the turn of the twentieth century and have motivated Tak, Chiang Mai, Chiang Rai, and Mae Hong Son in a good number of works on comparative Tai linguis- the north. Other Tai varieties are also spoken in scat- tics, particularly in the areas of pronunciation and tered communities, such as Tai Yai in Mae Hong Son, vocabulary. 1098  THAI AND TAI LANGUAGES Within the Tai family, Fang Kuei-Li distinguished (Cambodian) and subsequently developed to accom- three branches: Southwestern Tai (comprising the well- modate special characteristics of the Thai language as documented languages of Assam, Burma, Yunnan, it was spoken then. On the evidence of the Thailand, Laos, and northern Vietnam), Central Tai (the Ramkhamhaeng Inscription (1292 CE), Sukhothai most notable members are Nung and Tho spoken in Thai included four kinds of syllables. The first three northern Vietnam and areas of Guangxi adjacent to categories, ending in a vowel or sonorant (so-called northeastern Vietnam), and Northern Tai (e.g. Northern unchecked, live, or free syllables), used contrasting Zhuang, Buyi, Yay, Saek, and Li). Zhuang is spoken in tones (see below), whereas the fourth kind, ending in the Zhuang Autonomous Prefecture in Guangxi and stops /p t k/ (checked or dead syllables), exhibited no Wenshan County in Yunnan. The Zhuang constitute the tonal contrasts. In the Ayutthaya period (fourteenth to largest of China’s 56 official minority nationalities, eighteenth centuries CE), the language underwent two numbering nearly 20 million speakers. There are two major changes, which resulted from tonal splitting and main Zhuang languages: Northern Zhuang (already from borrowing words from Khmer, Pali, and Sanskrit. mentioned) and Southern Zhuang. The latter shows The tonal split, which also affected other Tai lan- affinity with Central Tai languages and is classified as guages, although at a much earlier date, was condi- a member of this group. Spoken in the upper Red River tioned by the nature of the initial consonant and split area in Vietnam, Yay (Giy) has been richly described by the system of three tones into a system of five in William J. Gedney. Saek (previously classified as a Modern Thai. Some Tai languages have been reported Mon-Khmer language) is spoken in Laos and Thailand. to have as many as nine tones. It has been reported to maintain the archaic final con- The inventories of sounds in Central and Northern sonant -l. Saek has also been investigated by Gedney. Tai are generally different from those in the In China, speakers of Tai languages are centered in Southwestern branch. For example, Nung and Tho autonomous prefectures and counties in the south- have peculiar initials such as py- or phy- for words that western provinces of Yunnan, Guizhou, Guangxi, share ancestry with those beginning with t- in the along the border region of Hunan, and Hainan Island. Southwestern varieties. Li (Hlai), a Northern Tai vari- They are officially recognized as the Dai (China’s ety from Hainan Island, has preglottalized stops and Latin-script equivalent of Tai, referring to speakers of contains the rare voiced stop g. Southern Zhuang con- the southwestern branch), Zhuang, Buyi, Dong, Li, tains an aspirated stop series but Northern Zhuang has and Gelao. The majority of the Dai speak Tai Lue and only unaspirated stops. are concentrated in Xishuangbanna Dai Autonomous A syllable in Thai and Tai languages consists of an Prefecture in Yunnan. Tai Nuea or Tai Luea is spoken initial consonant or consonant cluster, a vowel or diph- in Dehong-Dai-Jinpo Autonomous Prefecture along thong, and a tone, and may or may not have a final the Yunnan–Burma border. This language shares sev- consonant. When the final consonant is present, it is eral characteristics with Tai varieties in Shan of upper either a voiceless stop /p t k ʔ/, a nasal /m n/, or a semi- Burma. White Tai (Tai Don), Red Tai (Tai Daeng), and vowel /y w/. The glottal stop /ʔ/ is an innovation in Black Tai (Tai Dam), all belonging to the southwestern modern Tai languages; it probably did not exist in the branch, are spoken in various places in northern ancestor language Proto-Tai. Words that contrast in Vietnam. They are mutually intelligible and can be tones convey different meanings. For example, the fol- classified as dialects of a single language. Black Tai lowing words in Bangkok Thai mean different things speakers, numbering approximately 500,000, were depending on the pitch with which they are pro- originally centered in Muang Thaeng (Dienbienphu) nounced: naa1 ‘field’, naa2 ‘sentence-final particle’, in Vietnam but they have now settled in Laos and sev- naa3 ‘face’, naa4 ‘mother’s younger sibling’, naa5 eral provinces in Thailand. ‘thick’. The numbers 1–5 refer to different qualities of With a writing system based on Indic-derived pitches: 1 = mid level, 2 = low level, 3 = falling, 4 = scripts, several Tai languages have enjoyed long liter- high level, and 5 = rising. Checked syllables have ary traditions. Although Ahom has become virtually fewer tonal possibilities than unchecked syllables. extinct as a spoken language since the nineteenth cen- Black Tai, for example, chooses from six tones on tury, its literary legacy provides an important clue to unchecked syllables but only two tones on checked Tai-Ahom worldviews. The use of Lanna script in syllables. Weak syllables are also common in Thai and northern Thailand is today limited to religious texts. Tai languages. The tone on such syllables is usually The use of the present-day Thai script for the standard neutralized. The Thai kra- in krathaʔ4 ‘pan’, krathaŋ5 written language in Thailand can be traced back to the ‘flowerpot’, is a clear example of this type. reign of King Ramkhamhaeng during the Sukhothai Most inherited Tai words are monosyllabic. The sit- period in the mid thirteenth century. The Sukhothai uation in Thai is more complicated due to words bor- writing system was borrowed from Khmer rowed from other languages. Pali loans and compounds 1099 THAI AND TAI LANGUAGES greatly increase the number of polysyllabic words in classifiers in Bangkok Thai to be of limited use and the language and distinguish it from the Tai languages some to be used more extensively. For example, of China, which are heavily influenced by Chinese. chɯak3 ‘classifier for elephants’ is being replaced by Contacts with other languages thus result in finer lexi- tua1 ‘classifier for animals’ in young people’s speech. cal distinctions between dialects of a single language. In the same way, ʔan1 ‘classifier for things’ has been For example, Tai Ya speakers from Muang Ya in extended to cover other nouns that formerly required Yunnan have difficulty communicating with their rela- special classifiers. tives in a transplanted community in Chiang Rai. Pronouns in Thai and Tai languages are of special Another common feature of Thai and Tai as well as interest to linguists, as there are many ways to express non-Tai languages in Southeast Asia is the use of dou- ‘I’ and ‘you’ depending on the sex of the speaker, the blets and elaborate expressions. These are two- to four- relationship between the speaker and the addressee, syllable words which contain repeated or synonymous and the formality of the situation. Kinship terms, information. They usually convey idiomatic or figura- words expressing occupations, and nicknames can be tive meanings and characterize playful speech. For used in this context. Furthermore, these languages example, the expression in Bangkok Thai kin1 khaaw3 employ a number of sentence-final particles, another kin1 plaa1 (eat + rice + eat + fish) simply means ‘to areal feature in Southeast Asia. These particles are eat’ but it is composed of two identical words (first and used to perform different kinds of speech acts and third syllables) and two words with similar meanings convey the speaker’s attitudes toward the utterance. (second and fourth syllables). Like pronouns, they are used in accordance with the Grammatical relations in Thai and Tai languages social status of the speaker with respect to the are indicated by word order, not by word inflections. addressee, degrees of intimacy, and politeness. Subject–Verb–Object is the usual pattern, although According to Gedney, pronouns and particles are other patterns occur: Object–Subject–Verb (for object what distinguish languages of this family which are emphasis), Verb–Subject (stating existence), and far apart from one another, rendering them mutually Subject–Object–Verb (as in Khamti, a Tai variety in unintelligible. Ahom). An important characteristic of spoken and written texts is that virtually everything except verbs or adjectives can be omitted. Pronouns are especially References prone to deletion. Nouns do not have plural forms, and Benedict, Paul K. 1975. Austro-Thai language and culture, with verbs express no tense. To indicate whether actions a glossary of roots. New Haven, CT: Human Relations Area have already taken place or will take place, the speak- Files Press. er counts on the addressee’s interpretation and makes Bickner, Robert J., John Hartmann, Thomas J. Hudak, and Patcharin Peyasantiwong (eds.) 1989. Selected papers on use of temporal adverbs, preverbal auxiliaries, as well comparative Tai studies, Ann Arbor: University of as serial verb constructions (series of verbs that occur Michigan. in concatenation without intervening words). It is still Compton, Carol J., and John F. Hartmann (eds.) 1992. Papers controversial as to whether there are true adjectives on Tai languages, linguistics, and literatures in honor of and prepositions in these languages, as the former William J. Gedney on his 77th birthday. Monograph Series on Southeast Asia Occasional Paper No. 16. DeKalb: Center generally behave in the same way as verbs and the lat- for Southeast Asian Studies, Northern Illinois University. ter can function as nouns. Diller, Anthony. 1994. Tai languages: varieties and subgroup Classifiers (category-words used in e.g. counting, terms. Thai-Yunnan Newsletter, No. 25, ed. by Gehan somewhat like English a slice of bread, a slice of Wijeyewardene. Canberra: Research School of Pacific pizza) are found throughout the family, suggesting that Studies, Australian National University. Edmondson, Jerold A., and David S. Solnit (eds.) 1997. this is an inherited trait from Proto-Tai. However, there Comparative Kadai: the Tai branch. Publications in are variations among the languages and dialects. For Linguistics No. 124. Dallas: Summer Institute of Linguistics example, Bangkok Thai and Tai Lue have the and the University of Texas at Arlington. Southeast Asian pattern with the following order: noun Gedney, William J. 1991. The Yay language glossary, texts and + numeral + classifier, whereas Black Tai and White translations, ed. by Thomas J. Hudak. Ann Arbor: Center for South and Southeast Asian Studies, University of Michigan. Tai follow the Chinese pattern with the order numeral ––––––. 1992. The Saek language glossary, texts and transla- + classifier + noun. Classifiers can substitute for mod- tions, ed. by Thomas J. Hudak. Ann Arbor: Center for South ified nouns in both Thai and less well-known varieties and Southeast Asian Studies, University of Michigan. such as Tai Lue and Tai Nuea. These linguistic forms Hartmann, John F. 1984. The linguistic and memory structure are generally used to indicate shape, size, flexibility, of Tai-Lue oral narratives. Monograph Series B-90. Canberra: Australian National University. and animacy. Although classifiers are similar in gram- Hudak, Thomas J. 1987. Thai. The world’s major languages, ed. mar and meaning in various Tai languages, they differ by Bernard Comrie. London: Croom Helm, and New York: in number and usage. There is a tendency for some Oxford University Press. 1100  THEMATIC STRUCTURE Li, Fang-Kuei. 1977. A handbook of comparative Tai. Oceanic Wijeyewardene, Gehan. 1990. Thailand and the Tai versions of Linguistics Special Publications No. 15 Honolulu: ethnic identity. Ethnic groups across national boundaries in University of Hawaii Press. mainland Southeast Asia. Singapore: Institute of Southeast Terwiel, B.J., and Ranoo Wichasin (translators and editors) Asian Studies. 1992. Thai ahoms and the stars: three ritual texts to ward off danger Ithaca: Cornell University Southeast Asia Program. KRISADAWAN HONGLADAROM Thematic Structure Thematic structure is a list of specifications in the lex- The notion ‘thematic relation’ was first introduced icon about the thematic relations that hold between a in Gruber’s Lexical structures in syntax and semantics predicate and its arguments. In some approaches, ‘the- (1976) and in Fillmore’s The case for case (1968), matic structure’ is referred to as ‘thematic grid’, ‘theta who called them ‘deep cases’. Fillmore assumed that grid’, or ‘argument structure’. each argument is associated with a preposition that ‘Argument’ and ‘thematic relation’ are lexical corresponds to its deep case. Every argument is insert- semantic notions. The event or state of affairs a verb ed with its preposition into a unique syntactic position denotes entails a fixed number of participants, which in deep structure. In the examples in (1), John has the mostly have to be expressed in every well-formed sen- deep case ‘agentive’, the door the deep case ‘objec- tence in which the verb is contained. These are the tive’, and the wind the deep case ‘instrumental’. The arguments of the verb. Although obligatory syntactic agentive and the instrumental take the prepositions by realization is not a necessary criterion for the identifi- and with, respectively, cf. (1a), (1c), and the preposi- cation of arguments, it is true that an argument is still tion of the objective is zero, cf. (1a–d). implicit in the meaning of the verb even if it is not (1a) John opened the door with the key. expressed in a sentence. In Fred eats an apple with (1b) The key opened the door. great relish, for example, the expressions Fred and an (1c) The door was opened by John. apple realize the arguments of the verb eat. If eat is (1d) The door opened. used intransitively, e.g. Fred is eating with great relish, we still understand that Fred is relishing something he ‘Agentive’ is, in Fillmore’s theory, ‘the case of the typ- eats. The lexical entry of a verb specifies for each argu- ically animate perceived instigator of the action iden- ment the way in which it is involved in the denoted tified by the verb’; ‘instrumental’ is ‘the case of the event. In the example above, Fred is the instigator of inanimate object or force causally involved in the the eating event and the apple is the object that is action or state identified by the verb’, and ‘objective’ affected. The roles participants may have in events ‘should be limited to things that are affected by the denoted by verbs are the thematic relations. Other com- action or state identified by the verb’; cf. Fillmore mon names in different theories are ‘thematic’ or ‘theta (1968:24f.). roles’ (‘Θ roles’), ‘semantic roles’, ‘participant roles’ In the course of the derivation of an active sentence, and so on. They are generally subsumed under labels an agentive moves to subject position, where its pre- like ‘agent’, ‘patient’, ‘goal’, ‘beneficiary’, and ‘expe- position is deleted, cf. (1a). If no agentive is present at riencer’. Theories differ with respect to the number of deep structure, the instrumental will move to subject labels they assume and the way in which they define position where its preposition is deleted. Even objec- the corresponding thematic relations. tives can move to subject position, if they are the only In linguistic theories, thematic relations serve as an argument in active sentences, cf. (1d), or in passive interface between lexical semantics and syntax. In dif- sentences, then the agentive remains with its preposi- ferent theories, there are various syntactic generaliza- tion in its original deep structure position, cf. (1c). tions that refer to thematic relations, but almost every Later approaches in Generative Grammar departed theory uses them overtly or covertly to predict the syn- from the postulation that arguments are inserted into tactic realization of arguments, e.g. as subject, object deep structure with prepositions according to their the- or indirect object, or with respect to case marking or matic relations. They nonetheless continued to adhere syntactic alignment of the corresponding noun phrase. to the idea that there is a one-to-one correspondence 1101 THEMATIC STRUCTURE between thematic relations and syntactic positions. agent effector experiencer locative theme Chomsky (1981) introduced in his Lectures on gov- patient source path goal recipient ernment and binding (GB) the ‘theta criterion’, which poses the following twofold restrictions on thematic Figure 1 structure: (1) each argument is assigned one and only one theta role; and (2) each theta role is assigned to one and only one argument. In (1a) and (1b), we have the choice of subject and the application of other syn- two different variants of the verb open. In (1a) the tactic rules is reflected by thematic hierarchies, cf. agent role is assigned to the subject, the direct object Van Valin’s (1993) Role and Reference Grammar is assigned the theme role, and the prepositional (RRG: 41) (see Figure 1). phrase is an instrument(al). (1b) differs from (1a) in On this continuum, thematic relations that are near- that the subject bears the instrument role. Both sen- er to the left end share more properties with the agent tences meet the theta criterion because there is no (‘the willful, volitional, instigating participant’, cf. argument without a theta role and no theta role is RRG: 42), while thematic relations that are nearer to assigned twice in one sentence. The theta criterion cor- the right end are more patient-like (‘the nonwillful, rectly rules out sentences like The key opened the door noninstigating, maximally affected participant’, cf. with the bit, where the instrument role is assigned to RRG: 42). ‘Effector’ is a generic term for instruments the key and the bit (unless the subject is interpreted as and forces, experiencers are, for example, nonwillful a character in a fable or cartoon), or John and the key instigators of perceiving events, and themes undergo a opened the door, where the collective subject is change of location but not a change of state or condi- assigned the instrument and the agent role at the same tion. Sources, paths, goals, and recipients are grouped time. In GB it was assumed that the verb in (1d) together, because they do not compete with each other assigns the theme role to the direct object, but that it for subjecthood or other syntactic relations. They do, fails to assign case to this argument. It was postulated however, compete with the theme. The mapping of that a verb’s failure to assign case to the direct object thematic relations to syntactic relations proceeds in the always correlates with its failure to assign a theta role following way: (1) The highest ranking thematic rela- to its subject. In other words, in the underlying struc- tion in a thematic structure is assigned the macrorole ture of (1d) the subject position is empty and the direct ‘actor’ and the lowest ranking thematic relation is object lacks a case feature. In order to receive case, the assigned the macrorole ‘undergoer’. If a verb has more phrase moves to the subject position and thus yields than one argument, in accusative languages the actor is the surface structure in (1d). Passive sentences like mapped to the subject and the undergoer to the direct (1d) were also accounted for by movement of the object. Any remaining arguments are assigned cases or direct object to the subject position, but in this case the prepositions according to their thematic relations. The theta role of the subject and the case of the direct choice of the particular case or preposition is lan- object was assumed to be absorbed by the passive guage-specific. While in example (1a) the agent John verb. (The apparent agent by John in (1d) does not rep- outranks the theme the door, in (1b) it is the effector resent an argument of the passive verb. Instead it is the key. Thus, John and the key are actors in (1a) and assumed to be an adjunct.) (1b), respectively, and the door is undergoer in both While in GB for most arguments a thematic and a sentences. In (1c) and (1d) open has only one argu- grammatical relation is specified in the lexical entry of ment, the theme the door. Since the verb does not the verb, other theories provide algorithms that predict express an activity in either case, the argument is an the grammatical relation or the morphosyntactic real- undergoer, but this does not affect the choice of the ization of arguments from their thematic relations. grammatical relations, since only arguments are Such theories are called ‘linking theories’. The exam- always realized as subjects. ples in (1) illustrate the fact that, in English, agents, One advantage of RRG is that it captures distinc- instruments, and themes or patients can all become tions between thematic relations like agent and effec- subject of active sentences. However, an agent will tor and at the same time allows for generalizations always be realized as subject while an instrument can- over a wide range of thematic relations, for example, not become subject if the thematic structure contains with respect to linking. Still a disadvantage of RRG is an agent, and a patient or a theme can only become that it operates with role labels. A thematic relation subject if it is the only argument in the thematic struc- that is covered by a role label is defined by a number ture. These generalizations seem to hold cross-linguis- of related properties (e.g. Van Valin’s definition of the tically for languages that have accusative case systems agent above), which may not all apply to every argu- like English. In some contemporary approaches, the ment for which the role label was intended. Because in different ranking of thematic relations with respect to every theory there is a residue of arguments that do not 1102  TIBETAN fit in any role label category or that do fit equally in structures with three arguments, it holds that the argu- two or more categories and whose assignment to one ment having less proto-agent properties than the sub- or another label often seems to be subject to theoretic ject and less proto-patient properties than the direct bias, many critics have rejected the theoretic value of object will be the indirect object. thematic relations altogether. Dowty’s (1991) Because of its obvious advantages, Dowty’s model Thematic proto roles and argument selection (TPR) has received wide acknowledgment in the literature. offers a solution to this problem: in this model, the- However, his linking rules have been subject to the matic relations are characterized by already familiar criticism that they refer to grammatical relations and properties like volitionality, etc. that are entailed by not to morphosyntactic features. As Primus’s (1999) the verb of its arguments. These properties come from Cases and thematic roles exhaustively shows, notions two finite sets that define the prototypical agent and like subject and object only apply consistently to lan- the prototypical patient, respectively, and combine guages with accusative case systems, but not to erga- freely with each other. The proto-agent properties are tive and split intransitivity languages. As a (1) volitional involvement in the event or state, (2) sen- consequence, Primus modified Dowty’s model in tience/perception, (3) causing an event or state in order to achieve universal validity, like it was intended another participant, and (4) movement relative to by Dowty in the first place. another participant. The proto-patient properties are (1) undergoes a change of state, (2) incremental theme References (a participant that successively undergoes a change of Chomsky, Noam. 1981. Lectures on government and binding. state in an event that does not end before every part of Dordrecht, The Netherlands: Foris. the participant has been affected, e.g. a pizza in Bert Dowty, David. 1991. Thematic proto-roles and argument selec- ate a pizza), (3) causally affected by another partici- tion. Language 67. pant, and (4) stationary relative to movement of anoth- Fillmore, James. 1968. The case for case. Universals in linguis- er participant (TPR: 572). Dowty’s model permits a tic theory, ed. by Emmon Bach and Robert Thomas Harms. New York: Holt, Rinehart and Winston. very fine-grained distinction of thematic relations Gruber, Jeffrey. 1976. Lexical structures in syntax and seman- since arguments can be characterized by a great vari- tics. Amsterdam, The Netherlands: North-Holland ety of properties from one or both concepts, or even by Publications. the absence of proto-role properties. On the other Jackendoff, Ray. 1983. Semantics and cognition. Cambridge, hand, due to the limited number of proto-role proper- MA: MIT Press. ––––––. 1987. The status of thematic relations in linguistic the- ties and the mutual independence of the properties, it ory. Linguistic Inquiry 18. is possible to generalize over very large classes of Primus, Beatrice. 1999. Cases and thematic roles: ergative, accu- arguments that share a certain property, regardless of sative and active. Tübingen, Germany: Niemeyer. other possible proto-role properties these arguments Van Valin, Robert. 1993. Advances in role and reference gram- may have. The linking rules in TPR correctly predict mar. Amsterdam, The Netherlands and Philadelphia, PA: Benjamins. that the only argument or the argument that has the greatest number of proto-agent properties in the the- KERSTIN BLUME matic structure of a verb is lexicalized as the subject. See also Chomsky, Noam; Fillmore, Charles; Gen- The argument that bears the greatest number of proto- erative Grammar; Grammatical Function; Lexi- patient properties becomes direct object. For thematic con: Overview. Tibetan Tibetan belongs to the Bodish branch of the Tibeto- by approximately five million people in six different Burman division of the Sino-Tibetan language family. nation-states: China, Burma, India, Nepal, Bhutan, and It is closely related to Newar and Tamang, indigenous Pakistan. Only in Bhutan is it the national language. languages of Nepal; distantly related to other Tibeto- Significant groups of Tibetan speakers are also found Burman languages such as Akha, Lahu, Karen, and in exile communities in India, Nepal, Bhutan, and sev- Burmese; and remotely related to Chinese. It is spoken eral western countries: at least 100,000 Tibetans fled 1103 TIBETAN Tibet in 1959 when China took it over. Of the five mil- Nepal who are influenced by Tibetan Buddhism such lion Tibetans, 2,100,000 speakers are concentrated in as the Tamang, Gurung, Thakali, and Manang. It has the Tibetan Autonomous Region, with Lhasa as the served as the standard written language for all learned main municipality, and 2,500,000 speakers are found in documents in Bhutan. The written variety from the several Tibetan autonomous prefectures and counties in nineteenth century, which was developed from four western provinces of China: Qinghai, southwest Classical Tibetan with influence from the Lhasa collo- Gansu, western Sichuan, and northwest Yunnan. The quial, is called Modern Literary Tibetan or Newspaper remainder are in the Himalayan regions covering the Tibetan. It is used widely in Tibetan communities in northern tip of Burma, northern Nepal, Bhutan, the China and elsewhere as the standard medium for Indian states of Arunachal, Sikkim, Himachal Pradesh, newspapers, magazines, and other kinds of modern and Uttar Pradesh, the states of Jammu and Kashmir writings, including radio broadcasts. Other varieties (divided between India and Pakistan), and a Pakistani that have developed from Classical Tibetan but are district of Baltistan. based on regional idioms also emerged, but their use is Tibetan-speaking people in the Himalayas are limited to particular regions. Written Tibetan, specifi- known as Bhotia (from the word bod, the name of cally Classical Tibetan, still represents the pronuncia- Tibet). A minority group in eastern Nepal numbering tion of the language as it was in the ninth century, approximately 14,000, more commonly known as the when it underwent a major reform. Hence, there are Sharpa (Sherpa), speaks the southern variety of considerable differences between Written Tibetan and Central Tibetan. The Bhutanese call their national lan- modern spoken dialects. For example, the greeting guage Dzhongkha and do not classify themselves as expression bkra-shis bde-legs ‘Good luck’ is pro- Tibetans. Ethnic Tibetans in China are officially desig- nounced trashi tele in Lhasa speech. nated as the Tibetan nationality (zàngzú in Chinese or Spoken Tibetan is divided into several dialects and bod-rig, a newly coined term in Tibetan). The Jiarong subdialects that are quite different from one another. (Rgyarong), Baima, and other ethnic groups of west- The number of dialects varies, depending on the geo- ern Sichuan consider themselves Tibetans and are offi- graphical areas under investigation. Most authorities cially included within the Tibetan nationality, although agree that there are at least four major dialects: Central they speak non-Tibetan languages as their mother Tibetan (Ü-tsang), Northeastern Tibetan (more com- tongues. The variety of Tibetan spoken in Baltistan is monly known as Amdo), Eastern Tibetan (Kham), and called Balti, but it acquires another name (Purik) when Western Tibetan (Töö). Central Tibetan, which spoken on the Indian side. includes Lhasa, the standard dialect, Shigatse, and With historical evidence dating back to the seventh other subdialects, is spoken in the Tibet Autonomous century, Tibet was an important empire of Central Asia Region (TAR) and along the Tibet–Nepal border. extending its influence to the Tang dynasty of Imperial Amdo is spoken in Gansu and Qinghai provinces. China and ruling over small kingdoms such as the Nan Kham is found in Sichuan, Qinghai and Yunnan Chao Kingdom (present-day Dali) in southwest China. provinces, and some parts of TAR. Western Tibetan It was in contact with several nation-states of Asia, includes Ngari in TAR, Ladakhi in India, and Balti in including India, from where it borrowed a writing sys- Pakistan. Chinese linguists pay attention only to the tem. The Tibetan script––with 30 consonant and four first three groups, as they are the major dialects spoken vowel signs and a few punctuation marks––was mod- in China. eled after the Brahmi script during the reign of King Within Kham Tibetan, two groups are distinguished: Srongtsan Gampo (617–650 CE). The language from valley or sedentary dialects (Rongke) and nomadic this period through the ninth century, known as Old dialects (Drokke). Each category is further divided into Tibetan, was used mainly to record royal documents several subdialects. Amdo too is divided into valley and historical chronicles. A large number of Old dialects and nomadic dialects, although the difference Tibetan manuscripts found in the Dunhuang caves are between these two categories is not so great as in among the most important materials for the study of Kham. There is a continuum of mutual intelligibility the proto-history and proto-language of Tibet. The among the speakers of these dialects. Amdo speakers majority of these materials have been cataloged and are reported to have problems understanding Lhasa preserved in national libraries in France and England. speakers and vice versa, and thus it is common for the Classical Tibetan (sometimes called Chöke by two groups to resort to Chinese as a lingua franca. For western scholars) has developed since the eleventh each region, one subdialect associated with cultural century. It remains in use as the medium for religious heritage, education, or political power is generally con- texts and other kinds of learned discourse. This literary sidered to be more prestigious than the others and variety was adopted as the liturgical language by serves as a regional lingua franca. Lhasa Tibetan is a Mongols and other non-Tibetan-speaking groups in subdialect of Central Tibetan, but because it is spoken 1104  TIBETAN in the metropolitan and pilgrimage center and former indicates absolutive case. Lhasa Tibetan is often cited seat of government, it is generally regarded by Tibetans by linguists as a language with a ‘split’ ergative-mark- as the most prestigious form. Lhasa Koiné, with loan- ing pattern: ergative subjects are required only when words from Nepali, Hindi, or English, is spoken in reported actions have already taken place. Therefore, exile communities by refugees who came from various the subject in kho¯ŋ-ki tà t-¯ sε`ε so¯ŋ ‘He killed a tiger’ linguistic backgrounds. Other regional koinés in requires an ergative case particle because the verb indi- Tibetan territories in China include the Labrang dialect cates a past event, whereas the subject of kho¯ŋ tà t-¯ of Amdo and the Derge dialect of Kham. sε`ε-ki rè ‘‘He will kill a tiger’ does not. The majority of words in Tibetan are disyllabic. A Tibetan nouns express neither gender nor number. basic monosyllabic word consists of an initial and a Plurality is indicated by the plural marker tsho as well vowel and may or may not have a final consonant, e.g. as dag and rnams (the latter two are more common in bod ‘Tibet’ (phöö in the Lhasa dialect; wot in the written texts). In modern dialects, these plural particles Amdo dialect). Consonant clusters are common in are generally employed only with pronouns. Adjectives written Tibetan and archaic dialects––those containing follow nouns. Determiners (words that modify noun archaic features and therefore closer to written phrases) include the number ‘one’ gcig, which func- Tibetan––namely Amdo and western dialects. The tions as an indefinite article, and the spatial demonstra- number of initial consonants varies from one dialect to tives di ‘this’ and de ‘that’ functioning as definite another. There are 28 consonants in Lhasa Tibetan. articles. Negation is indicated by mi- or ma-, which is Kham valley dialects have about 40 consonants. prefixed to the main verb. In Lhasa Tibetan, it is more Tibetan presents an interesting case for the study of the common to use sentence-final negative verbs, i.e. min origin of tones (i.e. intonational differences that dis- and me, when the subject is first person ‘I’, and the tinguish word meanings), as this innovative feature of negative prefixes when the subject is second or third distinguishing meanings by musical pitch alone is person, i.e. ‘you’ or ‘he or she or it’. A sentence may found only in some modern dialects. Central Tibetan consist of a single clause or a series of clauses chained and Kham Tibetan exhibit a well-developed system of together with the main clause as the last element. lexical tones, whereas Amdo and Western Tibetan do Because of this characteristic, Tibetan is typologically not have tones. The number of tones ranges from two classified as a ‘clause-chaining’ language. to four. Tibetan dialects in Nepal display a two-tone Verbs in Old Tibetan and Classical Tibetan, most of system (high tone vs. low tone). Most Kham dialects which are monosyllabic, have variant forms corre- contain four tones (high tone, low tone, rising tone, sponding to tenses (future, present, past) and mood and falling tone). Initial and final consonants are rela- (imperative). In Central Tibetan and Kham Tibetan, tively simple in these tonal dialects. In addition to this distinction is neutralized: there is only one form tones, some of these dialects also developed a rich for each verb. Instead, sentence-final auxiliary verbs inventory of diphthongs. are employed to convey, for example, tense. These Like the majority of Tibeto-Burman languages, auxiliary verbs and verbs equivalent to ‘to be’ or ‘to Tibetan has the subject–object–verb word order. exist’ are used according to person (whether or not the Grammatical relations are expressed by means of case subject is first person) and evidentiality (whether or particles. Five cases are distinguished for most spoken not the person has witnessed the event described in the and written varieties: ergative–instrumental, ablative, utterance). Person marking in Tibetan is different from absolutive, genitive, and dative–locative. The genitive person agreement, a common feature of languages of case particle also functions as a relative clause marker. the Himalayan branch of the Tibeto-Burman group. The ergative and the instrumental case markers are The emergence of person and evidential markings in homophonous but they are attached to nouns with dif- modern spoken dialects of Tibetan has attracted a great ferent semantic roles: the ergative particle marks an deal of attention from linguists. These phenomena are actor, which is chiefly animate, whereas the instrumen- also found in other Bodish languages. tal particle marks an instrument. Likewise, the dative Tibetan has a well-defined and productive system and the locative case particles are identical in form, but of honorifics (zhesa). These are special words (mainly they have different functions: the former marks a recip- nouns and verbs) used when referring to Buddhas and ient or a patient (an entity, generally an animate being, deities and in everyday conversation when one talks affected by the action); the latter is attached to a place with people of higher social standing, elderly and or time. Written Tibetan is a prototypical example of an respectable people including monks, officials, teach- ergative language: subjects of verbs that can occur with ers, and one’s own parents and elder siblings. objects require ergative case marking, whereas objects Honorific vocabulary, which stands in parallel with of such verbs and subjects of verbs that never take ordinary vocabulary (phal-skad), is common in Old objects behave differently: they are unmarked, which Tibetan texts and has developed into a complicated 1105 TIBETAN system in Lhasa Tibetan. It is found in most of the Eric Shiller. Tempe: Arizona State University Program for dialects but in a relatively less sophisticated manner. Southeast Asian Studies. Beyer, Stephen. 1992. The classical Tibetan language. Albany, Contrary to stereotypes among native speakers, hon- NY: State University of New York Press. orifics are used among Kham speakers, but mainly Bielmeier, Roland. 1985. A Survey of the development of when they engage in conversation with monks, high Western and Southwestern Tibetan dialects. Soundings in officials, or the elderly. The following expressions rep- Tibetan civilization: Proceedings of the 1982 seminar of the resent two ways of saying ‘his hat’, with the structure International Association for Tibetan Studies held at Columbia University, ed. by Barbara Nimri Aziz and [he-genitive hat]: kho¯-ki zha_mo (ordinary speech) and Matthew Kapstein. New Delhi: Manohar. kho¯ŋ-ki üzha_ (honorific speech). The honorific com- Bradley, David. 1996. Tibetan. Atlas of languages of intercul- pound üzha _ , which consists of ü, the honorific root tural communication in the Pacific, Asia, and the Americas, for ‘head,’ and zha_ , the ordinary root for ‘hat,’ is typi- ed. by Stephen A. Wurm, Peter Mühlhäusler and Darrell T. cal of how honorific nouns are formed in Tibetan. Tryon, Vol. 2. Berlin: Mouton de Gruyter. DeLancey, Scott. 1998. Semantic categorization in Tibetan hon- Given that most Tibetan-speaking areas are inac- orific nouns. Anthropological Linguistics 40. 1. cessible for fieldwork, both a basic description and an Denwood, Philip. 1999. Tibetan. Amsterdam and Philadelphia: investigation of contemporary aspects of Tibetan lan- John Benjamins. guages and dialects still await linguists today. With Goldstein, Melvyn C., Gelek Rinpoche, and Lobsang Phuntsog. continuing sociocultural and political changes, Tibetan 1991. Essentials of modern literary Tibetan: a reading course and reference grammar. New Delhi: Munshiram remains one of the most challenging areas of study for Manoharlal Publishers. the twenty-first century. Makley, Charlene, Keith Dede, Hua Kan, and Wang Qingshan. 1999. The Amdo dialect of Labrang. Linguistics of the Tibeto-Burman Area 22. 1. References Tournadre, Nicolas. 1998. Manuel de tibétain standard. Paris: Beckwith, Christopher, I. 1992. Deictic class marking in Tibetan Asiatheque. and Burmese. Proceedings of the first annual meeting of the van Driem, George. 1998. Dzongkha. Leiden: Leiden University. Southeast Asian Linguistic Society, ed. by Martha Ratliff and KRISADAWAN HONGLADAROM Time and Tense The tense of a sentence is a grammatical category that analysis of John goes to the cinema. However, the specifies a time at which the situation described by the treatment of the complex tenses like perfect and plu- sentence is the case. In a sentence like I was often in perfect is not that straightforward. Some analyze a Venice, the past tense was indicates that the situation complex tense as a single modifier. Thus, John has was the case before the time at which the sentence was gone to the zoo would be represented as PERF (John go uttered. Thus, tense mostly gives us information about to the zoo), whereas John had gone to the zoo would the distance between a certain situation and the utter- be PLUPERF (John go to the zoo). Others assume that ance time. Besides the grammatical category of tense, the complex tenses are combinations of modifiers, there are also lexical means of expressing location in where the perfect combines PRES and PERF, and the time: temporal adverbs (last year, yesterday, etc., cf. pluperfect PRET and PERF. In these combinatorial John left home yesterday), prepositions of time (in, at, accounts, John has gone to the zoo is analyzed as PRES on, etc., cf. John left home at lunch time), and tempo- (PERF (John go to the zoo)), whereas John had gone to ral conjunctions (when, after, before etc., cf. John left the zoo would be PRET (PERF (John go to the zoo)). At home when the phone rang). first glance, the combinatorial account seems more Although tense is marked on the verb in many lan- attractive because it mirrors the actual words being guages, many researchers assume that tenses modify used: present tense (has) and past tense (had) combine sentences as a whole. Thus, a common analysis of I with perfect (gone). However, not all semanticists are was often in Venice is PRET (I often be in Venice), where convinced that tense on an abstract level necessarily PRET is the abstract marker for preterite tense (simple keeps these modifiers separate. The theory of tense past). Sentences in the present are analyzed in the outlined here was heavily influenced by the work of same manner: PRES (John go to the cinema) is the Hans Reichenbach. 1106  TIV AND TIVOID LANGUAGES Another very influential source for tense theory was succession: Jameson entered the room. He switched on tense logic. The treatment of tenses as sentence modi- the light. If both sentences contain states or activities, fiers, which was sketched above, goes back to tense then the events overlap each other: Jameson cleaned logic. Tense logicians assigned indefinite meanings to the room. It was very cold in there. However, if one of the tenses. Compare the sentence I turned off the stove, the two sentences contains a state or an activity, and the represented as PRET (I turn off the stove). This is con- other an accomplishment or achievement, both overlap sidered true if and only if there is an indefinite time and succession are possible. The following is an exam- before the time of the utterance at which ‘I’ turn off the ple of succession: He switched off the light. Now it was stove. But this indefinite meaning of the preterite is pitch-dark around him. The darkness is due to the counterintuitive: anyone uttering the sentence will switching off and thus cannot overlap it. The following have a certain (definite) time of turning off in mind is an example of overlap: He switched off the light. It (for instance the time after the meal was ready). The was very cold in the room. Interestingly, the progres- point becomes even clearer with negation: I didn’t turn sive tense (marked with -ing in English) behaves like off the stove does not mean that there is no turning off states or activities: if one of the two sentences contains in the past. Instead, it means that there is a definite a progressive, and the other an accomplishment or time in the past, and at this time ‘I’ did not turn off the achievement, both overlap and succession is possible: stove. To say that tenses are definite means that a con- He awoke to the sound of her screeching. She was text is required for the listener to determine which shaking him (overlap); He opened the door again. The exact point or stretch in time is being talked about. man outside was smiling at him (succession). Relevant contexts are the immediate situation of utterance (in the case of the present tense) and times mentioned in previous discourse (in the case of the References past tense). Benthem, J.F. van. 1983. The logic of time. Dordrecht: Reidel. Whereas ‘classical’ tense logic confined itself to Comrie, B. 1985. Tense. Cambridge: Cambridge University Press. single, isolated sentences without context, more recent Hinrichs, E. 1986. Temporal anaphora in discourses of English. work on tense has also focused on tense in discourse. Linguistics and Philosophy 9. 63–82. Kamp, H., and U. Reyle. 1993. From discourse to logic. Discourse may be ‘temporally connected’ or ‘tempo- Dordrecht: Kluwer. rally free’. A sequence of two sentences forms a tem- Klein, W. 1994. Time in language. London, New York: Routledge. porally connected discourse if the second sentence is Nerbonne, J. 1985. German temporal semantics: three-dimen- temporally interpreted by taking elements from the sional tense logic and a GPSG fragment. Ann Arbor, MI: first one. An example is: Al went to New York. The oth- University Microfilms. Ogihara, T. 1996. Tense, attitudes, and scope. Dordrecht: Kluwer. ers were there, too. In contrast, temporally free dis- Partee, B. 1973. Some structural analogies between tenses and course lacks a direct temporal connection. Compare: pronouns in English. The Journal of Philosophy 70. 601–9. Al went to New York. The others were there once, too. Prior, A. 1967. Past, present and future. London: Oxford The division between sentences in the preterite University Press. whose events overlap each other and sentences where Stechow, A. von. 1991. Intensionale Semantik—Eingeführt anhand der Temporalität. Konstanz: Arbeitspapier Nr.40, the events follow each other is related in a very sys- Universität Konstanz. tematic way to the types of the events being described. Stechow, A. von. 1995. On the proper treatment of tense. If both events are accomplishments or achievements, Proceedings of SALT V, ed. by T. Galloway and M. Simons. then the events are understood as happening in MONIKA RATHERT Tiv and Tivoid Languages Tiv (Mitshi, Munshi) is spoken in Eastern Nigeria in an and Igede; and to the south one finds a number of relat- area all along the Katsina Ala and its tributaries, crossing ed languages/dialects, all of which are listed below. In the Benue where the Katsina Ala joins it and north of the standard reference works it is classified as Bantoid— Makurdi, and in a few isolated villages across the border Benue-Kwa—Niger-Congo. Still, there is well-justified, in Cameroon. Tiv’s neighbors to the east are the Jukun; long-standing discussion on just how close Tiv is to to the north Alago is spoken; to the east Eloyi, Agatu, Bantu itself. Munshi is a derogative name for the Tiv 1107 TIV AND TIVOID LANGUAGES used by their neighbors. The term is a Hausa corruption While Tiv is relatively well documented, for the of the Jukun term for the Tiv ‘Mbitse’ or ‘Mbiche’. other Tivoid languages the documentation is restricted Within Tiv proper very little dialect variation is to a few short word lists. known, even though it might be possible, based on dif- ferences in the realization of front vowels, to speak of Some Structural Facts the Iharev, Kparev, Kunav, Shitire, and Utisha dialects. Phonology. There are 26 consonants, six vowels, and The Shitire ‘dialect’ is used in most of the vernacular three tones, low, mid, and high [` -´] (Table 1). Tones published literature. are normally not marked in the official orthography. A To the Tivoid subgroup of Bantoid belong a series number of morphophonemic rules are to be observed. of very closely, potentially mutually intelligible lan- Particularly interesting are the palatalization and labi- guages spoken on either side of the Nigerian alization of stem-initial consonants under the influ- (N)–Cameroonian (C) border along the 6th latitude ence of prefix vowels (see below) and stem north. Included in this group are the following: alternations, in verbs due to the conjugational type and Abon (N), Avande (C), Batomo (C), Batu cluster in nouns due to noun class prefixes. (N), Bitare (N), Evant (N), Iceve-Maci (N), Iyive (N), Mesaka (C), Ndii (C, N) Oliti (C), Otank (N), Tiv (N), Morphology, nouns. There are 12 noun classes Tivi (C), and Yiive (C). grouped into 16 genders: 1|2; 1|4; 1|6; 5|4; 5|6; 5|6; 7|6; While there are around 2,500,000 speakers of Tiv, 7|6.a; 7|8; 9|4; 9|6; 14|8; 15|4; 15|6; 15|6.a; 6.a. most of the other members of this subgroup are spoken Morphological agreement is shown in all categories of by only a few thousand speakers. modifiers, in the pronominal system, and in the form Tiv was first documented by Clarke 1848––some of subject agreement on the verb. Noun class prefix as number names and basic nouns––under the name ‘Appa’ well as suffix markings are to be observed on the and by Koelle 1854, who gave a lengthy word list. nouns and the adjectives. In Table 2, samples of each Significant linguistic works include the dictionary and noun class each with one modifier category are given. grammar of Abraham and Malherbe, and the sketches of The numbers to the left (and those given for the gen- Lukas, Arnott, Sibomana, and Jockers. The language is ders above) refer to the corresponding numbers of the used in primary schools and there are some readers for Benue-Congo noun class system. adult education. There is a Bible (1964) and a modest While only class (6a) nouns show clear prefixes and amount of liturgical literature. Tiv is used in regional suffixes, e.g. m´-kúlé-m ‘oil’, remnants of class prefixes radio broadcasts; newspapers or journals in Tiv existed and/or suffixes can be established for all nouns. In some until the early 1980s, but have since been discontinued. cases, the noun class prefixes have been incorporated Bohannan and Bohannan (1953:11) report that some into the stem: historical *u- is realized as labialization on people as far away as Bamenda, Cameroon used Tiv as the first consonant, e.g. kwa¯sé ‘woman’ compare kàsév a trade language. Today this no longer seems to be the ‘women’, while prefix i- causes palatalization of the first case. Tiv does, however, enjoy a rapid expansion to the consonant, e.g. ínyáhá ‘paddles’, cf. náhá ‘paddle’. north, much to the dismay of its neighbors. Modifiers always follow the head noun. TABLE 1 Consonants and Vowels Bilabial Labio-dental Alveolar Postalveolar Palatal Labio-velar Velar Glottal Nasal m n ny [I] ng [ŋ] Plosive p, b t, d kp, gb k, g Affricative ts, dz c [t], j [d ] Fricative f, v s, z sh [] gh [γ] Lateral l~r Approximant w y h Front Back High i u e o o.[ɔ] Low a 1108  TIV AND TIVOID LANGUAGES TABLE 2 Sample Meaning my this, one two bad PRO subject REL Noun these concord 1 kwa¯sé woman wàm´ ngù mò m´ ù bó ún` á ù 2 kàsév women áv- mbá` ú hál- mbà bóv´ vé vé mbà 3 náhá paddle wám´ ngú` mó m- ú bo¯ ú ú úu¯ 4 ínyáhá paddles yám´ ngí` í hyál- í bo¯ í í íà 5 íkyésé basket yám´ ngí` í mó m- í bo¯ í í íà 6 ákésé baskets ám´ ngá` á hál- á bo¯ á á áa¯ 6a m´kúlém oil ám- má` m´ hál- mà bóm´ má má mà 7 íny’álégh´ money yágh- kí` í mó m- kì bógh´ kí kí kì 8 íkyáv´ loads yáv- mbí` í hyál- mbì bóv´ mbí mbí mbì 9 ìwyá dog yàm´ ngì ì mò m´ ì bó ì ì ì 14 ángév´ disease áv- mbúù mó m- mbù bóv´ mbú mbú mbù 15 gbógh´ stick ágh- kú` mó m- kù bógh´ kú kú kù Verbs. Verb stems (and also some nouns) may show PAST PERFECT á vénde he refused internal alternations: etc. tèmà tùmà sit Syntax. Tiv is a basic SVO language. Sample sen- nyímè nyúmè refuse tences taken from the Tiv dialect spoken just across the sèngè sòngu slaughter Cameroonian border giving core participants show a strict S V, S V O, S V DAT O order: Since not all verbs undergo these alternations, there is no immediate explanation for these changes Wán lá à-kpé although they appear to be remnants of an earlier Child DEF it-died (ATR) vowel harmony system. The child died. There are traces of verb extensional suffixes. Five morphemes have been identified. It is not clear to what Wán lá à-yá àyàbà Child DEF it-eat plantain extent these processes are productive; they do not The child ate plantain. seem to play a role in marking participants. Wán lá à-ná tùɔ` àyàbà sV yóò announce yóò-sò talk a great deal ™ r` búgh` open búghù-r` burst open Child DEF it-give chief plantain gh` pér` cross (a river) pérè-gh` cross over (road) The child gave the chief plantain. m` lègh` soft lèghè-m` be soft n` pùù despise pùù-n` find fault with Additional participants are expressed by a series of prepositions: The tense–aspect–modality (TAM) system is extremely complex and only superficially described. A Wán lá à-gbé shìn ngèr selection of the many forms that exist for each verb is Child def it-fall into water given below. It should be noted that TAM is frequent- The child fell into the water. ly marked with tones as well as TAM markers before the verb and vowel alternations after the stem. Negation, even in complex sentences, is marked at the very end of a sentence: vèndà Refuse INFINITIVE ù-vèndá-n´ to refuse m` .fá àlù wán lá à-kpé gá. GERUND vèndà-n` ‘refusing’ I-know if child DEF it-die NEG RESTRICTED PAST PERFECT m´-`vèndé I refused I don’t know whether the child died. PRESENT HABITUAL I mbá vé `vendá they usually refuse In some cases, as in the following focus construc- PRESENT HABITUAL II Dàm ` `vèndé Dam usually refuses tion, negation is additionally marked on the verb: PRESENT HABITUAL III mbá `vé véndá- they ususally n´ refuse tùɔ` ù wán lá (à)-ne-ná àyàbà gá IMPERATIVE `vèndá refuse! ™ NEGATIVE IMPERATIVE dé `vèndá-n` gá don’t refuse! Chief REL child DEF (it)-NEG-give plantain NEG SUBJUNCTIVE, HORTATIVE à vénda let him refuse! It’s not a chief the child gave the plantain. 1109 TIV AND TIVOID LANGUAGES Relativization is marked by a RELative pronoun indi- gbúká lá gbà pùú cating gender and number at the beginning and a low gun DEF fire BANG The gun went off ‘bang’. tone at the end of the relative sentence. The low tone has the effect of down-stepping the following high- toned syllable. If the relativized noun is marked with In this case pùú is the sound made by the gun and the invariable definitizer lá, the relative clause appears only that. In the following two sentences, however, between the head noun and the lá. If, moreover, the rel- pùlùtùtù and cìcá can be used with any color to ativized noun is definite, a resumptive pronoun appears express intensity: in the ‘original’ syntactic position in the relative clause: àvámbè nìà pùlùtùtù cìcá Wán ù m`-méngé ` ndí blood red INTENSE Child REL i-see REL lost blood is absolutely red A child I saw is lost. àlísì yìlè pùlùtùtù Wán ù m`-méngé àná lá ` ndí cìcá Child REL i-see PRO DEF REL lost charcoal black INTENSE The child I saw is lost. charcoal is absolutely black In relativized objects of prepositions, a shortened Tivoid. Aside from a very large number of vocabulary form of the resumptive pronoun appears as well. The resemblances, all languages belonging to the Tivoid sub- preposition maintains its original syntactic position: group share a complex nominal system where a noun is marked both by a prefix and a suffix. A single exam- (à)-lú shìn yò ì wán lá gbé shìn mì `. ple––the word for oil––suffices to show the unity of (he) live in house REL child DEF fall into pro REL He lives in the house the child fell into. Tivoid on the one hand and the close, but non-Tivoid character of neighboring Bantoid languages on the other: There is no clear evidence of serialization. Tivoid other Bantoid Complex structures giving the appearance of serializa- tion are instances of simple conjoined sentences: BATOMO mákutúm AMANAVIL mètí BELEGETE mutum AMASI òmìlí ICHEVE mukurem AMBELE mùút wán lá à-té àyàbà, à-dzé tùɔ` ámì. NDII mkutem ASAKA ɔ´múú ™ OLITI mùkùlémí ASUMBO mɔ` ɔ´t child DEF it-take Plantain it-go to chief with them TIV m´kúrém ESIMBI mìri The child took the plantain (and) went to the chief with them. TIVI íŋkùlŋ KENYANG bawet YIIVE mə`kùtε`m MESAKA òmbúl Only in cases where verbs of motion that do not require a goal are involved does one obligatorily find References dze ‘go’ as the second verb. The latter then requires the expected preposition: Abraham, Roy Clive. 1933. The grammar of Tiv, Vol. 5(1), 213pp. Kaduna: Government Printer. wán lá à-yévísí dzé shìn yò ––––––. 1940a. The principles of Tiv, 102pp. London: Crown child DEF it-run go into house Agents. The child ran into the house. ———. 1940b. Dictionary of the Tiv language, ix, 331 pp.London: Crown Agents. [reprinted 1968 by Gregg]. ––––––. 1940c. A Tiv reader for European students, 82pp. Comparatives have the Africa-typical structure: X London: Crown Agents. [Appendix: Further points of gram- exceeds Y in Z, or X is Z, exceeds Y: mar for reference, pp. 69–82]. Arnott, David W. 1958. The classification of verbs in Tiv. unòr(-mban) bàìkɔ´ hèmbà ùnor-ngèr Bulletin of the School of Oriental and African Studies 21(1). elephant fat exceed hippopotamus 111–33. An elephant is fatter than a hippo. ––––––. 1964. Downstep in the Tiv verbal system. African Language Studies 5.34–51. unòr(-mban) hèmbá bàìkɔ´-n ùnor-ngèr ––––––. 1967. Some reflections on the content of individual elephant exceeds being fat hippopotamus classes in Fula and Tiv. La Classification Nominale dans les An elephant is fatter than a hippo. Langues Négro-Africaines, ed. by Gabriel Manessy, 45–74. ––––––. 1992. R. C. Abraham’s books on Tiv. In Jaggar, pp. 141–5. There are a good number of ideophones; not all of Bohannan, Laura, and Bohannan, Paul. 1953. The Tiv of Central them, however, are semantically ‘identical’ to the head Nigeria, viii, 9-100. London: International African Institute verb: (Ethnographic Survey of Africa, Western Africa, Part 8). 1110  TOCHARIAN Clarke, John. 1848. Specimens of dialects: short vocabularies vermehrt durch eine historische Einführung von P. E. H. of languages: and notes of countries & customs in Africa, Hair und einen Wortindex von David Dalby. Graz, 104pp. Berwick upon Tweed: D. Cameron. [Edited with a Akademische Druck- u. Verlagsanstalt, 1963. 19pp.]. commentary by Edwin Ardener, with a biographical note by Lukas, Johannes. 1952. Das Nomen im Tiv. Anthropos Shirley Ardener. Gregg Press, 1972, 76pp.]. 47.147–76. Jagger, Philip J. (ed.) 1992. Paper in Honour of R.C. Abraham Malherbe, W. A. 1932. Tiv–English dictionary, with grammar (1890–1963). London: School of Oriental and African Studies. notes and index, xxxix, 207 pp. Lagos: Government Printer. Jockers, Heinz. 1991a. Studien zur Sprache der Tiv in Nigeria, Sibomana, Leonidas. 1980. Aspects of Tiv morphophonemics. xv, 164pp. Frankfurt am Main: Peter Lang. (Europäische Afrika und Übersee 63(1).69–77. Hochschulschriften 21). ––––––. 1981. Tonal lexicalization of Tiv nouns. Festschrift ––––––. 1991b. The alphabetization of the Tiv: a case study. zum 60, ed. by Inge Hofmann, Hrsg. Geburtstag von P. Language standardization in Africa/Sprachstandardisierung Anton Vorbichler, 1. Teil, 155–73. Wien: Institut für in Afrika/Standardisation des langues en Afrique, ed. by Afrikanistik und Ägyptologie der Universität Wien. Norbert Cyffer et al., 145–56. Hamburg: Helmut Buske. ––––––. 1983. Tonal structure of Tiv verbs. Afrika und Übersee ––––––. 1992. Linguistic studies on Tiv––before, by, and after 66(1).77–89. R. C. Abraham. In Jaggar, pp. 163–73. ––––––. 1984. Displacement and lowering of Tiv tones. Afrika Koelle, Sigismund Wilhelm. 1854. Polyglotta africana; or a und Übersee 67(2).277–87. comparative vocabulary of nearly three hundred words and Terpstra, Gerard. 1968. English–Tiv dictionary, 120pp. Ibadan: phrases in more than one hundred distinct African lan- University of Ibadan. (Institute of African Studies, guages, vi, 24, 188 p. London: Church Missionary House. Occasional Publications No. 13). [Unveränderter Nachdruck der Ausgabe London 1854 F. K. ERHARD VOELTZ Tocharian Tocharian is the name applied to a small group of The first extensive analysis and decipherment of extinct Indo-European language varieties attested in the hitherto unknown Tocharian language were made fragmentary manuscripts first discovered in the last by Emil Sieg and Wilhelm Siegling in 1908. The deci- decade of the nineteenth century CE. The recovery of pherment was facilitated by the knowledge of other manuscripts in Tocharian as well as others written in languages written in Bra¯ hm¯ and related syllabaries Khotanese Saka, Sanskrit, and various Turkic and as well as the existence of bilingual Buddhist reli- Sino-Tibetan languages from the Taklamakan Desert, gious texts (in the original Sanskrit, accompanied by Tarim Basin, and region south of the Tian Shan range, Tocharian translations). It is still under debate all located in the modern-day Xinjiang Uygur whether the name Tocharian itself is the most proper Autonomous Region of northwestern China, was a name for this language (group); this topic alone has consequence of the many exploratory expeditions yielded much speculation about the origins of undertaken around the turn of the century along the Tocharian and its speakers. The name Tocharian is ancient Silk Road trade routes. The majority of extant associated with a people called the Tocharoi, who, as Tocharian texts originate, by most estimates, from the suggested in the writings of the ancient Greeks Strabo sixth to ninth centuries CE, after which the Tocharian- and Ptolemy, at one time inhabited the area formerly speaking culture is believed to have assimilated with known as Bactria in present-day Afghanistan and neighboring (e.g. Turkic) cultures or died out. Tajikistan. Although the existence of the Tocharoi is The texts, written on palm leaves, paper scrolls, and not in question, some scholars believe that they were wooden tablets, represent a diverse array of genres and of Iranian stock and bore no relation to the speakers subject matter, from scientific, legal, and commercial of the non-Iranian language in the Tocharian manu- documents (such as caravan passes) to dramas, narra- scripts. Still others who may not necessarily share this tives, poetry, and Buddhist and Manichaean religious belief maintain that the name Tuγ ri (Twγ ry/Twghry/ works. A modified form of the Bra¯ hm¯ script of north- Twqrï) found in the neighboring Turkish language ern India, the writing system employed in the Uygur would be more appropriate as a name for Tocharian texts, consists of a syllabary in which each Tocharian. Despite the conflicting views among consonantal letter is paired with an inherent (unwrit- scholars on these matters, however, the name ten) vowel. Vowels other than this inherent vowel are Tocharian has prevailed within the general communi- indicated with distinctive letters or diacritics. ty of linguists. 1111 TOCHARIAN Apart from their association with the Tocharoi, it has routes in the region among people who spoke mutual- also been suggested that the speakers of Tocharian were ly unintelligible languages. However, most of the one and the same with a culture that the Chinese called Tocharian manuscripts––in two distinct Tocharian the Yuezhi, a confederacy of several (possibly Indo- varieties no less (A and B)––have been concentrated European-speaking) tribes originally living at the edge within a single region and little to no evidence of a sig- of the Gobi Desert in present-day Gansu Province of nificant Tocharian presence beyond the confines of the northern China and not overly distant from the sites of Tarim Basin has appeared to date to confirm this view. the Tocharian discoveries. Having suffered a major mil- Whatever their origins, the speakers of Tocharian itary defeat in the middle of the second century BCE, appear to have filled many cultural and economic roles evidence from Chinese and Greek annals as well as in northwestern China for at least three centuries from from sites along the migration route exhibiting Yuezhi- the mid to late first millennium CE. They played a particular burial practices corroborates the view that considerable role, for example, in the spread of Yuezhi tribes journeyed west out of China, deeper into Buddhism and Manichaeism (with roots in northern central Asia. By the early 130s BCE, some of the Yuezhi India and Persia, respectively) into western China. tribes had arrived in Bactria and neighboring Sogdiana Trade and other forms of contact with the Chinese, surrounding present-day Samarkand in Uzbekistan. Persians, Turks, various cultures from India, and oth- These tribes that had reached Bactria are known as the ers secured for the Tocharian speakers of the Tarim Greater Yuezhi. At around 50 CE they may have estab- Basin sites a position of great influence, which lished the Kushan Empire that extended into northern endured for some time. India. Some generally accepted views are that neither From a linguistic standpoint, scholars were at once the Tocharoi of Greek sources nor the Greater Yuezhi amazed and perplexed after it was determined that were the speakers behind the Tocharian manuscripts. Tocharian was in fact Indo-European. A traditional Others suggest that both peoples were actually one and and previously straightforward way of grouping the the same since both ended up in Bactria and appeared to Indo-European languages was to refer to the have had otherwise similar histories according to the centum/satem divide. Languages of the centum group written records available. These are fair assumptions (so-called after the Latin word for ‘100’) such as given the fact that by the early second century BCE the Celtic, Germanic, Greek, and Italic have velar reflexes Greater Yuezhi had already migrated well past the sites (or later correspondents) of velar sounds (e.g. /k/ and of the Tocharian manuscripts first produced approxi- /x/) of the reconstructed Proto-Indo-European (PIE) mately 650–750 years later. language, while languages of the satem group (so- At least one more possibility for a Yuezhi- called after the Avestan word for ‘100’) such as Tocharian connection exists, however. The Lesser Armenian, Indo-Iranian, and Slavic have sibilant Yuezhi, who evidently halted their migration out of reflexes (e.g. /s/ and /š/). Since the word for ‘100’ is more easterly districts of China well short of the trek känt in Tocharian A and kante in Tocharian B, the once completed by their Greater Yuezhi kinsmen to Bactria, solid west/east grouping seen in the centum/satem may have settled south of the Tarim Basin and divide was disturbed. Taklamakan Desert for a time. Between the cities of With regard to geographical and social distinctions Niya and Loulan in this area (both buried by sand over among the varieties of Tocharian, which some consider the ages), documents in the Indic language Niya separate languages, others merely dialects of a single Prakrit came to light and have been claimed to demon- language, Tocharian A was found concentrated in the strate signs of influence from a Tocharian language, oasis town of Turfan and to a lesser extent further west which may have existed in the vicinity at least several at the central site of Karashahr, while Tocharian B was centuries prior to the period of the Tocharian manu- mainly concentrated in Kucha, the westernmost of the scripts found further north. It is possible that the three primary sites, and found to some extent in both (Lesser) Yuezhi, if not of Iranian origin, as commonly Karashahr and Turfan as well. Based on these geo- believed, were the speakers of this hypothetical south- graphical differences, Tocharian A has also been called ern Tocharian language and eventually migrated north East Tocharian, Turfanian, Karashahrian, and Agnean where they gave rise to the speakers of the attested (after the former state of Agni situated around manuscripts. If correct, this theory would serve to link Karashahr), while Tocharian B has also been referred to the Greater Yuezhi, as kinsmen of the Lesser Yuezhi, as West Tocharian and Kuchean (after the former state with one or more Tocharian languages. New manu- of Kuchi and its site around Kucha). Under one propos- script discoveries along the migratory paths of the al, both varieties of Tocharian were living languages, Greater Yuezhi to Bactria would clarify this view. with B having been imported to the two eastern sites by It has been suggested that Tocharian was employed Buddhist missionaries. A second suggests that A was a as a lingua franca along the Silk Road and other trade dead language used primarily for liturgical purposes 1112  TOCHARIAN while B was the living language used in everyday life groups of mummified humans first discovered during since it was common to all three sites. the Western expeditions of the late nineteenth and While the status of Tocharian as a member of Indo- early twentieth centuries CE. These mummies, howev- European remains uncontroversial, the exact relation- er, have only been widely known and studied in the ship between the Tocharian languages and other last quarter of the twentieth century CE. Due to their Indo-European branches has long been subject to blonde and reddish hair and other distinctive bodily speculation. Linguists have made attempts to establish features, some scholars claim that these mummies genetic ties between Tocharian and each of the were the ancestors of the Tocharian speakers. Clearly, Anatolian, Balto-Slavic, Celtic, Germanic, Greek, they are not easily identifiable with most of the cul- Indo-Iranian, and Italic branches at one time or tures from the past or present that are commonly another––in other words, between Tocharian and most linked to the region, such as the Chinese, Mongols, of the Indo-European subgroupings. This only serves and Turks, but rather have appearances more typical of to demonstrate the relatively modest consensus that Europeans. Additional clues such as wall paintings exists over the complex position occupied by found in caves around Kucha and Turfan depicting Tocharian within the Indo-European context in histor- humans with red hair, along with the proximity of the ical, linguistic, and social terms. mummy sites to the sites of the Tocharian manuscripts, The phonology (sound structure) of Tocharian is all serve to fuel the proposed connection of the mum- characterized by palatalized variants of almost all con- mies with the speakers of Tocharian. sonants, a single series of stop consonants (i.e. voice- The discovery of Tocharian has influenced many less /p/, /t/, /k/ but no voiced /b/, /d/, /g/ or voiced Indo-Europeanists to reconsider their views about the aspirated correspondents––all three series existed in history of the Indo-Europeans and the development PIE), the loss of final consonants in A and B (e.g. PIE and interrelations of the languages attributed to them. *tod
B te ‘this’/‘it’) and final vowels in A (e.g. klots Various forms of linguistic and archeological evidence ‘ear’ vs. B klautso, la¯nts ‘queen’ vs. B la¯ ntsa, wäl may serve to further unlock the secrets presented by ‘king’ vs. B walo), and some loss of PIE distinctions the speakers of Tocharian A and B (and perhaps other between short and long vowels. Singular and plural, as varieties of Tocharian that have left little to no trace). well as dual (e.g. B wi rsoñc ‘two spans’ vs. the plural Among the major figures whose works are not list- form rsonta) and further subtleties in number distinc- ed in the References section but who have nonetheless tion, are also characteristic of Tocharian. A unique fea- made substantial contributions to the advancement of ture of Tocharian among the Indo-European languages knowledge about the origins of Tocharian and its is its elaborate system of several primary cases (e.g. speakers are Elizabeth J.W. Barber, Gerd Carling, nominative and genitive) and a host of secondary cases Walter Couvreur, Viacheslaw V. Ivanov, Jay H. such as the perlative, allative, and comitative, each of Jasanoff, Sylvain Lévi, Victor H. Mair, James P. which has a specific grammatical function. The sec- Mallory, Antoine Meillet, Holger Pedersen, John H.W. ondary cases are formed by the suffixation of various Penney, Pavel Poucha, Ernst Schwentner, and Werner postpositional elements to forms in the oblique, one of Winter. the primary cases. Within the verbal system, both active and medio-passive voices are found. The latter is marked in the present tense by an -r ending (e.g. A References klyosamtär and B klyausemtär ‘we hear’) as in some Adams, Douglas Q. 1988. Tocharian historical phonology and other centum languages and Indo-Iranian from the morphology. New Haven, CT: American Oriental Society. satem group. Lexical borrowings from Sanskrit and Adams, Douglas Q. 1999. A dictionary of Tocharian B. Amsterdam, Atlanta: Rodopi. neighboring Iranian languages are common. Tocharian Krause, Wolfgang. 1952. Westtocharische Grammatik (West is thought to have been influenced lexically and struc- Tocharian Grammar). Heidelberg: Carl Winter. turally, to various extents, by the myriad of non-Indo- Krause, Wolfgang, and Werner Thomas. 1960–1964. European languages in northwestern China and its Tocharisches Elementarbuch (Elementary Tocharian periphery, such as Sino-Tibetan and Turkic, as well as book), 2 vols. Heidelberg: Carl Winter. Lane, George S. 1966. On the interrelationship of the Tocharian those encountered by the ancestors of the Tocharian dialects. Ancient Indo-European dialects, ed. by Henrik speakers along the eastward migratory paths out of the Birnbaum and Jaan Puhvel. Berkeley, CA: University of central Indo-European homeland and eventually into California Press. China (e.g. Finno-Ugric languages). Lane, George S. 1967. On the significance of Tocharian for The harsh, arid environment around the Indo-European linguistics. Studies in historical linguistics in honor of George Sherman Lane, ed. by Walter W. Arndt, Taklamakan Desert region, which was ideal for safe- et al. Chapel Hill, NC: University of North Carolina Press. guarding the Tocharian manuscripts for up to 1,500 Lane, George S. 1970. Tocharian: Indo-European and non-Indo- years, was also instrumental in preserving several European relationships. Indo-European and Indo-Europeans, 1113 TOCHARIAN ed. by George Cardona, Henry M. Hoenigswald, and Alfred Sieg, Emil, Wilhelm Siegling, and Wilhelm Schulze. 1931. Senn. Philadelphia: University of Pennsylvania Press. Tocharische Grammatik (=Tocharian Grammar). Göttingen: Pinault, Georges-Jean. 1989. Introduction au tokharien Vandenhoeck & Ruprecht. (Introduction to Tocharian). Lalies: actes des sessions de Windekens, A.J. van. 1976–1982. Le tokharien confronté avec linguistique et de littérature de l’École normale supérieure 7. les autres langues indo-européennes (=Tocharian confronted 5–224. with the other Indo-European languages), 2 vols. in 3. Ringe, Donald A. 1991. Evidence for the position of Tocharian Louvain: Centre international de dialectologie générale. in the Indo-European family? Die Sprache 34. 59–123. SEAN O’ROURKE Ringe, Donald A. 1996. On the chronology of sound changes in Tocharian, Vol. I. New Haven, CT: American Oriental Society. See also Indo-European 1: Overview Tok Pisin Tok Pisin (or New Guinea Pidgin) is the dialect of adjectives, e.g. bikpela haus ‘big house’ and Melanesian Pidgin spoken in Papua New Guinea. It -im on transitive verbs (see below). serves as the main language of wider communication Such a contact language is called a ‘pidgin’ when it in a country where more than 800 separate indigenous continues to be used primarily as a second language languages are spoken by a population of nearly five for intergroup communication, and it is called a ‘cre- million. The two other dialects of Melanesian Pidgin ole’ when it becomes the mother tongue of a particular are Pijin, spoken in the Solomon Islands (with more community of speakers. As will be seen below, there is than 80 indigenous languages and a population of some controversy about whether Tok Pisin is actually around 480,000), and Bislama, spoken in Vanuatu a pidgin or a creole. (more than 100 languages, population 192,000). Because Melanesian Pidgin is a contact language Torres Strait Creole (also known as Broken or that came into being fairly recently, we can describe Yumiplatok)––spoken by approximately 10,000 peo- almost its entire history. The first stage of the develop- ple around the northern tip of eastern Australia––is ment of Melanesian Pidgin dates from the early 1800s, closely related to Melanesian Pidgin but is usually when Melanesians began to have frequent contact with considered to be a separate language. Europeans (including Australians and Americans). Melanesian Pidgin is an example of a contact lan- This was the result of whaling in the region, followed guage––a new language that developed in a situation by trading in sandalwood and bêche-de-mer. Because in which speakers of different languages needed to Melanesia is one of the most linguistically diverse communicate but did not share a common language. areas of the world, it was impossible for Europeans to Most of the vocabulary comes from one language, learn the local languages for trading (as they did in called the ‘lexifier’ (English, in the case of other areas of the Pacific). Thus, to communicate, they Melanesian Pidgin), but the words often have differ- used simplified English or existing contact languages, ent meanings. For example, in Tok Pisin, spak, such as South Seas Jargon and various forms of derived from English spark, means ‘drunk’. Many Aboriginal Pidgin English from Australia. As a result words are also derived from other languages (see of these limited encounters, many Melanesians picked below). Furthermore, the phonology and morphosyn- up some English vocabulary and phrases from the tax of a contact language such as Tok Pisin are dis- existing contact languages. tinct from those of the lexifier. For example, in The second stage came with the beginning of the phonology, Tok Pisin has a five-vowel system (a, e, i, Pacific labor trade in 1863, when Melanesians started to o, u) and only 15 consonants (represented in writing be recruited (and in some cases kidnapped) to become as b, d, g, h, k, l, m, n, p, r, s, t, v, w, and y). The laborers for plantations in Queensland (Australia) and English phoneme inventory contains about twice as Samoa. Melanesians from diverse areas found them- many elements. For sentence structure and meaning, selves literally in the same boat, and their only common Tok Pisin uses separate words to show number and language was what they had acquired from earlier con- tense (see examples below) where English generally tacts with Europeans. Thus, they used this to communi- uses suffixes, e.g. -s and -ed. However, Tok Pisin has cate with one another on the ships and later on the a few suffixes not found in English, such as -pela on plantations. With continued use, new features were 1114  TOK PISIN added, norms began to emerge, and a stable pidgin lan- that nearly all of the Tok Pisin-speaking population are guage began to develop––early Melanesian Pidgin. bi- or multilingual, whereas creole-speaking popula- The earliest Queensland laborers were mainly from tions are most often monolingual. the New Hebrides (now Vanuatu) and the Solomon Today Tok Pisin is the lingua franca of the entire Islands. Some laborers from German New Guinea also country of Papua New Guinea; it is known by an esti- went to Queensland in 1883 and 1884, but many more mated three quarters of the country’s nearly five mil- went to plantations in German-controlled Samoa, from lion inhabitants. It is the most widely used language of 1879 to 1912. Laborers from the other countries also urban areas and is also widely spoken in rural areas in started going to Samoa in 1878, and many of these had the northern part of the country. already worked in Queensland. Thus, early Melanesian Tok Pisin is used to some extent in radio and televi- Pidgin was transported to Samoa. However, after 1885, sion broadcasting, especially in interviews and news laborers from the New Hebrides and Solomon Islands reports. (It is also used in some of Radio Australia’s were no longer recruited for Samoa, and early Papua New Guinea Service broadcasts.) The weekly Melanesian Pidgin began to diverge into two slightly Tok Pisin newspaper Wantok has a readership of more different varieties––one spoken in Queensland and one than 10,000, and many government publications are in Samoa. also written in Tok Pisin. The language is widely used The third stage of development began when laborers in religion, and there is a Tok Pisin translation of the returned to their home islands, after their contracts fin- New Testament of the Bible. ished, and brought the developing pidgin with them. The constitution of Papua New Guinea recognizes Previously, these islands had no language of wider Tok Pisin as one of the national languages. (The con- communication; because the pidgin served this func- stitution of Vanuatu names Bislama as the national lan- tion well, it spread rapidly. It was also used by the guage.) Although English is more widely used for large-scale internal labor force that worked on the plan- government business, much of the debate in tations of German New Guinea, the New Hebrides, and Parliament is in Tok Pisin. the Solomon Islands after the external labor trade had Until recently, English was the official language of ended at the turn of the century. In each of these coun- education in Papua New Guinea, and it was used in all tries, early Melanesian Pidgin further stabilized and government schools (although Tok Pisin was widely changed under the influence of the local indigenous used in community and church-run preschools and languages. Today, the three dialects differ mainly in vocational schools). However, with the recent educa- vocabulary and a few grammatical rules (see below). tion reform, communities can choose the language to After Tok Pisin had stabilized, it began to be used be used in the first three years of elementary educa- for new functions, such as religion. It was developed tion, and many have chosen Tok Pisin. Although many into a written language by missionaries in the 1930s people still feel that Tok Pisin is inferior to English, and was later used in newspapers and radio broadcast- most accept it as a separate language and an important ing. As its use was extended into new areas, it changed language of Papua New Guinean identity. linguistically to become more complex; i.e. it acquired Tok Pisin differs from the other two dialects more vocabulary and more grammatical rules and because, as mentioned above, most of the returned inflections. (The same process occurred with Bislama laborers had worked in Samoa rather than Queensland. and Pijin.) Thus, in both function and structure, Tok In addition, nearly all of the New Guinea laborers Pisin (and Melanesian Pidgin as a whole) became were from New Britain and New Ireland and the what is called an ‘expanded pidgin’. neighboring small islands, where the internal German- Finally, in recent years, especially in urban areas of owned plantations were also located. Thus, Tok Pisin Papua New Guinea, such as Port Moresby and Lae, has many words from the languages of these islands many people have been marrying outside their tradi- (especially Tolai), as well as some from Samoan and tional language groups. Thus, often the common lan- German. Some examples are kiau ‘egg’ and lapun guage of the parents is Tok Pisin, and this is what their ‘old’ from Tolai, malolo ‘rest’ from Samoan, and gumi children acquire as their first language. Because of this ‘rubber’, beten ‘pray’, and raus ‘get out!’ from nativization (the process of a pidgin becoming a native German. language), some linguists apply the term ‘creole’ to Bislama has many words derived from the lan- Tok Pisin (and Melanesian Pidgin in general), empha- guages of central Vanuatu, e.g. nakamal ‘meeting sizing that it has thousands of native speakers and has house’ and nabanga ‘banyan tree’. Bislama also dif- the functions and grammatical features found in typi- fers from both Tok Pisin and Pijin in having many cal creoles. Those who say that it is still a pidgin words derived from French. (Both France and Britain demonstrate that more than 90% of its speakers have were the former colonial powers in Vanuatu.) Some learned it as a second rather than a first language and examples are bonane ‘New Year’s celebration’, pima 1115 TOK PISIN ‘chilli’, and lafet ‘holiday’. Pijin differs from the other of the pidgin or creole (the substrate or substratum lan- two dialects in having very few words derived from guages) have affected its structure. Although many local languages or any language other than English. scholars have downplayed the role of substrate influ- An example of a grammatical difference between ence in pidgins and creoles, many features of the three dialects is in the way they indicate progres- Melanesian Pidgin are clearly derived from or rein- sive aspect (marked with -ing in English). Tok Pisin forced by those of the Oceanic Austronesian substrate most often uses wok long (literally ‘work at’) before languages. These include the following: the verb or the locative or existential verb stap ‘stay, (1) The transitive suffix (showing that the verb takes exist’ after the verb (with an intervening i); Bislama a direct object) (-im in Tok Pisin), e.g. meri i bin also uses stap but before the verb; and Pijin sometimes boilim wara, ‘the woman boiled the water’. uses initial consonant–vowel reduplication (but often (2) The subject referencing marker (i) in the verb does not mark progressive aspect). For example, ‘I’m phrase, e.g. Tom i wok long Mosbi, ‘Tom works drinking’ would be the following in the three dialects: in Port Moresby’. Mi wok long dring or Mi dring i stap (Tok Pisin); Mi (3) Exclusiveness and dual and trial number stap dring (Bislama); and Mi didring (Pijin). marked in the pronoun system, e.g. yumi (first- Tok Pisin has been the focus of research testing sev- person plural inclusive, i.e. ‘we, including eral theoretical questions in creolistics (the study of you’) vs. mipela (first-person plural exclusive, pidgin and creole languages). The first concerns ‘typi- i.e. ‘we, not including you’) and yu (second- cal’ creole features and developmental history. Creole person singular, i.e. ‘you, one person’), yutu- genesis (or ‘creolization’) in areas such as the pela (second-person dual, i.e. ‘you, two Caribbean is believed to have occurred very quickly, persons’), yutripela (second-person trial, i.e. with stabilization and expansion occurring along with ‘you, three persons’), and yupela (second-per- nativization. However, the creolization of Tok Pisin son plural, i.e. ‘you, many people’). has been taking place gradually, with stabilization and then expansion occurring before nativization. Nevertheless, it shares many features with other cre- References oles, such as the following: Crowley, Terry. 1990. From beach-la-mar to Bislama: the emer- (1) Tense, aspect, and modality are indicated by gence of a national language in Vanuatu. Oxford: Clarendon and New York: Oxford University Press. separate words occurring before the verb, e.g. ––––––. 2003. A new Bislama dictionary, 2nd edition. Suva, Em i bin wok asde, ‘he worked yesterday’ and Fiji: Institute of Pacific Studies, and Vila, Vanuatu: Pacific Em bai i wok tumora, ‘he will work tomorrow’. Languages Unit (University of the South Pacific). (2) Lack of a copula (a form of the verb ‘to be’), Heubner, Thom, and Stephen Rex Horoi. 1979. Solomon e.g. mi hanggre, ‘I’m hungry’. Islands Pijin. Brattleboro, Vermont: The Experiment in International Living. (3) Use of the third-person plural pronoun (ol in Jourdan, Christine. 2002. Pijin: a trilingual dictionary. Tok Pisin) as a plural marker, e.g. mi lukim ol Canberra: Australian National University (Pacific dok, ‘I saw the dogs’. Linguistics). Keesing, Roger M. 1988. Melanesian Pidgin and the Oceanic Second is the theoretical question of discontinuity. substrate. Stanford: Stanford University Press. It has been claimed that creoles are structurally dis- Mihalic, F. 1971. The Jacaranda dictionary and grammar of continuous from their pidgin predecessors in having Melanesian Pidgin. Milton, Queensland: Jacaranda. many more complex structures. These are supposedly Romaine, Suzanne. 1992. Language, education, and develop- ment: urban and rural Tok Pisin in Papua New Guinea. innovations of children that occurred during nativiza- Oxford: Clarendon. tion and could be attributed to their innate language Shnukal, Anna. 1988. Broken: an introduction to the Creole faculty. Tok Pisin differs from nearly all known cre- language of Torres Strait. Canberra: Australian National oles in that it is still acquiring native speakers, and University (Pacific Linguistics C-107). reprint, 1998. therefore it has provided a good opportunity for study- Siegel, Jeff. 1998. Substrate reinforcement and dialectal differ- ences in Melanesian Pidgin. Journal of Sociolinguistics 2. ing creolization as it occurs. Because it was believed Simons, Linda, and Hugh Young. 1978. Pijin blong Iumi: a that grammatical expansion in creoles is the result of guide to Solomon Islands Pijin. Honiara, Solomon Islands: nativization, researchers expected to find discontinuity Solomon Islands Christian Association. between the Tok Pisin spoken as a second language by Smith, Geoff P. 2002. Growing up with Tok Pisin: contact, cre- adults and the creolized Tok Pisin of children who olization, and change in Papua New Guinea’s national lan- guage. London: Battlebridge. acquired it as a first language. However, no substantive Wurm, S.A., and P. Mühlhäusler (eds.) 1985. Handbook of Tok differences have been found. Pisin (New Guinea Pidgin). Canberra: Australian National Last is the question of substrate influence––the University (Pacific Linguistics C-70). degree to which the first languages of the originators JEFF SIEGEL 1116  TONE LANGUAGES Tone Languages In addition to consonants and vowels, languages also may change, under certain circumstances, from one use the suprasegmental to convey different meanings. category to another. However, individual tonal compo- Through stress, English contrasts such pairs of words nents of a category do not spread freely in syllable- as /`insult/ vs. /in`sult/ and /`permit/ vs. /per`mit/. tone languages. Only rarely do we see the copying of Many languages employ tone for similar purposes, but an entire tonal category from one syllable onto an they differ from stress languages in some significant adjacent one in such languages. ways. First of all, tone languages regard tone as an Taking the simple word as its basic domain, word- important means to contrast lexical meanings. Thus, tone languages assign tonal categories to the word as Mandarin distinguishes unrelated words by tone alone a whole, irrespective of the number of syllables avail- (the raised numerals indicate relative pitch levels at able. Core tonal categories can exhaustively be iden- which a tone starts and ends, with 5 representing the tified in shorter words. Mende (an African language) highest and 1 the lowest): /aj55/ ‘sadness’ vs. /aj35/ ‘suf- has a word-tone system. The actual patterns for words fer’ vs. /aj214/ ‘(of height) short’ vs. /aj51/ ‘love’. (cf. carrying the tonal category of LH (a low tone plus a the English examples above, where stress contrasts high tone, to be signified with the diacritics ` and ´, only word categories––nouns vs. verbs). Second, the respectively) thus vary as follows: /mba / ‘rice’, occurrence of tone in a word is not predictable, where- /fàndé/ ‘cotton’, /ndàvúlá/ ‘sling’. The LH pattern is as the assignment of stress is rule-governed. observed only in disyllabic words; it becomes a rising Therefore, we may define tone language as one that tone in monosyllabic words, whereas in trisyllabic uses distinctive pitch to contrast concrete meanings of words the H tone spreads onto the final syllable. simple words, with the assignment of tone unpre- Tone-spreading within the basic domain of tonal cat- dictable by phonological rules. egories is common. Lexical tone and stress need not be regarded as a Pitch-accent languages such as Japanese generate pair of dichotomies, however. Indeed, some languages tonal patterns in an abstract domain, which may be make use of both. While no European language is pre- called ‘templet’. They do not directly contrast one dominantly tonal, most dialects of Norwegian, tone against another. Instead, tonal categories emerge Swedish, Slovene, Serbo-Croatian, and Lithuanian from the orderly placement of a marked tone at the have prosodic systems with a combination of tone and nth position in the templet, where n typically starts stress. In these Scandinavian languages, tone assign- from 0 or 1 and continues up to n. Tone-spreading ment is sensitive to stress, whereas the reverse applies may occur within the templet to increase the reper- to the South Slavic and Baltic languages––the stress toire of tonal categories. As the pattern of a tonal cat- assignment is said to be sensitive to tone. egory is predetermined in the templet, situational Based on distinctive organizations of tones, three neutralization of two tonal categories is possible in major types of tone language can be recognized, i.e. shorter words. For instance, the citation form of syllable-tone, word-tone, and pitch-accent languages. Tokyo Japanese /hàná/ is ambiguous between ‘nose’ Syllable-tone languages represent the best-known and ‘flower’, while in fact the two carry different type of tone system, in which the syllable constitutes tonal categories. The neutralization ceases when the the basic domain for each tonal category. Cantonese nouns are attached with the topic clitic /wa/: /hàná is an excellent example of a syllable-tone language, wá/ ‘as for noses’ vs. /hàná wà/ ‘as for flowers’. The e.g. /fu55/ ‘call aloud’, /fu35/ ‘tiger’, /fu33/ ‘rich’, /fu21/ major characteristics of these three types of tonal sys- ‘hold’, /fu13/ ‘woman’, and /fu22/ ‘father’. The tone tem are summarized in Table 1. TABLE 1 A Comparison of Syllable-Tone, Word-Tone, and Pitch-Accent Languages Syllable-Tone Word-Tone Pitch-Accent Basic domain of tonal categories Syllable Word Templet Manifest of all tonal categories in shorter words Yes Yes Not necessary Neutralization of tonal contrasts in the citation form Impossible Not possible Possible Spreading of individual tone component Rare Common Common 1117 TONE LANGUAGES Tone Languages of Asia exists in monosyllabic words, e.g. /ʃî/ ‘louse’ (high- falling) vs. /ʃí/ ‘hundred’ (high-level) vs. /ʃ / ‘new’ Tone languages are abundant in Asia, stretching from (rising), but the first two words are nearly neutralized Japan in the east to Nepal in the west, and from China in the citation form. Tonal distinctions in Prinmi are in the north to Thailand and Myanmar (Burma) in the made by specifying the locus of the H tone in the tem- south. Sporadic tone languages are also found in plet and whether it spreads to the next unit. Expressing Indonesia and Pakistan. Sino-Tibetan comprises the these two parameters as [n] [±spread], the full patterns largest family of tone languages in this region, but of the seven tonal categories of Prinmi are illustrated some Tibeto-Burman languages have no tones, e.g. as in Table 3. Amdo Tibetan and the Kiranti languages of Nepal. All of the Tai-Kadai (including Thai and Lao) and Tone Languages of Africa Hmong-Mien (Miao-Yao) languages, spoken in Southeast Asia and southern China, are tonal. The great majority of African languages south of the Nontonal Southeast Asian languages are mainly those Sahara have tone. Except for Fulani, Wolof, and of the Austronesian (including Malay, Indonesian, and Swahili, the rest of the Niger-Congo family are almost Pilipino) and Austroasiatic families, with the notable all tonal. So are the Nilo-Saharan and Khoisan fami- exception of Vietnamese. lies. Afroasiatic languages, other than Semitic (e.g. Tone is well developed in all Chinese languages. Arabic and Amharic) and Berber, are also tonal. With Classic works describe Middle Chinese as having four a wealth of complicated tonal data, African languages tonal categories: Level, Rising, Departing, and have contributed significantly to the advance of Entering. Later linguistic development has brought autosegmental phonology. tonal systems of Chinese to extreme diversity. Among While only a few Asian languages use tone to dis- Mandarin dialects, for example, some have more tones tinguish word categories, African languages frequent- than Beijing, while others have fewer. Furthermore, ly exploit tone to signify grammatical categories such pitch values for the same tonal category may also vary as tense, gender, number, and so forth. Shona (spoken from one dialect to another, as shown in Table 2. in Zimbabwe) contrasts lexical and/or grammatical Korean represents an unusual case of dialectal vari- meanings of the following set of words by tone alone: ation in tone. While the Central dialect of Seoul is no (1a) /vá-nò-pàr-à/ ‘they offend’, (1b) /và-no´ -pàr-à/ longer tonal, tone distinctions in Middle Korean have ‘the ones who offend’, (2a) /vá-nò-pár-á/ ‘they been retained in Northeastern (Hamkyeng) and scrape’, (2b) /và-no´ -pár-á/ ‘the ones who scrape’. Southeastern (Kyengsang) dialects. The latter has such Similarly, tone in Hausa (spoken in Niger and north- contrast as /mál/ ‘horse’ vs. /mÁl/ ‘measure’ vs. /màl/ ern Nigeria) differentiates lexical meanings, e.g. ‘word’. /go´ oràa/ ‘bamboo’ vs. /gooráa/ ` ‘large gourd’, and The majority of tone languages in China and grammatical meanings, e.g. /màatáa/ ‘wife’ vs. Southeast Asia are of the type of syllable-tone. Word- /máatáa/ ‘wives’, and /kíráa/ ‘call’ vs. /kíràa/ ‘call- tone languages are found mainly in Tibeto-Burman ing’. Indeed, the grammatical tone predominates in languages of Tibet and Nepal, including Ü (Lhasa most African languages. Another remarkable phe- Tibetan), Tamang, Thakali, and Gurung. In addition, nomenon observed in many African tone languages is Shanghainese has also been identified as a type of ‘downstep’, in which a floating low tone (resulted word-tone language. from deletion of segments) causes the pitch of a high The best-known pitch-accent language is Tokyo tone to be lowered. Japanese. Apart from toneless dialects such as Sendai, With typological similarity to Southeast Asian lan- all Japanese dialects have pitch-accent systems––some guages, some West African languages have developed simpler while others more elaborate. Another pitch- syllable-tone systems, e.g. Grebo (spoken in Liberia). accent language is Prinmi (Pumi), a Tibeto-Burman The language has four primary tonal categories (the language of southwestern China. A three-way contrast lowered numeral1 indicates the highest pitch): (A) TABLE 2 Varying Pitch Values of the Tonal Categories in Major Dialects of Mandarin Tonal Categories Beijing Wuhan Guiyang Chongqing Kunming (1) Yin-ping (High-level) 55 55 55 55 44 (2) Yang-ping (Low-level) 35 213 21 11 31 (3) Shangsheng (Rising) 214 42 53 42 53 (4) Qusheng (Departing) 51 35 24 24 212 1118  TONE LANGUAGES TABLE 3 Tonal Categories in the Pitch-Accent System of Prinmi (R=Rising) Category Parameters Monosyllable Disyllable Trisyllable Quadrisyllable A [1] [–spread] Falling H-L H-L-L H-L-L-L B [1] [+spread] High H-H H-H-L H-H-L-L C [2] [–spread] Rising L-H L-H-L L-H-L-L D [2] [+spread] L-H L-H-H L-H-H-L E [3] [–spread] L-R L-L-H L-L-H-L F [3] [+spread] L-L-H L-L-H-H G [4] [spread] L-L-R L-L-L-H High Rising: /na21/ ‘drink’, (B) High: /na22/ ‘fire’, (C) The San Juan Copala dialect of Trique (spoken in Mid Rising: /hε∼32/ ‘four’, and (D) Low: /tε∼44/ ‘test’. Mexico) shows a syllable-tone system with eight tonal Mende (spoken in Sierra Leone) has a word-tone categories, e.g. /yã 21/ ‘be sitting’, /yã 32/ ‘corncob’, system with five major tonal categories: H, L, HL, LH, /yã 33/ ‘he is sitting’, /yã 34/ ‘salt’, /yã 35/ ‘scar’, /yã 44/ and LHL. The following words belong to the final cat- ‘unmarried’, /yã 55/ ‘one (in certain number phrases)’, egory: /mba/ ~ ‘companion’, /nyàhâ/ ‘woman’, and and /yã 53/ ‘Spanish moss’. Similarly, minimal sets of /ni`k´ini`/ ‘groundnut’. Etung (spoken in Nigeria) also tonal contrast in Sarcee (spoken in Canada) may distin- appears to be a word-tone language. It has eight tonal guish word meanings or grammatical meanings: /m´i/ categories: H, L, HL, LH, HLH, HHL, LHL, and LLH. ‘moth’ vs. /m¯i/ ‘snare’ vs. /m`i/ ‘sleep’ and /-lo´/ ‘hurt Somali (spoken mainly in Somalia, Ethiopia, and (continuative)’ vs. /-lo¯/ ‘hurt (imperfective)’ vs. /-lò/ Kenya) is a pitch-accent language. According to ‘hurt (perfective)’. Stress plays an important role in the Saeed’s Somali (1999), its tonal system is organized tonal system of some Central American languages such into three major categories: (A) H on the last mora (a as the Chinantec languages. short vowel has a mora, whereas a long vowel or diph- Donohue reports in Linguistic typology (1997) that thong has two), (B) H on the penultimate mora, and many languages in New Guinea have tone. Iau (spoken (C) H on nowhere, i.e. L throughout. Note that the in Irian Jaya, Indonesia) has a syllable-tone system locus of H in the templet is counted from the right. with eight tonal categories: /du44/ ‘wild pig’, /u33/ Furthermore, Somali does not multiple its tonal cate- ‘tree’, /da45/ ‘dog’, /tai23/ ‘landed’, /tai42/ ‘fallen’, gories with H-spreading; H always spreads leftward to /tai423/ ‘falling’, /of43/ ‘arm’, and /ko32/ ‘breadfruit’. the vowel in the same syllable. The three tonal cate- Teléfól (spoken in Papua New Guinea) is a simple syl- gories are exemplified as follows: (A) /wéyl/ ‘female lable-tone language with a rising tone against a falling calf’, / `inán/ ‘girl’; (B) /wé`yl/ ‘male calf’,/´inàn/ ‘boy’; one: /ku l/ ‘hand’ vs. /kûl/ ‘frog’. Word-tone languages and (C) /àmà/ ‘or’, /kùlùlàà/ ‘hot (past tense)’. include Enga, Kairi, and Usarufa, etc. Kairi has four Although representative syllable-tone, word-tone, tonal categories: H, LH, HL, and LHL. The final one and pitch-accent languages can be identified, a large always contains a falling tone, e.g. /pâ/ ‘tree species’, number of African languages do not fit into these pro- /tu` hˆi/ ‘lid’, and /àràmèrâu/ ‘basket type’. Pitch-accent totypes; many have been regarded instead as ‘restricted languages are numerous in New Guinea, but tone- tone languages’. Some of them might have pitch-accent spreading is generally absent. Una (spoken in Irian systems. Many Bantu languages appear to be interme- Jaya) generates four tonal categories by placing the diate between word-tone and pitch-accent systems. We marked H on the nth syllable, where n varies from 0 to shall arrive at a better classification of these languages 3: (A) /kàl/ ‘marsupial species’; (B) /ka´l/ ‘tree when a thorough understanding of the fundamental species’, /kJ´nkàlyà/ ‘joint’; (C) /kJ` nkályà/ ‘pimple’; organization of their tonal systems is available. and (D) /bi`lbi`ly´i/ ‘building materials’. References Other Tone Languages Bogers, Koen, Harry van der Hulst, and Maarten Mous (eds.) Many indigenous languages of the Americas, especially 1986. The phonological representation of suprasegmentals: those of Central America, are tonal. Tone development studies on African languages offered to John M. Stewart on differs considerably among North American languages, his 60th birthday. Dordrecht and Cinnaminson, NJ: Foris e.g. while Achumawi (of northern California) has tone, Publications. Comrie, Bernard (ed.) 1987. The World’s major languages. no tone contrast exists in Atsugewi, a closely related Oxford and New York: Oxford University Press. language. Like African languages, American languages Comrie, Bernard, Stephen Matthews, and Maria Polinsky (con- employ tone for both lexical and grammatical contrasts. sultant editors) 1996. The Atlas of languages: the origin and 1119 TONE LANGUAGES development of languages throughout the World. New York: Typology, and Related Topics. Tokyo: ILCAA, Tokyo Facts on File. University of Foreign Studies. Donohue, Mark. 1997. Tone systems in New Guinea. Linguistic Mazaudon, Martin. 1977. Tibeto-Burman tonogenetics. Typology 1. Linguistics of the Tibeto-Burman Area 3. Fox, Anthony. 2000. Prosodic features and prosodic structure: Merrifield, William (ed.) 1977. Studies in Otomanguean the phonology of suprasegmentals. Oxford and New York: phonology. Dallas, TX: Summer Institute of Linguistics. Oxford University Press. Odden, David. 1995. Tone: African languages. The handbook Heine, Bernd, and Derek Nurse (eds.) 2000. African languages: of phonological theory, ed. by John Goldsmith. Oxford and an introduction. Cambridge and New York: Cambridge Cambridge, MA: Blackwell. University Press. Pike, Eunice. 1992. Analysis of tone systems. Notes on Hulst, Harry van der (ed.) 1999. Word prosodic systems in the lan- Linguistics 56. guages of Europe. Berlin and New York: Mouton de Gruyter. Saeed, John. 1999. Somali. Amsterdam and Philadelphia, PA: Kaji, Shigeki (ed.) 1999. Proceedings of the Symposium–– Benjamins. Cross-Linguistic Studies of Tonal Phenomena: Tonogenesis, PICUS SIZHI DING Translation Translation occurs when a message or text produced in person’s words mean the same when spoken by one language (the ‘source’ language) is converted into another. a message or text in another language (the ‘receptor’ or ● (Applied) linguistics: Nowadays, however, trans- ‘target’ language). When used in contrast to interpret- lation studies is usually seen as a discipline that ing (the conversion of spoken text), translation denotes explores the nature of cross-language communi- the conversion of written text. Conversion, however, cation. It draws on (and contributes to) a wide may often be across different modes, as in ‘sight trans- range of fields, such as semantics, pragmatics and lation’, in which a written source text is converted into discourse analysis, psycholinguistics, communi- a spoken target text. Here, therefore, we take the wider cation science, and culture and gender studies. definition. Messages can also be translated between A key concept in translation studies is that of varieties of one language (e.g. supplying standard-lan- ‘equivalence’: the degree of similarity between source guage subtitles for dialect film dialogue). and target text or text element (section, paragraph, Translation studies is a long-established discipline phrase, or word). In size terms, the target may be that examines both theory and practice. It has three roughly as long as the source; it may summarize the main concerns: source (e.g. an English précis of Japanese company ● Prescriptive: To give practical advice to transla- documentation); or it may be an exegetic (explanato- tors. This tradition has existed for millennia: ry) expansion of the source (e.g. when translating holy written advice dates back, e.g. to Cicero in scriptures). In qualitative terms, the target may ancient Rome, or to Zhi Qian in third century CE ● Parallel the source’s exact words and syntax (an China. More recently, it expresses itself in the ‘interlinear’ translation), often in order to explain search for reliable means of assessing translation or gloss the source. quality and in an academic interest in translator ● Reproduce the source’s semantic content (a liter- training issues. al translation). ● Philosophical: Scholars and philosophers, from ● Convey the source’s communicative effect (a the medieval Jewish cabalists to Jacques Derrida ‘communicative’ or ‘dynamic-equivalent’ trans- in modern France, have explored how transla- lation), e.g. when translating the Chinese Chi fan tion, as communication that transcends a single le ma? (literally, ‘Have you eaten yet?’) as ‘How language, might shed light on key issues of are you?’ human communication, for example, whether there might have been a single ancestor of all the Establishing ‘correct’ equivalence becomes espe- world’s languages, whether meaning is tied to cially problematic when source and target languages language or independent of it, or whether one are very different in terms of 1120  TRANSLATION ● Culture (e.g. how can one convey the sense of ● Translators regularly backtrack to modify the the Arabic word jihad to a reader with no under- emerging target text in terms of word choice, standing of Islam?). discourse, style, and so on. ● Time (e.g. should one translate Shakespeare into ● First drafts are usually revised, preferably at a modern or Renaissance French?). later date. ● Linguistic explicitness: For instance, classical In much technical or commercial translating (as Chinese poetry’s lack of pronouns causes prob- opposed to literature or advertising, for example), lems for the translator into modern English there may be no conflict between source- and target- (should the subject be ‘I’ or ‘she’?). oriented tactics, because many source and target gen- Recent models of translation see texts within the res (e.g. Dutch and American air-conditioning framework of a wider communicative act, in which the documentation) follow the same textual rules. At translator—as source-text receiver and target-text pro- word, phrase, or clause level, however, problems ducer—is a key actor: inevitably occur; much of the translator’s creativity Original writer/speaker → Writes/speaks → Source consists of selecting strategies for solving such prob- text → Read/heard by → Translator → Writes/speaks lems. Among the translation strategies identified by → Target text →Read/heard by → Target reader/lis- translation process research are the following: tener ● Consulting a dictionary or other reference mate- An expert translator has reading and writing (or lis- rials. tening and speaking) expertise in one or more special- ● Paraphrasing or adding an explanation. ist genres; moreover, one source text may have several ● Compensation (e.g. replacing a pun by an idiom, different equivalents, none of which is necessarily the or by a pun in a different place). best. Hence, the translator is now seen as less of a reproducer and more of a creator or re-creator, making The translator may also have to make ideological active choices on the basis of genre knowledge, intu- decisions. Choosing whom to translate and not to itive feel, and personal preference. At the whole-text translate in a region of conflicting identities (e.g. level, translators may have to choose between 1990s Bosnia and Serbia) inevitably means taking sides. Translating into a world language such as ● Source-oriented translation: Staying faithful to English, especially if domesticating tactics (see the original writer’s or speaker’s wording and above) are used, may risk contributing to a neoimpe- intentions. This may involve ‘foreignizing’ rialist McDonaldization of world culture, although it (deliberately retaining source-language or may equally be a way for ‘minor’ cultures to make source-culture features), ‘idiosyncratic’ transla- themselves heard on their own terms. Feminist trans- tion (reproducing the original’s stylistic individ- lation theory raises ideological issues that extend ualities), or both. beyond gender, such as whether translators should ● Target-oriented translation: Keeping the target ‘improve’ offensive (e.g. sexist) texts or the role of text’s new purpose and readership uppermost in the translator in rediscovering forgotten (female) mind. For example, the great Arab translator writers. Hunayn Ibn Ishaq (809–873 CE) chose to con- vert the Greek specialist medical manuscripts of References Galen into popular-medicine handbooks. This may also involve ‘domesticating’ (adapting the Baker, Mona. 1992. In other words: a coursebook on transla- text to target-culture norms), ‘smoothing’ tion. London: Routledge. Baker, Mona. (ed.) 1998. The Routledge encyclopedia of trans- (removing the source’s stylistic individualities), lation studies. London: Routledge or both. Chan, Sin-Wai, and David Pollard (eds.) 1995. An encyclopae- In recent years, translation process studies have dia of translation: Chinese-English, English-Chinese. Hong examined how translators translate. Among the typical Kong: Chinese University Press. features of translating written prose revealed by these Fawcett, Peter. 1997. Translation and language. Manchester: St. Jerome. studies are the following: Steiner, George. 1992. After Babel: aspects of language and ● During translation, the translator reads ahead a translation, 2nd edition. Oxford: Oxford University Press. Venuti, Lawrence. 2000. The translation studies reader. New sentence or so at a time, analyzing sentences into York: Routledge. clauses and clauses into phrases. Von Flotow, Luise. 1997. Translation and gender. Manchester: ● By default, the actual translating tends to occur St. Jerome. phrase by phrase. FRANCIS R. JONES 1121 TRUBETZKOY, NIKOLAI SERGEYEVICH Trubetzkoy, Nikolai Sergeyevich Nikolai Sergeyevich Trubetzkoy was an important units that distinguish one word from another. member of the Linguistic Circle of Prague, a group of Trubetzkoy was a structuralist thinker, i.e. he did not linguists interested mainly in the analysis of the sound think that phonemes could be defined by themselves, systems of languages. Following the French linguist but only in relation to each other. The basic measure of Ferdinand de Saussure, these linguists based their difference is opposition; a sound stands in opposition work on a distinction between parole, language being to another insofar as the presence of one means the used, and langue, language as knowledge. In terms of absence of the other. If the presence or absence of a sound systems, this distinction appears as a clear divi- given sound unit creates differences in meaning in a sion between phonetics, the science of speech sounds given language, it is a phoneme of that language. as actually produced by speakers, and phonology, the This concept of phoneme developed by Trubetzkoy science of language sounds as mental representations. is typical of the so-called functionalist approach of the Trubetzkoy published extensively during the 1920s Linguistic Circle of Prague. For Trubetzkoy, the and his writing defines the Prague School views on phoneme can be defined neither by its psychological phonology. The most comprehensive and explicit for- character nor by its relationship with its phonetic vari- mulation of his theory can be found in his Grundzüge ants, but by its function. This is where Trubetzkoy’s der Phonologie, published posthumously in 1939 thinking differs from that of his contemporaries such (English translation 1939, Principles of phonology). as Leonard Bloomfield or Daniel Jones. Phonetics, as Trubetzkoy sees it, is the study of Once the phoneme inventory of a language is estab- speech sounds as they are physically instantiated, i.e. lished, the phonologist must determine the distinctive it deals with the production or articulation of speech features that differentiate the phonemes from one sounds and their acoustic features that are relevant for another. /p/ and /b/ differ, for example, in voicing, perception. Phonology is a level of analysis that is one while /p/ and /t/ differ insofar as they are articulated in step removed from the physical dimension: it abstracts specific different locations in the mouth. In his work on away from the variability of actual pronunciations and distinctive features, Trubetzkoy developed two impor- attempts to filter out the constant characteristics or tant concepts: ‘neutralization’ and ‘archiphoneme’. features of speech sounds that, for example, allow a Sometimes, different phonemes of a given language are hearer to perceive equivalent utterances of different pronounced exactly alike, i.e. their distinctive feature speakers as essentially the same. Thus, phonology has been suppressed. If this occurs in a systematic fash- describes the abstract mental representations of speech ion, e.g. in a particular position within a word, sounds that must be more or less the same for all Trubetzkoy says that the distinctive feature has been speakers belonging to a given speech community. neutralized. German, for example, distinguishes the While the one is more concrete and the other more phonemes /t/ and /d/, i.e. voicing is a distinctive feature abstract, phonetics and phonology nevertheless inform just as in English. German, however, devoices sounds on and borrow concepts from each other. at the end of words: Rad ‘wheel’ is pronounced exact- A phonological description starts with the discov- ly like Rat ‘counsel’. Thus, German neutralizes the ery of those phonetic features that differentiate mean- voicing distinction in word-final position. Trubetzkoy ings in a given language. Pit and bit mean different calls the representation of such neutralized elements things in English and differ only insofar as the first ‘archiphonemes’. Thus, an archiphoneme is a phono- sound bit is produced with vibrating vocal cords ( [b] logical unit characterized by only those distinctive fea- is ‘voiced’), while the first sound in pit is not ( [p] is tures that have not been neutralized. Thus, ‘voiceless’). Thus, the voiced/voiceless distinction is Trubetzkoy’s idea of the archiphoneme is related to the crucial in English (it is not necessarily so in other lan- concept of phonemic ‘underspecification’ in modern guages), and voicing is thus referred to as a ‘distinctive distinctive feature theories. feature’ in the phonology of English. Principles of Phonemes and distinctive features are the main phonology concentrates mainly on such distinctive phonological elements that Trubetzkoy used to describe features. the phonological system of a given language. These ele- Building on a concept elaborated by the Russian ments are defined through their relationships with the linguist Jan Baudouin de Courtenay, Trubetzkoy other elements of the phonological system of that lan- argued that words contain ‘phonemes,’ basic sound guage. These ideas are the basis of the structuralist (and 1122  TUAREG AND BERBER LANGUAGES functionalist) phonology for which Trubetzkoy is most- He was a member of the Prague Linguistic Circle ly known. since 1926 and a member of the Academy of Sciences However, Trubetzkoy is also well known as the first at Vienna, in 1930. Trubetzkey died on June 25, 1938. linguist who called for a new discipline, namely ‘mor- phonology’ (or ‘morphophonology’), to describe the References interaction of phonological and word structures (mor- Anderson, Stephen R. 1985. Phonology in the twentieth centu- phemes). In his principal work on morphonology ry: theories of rules and theories of representation. Chicago: (‘Gedanken über Morphonologie’), Trubetzkoy dis- University of Chicago Press. cusses the phonological structure of morphemes and Fischer-Jorgensen, Elli. 1975. Trends in phonological theory. develops a theory of how the phonological shape of Copenhagen: Akademisk Forlag. Jakobson, Roman. 1939. Nécrologie Nikolaj Sergejevic morphemes changes under particular conditions. Trubetzkoy.’ Acta Linguistica 1; Reprinted in Sebeok, Thomas A. (ed.) 1966. Portraits of linguists. A biographical source book for the history of western linguistics, Biography 1764–1963. Bloomington: Indiana University Press. Nikalai Sergeyerich trubetzkoy was born in Moscow, Jakobson, Roman (ed.) 1989. N. S. Trubetzkoy´s letters and notes. Berlin: Mouton [Mostly in Russian]. Russia on April 4, 1890. He began studying Russian Trubetzkoy, Nikolai Sergeyevich. 1929. Zur allgemeinen ethnography and folklore at the age of 13 and pub- Theorie des phonologischen Vokalsystems. Travaux du lished on Slavic languages and Finnish folklore before Cercle Linguistique de Prague [TCLP], 1. entering university. He studied at the University of ———. 1931. Gedanken über Morphologie. TCLP, 4. Moscow in 1908–1912. He received his doctorate ———. 1933. La phonologie actuelle, Journal de psychologie. 30. ———. 1934. Das morphonologische System der russischen from Moscow in 1913 with a dissertation on the future Sprahe. TCLP, 5. tense in Indo-European. He became Assistant ———. 1939. Grundzüge der Phonologie, TCLP, 7 (1939, Professor, University of Moscow in 1915; he fled to posthumous). English translation, Principles of Phonology, Rostov after the 1917 Russian Revolution, where he Los Angeles, CA. taught Slavic Philology at the University. He was Trnka, Bohumil1958. ‘Prague structural linguistics’. Philologica Praguensia 1. Professor of Comparative Linguistics at the University ELENA BATTANER-MORO of Sofia (Bulgaria), 1920–1922, and Professor of Slavic Philology, University of Vienna, 1922–1938. See also Phonetics; Saussure, Ferdinand de Tuareg and Berber Languages Berber, which is the mother tongue of the first inhabi- Algeria. Third, we have the Tuareg populations, which tants of North Africa, is a Hamito-semitic language. It represent around 3 million people scattered in the sub- is spoken chiefly in North Africa, namely: Morocco, Saharan countries, namely: Niger, Mali, and Libya. Algeria, Tunisia, Libya, and Egypt. Berber is also spo- The fourth population group totals about 140,000 peo- ken by the Tuareg populations in Mali, Niger, and ple; these are generally in small and isolated areas, par- Burkina-Faso. The regions where Berber is used are ticularly in Tunisia (less than 100,000) Mauritania discontinuous in nature; that is, they are surrounded by (about 10,000 people), and Siwa in Egypt (30,000 peo- Arabophone or other populations. These discontinu- ple). cf. Chaker (1990) and Mustapha (1993). ous groups may range in size from a tribe, a village, or even a greatly populated center. Tuareg The Berber population can be divided into four major groups. First, we have Morocco, where the Tuareg is a variety of Berber spoken in Mali and Niger Berberophones are estimated to comprise about 40% of in West Africa; it has an alphabet called Tifinagh. the overall population. Second, there is Algeria, in Tuareg people used to be called ‘hommes bleus’ (blue which approximately 25% of the population is Berber people) because they lived in an immense desert and (cf. Grandguillaume 1983; Chaker 1990 and Ennaji were brave and respectable people. The Tuareg were 1991). This means that the Berber-speaking population colonized by France until 1960 when Mali and Niger is around 9.5 million in Morocco and 5.5 million in obtained independence. 1123 TUAREG AND BERBER LANGUAGES Today, Tuareg people are scattered over Mali and the other countries (Algeria, Libya, and Burkina-Faso). Niger; after four decades of independence, Tuareg peo- Today, the Tuareg people, who are generally proud of ple living in these two countries feel excluded and their their langue, struggle to freely use their mother tongue language and culture marginalized. Many cultural and their culture. But because of the pressure put by the associations in the region militate for the official recog- authorities in both Niger and Mali on Tuareg people, nition and revitalizing of Tuareg language and culture. some young native speakers feel it is degrading to In a letter addressed to the United Nations on August 1, speak Tamacheq in public (see Elékou 1998). 1997, Abdoulahi Attayoub, president of the association ‘Survie Touarègue’ asked the UN to protect the Tuareg Berber Languages minorities in Mali and Niger whose rights were violat- ed and whose lives were in danger because of racial Of all the North African countries, Morocco has the discrimination instituted by the regimes in place. Thus, largest Berber-speaking population, followed by thousands of Tuareg people were killed by the army in Algeria. The above estimates are only tentative, as both countries between 1980 and 1995. Others fled exact figures are impossible to provide given that the their country and settled in Algeria, Mauritania, and old censuses in Morocco (1982) and in Algeria (1986) Burkina-Faso, where they still live in hard conditions did not take into account the linguistic situation. (see the journal Tifinagh 1998, issue 3). However, the most recent Moroccan census (organized Since the periods of independence, the different gov- in 1994), which was also concerned with the linguistic ernments that took power in Niger and Mali marginal- factor, reveals that approximately 11 million ized the Tuareg people and often attempted to wipe out Moroccans speak Berber today. their language and culture. The Tuareg regions have Nearly half of the Moroccan population is been deprived of all hope of socioeconomic develop- Berberophone; i.e. they are monolingual Berbers, and ment, and their cultural space has been denied to them. the majority of them are bilingual because they also The Tuareg people live in Central Sahara and on the speak Dialectal Arabic. The monolinguals are basical- borders with the Sahel. They call themselves ly very small children and old people who do not par- Imajighen, Imoushagh, or Imouhagh depending on the ticipate in active life, whereas the bilinguals are regions; their language is called Tamajeq, Tamasheq, usually the young and the educated. Berberophones or Tamaheq. The Tuareg language is generally consid- are considered illiterate as long as they do not master ered the most well-preserved and the richest of the Standard Arabic or French. Berber languages. Tuareg people have retained an Berber is a spoken language with virtually no real original alphabet called Tifinagh, which is derived written history. Unlike the North African populations, from an ancient Berber writing system that was used the Tuareg populations in the south of Algeria and in all North Africa and in the Sahara. Mauritania –– as mentioned above –– have preserved an The oral tradition in Ahaggar shows that Tuareg old writing system called Tifinagh, which has been hand- people are the descendants of Tin-Hinan, queen and ed down from generation to generation through cen- mythical ancestor of the north. Their political system turies. However, there is no evidence that Tifinagh has was based on a confederation, where power belonged ever been utilized to write down history or literature. to the direct descendants of this queen mother, who is Today, the Classical Arabic script is at times used by the founder of the linguistic community of the Tuareg creative writers, close friends, or members of a family language. All Tuareg people have a common history, to write personal letters. Resorting to writing Berber in the same language, and the same lifestyle, which are Arabic script is often due to the lack of mastery of the the bases of a solid feeling of belonging to their land Classical Arabic language. The Moroccan Association and of a common cultural identity called Temoust. of Cultural Exchange publishes a periodical in Berber Traditionally, there were seven Tuareg confedera- entitled Amud, where Berber is written by means of the tions, which are today distributed as follows: Ahaggar Arabic writing system. In Algeria, Berber textbooks and (in Algeria and Niger), Ajjer (in Libya and Algeria), novels are written in Latin script (cf. for example, the Aïr (in Niger), Azawad (in Niger and Mali), Adghagh works of Mouloud Mammeri, Rachid Aliche, and (in Mali), Tadamakkat (in Mali), and Oudlan (in Ramdane Achab). The Tifinagh script, which has been Burkina-Faso). Today’s political borders are inherited recently revived and modernized, is widespread in the from the days of colonization; they have cut the Berber populated area of Kabylie (see Chaker 1984:37). Tuareg country into many parts that are integrated into In Morocco, there are three main Berber dialects. the five countries cited above. First, there is Tarifit used in the Rif, north of the coun- Because of the lack of a census where the linguistic try. Second, there is Tamazight, spoken in the Middle factor is taken seriously into account, the population of Atlas, and east of the High Atlas. Third, there is Tuareg people is estimated to be around three million: Tashlhit, spoken in the High Atlas and the Anti-Atlas 1.5 million in Niger, 1 million in Mali, and 500, 000 in Mountains, south of Morocco. 1124  TUAREG AND BERBER LANGUAGES The Berber dialects spoken in Morocco are gener- On the other hand, Berber itself has affected ally mutually intelligible unless the two varieties in Dialectal Arabic in many ways. A multitude of Berber question are situated at extreme geographical points. terms are incorporated in the latter language. If we For example, a native speaker of Tarifit in the north take the Moroccan context for illustration, we notice can hardly understand or communicate with a speaker that Berber intonation has infiltrated the Moroccan of Tashlhit in the south. Additionally, the Berber Arabic sound pattern, especially in the cities neigh- dialects used in Morocco and Algeria are usually boring Berber villages. Moroccan Arabic is also full of mutually incomprehensible. Berber loans; for instance, the following Moroccan In Algeria, there exist four major Berber dialects. Arabic words originate from Berber: brra (outside), First, there is Kabyle, which is spoken in the area of berrm (turn), atay (tea), taRunja (prayer for rain), and Kabylie, near Algiers and Bougie. Second, there is the tiwizi (collective land ploughing). Shawia dialect spoken by the populations of Aures, Most of these lexical items are fully integrated in north of Constantine. Third, there is the Tuareg dialect, the Dialectal Arabic inflectional system and so wide- which is a variety that is largely used in the extreme spread that few Moroccan Arabic speakers today will south of Algeria. Fourth, there is the Mzab dialect in recognize them as Berber. However, the more a Berber the Ibadhits, notably in Ghardaia. dialect is geographically distant or isolated from the In Tunisia, the Berber population is less important. urban centers, the less affected it is by the Arabic lan- Berber is spoken only in the extreme south of the guage. This is evidenced by the fact that in towns and country. There exist a handful Berberophone villages, cities, Berberophones tend to use more Arabic loans chiefly situated in Djerba, in the west of Matmata, than in the rural Berber areas. Nonetheless, the young and in the east of Gafsa (cf. Payne 1983; Chaker generation of Berberophones is much more influenced 1984). by Arabic than the old generation. This is mainly due The traditional segmentation of Berberophone to the importance of urbanization and the expansion of regions into zones does not genuinely reflect the con- schooling and Arabization in the educational system siderable effect of rural exodus that the Maghreb (see Ennaji 1988; Elbiad 1991). (Morocco, Algeria, Tunisia, and Libya) has witnessed Linguistically, Berber has the following intrinsic since Independence. As a result, there are nowadays features. First, phonologically, it resembles Dialectal vast Berberophone populations in the urban centers of Arabic in that most of the consonants and vowels the Maghreb. Thus, Berber is also spoken in the cities existing in Dialectal Arabic do exist in Berber. of Algiers, Oran, Tizi-Ouzou, Agadir, Casablanca, Nevertheless, there is a difference as far as the verb Rabat, Fès, Nador, Tangier, and, in fact, in most large system is concerned. The latter is not much affected by cities (see Chaker 1990; Ennaji 1991). Dialectal Arabic. For example, most of the loan verbs Although there are many Berber dialects, Berber is are integrated in Berber verb morphology with no con- fundamentally one language (cf. Boukous 1975; sequence on the verb group pattern. By way of illus- Chaker 1984). Although most Berber dialects are gen- tration: xdm (work) --> i-xdm (he worked); jawb erally mutually unintelligible, they share similarities at (answer) ---> t-jawb (she answered). This shows that the phonological, morphological, syntactic, and lexi- the Berber verb inflection is essential for sentence for- cal levels, as will be seen below (cf. Basset 1959; mation and meaning, which accounts for the verb sys- Applegate 1970 for details). tem’s resistance to Arabic influence. The Linguistic Features of Berber Evolution of Berber Historically, Berber has borrowed mainly from Latin, Berber has experienced a gradual regression in Arabic, and French. The remnants of Latin loanwords Tunisia, Morocco, and Algeria. As we have already are limited to a dozen or so. Nonetheless, the most mentioned, the number of Berber speakers in Tunisia well-known loans come from Dialectal Arabic and has dramatically dwindled to about 100,000, while in French. Most of these borrowings are nowadays com- Morocco and Algeria the number of competent native pletely adapted to the patterns of Berber. For instance, speakers of Berber is also regressing, especially in the Latin words asinus (little donkey), burrhus (coat), urban areas where Dialectal Arabic is more powerful tussis (cough) are pronounced asinus, abernus, and and commonly used. tusut in Tamazight, the Berber dialect of the Moroccan Four major reasons lie behind the regression of Middle Atlas; The Dialectal Arabic words sa:’a Berber. First, there is the phenomenon of rapid (watch) and Tbib (doctor) become tassa’t and adbib, urbanization, which is itself the result of migration respectively, in Berber. The French loans veste (vest) from rural areas to urban centers. Thousands of and automobile (car) become lfista and Tomobil, Moroccan Berberophones have quit their homes and respectively, in Berber (see Ennaji 1991). villages in the last four decades to settle down in cities 1125 TUAREG AND BERBER LANGUAGES like Casablanca, Agadir, Rabat, Nador, El-Hoceima, and vitality. They are historically among the oldest lan- Fès, and Marrakesh to better their lives. guages in the world. Tuareg Berber is marginalized Second, there is the expansion of free education although it is spoken by approximately 3 million people since Independence. Thus, the schooling of Berber- scattered around five different countries: Mali, Niger, speaking children meant their Arabization and their Burkina-Faso, Algeria, and Libya. It is the Berber cul- gradual loss of competence in Berber for the latter is tural influence that gives Morocco and Algeria their his- excluded from education. This is a factor that can be tory, and that makes them so different from the rest of remedied if Berber is seriously taught as a language in the Arab world. Berber remains vital because it is still public schools. spoken by a great part of the population in North Africa. Third, there is the fact that most Moroccans are The recent revival of interest in Berber language and cul- aware that Berber is pragmatically a local language ture is a good indication that its vitality will be strength- whose use is not widespread, unlike Arabic or French, ened to the extent that it may be raised to the level of a and hence it is not useful in business, science, tech- standardized and codified language in the future. nology, and international affairs. Its role is merely con- fined to the cultural, social, and historical domains (cf. References Ennaji 1991). Applegate, J.R. 1970. The Berber languages. Current Trends in Finally, the fact that Berber is officially treated as ‘a Linguistics 6. 586–61. dialect’ discourages people from using it in all walks Basset, André. 1959. Articles de dialectologie berbère. Paris: of life. Many Berberophone parents no longer interact Klinksiek. with their own children in Berber, simply because they Boukous, Ahmed. 1995. La langue berbère: maintien et changement. International Journal of the Sociology of feel that Berber lacks prestige and does not enjoy the Language 112. 9–28. status of a language (see Boukous 1997). One way of Boukous, Ahmed. 1997. Situation Sociolinguistique de verifying the regression of Berber is to examine the l’Amazighe. International Journal of the Sociology of domains in which it is normally used. Unlike in the Language 123. 41–60. 1940s and 1950s when Berber was spoken not only at Chaker, Salem. 1984. Textes en linguistique berbère. Paris: CNRS. home but also in the public administration and in polit- Chaker, Salem 1990. Langue berbère: une planification lin- ical circles, nowadays Berber is virtually limited to guistique extra-institutionnelle. Maghreb linguistics, ed. by home and to conversation among close friends. J. Pleines. 237–63. Dialectal Arabic is rapidly gaining ground to the detri- Elbiad, Mohamed. 1991. The role of some populations sectors ment of Berber even in homes of Berberophones and in the progress of Arabization in Morocco. International Journal of the Sociology of Language 87. 27–44. even in rural areas where Berber was hitherto the pre- Elékou, Moussa, Ag. 1998. Les Touareg au Niger: Repères dominant language of communication. For instance, Historiques. Tifinagh 13. 95–7. Berber is implicitly rejected in administrative offices, Ennaji, Moha. 1988. Language planning in Morocco and in the private sector, and in political spheres (cf. Sadiqi changes in Arabic. International Journal of the Sociology of 1997; Boukous 1997). Because of this rapid decline, Language 74. 9–39. Ennaji, Moha. 1991. Multilingualism in the Maghreb. intellectuals and progressive forces in the region have International Journal of the Sociology of Language 87. 7–25. recently begun a campaign seeking the revitalization Grandguillaume, Gilbert. 1983. Arabisation et politique lin- of the Berber language and culture. This campaign has guistique au Maghreb. Paris: Maisonneuve. resulted in positive changes in officials’ and people’s Mustapha, A.A. 1993. Siwa: a fight between death and life (in attitudes toward Berber, as we will see in the section Arabic). Al-Arabi 411. 134–51. Payne, R.M. (ed.) 1983. Language in Tunisia. Tunis: Bourguiba below. Institute of Modern Languages. Tuareg and the other Berber languages are spoken in Sadiqi, Fatima. 1997. Grammaire du Berbère. Paris: L’Harmattan. the Maghreb and the Sahel. They enjoy both historicity MOHA ENNAJI Tungusic The Tungusic languages are spoken by small, mostly region of Siberia between the Lena and Anadyr’ rivers, rural populations scattered across East Asia. The north- although one Even speech community is located farther ernmost language, Even, is found in the northeastern east on the Kamchatka peninsula. The southern 1126  TUNGUSIC geographic boundary of the Tungusic languages is in hunting or reindeer herding. Their mode of subsis- defined by Solon and Orochen, which are situated in tence made it necessary for individual clans to be small and around the Hinggan mountains of China. To the and for the clans to be spatially dispersed so as not to west, one finds Evenki speakers along the Yenisey overgraze or overhunt particular regions. Although indi- River in Russia and Sibe communities in Xinjiang vidual clans would remain in contact with each other, Province in China. To the east, speakers of Orok and the degree of interaction was limited; yet it was not Evenki inhabit Sakhalin Island. Despite the large area uncommon for individual or families to switch from one over which the Tungusic languages are spread, the clan to another, particularly for the purpose of marriage. highest density of Tungusic languages is within the The linguistic consequence of these social characteris- Amur River basin in southeastern Siberia, where there tics was extensive dialect differentiation. These dialects are speakers of Oroch, Negidal, Evenki, Nanai, Udege, frequently form a dialect chain: speech varieties that are Orok, Kili, and Ulch. close to each other along the chain are mutually intelli- Although scholars agree that the Tungusic lan- gible and are clearly dialect-like in their differences. guages constitute a distinct language family, some Varieties farther apart on the chain, however, cease to be uncertainty surrounds exactly how individual lan- mutually intelligible and are thus different enough to be guages are related. The family is commonly divided called separate languages. As a result, it is a matter of into three main branches: Southern Tungusic (com- some difficulty today to make a clear demarcation prising Manchu, Sibe, and the dead language Jurchen), between languages and dialects; thus, lists of Tungusic Central Tungusic (comprising Kili, Nanai, Oroch, languages can vary in content from as few as six mem- Orok, Udege, and Ulch), and Northern Tungusic (com- bers to as many as 17. prising Even, Evenki, Negidal, Orochen, and Solon). An added complication is that Tungusic groups are Two points of contention surround this classification. commonly referred to by names that have been given First, disagreements have persisted about how to clas- to them by political authorities and other outsiders. sify certain languages and dialects, particularly those These names do not always accurately reflect linguis- listed above as Central Tungusic. Second, some lin- tic realities. For example, Ewenki (Chinese Ewenke) guists argue that Central Tungusic is sufficiently close and Orochen (Chinese Elunchun) are the names given to Southern Tungusic to warrant combining these two to two of China’s official minorities. The former sug- branches into one. Indeed, this notion also character- gests a connection to the Evenki, one of the official ized many of the early efforts to classify the Tungusic minorities of the Russian Federation. However, some languages, notably the pioneering work of V.I. Cincius of the Orochen groups in China have a closer linguis- and Johannes Benzing. This two-way classification of tic connection to the Russian Evenki than do the Tungusic accounts for why the family is commonly Ewenki. Even more confusing is that what the Chinese referred to as Manchu-Tungusic, with Manchu repre- have designated as a single unit, the Ewenki, turns out senting the Southern/Central branch and Tungusic rep- to be three linguistically distinct groups, the Solon, the resenting the Northern branch. Khamnigan Evenki, and the Yakut Evenki. The last of Regardless of whether a two-way or a three-way these groups is intimately connected to the Evenki of classification of Tungusic languages ultimately proves Russia, the first two far less so. correct, there are certain criteria that have played, and Although some points of uncertainty remain will continue to play, a crucial role in determining the regarding the internal classification of Tungusic, much historical relationships among these languages. For greater debate has emerged over the relationship of the example, most Tungusic experts place great signifi- Tungusic family to other language families. In partic- cance on the correspondence of word-initial f- in ular, many scholars argue that Tungusic is a member Manchu and Sibe with word-initial p- in many (but not of a macrofamily called Altaic, along with at least all) of the Central Tungusic languages and word-initial Turkic and Mongolic, and perhaps Japanese, Korean, h-, x-, or no consonant in Northern Tungusic (e.g. and Ainu as well. Despite a superficial similarity in the Manchu faxun ‘liver’, Nanai pa, Orok paka, Negidal grammars of all these languages and suggestive evi- xaxin, Evenki hakin, Orochen axin). Another useful dence based on shared vocabulary and sound corre- characteristic commonly examined for classification spondences, proponents of Altaic have not yet been purposes is the presence vs. absence of -g- between able to demonstrate their hypothesis to the satisfaction vowels. Where many (but not all) Northern Tungusic of many linguists. Therefore, the question of the con- languages use the sound intervocalically, it tends not nection between Tungusic and other language families to appear in Southern and Central Tungusic (cf. remains an exciting area of current research. Orochen buga ‘land’ vs. Manchu bua vs. Ulch bua). The most familiar Tungusic language is Manchu, Traditionally, speakers of most Tungusic languages which served as the language of the Qing dynasty of were organized into small nomadic groups that engaged China (1644–1911), one of the largest empires in his- 1127 TUNGUSIC tory. Manchu and Jurchen hold a special place within Therefore, it is likely that all the Tungusic languages, Tungusic as the only two languages for which we have with the exception of Sibe, will become moribund written records that predate the twentieth century. The within the next generation. system used to write Manchu can be described as In terms of their structural features, Tungusic lan- either an alphabet or a syllabary, i.e. it contained sym- guages tend to organize their basic clause structure bols both for individual sounds and for syllables. It around the pattern Subject–Object–Verb, although was derived from Mongolic script at the very end of individual languages vary in how rigid this word order the sixteenth century, and the language was used for is. They have large numbers of suffixes, which are all manner of official documents and for the transla- used not only on verbs, nouns, and adjectives, but on tion of many Chinese classics; consequently, a large closed class items such as numerals. For example, corpus of written Manchu exists, which has proved Oroqen builds a number of words from numeral roots: invaluable for the reconstruction of proto-Tungusic, ilan ‘three’, ila-Nna ‘three animals’, ila-la ‘three the ancestor language, and for attempts to link days’, ila-na ‘a group consisting of three animals’, ila- Tungusic to other language families. tal ‘three each’, ila-ra ‘three times’, ila-kan ‘only During the reign of the Qing emperors, the use of three’, and ila-ki ‘third’. Manchu steadily declined as more and more ethnic Another common characteristic of Tungusic is vowel Manchus employed Chinese for their everyday activi- harmony. Most significantly, the vowels in certain suf- ties. By the time the Qing empire was overthrown in fixes must match the properties of vowels that are in the the early twentieth century, the Manchu people had word root. For example, in Sibe many derivational and been so thoroughly assimilated into the Chinese lan- inflectional suffixes have two forms, one with the vowel guage and culture that a linguistically distinct Manchu [u] and one with the vowel []. The choice of suffix people had ceased to exist. Despite nearly 10 million form is determined by the last vowel in the root. If this people claiming Manchu ethnicity in China today, the root vowel is pronounced with rounded lips, then the best estimates of fluent speakers of the language place suffix containing [u], also a rounded vowel, is the number between five and 30. employed. If the last root vowel is not rounded, the suf- The oldest records of a Tungusic language come fix containing [I], a non-rounded vowel, is used from Jurchen, a language spoken by the founders of the (e.g. batur ‘hero’ yields batur-lu ‘to be heroic’, while Jin dynasty of China (1115–1234). These records take ɔldi ‘a capturing’ becomes ɔldi-l ‘to capture some- the form of inscriptions found in northeastern China one’). Although all Tungusic languages exhibit some and in Korea. The Jurchen writing system, which harmonizing processes, there is tremendous variation in appears to combine symbols for words as well as for the form that the harmony takes and in how productive syllables, has not been fully deciphered. However, it is. Sibe possesses only rounding harmony, as just enough evidence from Chinese lists of Jurchen words illustrated. Other Tungusic languages also have harmo- exists to establish that the language is closely related to ny based on the feature of vowel tenseness (e.g. Nanai, Manchu, either as a historically earlier variety of where suffixes with a tense vowel, such as the adjec- Manchu or as a dialect parallel to Manchu. tive-forming -su, as in densi- ‘to care for’/d´nsi-su In terms of the number of speakers, the largest ‘careful’, have an alternative form that contains a lax Tungusic language is Sibe (Chinese Xibe), which, vowel -s•, as in sa- ‘to know’/sa-s•· ‘well-informed’). according to the 1982 Chinese census, had 26,760 Still other Tungusic languages, such as dialects of speakers. Sibe is a dialect of Manchu but has been des- Oroqen and Solon, have both types of harmony. ignated separately within China, primarily because of Tungusic languages make frequent use of converbs, the presence of a Sibe group in the Jungaria region of which are verb forms used for subordinate structures. Xinjiang Province. Over time this group, which has its Consider the following example from Evenki: si: origins in Manchu military units sent to protect the ↔m↔-tΣ↔l↔⎤-s nuŋartin ŋ↔n↔-d↔ŋ↔⎤-tin frontiers of the Qing dynasty in the mid-eighteenth ‘After you come, they will go.’ (literally: ‘you come century, has developed independently from other eth- they go’). The converb is ↔m↔tʃ↔l↔⎤s, which has nic Manchu groups. Moreover, the Sibe have been lin- been created by attaching -tʃ↔l↔ to the verb root guistically and culturally influenced by Mongolic and ↔m↔- ‘come’. This particular suffix indicates that Turkic groups living in Xinjiang. the event expressed by the converb preceded the event The only other Tungusic languages that currently expressed by the main verb. Evenki possesses 12 other have a critical mass of fluent speakers are Solon (esti- converb-forming suffixes. The suffixes allow speakers mated 14,000 speakers), Siberian Evenki (estimated to express many different meanings including simul- 10,000), Even (estimated 7,400), and Nanai (estimated taneity, posteriority, condition, and purpose. 6,000). In very few of the communities where these Another characteristic property of Tungusic lan- language are spoken, however, do children learn them. guages is that they employ case suffixes on nouns to 1128  TURING, ALAN indicate grammatical functions. In this regard, the References Northern Tungusic languages tend to have far more Bulatova, Nadezhda, and Lenore A. Grenoble. 1999. Evenki. case inflections than the Southern Tungusic languages. München: Lincom Europa. For example, Even differentiates at least 12 cases Doerfer, Gerhard. 1978. Classification problems of Tungus. (etiken ‘old man’, etike-m ‘old man’ (object), etiken- Tungusica, Vol. 1, ed. by Gerhard Doerfer and Michael du ‘old man’ (indirect object), etike-n’ ‘with the old Weiers. Wiesbaden: Otto Harrassowitz. Janhunen, Juha. 1991. Material on Manchurian Khamnigan man’, etike-n’un, etiken-duk, etiken-dula ‘at the old Evenki. Castrenianumin toimitteita 40. Helsinki: University man’, etiken-duli, etiken-taki ‘to the old man’, etike- of Helsinki. ŋitʃ, etike-kla, etike-kli), whereas written Manchu has Li, Bing. 1996. Tungusic vowel harmony: description and only five (ama ‘father’, ama-i ‘of the father’, ama de analysis. Amsterdam: University of Amsterdam. ‘father’ (indirect object), ama be ‘father’ (direct Li, Gertraude Roth. 2000. Manchu: a textbook for reading doc- uments. Honolulu: University of Hawaii. object), and ama ci). Malchukov, Andrei L. 1995. Even. München: Lincom Europa. Unlike Southern Tungusic, Northern Tungusic lan- Nedjalkov, Igor V. 1997. Evenki. London: Routledge. guages also contain a set of nominal suffixes that indi- Norman, Jerry. 1974. A sketch of Sibe morphology. Central cate possession (e.g. Oroqen murin-iw ‘horse-my’, Asiatic Journal 18. murin-iy ‘horse-your (sing.)’, murin-in ‘horse- Whaley, Lindsay J, Lenore A. Grenoble, and Fengxiang Li. 1999. Tungusic classification from the bottom up. Language 75. his/her’, muri-mun ‘horse-our/excluding the listener’, Zhang, Yan-chang, Bing Li, and Zhang Zi. 1989. The Oroqen murin-tir ‘horse-our/including the listener’, murin-ʃun language. Changchun: Jilin University Press. ‘horse-your (pl.)’, murin-tin ‘horse-their’). LINDSAY WHALEY Turing, Alan Alan Turing made foundational contributions to phi- tended to assume that it meant ‘explicit, step-by-step, losophy, mathematics, biology, artificial intelligence, requiring-no-creativity’ and left it at that. Turing and computer science. He, as much as anyone, invent- saw that if a procedure were mechanical, it could be ed and showed how to program the digital electronic automated. computer. From September 1939, his work on compu- Turing’s answer is to imagine a starkly minimal tation was war-driven and practical. He developed machine. He called it a ‘theoretical computing high-speed computing devices needed to decipher engine’, because he had no intention of building one. German Enigma Machine messages to and from U- This is a machine with which to think. It is fed by an boats, countering the most serious threat by far to indefinitely long tape divided into frames like those on Britain’s survival during World War II. a roll of film. The machine has a ‘read head’, which Because of official secrecy, his war time exploits can tell whether the frame under it has a ‘/’, a ‘\’, or is were unknown until the 1980s. By then some of his blank. The read head is also a ‘write head’ that can inventions no longer seemed connected to a real erase, write in a ‘/’, or ‘\’, or do nothing. The machine human being. For example, Turing’s 1936 paper, ‘On can also move one frame forward or backward or stay computable numbers’, was soon seen as the most in place. Then there is another move and so on. At the important theoretical paper ever written on computa- beginning of each move, the machine is in one of a tion. Hence, mathematicians, engineers, and computer small number of ‘internal states’ and this may switch scientists came to write of ‘turing machines’ and ‘uni- to another after the move. The machine is built to enact versal turing machines’ almost forgetting that there instructions in its machine table of the form ‘if / is read was an Alan Turing. and the internal state is 1, then erase, move one frame In his 1936 paper, Turing answered the deepest forward, and go into state 2’. A turing machine for computational question –– whether there is a finite adding would get, say, the input sequence ‘// //’ and mechanical procedure for deciding whether any given then automatically change it into ‘////’ through a long mathematical statement is true or false. Turing real- series of steps and then stop. For us this is the compu- ized that the key was to specify what a ‘mechanical tation ‘2+2=4’. Because Turing was thinking about a procedure’ (or computation) was. Mathematicians theoretical device, he did not mind that a million 1129 TURING, ALAN would be represented by a like number of ‘/’s. Indeed, In his 1950 paper that created the field of artificial he insisted that an actual and literal universal turing intelligence, Turing asks you to imagine the following machine would be much too slow for practical compu- ‘imitation game’. We have some judges who commu- tation. nicate by a terminal to A and B terminals. One of these Turing goes on to show that anything mathemati- terminals is operated by a woman, one by a man. Under cians call a mechanical problem or computation can be the judges’ questioning, the woman tries to convince represented by some turing machine. Further, he them that she is the woman, while the man tries to con- shows that there is a universal turing machine. vince them that he is the woman. For example, the Depending on the input tape sequence, this turing judges might ask, ‘How do you do your hair?’ Turing machine can turn itself into any particular turing comments that the best strategy for the woman is to tell machine, do a computation, and then turn itself into the truth. Interrogators look for inconsistencies. Telling another turing machine and do a computation, and so the truth is the simplest way to avoid them. The man is on. In 1936, Turing thought up the general-purpose going to have a lot to keep track of. If the man manages digital computer and gave its definitive abstract to win, you might say that he can think like a woman. description. The tape can also store data and programs Turing then proposes that we substitute a computer (‘memory’, ‘skills’, ‘plans’), represent incoming data for the man. If the computer ‘passes’, that proves that (‘sensory input’), and issue output instructions (‘motor the computer can think like a human being and that outputs’). Thus, Turing has also given the framework proves that the computer can think period. Turing in which to describe any sort of individual thinker. As remarked that the test draws a clear line between the many scientists have said since the 1960s, each human physical and mental. He likened putting the contest- is, more or less, a universal turing machine and so are ants in separate rooms, so the judges cannot see them, our digital electronic computers; for the same reason, to musical contests where the players perform behind we now think of thinking as computing or data-pro- a screen, so the judges would not be biased by their cessing. To return to his original question, Turing also physical appearance. This is what we now call the showed that no possible computer (digital electric, ‘Turing test’. Turing did not think it was going to be human, or alien) can decide the truth or falsity of every easy to create a passer. He provided a vivid and mathematical formula. Turing broke the species barri- testable goal for artificial intelligence research. er: given his minimalist physical description of com- On the other hand, some of Turing’s work was so putability, anything could be a mind whatever its far ahead of his time that it earned credit only in retro- physical composition. spect. His last published paper (1952) anticipated the His 1950 paper, ‘Computing machinery and intelli- most important new approach—chaotic, structure- gence’, set the agenda for cognitive science. From it, spawning reactions—in the last half century of devel- scientists and philosophers extracted the goal of writ- opmental biology. But it was not until the late 1970s ing ‘intelligent’ computer programs good enough to that scientists began to refer to ‘turing structures’. pass the ‘turing test’ for simulating human intelli- gence. Although some computer scientists boasted that Biography they would program a ‘passer’ within a decade, they have not come close after five decades. Turing also Alan Turing was born on June 23, 1912, in London. He suggested several other ways, aside from program- studied in Sherborne School, 1926–1931 and Wrangler, ming, that might be used to simulate human intelli- Mathematics Tripos, King’s College, Cambridge, 1931. gence. When he did so, he anticipated approaches that He received his Ph.D. from Princeton University in were explored in the 1980s and 1990s –– training con- 1938; he was Fellow, King’s College, 1935–1945; nectionist nets, sending a ‘child machine’ to school, Princeton University, 1936–1938; British Foreign perhaps equipping it with eyes, ears, and hands. The Office, Bletchley Park, 1939–1945; National Physical centrality of natural language in Turing’s test, and in Laboratory, 1945–1948; and University of Manchester, human intelligence, is reinforced by the continuing 1948–1954. He received the Smith’s Prize from failure of machine translation from one human lan- Cambridge University in 1936 and the Order of the guage to another. Computers now easily exceed the British Empire in 1946. He was Fellow, Royal Society most talented humans in arithmetic calculation and in 1951. In 1952, Alan Turing was convicted of ‘homo- chess playing, but none come anywhere close to the sexual acts’ and legally forced to take huge hormone performance of ordinary human translators (in merely doses that rendered a man who nearly represented taking dictation, our most sophisticated voice recogni- Britain in the 1948 Olympic marathon impotent, femi- tion programs do not come close to matching the per- ninely breasted, and obese. He died of self-administered formance of ordinary human secretaries). cyanide on June 7, 1954, in Manchester, England. 1130  TURKEY References ———. 1950. Computing machinery and intelligence. Mind. 59. ———. 1952. On the chemical basis of morphogenesis. Hodges, Andrew. 1984. Alan Turing: the enigma. New York: Philosophical Transactions of the Royal Society of London, Simon & Schuster. Series B. 237. 37–72. Leiber, Justin. 1991. An invitation to cognitive science. Oxford: ———. 1950. Computing machinery and intelligence. Mind. Blackwell. Also available in many other places including Hofstadter, D., Turing, Alan. 1936. On computable numbers, with an applica- and Dennett, D., The mind’s I. New York: Basic Books, 1984. tion to the Entscheidungsproblem. Proceedings of the ———. 1952. The chemical basis of morphogenesis. London Mathematical Society 42. Philosophical Transactions of the Royal Society. ––––––. 1947/1970. Intelligent machinery. Machine intelli- gence, ed. by B. Meltzer and D. Michie. New York: JUSTIN LEIBER American Elsevier Publishing Co. See also Artificial Intelligence Turkey The name Turkey, as historical Anatolia (and a small languages (especially Kurdish) are not allowed. Only part of the European mainland) was officially renamed the officially recognized religious minorities (Greeks, in 1923, incorrectly implies that the land is home to Armenians, and Jews) are permitted to teach in their the Turkish language. In fact, Anatolia was the tradi- respective languages at the primary and secondary tional home of a number of Indo-European languages school levels. As a result of these constraints and (notably Ionic Greek in the west, Hittite in the center, increasing urbanization over the last few decades, the Pontic Greek in the north, and Armenian in the east) many minority languages of Turkey have been severe- and several autochthonous languages including ly eroded, and many have disappeared entirely. There Hurrian, Hattic, Urartian, and Laz. Turkish and its are at least 37 languages currently or formerly spoken closest relatives only began to appear in Anatolia in natively in Turkey, but most of these are in danger of the eleventh century AD, and during the six centuries disappearing in the near future. of Ottoman rule (thirteenth to twentieth centuries) The largest language at present is Turkish, spoken expanded far beyond the borders of present-day by more than 60 million residents of Turkey, about Turkey and largely supplanted its linguistic predeces- 90% of the population. As of 1995, there were some sors. Dialects of Turkish are still spoken in small pock- 13 million additional speakers in other countries, pri- ets in the Balkans, especially Greece and Bulgaria, as marily Germany, Bulgaria, Greece, Macedonia and well as in Syria. More recently several other language Yugoslavia, the Netherlands, France, Romania, and families have become established in Turkey, notably Cyprus. Other Turkic languages are represented in Kurdish in the southeast and various northwest Turkey as well, including (as of 1982) Azeri (530,000 Caucasian languages in the north and west. in the Kars province), Gagauz (327,000), Kazakh These minority groups and their languages and cul- (600+), Kirghiz (1,137 in the provinces of Van and tures were tolerated to some extent during most of the Kars), Tatar (several thousand, including the Crimean Ottoman period, but the Empire in its last days and the Tatar village of Karakuyu in the province of Ankara), Republic that replaced it (1923–present) has imple- Turkmen (925 in the province of Tokat), Uyghur mented a number of military and legal actions designed (500+ in Kayseri and Istanbul), and Uzbek (1,981 in to minimize their presence, most importantly the elim- the provinces of Hatay, Gaziantep, and Urfa). ination of the sizeable Armenian and Greek popula- Within the Indo-European family, the largest repre- tions in 1915–1922, the name substitution program, sentatives in Turkey until the early twentieth century which replaced more than 30,000 traditional place were Greek and Armenian, which had about two million names (primarily Greek and Armenian) with new speakers each in Turkey in 1900. At the present time, Turkish names, and Law 2932, passed in 1983, which there remain a few tens of thousands of Greek speakers prohibits publishing in languages other than Turkish. in Turkey (mostly Muslims from the eastern Black Sea Children, moreover, are generally forbidden to speak a area; Orthodox Greeks now number only 3,000 in the language other than Turkish at school, and with few entire country), and less than 50,000 speakers of exceptions, radio and television broadcasts in minority (Western) Armenian, almost all of them in Istanbul. 1131 TURKEY There are also dozens of villages in Artvin, the At least two Balkan Slavic languages are represent- northeasternmost province of Turkey, in which vari- ed in Turkey as well: Bulgarian (270,000 speakers in eties of Homshetsma, a western Armenian dialect, are 1993, scattered in Edirne and other western provinces) still spoken. The exact number of Hemshinli who still and Bosnian (20,000 first language speakers in 1980, speak their ancestral language is unknown; 20–30,000 out of an estimated 61,000 Bosnian Turks scattered would be a rough estimate. All Hemshinli speak through western Turkey). It is not entirely clear how Turkish as well, and the western Hemshinli (who live these populations have been affected by the upheavals in the Rize province) now speak only Turkish, pep- in Bosnia and Kossovo in the 1990s, but there was at pered with remnants of their original Armenian dialect least a temporary exodus to Turkey during that period; (Bläsing 1992). This change was precipitated by the the same holds for the Albanian speakers in Turkey, conversion of the Hemshinli from the Armenian who numbered about 65,000 in 1993, scattered over Apostolic church to Islam beginning in the sixteenth western Turkey. century; those Armenians who did not wish to convert The Anatolian branch of Indo-European was locat- fled primarily to the Black Sea coastal regions of ed entirely in what is now called Turkey. The family Abkhazia and Russia, where they still reside today. includes Hittite, Luvian, Lycian, Lydian, Palaic, and Most Hemshinli at present do not consider themselves Carian, and appears to have completely died out by to be Armenian, and many of those who speak about the end of the first millennium BC. Extensive Homshetsma are unaware that it is a form of collections of cuneiform tablets (in Hittite and Armenian, a belief that is facilitated by the large num- Luvian), heiroglyphic inscriptions (in Luvian), and ber of Turkish loanwords and the Turkic-like structure inscriptions in Greek letters (in Lycian, Lydian, Palaic, of the grammatical system. and Carian) were unearthed and deciphered in the The Boshas or Poshas, sometimes referred to as twentieth century. The Indo-European language Armenian gypsies (because of their language, not their Phrygian was also spoken in the western half of ethnicity), also reside in Turkey (chiefly Istanbul; Turkey, and it too is now long extinct. there are also communities in Erevan, Armenia, Two Caucasian language families are fairly well Javaxk, Georgia, along the Black Sea, and in other represented in Turkey, partly due to spillover at the areas fomerly inhabited by Armenians) and speak a Georgian border in northeastern Turkey, and partly language related to Armenian. This language, thanks to the massive influx of refugees from the north- Lomavren (containing Lom ‘gypsy’ + the Armenian west Caucasus following the defeat of the Caucasian language suffix -(e)re¯n), is a ‘mixed language’, where- confederation by Russia in 1864, as a result of which in the lexicon is largely Indic (as with all gypsy lan- the Czar deported roughly half of the Northwest guages) and the grammar and phonology are based on Caucasian population to the Ottoman Empire. Some the Armenian dialect of Erzerum in northeast Turkey, Northwest Caucasian languages survived only in the where the Boshas once resided. As of 1982, there were diaspora, such as Ubykh (now extinct) and the Sadz also at least two other gypsy languages spoken native- dialects of Abkhaz. Other Northwest Caucasian lan- ly in Turkey: approximately 20,000 speakers of guages still to be found in Turkey today include Abaza Domari, mainly in western Turkey, and some 40,000 (10,000 speakers in Turkey as of 1995), Abkhaz speakers of Balkan Romani (Arlija). (35,000 speakers in Turkey as of 1993, mainly in The Indo-European branch that is best represented Artvin in northeast Turkey and villages in the provinces in Turkey today is Iranian, thanks primarily to the sev- of Bolu and Sakarya), Adyghe (130,000 in 1965, in eral million speakers of Kurmanji and other dialects of various provinces in central and western Anatolia), and Kurdish, but also Zaza (or Dimli; one million speakers Kabardian (202,000 speakers in Turkey as of 1993, pri- in 1992), Herki, and Shikaki. Bilingualism with marily around Kayseri). Turkish among the Iranian populations in Turkey is Northeastern Turkey contains the westernmost not as common as it is with the other minority lan- component of the South Caucasian language family, guages, presumably due to the significantly greater notably Laz (92,000 speakers in Turkey in 1980 along size and viability of the community, geographical the Black Sea Coast of the provinces of Rize, Artvin, obscurity, and the lower literacy rate (28%) than in the Sakarya, Kocaeli, and Bolu) and Georgian (some population as a whole (about 70%). 40,000 first language speakers in 1980 out of 91,000 Speakers of the northeast Iranian language Ossetic ethnic Imerxev, a Georgian minority, located primari- have been reported in Bitlis (Digor dialect) and ly in the provinces of Artvin, Ordu, and Sakarya). Sarkams¸ and Erzerum (Iron dialect). Speakers of var- Several Semitic languages survive in Turkey, ious Ossetic dialects have also been reported in MugFla, including Arabic (400,000 speakers as of 1996), Kars, and Antalya. Turoyo/Suryani (3,000 speakers in 1994, originally in 1132  TYPOLOGY the Mardin province), and Hertevin (250–300 speakers References in 1994, originally in the village of Ekindüzü in the Benninghaus, Rüdiger, and P. Andrews (eds.) 1989. Ethnic Siirt province). Turoyo is a modern descendant of groups in the Republic of Turkey. Wiesbaden: Ludwig Syriac, the language of the Syrian church, which was Reichert Verlag. centered in Edessa (modern S¸anlurfa, in southeastern Bläsing, Uwe. 1992. Armensiches Lehngut im Türkeitürkischen Turkey) and spoken until about the twelfth century am Beispiel von Hems¸in. Amsterdam: Rodopi. Barak, Veysi and Joyce Blau. 1999. Manuel de Kurde: AD, although it is still used as a literary secular lan- Kurmanji. Paris: L Harmattan. guage by some members of the Syrian church. Chirikba, Viacheslav. 1996. Common West Caucasian. Leiden: One may also classify the language of the CNWS Publications. Sephardic Jews, Ladino (or Judeo-Spanish, as it is Dumézil, Georges. 1963. Notes sur le parler d’un arménien sometimes called) as partially Semitic, although its musulman de Hemshin. Mémoires de l’Académie Royale de Belgique. Classe des Lettres 57.4. Brussels. core is an Iberian Romance language closely related to Finck, Franz Nikolaus. 1907. Die Sprache der armenischen Spanish. As of 1980, it was estimated to have 8,000 or Zigeuner. St. Petersburg. fewer first language speakers, mainly in Istanbul but Hadank, Karl. 1932. Mundarten der Za^za^. Berlin: Verlag der also in Izmir. The Dönme, adherents of Shabbetai Zevi Preussischen Akademie der Wissenschafter in Kommission who converted to Islam, also speak Ladino; there were bei Walter de Gruyter & Co. Kornfilt, Jaklin. 1997. Turkish. London: Routledge. 15,000 of them residing in Turkey in 1976. Almost all Newton, Brian. 1972. The generative interpretation of dialect: a speakers of Ladino are bilingual in Turkish. study of modern Greek phonology. Cambridge: Cambridge Turkey also possesses at least one distinct signed University Press. language, Turkish Sign Language. To the best of my Vaux, Bert. 1998. The phonology of Armenian. Oxford: Oxford knowledge, no linguistic studies have been carried out University Press. on this language. BERT VAUX Typology At its core, typology is the classification of languages disclose fundamental properties of human language on the basis of shared formal characteristics. In this more generally. way, it is distinct from genetic classification, which attempts to determine shared ancestry among lan- Universals guages, and from demographic classification, which categorizes languages based on geography, the num- One of the core notions of typology is that of the ‘lan- bers of speakers, or other social variables. The ulti- guage universal’, a statement about what is always or mate goals of typology are to ascertain the ways in nearly always true about language structure. which languages are similar in structure and to deter- Universals that are without exception, such as ‘all lan- mine just how different human languages can be. guages possess both vowels and consonants’, are Concomitant to these goals is the discovery of traits in called ‘absolute’. Universals that admit exceptions but language that are logically distinct but actually occur hold true to a statistically significant degree are together. referred to as ‘probabilistic’ universals. An example of It should be stressed that typology is not a theory a probabilistic universal is ‘Languages nearly always of language structure. It does not attempt to provide distinguish between nouns and verbs’. While a formal model of sentence structure or sound sys- noun–verb polarity is a dominant characteristic in lan- tems. Rather, typology is an approach that one can guages, there are a few known languages, such as take in investigating the composition of human lan- Tongan, which lack it; hence, the property does not guages, an approach that is driven by a method of hold as an absolute universal. crosslinguistic comparison and rests on the assump- The universals noted so far assert the presence of a tion that structural similarities between languages single structural property, and as such they logically 1133 TYPOLOGY define two types of languages: those that have the into a sample, any shared areal features will be over- property and those that lack it. The latter is only pos- represented in the data. sible in the case of probabilistic universals. However, many typological universals are stated in the form of Word-Order Typology ‘implications’.That is, the statement of the universal carries a condition. An example of an implicational Although typological studies have probed into many universal is: ‘Languages with dominant Verb Subject facets of language structure, the most prominent focus Object order are always prepositional’ (Greenberg of research in the last 50 years has been on word order. 1966). Here the property of having prepositions (as Joseph Greenberg’s seminal publication in 1966, opposed to postpositions) is said to depend on another ‘Some universals of grammar with particular reference property, namely having a certain word order in the to the order of meaningful elements’, identified 45 sentence. This implication is absolute, but many impli- universals, 28 of which dealt with the relative order of cational universals are probabilistic as well. sentence elements. Greenberg’s universals were of the Both simple and implicational universals reveal implicational sort; e.g. he proposed, ‘If the nominal something significant about the form language takes, object always precedes the verb, then verb forms sub- but in different ways. Simple universals bring language ordinate to the main verb also precede it’ (1966:110). structure into relief against ways we might imagine it In Greenberg’s universals, there is repeated refer- could be different. We can create a mental construct of ence to the ordering of subject, object, and verb in the a language that lacks consonants, for instance, but such precondition of the universal. This striking fact led a language simply does not exist. Implicational univer- other linguists, in particular W.P. Lehmann, to propose sals, on the other hand, underscore connections that the relative ordering of verbs and objects was in between language properties that are logically distinct. some sense basic to language structure, and the rela- There is no a priori reason that the order of subjects, tive ordering of other constituents was largely pre- objects, and verbs should be tied into a language using dictable once the order of verb and object was known. prepositions, yet there the correlation exists. For instance, according to Lehmann, in languages with the most common order Verb–Object, one would more than likely find prepositions before the nouns they Sampling governed, nouns before modifying adjectives, nega- For many reasons, not all human languages can be tives before verbs, and comparative adjectives before simultaneously compared. Many languages have died the standards of comparison. In languages where the leaving behind no trace, and many languages have not most common order is Object–Verb, just the opposite yet come into being. Detailed descriptions are only orderings are predicted to occur. available for a fraction of known languages. Therefore, There have been four further developments of note by necessity, typologists must create a sample of lan- in word-order studies. First, efforts have been made to guages when they carry out their research. recast the correlations like those explored by Lehmann Since most typological universals are probabilistic, in terms of more abstract grammatical concepts such the structure of the sample is of great import. If the as ‘head’ and ‘modifier’ instead of treating verb vis-à- languages in the sample are not selected to protect vis object as the dominant organizing principle. against certain biases, then the information that is Second, there has been much discussion regarding gleaned will be unreliable. Two main types of bias why such correlations between sentence elements must be controlled for when constructing a sample: a exist and regarding the ways individual languages genetic bias and an areal bias. Languages that are in depart from the word-order patterns that normally hold the same language branch or language family may be sway. Third, in 1983 the linguist John Hawkins similar because they have retained features from a attempted to provide a fuller set of absolute universals common ancestor. If a particular family is represented of word order that would allow linguists to get a better in the sample disproportionately, then a pattern might idea of possible, typical, and impossible language emerge that reflects the properties of this language types. In order to formulate his universals in absolute family rather than languages in general, and thus the terms, Hawkins used multiple preconditions. For sample would be genetically biased. An areal bias can example, one of Hawkins’s universals claims that if a arise in a language sample when a disproportionate language has Subject–Object–Verb order, AND the lan- number of languages from one geographical region is guage puts adjectives before nouns, then the language used. Since languages that are in close proximity tend will also put genitives (possessors) before their gov- to influence each other’s structure, certain structural erning nouns. Finally, by improving on the sampling features come to be shared by all languages in a given methods employed in earlier word-order studies, lin- area. By placing too many languages from one area guists have provided a much better basis for claims 1134  TYPOLOGY about correlations. Notable in this area is the work of based on the general preference they display for head Matthew Dryer, who has demonstrated, among other marking, dependent marking, double marking, or no things, that earlier claims about the correlation marking. Once so categorized, links between this between verb/object ordering and noun/adjective are parameter and other structural traits can be explored. suspect, as is the correlation between verb/object and negative/verb. Explanations for Universals Although typology is not a theory of language per se, Morphological Typology the discipline has traditionally been aligned with func- Typologists of the nineteenth century wondered tionalist and historical approaches to language. The whether languages could be categorized on the basis of link is most clearly revealed in the kinds of explana- their word structure (morphology). This early work tions that typologists offer for the universals they have was revised by the great American linguist Edward discovered. Rather than provide explanations based on Sapir, who proposed that languages could be classified the system of language itself, typological explanations on the basis of two independent parameters: how many are typically external to the language system. That is, word elements are permitted within individual words, the explanations rest on considerations about how lan- and how much the pronunciation changes when word guage is used, how it is perceived, how it is processed, parts are put together. With respect to the first parame- and how it is shaped by human cognition. Moreover, ter, languages are said to be ‘analytic’ (or ‘isolating’), the explanations tend to be cast in a historical per- i.e. words are not divisible into parts; ‘synthetic’, i.e. spective. It is assumed that a language comes to have manifesting a limited number of word parts per word; the shape it does because of the interaction of regular or ‘polysynthetic’, i.e. employing many parts per word. processes of historical change with functional pres- With respect to the second parameter, nonisolating lan- sures that nudge speakers to maximize the efficiency guages are either ‘fusional’, where the word parts are of their languages for communication. relatively difficult to isolate from each other and show To better exemplify the nature of typological expla- considerable variation in their sound structure depend- nations, consider the phenomenon of ‘pro-drop’. In ing on the particular combinations; or ‘agglutinative’, many languages where the verb agrees with the subject, meaning that they have easily identified word parts. a subject pronoun need not be included. In Spanish, for Because these labels appear to be impressionistic instance, the sentence Yo bailo ‘I am dancing’ is more rather than categorical, and because no language is of commonly rendered Bailo, without the subject pronoun a pure type (for example, agglutinative languages Yo ‘I’. Because information about the subject is includ- invariably have some fusional aspect), contemporary ed in the verb (the ending -o indicates a first person sin- typologists have paid relatively little attention to gular subject), the dropping of Yo causes no loss of developing Sapir’s morphological typology. Even so, information. In contrast, English verbs show very little the terminology is widely employed to provide broad agreement with subjects, and the dropping of a subject characterizations of languages in terms of their overall pronoun is not permitted, presumably because informa- word structure. tion about the subject cannot be recovered from the In recent years, far more attention has been paid to a verb. This link between subject agreement and the abil- different parameter related to word structure, ‘head ity to drop subject pronouns is a strong one. Typologists marking’ vs. ‘dependent marking’, a distinction see the link as part of a larger functional pressure on explored most thoroughly in the work of Johanna language referred to as ‘economy’. Languages tend to Nichols. In constructions involving a head element and be economical, eliminating redundant information an element that depends on it, languages vary with where possible. Over time, in a language where infor- respect to how their relationship is indicated. Consider mation is being expressed redundantly in the grammar, possessive constructions, where the possessor is the the grammar changes to eliminate the redundancy. dependent and the possessed element is the head. The functional pressures operating on language are English, for example, is said to use dependent marking often expressed in terms of ‘markedness’. The core to express the dependency, because the possessor idea behind markedness is that there are asymmetrical appears either with the possessive -s (John’s house) or relationships between grammatical elements that are in the form of a prepositional phrase (the house of otherwise equal. For example, the sounds [p], [t], and John). Other languages mark the head in this construc- [k] (as well as several other sounds) form a natural tion (e.g. Hungarian az ember ház-a ‘the man’s house’, class in that they are all voiceless stops. As members literally ‘the house man-his’). Still other languages of this set, they all have equal status. However, when mark both the head and the dependent, or do not mark one examines the set of voiceless stops that occur in either of them. Languages can then be placed into types the sound systems of languages, it becomes apparent 1135 TYPOLOGY that there is something special about [t]: If a language surprising asymmetry that arises between grammatical only has only one voiceless stop, it is very likely to be elements in language. [t]. And when one considers [p] and [k], one finds something special about [k]: if a language has [p], then References it also has [k]. Therefore, although these sounds are Comrie, Bernard. 1989. Language universals and linguistic equal in terms of their membership in the natural class typology, 2nd edition. Chicago: University of Chicago. of voiceless stops, they are unequal in the way that Croft, William. 1990. Typology and universals. Cambridge: they are distributed in languages. [t] is then said to be Cambridge University Press. ‘unmarked’ relative to [k] or [p]. Furthermore, [k] is Dryer, Matthew S. 1992. The Greenbergian Word Order Correlations, Language. 68. 81–138. unmarked relative to [p]. When a series of these Greenberg, Joseph, H. 1966. Some universals of language with markedness pairs occurs within a set of otherwise particular reference to the order of meaningful elements. equal elements, the serial relationships can be Universals of grammar, ed. by Joseph H. Greenberg. expressed in terms of a ‘markedness hierarchy’. For Cambridge, MA: MIT Press. the case at hand, the hierarchy is t<k<p, with the least Hawkins, John A. 1984. Word order universals. New York: Academic Press. marked member furthest to the left on the hierarchy. Mallinson, Graham, and Barry J. Blake. 1981. Language typol- Such hierarchies are commonplace in typology. ogy. Amsterdam: North-Holland. Some of the better known hierarchies include the Nichols, Johanna. 1986. Head-marking and dependent-mark- grammatical relations hierarchy (subject < object < ing grammar. Language. 62. 56–119. indirect object < other), the number hierarchy (singu- Sapir, Edward. 1921. Language. New York: Harcourt, Brace and World. lar<plural<dual<trial/paucal, i.e. ‘three/few’), the ani- Shopen, Timothy (ed.) 1985. Language typology and syntac- macy hierarchy (first and second person pronouns < tic description, 3 vols. Cambridge: Cambridge University third person pronouns < proper names < human nouns Press. < animate nouns < inanimate nouns), and the color Whaley, Lindsay J. 1997. Introduction to typology: the unity term hierarchy (white/black < red < green/yellow and diversity of language. Thousand Oaks, CA: Sage. <blue < brown). While all these hierarchies obviously LINDSAY WHALEY have quite different implications for what the structure See also Greenberg, Joseph Harold; Lehmann, of language is like, they all highlight the sometimes Winfred Philipp; Sapir, Edward 1136  U Udmurt Udmurt, also known as Votyak, is a language with an to be the precursor of any one of the modern Permian old written literature, which has long been neglected languages; however, it is considered to be closer to by scholars, politicians, and even the inhabitants of the Komi-Zuryanskiy (also known as Komi, the language Udmurt Republic itself. The alternative name was of the Komi Republic) than to Udmurt. According to promulgated via Russian and is now considered old- scholars, the relationship between Udmurt and Komi is fashioned; indeed, many ethnic Udmurts also consider similar to that which holds between Russian and it disparaging. The word Udmurt is formed by the Polish. The phonetic system and grammatical struc- compounding of murt ‘human being’ and *odo (the tures of these languages are almost identical, and they latter is of obscure origin). have much common vocabulary. Through the centuries, The earliest extant literary texts in Udmurt date the Udmurt people have come into contact with many from the sixteenth century and are typically word lists. different peoples and their languages and this has The first grammar of Udmurt was published at the end affected their language. During the first half of the sec- of the eighteenth century, while in the mid-nineteenth ond millennium of the Christian era, the Tatars were century the Gospel texts of the New Testament were the most significant contacts; however, during recent translated. Following the Great October Revolution years, the influence of ethnic Russians has proved more and the inception of the Communist state, many other decisive. As a result, loanwords from Russian, or works of literature have been translated into Udmurt, through Russian, are common: for example, tsement including the leading texts of Marxism–Leninism. ‘cement’, t’ipovoj ‘typical’, or sborn’ik ‘collection’. The Udmurts and their language developed between Udmurt is spoken principally in the Udmurt the Volga-Kama and Vyatka Rivers during the second Republic, which lies some 1,325 kilometers east of half of the first millennium of the Christian era and Moscow in the Pre-Ural region of the Russian there has been a distinct Udmurt nation since c. 1100. Federation. Approximately, two thirds of all Udmurts In terms of genealogical classification, the Udmurt lan- live in the Republic, where they constitute c. 30% of guage belongs to the Permian branch of the Finno- the population. The remainder of the population of the Ugric languages of the Uralic family, and is most Udmurt Republic are predominantly ethnic Russians. closely related to the Komi languages: Komi- The final third of Udmurts live chiefly in the Perm, Zuryanskiy and especially Komi-Permyalskiy. The two Kirov, and Yekaterinburg provinces of the Russian peoples separated between c. AD 900 and 1100 and Federation and also Tatarstan and Bashkorstan. It is today the two languages (Udmurt and Komi) are not difficult to quote a precise figure for the total number mutually comprehensible; however, they occasionally of ethnic Udmurts; however, a population of between present noticeable similarities in vocabulary. For exam- approximately 600,000 and 750,000 is generally ple, night is Udmurt uj and Komi voj. The oldest agreed upon and this concurs with the 1989 Russian Permian language, Old Permian, cannot really be said Federation census statistics, which indicated a poten- 1137 UDMURT tial population of c. 740,000 Udmurts. Where agree- by (clusters of) consonant sounds. The one proviso is ment ends, however, is on the question of the number that consonant clusters may not exceed two consonant of Udmurt speakers. sounds. Word stress typically falls on the final syllable Udmurt had never been an official state language, of the word; however, in certain circumstances the first and by the beginning of the twentieth century it had all syllable carries the main stress. This is usually the but died out. It was not until the 1920s and 1930s that case, among others, with imperatives and adverbs and there was any concerted effort to preserve the language, adjectives, which have been created through a process and the culmination of this may be seen in the 1994 of reduplication: for example, gordgord ‘bright red’. constitution. In this document, Udmurt is presented as Udmurt has six parts of speech: adjective, noun, one of the two official languages of the Udmurt numeral, particle, pronoun, and verb. These categories Republic, along with Russian. However, according to may then be subdivided further. Nouns in Udmurt are some sources, there has been considerable opposition inflected for case, number (singular and plural), and to the state sponsorship of the language, and now those possession. Udmurt has 15 grammatical cases, which areas of life where Udmurt is used and where it holds are marked via affixes. Besides indicating whether a prestige are limited. There is considerable debate as to noun represents the subject or object of a clause, the the current number of speakers of Udmurt, with figures cases of Udmurt are furthermore used to encode infor- ranging from in excess of 500,000 at one extreme to mation such as ‘from…’, ‘to…’, ‘with…’, etc. under 50,000 at the other end of the spectrum. Both Possessive inflectional affixes may also be added to sets of figures should be viewed with caution until reli- adjectives, nouns, numerals, and pronouns to provide able statistical data can be researched. However, information on number (i.e. singular and plural) and UNESCO classifies Udmurt as an ‘endangered’ lan- person (first, second, or third). Interestingly, there is no guage and it would be safe to say that Udmurt is not a case that corresponds to the Indo-European vocative language enjoying a renaissance at the present time. (used to mark the addressee of an utterance). Udmurt Udmurt is written using a modified version of the verbs fall into two conjugation patterns, depending on Cyrillic alphabet consisting of 38 characters. However, the last vowel of the stem. Verbs in Udmurt are inflect- prior to 1930 both the Latin and the Cyrillic alphabets ed for mood, negation, nonfinite categories (gerund, had been used. The language can be divided into four infinitive, and participle), person, and tense. main dialects and these are typically referred to as the In affirmative intransitive sentences, the typical Northern, Southern, Central, and Peripheral dialects. word order is: subject–(adverb)–verb, or in short The first three of these groups refer to the geographical S–(A)–V. Since the copula (i.e. ‘to be’) is not required location, within the Udmurt Republic, of the speakers, to be used in the present tense, this structure may be while the final group is the term used to refer to those reduced to S–A. In affirmative transitive sentences, the speakers who live outside the Republic, notably in typical unmarked word order is: subject–(adverb)– Tatarstan and Bashkorstan. These dialects are mutually object–verb, or S–(A)–O–V. comprehensible, although they do differ to some Negation in Udmurt is typically expressed via a degree from each other in terms of vocabulary and, to negative auxiliary or negative particle, which is placed a lesser degree, pronunciation. before the verb; thus, S–(A)–(O)–Neg–V. Udmurt has seven vowels and 26 consonants. The word order in imperative sentences is usually Syllables can be formed by combining consonant (C) verb initial, except in those cases where a personal and vowel (V) sounds in a way that can be presented pronoun is present. In the latter case, the personal pro- schematically as follows: (C(C))V(V)(C(C))—segments noun is initial: (PersPro)–V. When the imperative is in parentheses are optional. negated, the negative is placed in clause initial posi- tion: i.e. Neg–V. V: api (noun) ‘elder sister’ Interrogative sentences in Udmurt appear to have CV: ju (noun) ‘grain’ no set word order pattern, since interrogation may be CVC: tus (noun) ‘appearance’ marked in a number of ways: an interrogative particle CCV: visjas’kyny (verb) ‘to detach oneself’ may be affixed to the word which forms the focus of VC: an (noun) ‘chin’ the question; or this word may be stressed to mark the VCC: akt (noun) ‘document’ (loanword sentence as a question. However, some tendencies may from Russian) be discerned. Interrogation sentences with a question CVV: tue (adverb) ‘this year’ word normally follow the pattern Q–S–V (where Q is CVCC: šal’kketyny (verb) ‘to flap’ the question word). CVVC: kuat’ (numeral) ‘six’ Adverbial phrases in Udmurt may appear at the Thus, it can be seen that syllables require a vowel beginning, middle, or end of a sentence. Adverbial sound, which may be either long or short. The core positioning may be to an extent determined by the vowel sound may be preceded, succeeded, or framed adverbial category; thus, adverbs of time and place 1138  UNITED STATES often occur in sentence-initial position, while certain Lallukka, Seppo. 1990. The East Finnic minorities in the Soviet other adverbs may appear immediately before the Union. Suomalainen tiedeakatemia: Helsinki. Suihkonen, Pirkko. 1990. Korpustutkimus kielitypologiassa object or the verb. sovellettuna udmurtiin (Computer corpus analysis in language typology applied to Udmurt), Suomalais–Ugrilaisen Seuran Toimituksia 207. Helsinki: Suomalais–Ugrilainen Seura. References ———. 1995. Udmurt–English–Finnish dictionary with a basic Alatyrev, V.I. (ed.) 1970. Grammatika sovremennogo udmurt- grammar of Udmurt (Lexica Societas Fenno-Ugricae skogo yazyka, Sintaksis prostogo predlozheniya. Izhevsk: XXIV). Helsinki: Suomalais - Ugrilainen Seura. Izdatelstvo Udmurtiya. UNPO. 2002. Udmurt (www.unpo.org/member/udmurt/ udmurt. Campbell, George L. 1991. Compendium of the world’s lan- html). guages: Vol. 2, Maasai to Zuni. London: Routledge. Vaba, Lembit, and Juri Viikberg. Udmurts or Votyaks, Comrie, Bernard. 1981. The languages of the Soviet Union. Endangered Uralic peoples: short reference guide, ed. by Cambridge: Cambridge University Press. Andres Heipanuu, Information Centre of Finno-Ugric peo- Government of Udmurtia. 1995. Konstitutsiya Udmurtskoj ples: Tallinn (http://www.suri.ee/eup/udmurts.html). Respubliki. Izhevsk: Udmurtia Press. Winkler, Eberhard. 2001. Udmurt (Languages of the world/ Katzner, Kenneth. 1995. The languages of the world, new edi- materials 212). Munich: Lincom. tion. London: Routledge. KARL BERNHARDT United States The United States is a country of some 250 million peo- Language families with the widest geographical ple spanning a large portion of the North American con- distribution in the United States include Algonquian, tinent. Language has been a controversial political Aztec-Tanoan, Eskimo-Aleut, Macro-Siouan, Na- subject in the United States since the time of Noah Dené, and Penutian. Hawaiian, an Austronesian lan- Webster, and the controversy continues in debates about guage, is spoken by fewer than 1,000 people today. language in the schools and the unofficial status of The serious study of aboriginal languages of the English today. Although today it is an essentially mono- Americas, following the lead of Franz Boas, in combi- lingual English-speaking country, the United States has nation with developments stemming from European a rich and varied linguistic heritage which includes philology, formed the basis for modern linguistics. The indigenous languages, immigrant languages, American following statement is from the resolution on Language Sign Language, and a wide variety of English dialects. Rights passed by the Linguistic Society of America. Many past and present members of the Society have Indigenous Languages devoted their professional lives to documenting and ana- lyzing the native languages of the United States. At the time when settlers of European descent arrived Unfortunately, most of the indigenous languages of the in the territory that is now the United States, there United States are severely threatened. All too often their were at least 300 languages spoken there. Most of eradication was deliberate government policy…. The these languages are either extinct or endangered today, decline of America’s indigenous languages has been with fewer than 50 having more than 1,000 speakers. closely linked to the loss of much of the culture of their Those with more than 10,000 speakers (data from speakers. Because of this history, the Society believes that the government and people of the United States 1977–1990) are listed below: have a special obligation to enable indigenous peoples Indigenous Languages with More Than 10,000 Speakers to retain their languages and cultures. The Society strong- ly supports the federal recognition of this obligation, as Language Number of Family expressed in the Native American Languages Act. Speakers This Act of Congress states that ‘It is the policy of Papago-Pima 12,000 Uto-Aztecan Western Apache 13,000 Na-Dené the United States to… preserve, protect, and promote Central Yupik 15,000 Eskimo-Aleut the rights and freedom of Native Americans to use, Dakota 15,000 Siouan practice, and develop Native American languages’. This Cherokee 22,500 Iroquoian is a major reversal in policy from the days when Indian Western Cree 35,000 Algonquian children were removed from their tribes and sent Western Ojibwa 35,000 Algonquian to English-only boarding schools where they were 1139 UNITED STATES punished for speaking their native languages. These ear- e.g. Yiddish in New York City, Scandinavian lan- lier educational policies led to the loss of languages of guages in the Upper Midwest, and, of course, Spanish tribes that had survived the warfare and epidemics that in the Southwest, although the latter predates the had destroyed so many other tribal languages. There are founding of the United States and cannot be consid- few monolingual speakers of indigenous languages in ered an immigrant language. the United States today, a fact that suggests their Main Sources of Borrowings in American English impending replacement by English. Although some Native American languages are now being taught in trib- Languages/ Predominant Examples al schools, it remains to be seen whether these languages Language Region/Era can survive as a means of everyday communication. Groups Algonquian Eastern New Moose, raccoon, England, 17th squash, pecan History of American English century At the end of the eighteenth century, in the midst of West African South, 18th Yam, okra, banjo, century jazz, tote anti-British sentiment born of revolutionary fervor, Dutch Hudson Valley, Cookie, coleslaw, Americans declared verbal independence and the lan- 18th century boss, sleigh guage that was formerly just ‘English’ became British German Pennsylvania, Cluck, dunk, English and American English. This was the first of 19th century hamburger, pretzel many such splits into national varieties that were yet to French Great Lakes, Prairie, portage, Louisiana, 18th, armoire, come as a result of the far-reaching colonialism of the 19th centuries jambalaya British Empire. Despite the fact that the English Spanish Southwest, Coyote, marijuana, believed the speech of the American settlers to be cor- 19th century ranch, cafeteria rupt and full of ‘barbarisms’, Webster defended it in Yiddish New York City, Schmuck, kibitzer, his Dissertations on the English language by declaring 20th century phooey, schmooze that ‘the people of America… speak the most pure English now known in the world’. His spelling When the United States took over, first, the reforms and insistence on middle class rather than Louisiana Territory from the French and, later, Texas, aristocratic norms were intended as a contribution to California, and the Southwest from the Spanish, the the formation of an American identity. nation thereby incorporated the many speakers of lan- In colonial times, however, a number of European guages other than English who were already living in languages were spoken in North America. English came these areas. The US government allowed French and to predominate, but not without feeling the influence of Spanish use to continue for a number of years in the these, and some of the indigenous languages, upon its local governments of states like Louisiana and vocabulary. Even before the arrival of the Mayflower, California, although in at least two instances (New words were borrowed into English from the indigenous Mexico and Hawaii), the preponderance of non- people of the Americas, especially the Caribbean, via English speakers in annexed territories became an Spanish; other Native American words were introduced obstacle to acceptance into statehood. During the into English by French-speaking explorers. The many debates over the years about the status of lan- Algonquian languages were the source of most borrow- guages other than English (and their speakers), critics ings from Native American languages, as this language have often overlooked the fact that many non-English family included the speech of the tribes along the speakers did not come to the United States voluntarily. Eastern Seaboard that were the first ones encountered In addition to the cases of French, Spanish, and by the English settlers. The settlers coexisted fairly Hawaiian and other indigenous languages noted peaceably with the Indians for the first few years, and above, many Chinese speakers were brought to the they had need of words to describe native flora and West Coast as forced laborers. The largest group of fauna that were not found in the Old World. As hostili- non-English speakers in US history also came invol- ties worsened and settlement moved westward, most untarily, brought as slaves from West Africa. borrowings were limited to place names. The debate among linguists over the history of Early on, words entered English from contact with African American Vernacular English (AAVE) has still the Dutch in New York and the Germans in not been resolved, although it has been one of the most Pennsylvania. More lexical and even grammatical productive areas of linguistic study in the United influences were to come from the many immigrant States for over 30 years, spurring a vast amount languages that arrived in later years. The impact of of research on Creole languages and on nonstandard these immigrant languages helped to create some of varieties of English in the United States. The sides the differences between regional dialects of English, have been characterized as ‘Anglicist’ vs. ‘Creolist’ 1140  UNITED STATES hypotheses, and the most simplistic description of occurred, although some scholars believe that even in these points of view is that AAVE originated either colonial times, regional dialect differences existed. It from the nonstandard British-based varieties of is possible to distinguish four streams of settlement English learned by the slaves or from a widespread from different parts of the British Isles that may have Creole language that once existed in the slave-owning played a role in the formation of the major dialects of areas of the South. Most linguists today agree that a American English. The eastern part of England was multiplicity of factors shaped the language of African the origin of many New England settlers, while Americans, including African substrate languages like Pennsylvania and Delaware were settled by people Wolof and Yoruba, English-based Caribbean Creoles, from northern England, and the South by colonists the British dialects of white indentured servants, and predominantly from the southwest of England. Inland developing American English dialects, both standard areas, including Appalachia, were settled slightly later and nonstandard, as well as language-learning features by the people known as the Scots-Irish, or Ulster that may be attributable to Universal Grammar. It is Scots, who were of Scottish descent but who came to likely that the first slaves to be brought to the United the United States after a sojourn in northern Ireland. States, who worked as domestic servants or on small Four major dialect areas of the eastern United States farms where there was ample exposure to native were defined by Hans Kurath based on data collected speakers of English, were able to learn English just as for the Linguistic Atlas of the United States and other immigrants did. As more and more slaves were Canada, an ambitious undertaking begun in the 1920s brought in, both directly from Africa and via the that was designed to map the dialect areas of English- Caribbean, opportunities to learn from native speakers speaking North America. Two influential publications diminished as the ratio of blacks to whites grew. It is based on the atlas data (a ‘word geography’ and a vol- likely that a true Creole only developed along the coast ume describing the pronunciation of influential citizens of South Carolina and Georgia, where large, labor- in various regions) provided the basic outline of dialect intensive rice plantations produced demographics sim- areas in the United States for many years. Kurath pro- ilar to those on the sugar plantations of the Caribbean. posed two primary boundaries and a secondary bound- (This now mostly decreolized variety is known as ary as shown in the map (Figure 1), forming the North, Gullah or Geechee, and it is the source of the Afro- North Midland, South Midland, and South dialect Seminole Creole in Texas.) Elsewhere in the South, a areas. Today, many linguists believe the primary and variety developed among African Americans that was secondary boundaries to be reversed, i.e. that the major very similar to white varieties of English in the region, division is North vs. South, with two main subregions due to mutual influence, and which had a Creole-like within each, corresponding to essentially the same tense and aspect system (see below). This variety was areas Kurath identified but now known as Upper and taken northward to cities such as Detroit, Chicago, Lower North and Upper and Lower South (Figure 2). New York, and Philadelphia when many blacks left the The dialect areas remained fairly distinct as settlement South during the first decades of the twentieth centu- proceeded westward as far as the Mississippi River, but ry, fleeing poverty and racially motivated violence. west of there, their features were mixed together. Although there are some identifiable dialect regions in Features of AAVE the West, as the most recently settled area of the United Copula absence States, dialect differentiation is not as sharp as in the She pretty. They in the house. East. Likewise, dialect differences in the United States Invariant/nonfinite be (habitual aspect) in general are more subtle than those in the long-popu- He be in school every day. lated countryside of Europe and England. Stressed been (remote tense) The vocabulary of American English is expanding at He BEEN ate it. She BEEN gone. Perfect done (completive aspect) a tremendous rate, coinciding with the twentieth centu- They done left the house. ry developments of universal public education, Lack of verbal morphology improvements in transportation, mass media, and a She always go to that store. The building burn last week. consumer culture. Words for archaic features of the culture, such as premechanized agriculture, are disap- While AAVE thus spread from south to north, other pearing, along with the many lexical items that simply dialects spread with the population from east to west. drop out of fashion over the years. On the other hand, Quite a lot of dialect mixing occurred in the early set- a great deal of vocabulary has been added through tech- tlement period, as colonists in any one town might nological change, especially in the field of computers. have come from various parts of England, Ireland, Cultural changes have also given us brand names that Wales, Germany, or elsewhere. There was probably a are widely known and used. New words are added from period of ‘koineization’, when a leveling process many sources. Slang is spread through youth culture, 1141 UNITED STATES 4 5 3 2 10 8 THE SPEECH AREAS OF THE EASTERN 7 STATES 11 THE NORTH 9 1. Northeastern New England 14 2. Southeastern New England 3. Southwestern New England 4. Upstate New York and W. Vermont 5. The Hudson Valley 12 6. Metropolitan New York 15 THE MIDLAND 7. The Delaware Valley (Philadelphia Area) 8. The Susquehanna Valley 9. The Upper Potomac and Shenandoah Valleys 10. The Upper Ohio Valley (Pittsburgh Area) 13 16 11. Northern West Virginia 12. Southern West Virginia 13. Western North and South Carolina THE SOUTH 17 14. Delamarvia (Eastern Shore of Maryland and Virginia, and Southern Delaware) 15. The Virginia Piedmont 16. Northeastern North Carolina (Albemarle Sound and Neuse Valley) 18 17. The Cape Fear and Peedee Valleys 18. South Carolina Figure 1. Major dialect areas of the US: Kurath (1949). and through mass-distributed music and other electron- important recent lexicographical work on American ic media; some slang words are ephemeral, but others, English is the Historical dictionary of American slang. such as cool and dude, have persisted for many gener- These two projects expand our resources on American ations. And borrowings from immigrant languages English far beyond the reach of the best current continue, although these terms may be regionally lim- unabridged dictionaries with an American perspective, ited. Despite increasing geographical mobility and Webster’s third and Random House. exposure to other dialects, regionalisms in the vocabu- American English grammar has diverged from that lary do remain, as documented in one of the most sig- of British English in a number of ways, for example, nificant American dictionary projects, the multivolume in retaining many uses of the subjunctive and in the Dictionary of American regional english. Another use of auxiliary verbs. 1142  UNITED STATES NORTHWEST NEW ENGLAND Western New England EASTERN UPPER MIDWEST NEW ENGLAND UPPER NORTH Northern California Central West Utah West Southeastern Pennsy TH lvania R NOR Colorado West LOWE Hoonet Apex Virgnia Piedmont Missouri Apex Upper TH Southern California SOU Atlantic South R Eastern North Carolin SOUTHWEST UPPE a Atlantic South TH LOWER SOU Lower Atlantic South West Texas Northern Louisia na Delta South Alabama Southeastern Louisia na Southeastern Florida Figure 2. Major dialect areas of the US: Carver (1987). Some Grammatical Differences between US English from a ‘wave’ model of understanding change as mov- and English English ing outward in an even pattern from a geographic ori- US English English English gin (like waves in a pond) to a model where change moves from one large urban area to another before dif- The salad has olives in it. The salad has olives in. fusing into suburbs, smaller cities, and finally rural In the future, we will In future, we will change change the plan. the plan. areas. One important survey conducted by William John gave it to me. John gave me it. Labov and colleagues, the Phonological Atlas of the Do you have any apples? Have you any apples? United States, has documented several sound changes I won’t be able to go. I shan’t be able to go. in the form of vowel shifts and mergers. In most of the If I were you, If I was you, I wouldn’t do it. United States, the vowels in lot and law are merging, I wouldn’t do it. and other shifts are taking place in some Northern cities (e.g. the raising of the vowel in bag toward the Standard English grammar is that which is used by vowel of beg) and in the South (e.g. the merger of the the most educated, upper social classes in all regions vowels of pen and pin before nasal consonants) and which carries the prestige of being regarded as (Figures 3 and 4). ‘correct’ or ‘good’ English by all. English syntax and There is no single standard of pronunciation in the morphology vary in nonstandard dialects across the United States, but rather a number of ‘regional stan- United States, with much of the variation involving dards’, characterized by the accents of the upper-mid- verb forms in the past and past participle. Nonstandard dle class. This includes people like lawyers, doctors, speech is usually stigmatized because it is typical of bankers, and politicians, economically successful and the working class or minority ethnic groups. These highly visible members of the community who often nonstandard dialects of English in all parts of the speak with a regionally identifiable accent that is the United States exhibit many of the same features, local prestige norm. Although people who move from including lack of subject–verb agreement, as in he one region to another may drop features of their native don’t, we wasn’t, and multiple negation (they don’t dialects that are stigmatized in other regions, and know nothing), often including ain’t. broadcasters are trained to do the same thing, this Pronunciation changes continue to spread across accentless type of speech is not required for success the United States, although dialectologists have moved even at the national level: witness the speech of recent 1143 UNITED STATES /i/ (beet) /u/ (boot) tics is known as language variation study, and it Ω includes what has traditionally been called dialectology /ι/ (bit) / / (put) as well as much of the subject matter of sociolinguis- /e/ (bail ) /o / (boat) tics. Sociolinguistics also includes the study of lan- ν guage attitudes, sometimes called perceptual /ε/ (bet) / / (but) dialectology. In many ways, the attitudes toward low- /æ/ (bat) c / / (brought) status varieties of English, especially ethnically-based varieties like AAVE, parallel attitudes toward lan- /a / (father ) guages other than English in the United States. These attitudes have led to political controversy over the use Figure 3. The northern cities vowel shift (adapted from Labov 1991). of languages and varieties other than Standard English, especially in the schools. /i/ (beet) /u/ (boot) /ι/ (bit) Ω / / (put) Immigrant Languages and Language in the Schools /e/ (bail ) /o / (boat) ν Although the United States has no official language, /ε/ (bet) / / (but) English serves as the de facto national language, with /æ/ (bat) c / / (brought) 87% of the adult population speaking only English, and 80% of those who speak another language at home also /a / (father ) speaking English ‘well’ or ‘very well’. Bilingualism in the United States has mostly been ‘subtractive’, leading Figure 4. The southern vowel shift (adapted from Labov 1991). to loss of the language other than English within three generations for immigrants. The rate at which immi- presidents, the strong New England accent of the elder grants are learning English has not changed, but public Bush and the Texas sound of the younger one, the perception today is that Spanish-speaking immigrants Southern accents of Clinton and Carter, and the are not learning English. This is due to continued high Boston accent of the Kennedys. numbers of new immigrants from Mexico and else- The speech patterns of traditionally economically where in Latin America. Early in the twentieth century, depressed regions of the United States (e.g. the South), there were complaints that Italian, Jewish, Greek, and or areas with larger concentrations of working class Slavic immigrants were not learning English; the eth- speakers (e.g. New Jersey, the Appalachians), like the nic languages spoken by these groups are disappearing speech patterns of poor people in general, have been quickly less than 100 years later. Demographic projec- stigmatized on a national level. The most prestigious tions predicting that Caucasians will comprise less than US dialect is one that dialectologists have labeled 50% of the population by the mid-twenty-first century Inland North, which encompasses upstate New York; have fueled fears that English speakers will become a the northern parts of Ohio, Indiana, and Illinois; minority in the United States. Opposition to the use of Michigan; and Wisconsin. Studies of language atti- languages other than English in the public sphere tudes have shown the speech of the South to be con- by anyone besides tourists has often been linked his- sistently labeled as the ‘least correct’, although not torically to xenophobia and nativism, as when racist, necessarily the ‘least pleasant’. Sociolinguists call this anti-Catholic, and anti-immigrant political groups in type of positive assessment of speakers of nonstandard the late nineteenth century fought to enact language dialects, as being more friendly and sincere, ‘covert restrictions. prestige’. The idea that nonstandard speech can be val- Germans, who were part of the earliest European ued as sounding more masculine has led researchers to settlements in the United States, were probably more invoke covert prestige to explain the tendency of men successful at maintaining their language than any in the United States to use more nonstandard features other non-English-speaking group. (However, the than women of the same social class. Other factors claim that a proposed law making German the official could explain this finding as well, especially the high- language of the United States failed by just one vote is er educational levels and Standard English skills a myth: rather, the Congress declined, in a close vote, expected of women in many low-paying jobs. to take up a proposal to print laws in German as well Correlational studies of how nonlinguistic variables as in English.) Although Benjamin Franklin and others such as social class, sex, age, region, and race influence complained about the use of German in Pennsylvania, speech patterns have taught us much about the com- German communities there and later in the Midwest plexity of American English. This subfield of linguis- continued to use German as the language of instruc- 1144  UNITED STATES tion in both public and parochial schools up until ent interpretations of what this would mean, but most World War I, when nationalistic sentiment put an end of them involve curtailing government services or doc- to the teaching of German even as a foreign language. uments in languages other than English. In response to In some places, German textbooks were burned by this and similar initiatives that have been passed in a mobs. Many Americans came to view bilingualism as growing number of states, the Linguistic Society of unpatriotic and deviant, and some states even forbade America passed the following resolution, mentioned the teaching of any foreign language for a time. above and excerpted below: Beliefs that bilingualism is somehow harmful persist, All residents of the United States should be guaran- so that the Linguistic Society of America found it nec- teed the following linguistic rights: essary to include the following in a press release on ● To be allowed to express themselves, publicly or language rights: ‘There is no convincing evidence that privately, in the language of their choice. bilingualism by itself impedes cognitive or education- ● To maintain their native language and, should al development. On the contrary, there is evidence that they so desire, to pass it on to their children. it may actually enhance certain types of intelligence.’ ● When their facility in English is inadequate, to The only places where a form of German remains be provided a qualified interpreter in any pro- viable today are the Amish communities, principally ceeding in which the government endeavors to in southeastern Pennsylvania. There are, however, two deprive them of life, liberty or property. other European languages that have been spoken in the Moreover, where there is a substantial linguistic United States for centuries: French varieties in minority in a community, interpretation ought to Louisiana and Maine, and Spanish, which has spread be provided by courts and other state agencies in from the Southwest into many urban areas. any matter that significantly affects the public. Most Spanish speakers in the United States today ● To have their children educated in a manner that are Mexican-Americans, or Chicanos. The linguistic affirmatively acknowledges their native lan- competence of Chicanos may include a plethora of guage abilities as well as ensures their acquisi- languages and varieties, as listed by Anzaldúa: tion of English. Children can learn only when ● Standard English they understand their teachers. As a conse- ● Working-class and slang English quence, some use of children’s native language ● Standard Spanish in the classroom is often desirable if they are to ● Standard Mexican Spanish be educated successfully. ● North Mexican Spanish dialect ● To conduct business in the language of their ● Chicano Spanish (Texas, New Mexico, Arizona, choice. and California have regional variations) ● To use their preferred language for private con- ● Tex-Mex [a code-switched variety] versations in the workplace. ● Pachuco [an argot] ● To have the opportunity to learn to speak, read, and write English. In addition, there are at least as many varieties of Spanish in the United States as there are countries of This statement on linguistic rights refers to the ben- origin for Hispanics, including the prominent efits of at least some instruction in languages other Caribbean dialects of Cuba and the US-governed than English, or bilingual education. Bilingual educa- colony of Puerto Rico. The first European settlements tion is not common in the United States and where it in what is now the United States were the Spanish set- does exist, it is usually ‘transitional’, meant to help tlements in Florida and Georgia. Spaniards also estab- students keep up with their other subjects, which they lished missions in California very early; however, the study in their native language, while they study Spanish variety that has been continued from colonial English intensively in order to join a mainstream class- times is that which was brought into the Southwest, room as soon as possible. Linguists would like to see especially Arizona, New Mexico, and Colorado by set- bilingual education also used for the purposes of lan- tlers from ‘old’ Mexico before those states were part of guage maintenance in order to foster a plurality of lan- the United States. Ironically, due to demographic guages and cultures in the United States, especially changes and the economic power of Anglo retirees, where endangered languages are spoken, e.g. indige- Arizona has been one of the most active states in push- nous languages or Cajun French. ing for ‘English Only’ restrictions. In Louisiana, there is a state-sponsored program, The English Language Amendment is a bill that has the Council for the Development of French in been introduced in Congress in several forms but that Louisiana (CODOFIL), which is trying to raise the has never passed. It would make English the official status of French there and to keep it from dying out. language of the United States. There are many differ- This program sponsors activities ranging from radio 1145 UNITED STATES and television broadcasts in Cajun French to recruiting al opportunity is clearly and directly tied to respect for, Francophone teachers from around the world to offer and support of, the first language of the child or stu- second-language classes, including immersion pro- dent.’ This same rationale for an ‘additive’ approach in grams, in French. Unfortunately, it has been difficult the classroom can be applied to any minority language to bridge the gap between the standard Metropolitan in the United States. African American Vernacular French of these teachers and the vernacular variety English, in particular, has been at the center of heated spoken by the descendants of the Acadians (who came debates over the role of languages other than Standard to the United States after being persecuted and forced English in the classroom. In 1972, linguists were out of Canada). About 10% of French speakers in involved in a court case in Ann Arbor, Michigan, where Louisiana speak Creole French rather than Cajun the judge ruled that AAVE was sufficiently different French. The CODOFIL mission statement sums up the from Standard English to merit special treatment. political issues surrounding the use of minority lan- Teachers were to give AAVE speakers additional help guages in the United States: in acquiring Standard English, partly by learning about While we would be the last to argue that a U.S. citizen the syntactic and phonological rules of AAVE them- or resident does not need to speak English in order to sur- selves so that they could point out the systematic dif- vive and thrive in our society, we should not predicate ferences to their students. A similar situation arose in fluency in English upon the elimination of other lan- Oakland, California, in 1996, when the school board guages. We in Louisiana have seen first hand the effects adopted a policy designed to help speakers of of this linguistic genocide. Although our parents and ‘Ebonics’ (AAVE) learn Standard English by educating grandparents were beaten, punished and humiliated for teachers about dialect differences. The public interpret- speaking French on the schoolgrounds, we have come to ed both of these attempts as a requirement to force realize that full participation in American democracy teachers and non-AAVE-speaking students to use does not necessarily mean forgetting our roots. AAVE in the classroom (not the case), and as a plan For many minority groups in the United States, intended to keep African American students from there is a tension between the need to assimilate to the learning Standard English and thus denying them the majority culture to attain economic and professional social benefits that come with a fluent command of success and the need to maintain their ethnic identity. Standard English (also not the case). This is the issue that educators have tried to address by Similar issues revolve around the question of the proposing that schools allow both bilingualism and role of American Sign Language (ASL) for Deaf stu- bidialectalism in classrooms (as occurs, for example, dents. ASL was developed by educators Thomas in Hawaii, where half of the population speaks Gallaudet and Laurent Clerc, who in the early nine- Hawaiian Creole; students read and write in English, teenth century introduced signs from a visual–spatial but the teachers explain things in Creole). For years, language used in France, combining them with signs schools have taken the subtractive approach to teach- already in use in the United States. ASL is not simply ing Standard English, both to students whose native a signed form of English. In fact, it is grammatically language is not English and to students who speak a unrelated to English; British Sign Language and ASL nonstandard variety of English. This ‘replacive’ view are not mutally intelligible, ASL being more closely assumes that Standard English is inherently superior to related to French Sign Language. Linguists have all other varieties, which are given labels ranging from described ASL in terms of its phonology (hand- ‘substandard’ and ‘bad’ to ‘gibberish’. The stigmatiza- shapes), morphology (location of sign in relation to tion of other varieties often evokes two types of the body, movement of articulators, and facial expres- responses from students: a defensive, resistant stance sions), and syntax (sequence of signs). Children in an toward a teacher who is trying to take away the lan- ASL home environment acquire ASL as their native guage of their family, neighborhood, and church; or language in the same way as other children acquire the development of low self-esteem and the internal- spoken languages, e.g. ‘babbling’ with their fingers at ization of the belief that their native language (and by the appropriate stage of development. These children extension, culture) is inferior. By fostering bidialectal- must become bilingual if they are to learn to read and ism and bilingualism, encouraging students to be flu- write in English. Some members of the deaf commu- ent in both Standard English and their vernacular, and nity feel that it is an unnecesary burden on children if teaching them which is appropriate in which situa- they are also required to learn an oral approach, which tions, schools can lower failure and dropout rates that involves lip reading and the production of speech are tied to language difficulties. sounds. It is difficult for a deaf person to learn to pro- The Native American Language Act states, ‘There is nounce a spoken language well enough to be under- convincing evidence that student achievement and per- stood by people who are not immediate family. The formance, community and school pride, and education- same issues of opportunity, identity, and community 1146  UNIVERSAL GRAMMAR that affect other speakers of minority languages in the Baron, Dennis. 1990. The English-only question. New Haven, United States confront bilingual ASL/English speak- CT: Yale University Press. Boas, Franz. 1911. Handbook of American Indian languages. ers, both negatively in terms of language prejudice and Washington: Smithsonian. positively in terms of having a wide range of commu- Carver, Craig M. 1987. American regional dialects. Ann Arbor: nicative abilities. University of Michigan Press. Cassidy, Frederic G., and Joan Houston Hall (eds.) 1985-96. Languages with more than 100,000 speakers in the Dictionary of American regional English, vols. 1–3. United States, from 1990 Census and Ethnologue. Cambridge, MA: Harvard University Press. Council for the Development of French in Louisiana website, Spanish 17,339,172 Arabic 355,150 http://www.codofil.org French 1,702,176 Hindi (Urdu) 331,484 Crawford, James. 1999. Bilingual education: history, politics, German 1,547,099 Russian 241,798 theory, and practice, 4th edition revised. Los Angeles: Italian 1,308,648 Yiddish 213,064 Bilingual Education Services. Chinese 1,249,213 Thai (Laotian) 206,266 Ethnologue website, Languages of the world, 13th internet edi- Tagalog 843,251 Persian 201,865 tion, http://www.sil.org/ethnologue/ethnologue.html Polish 723,483 French Creole 187,658 Kurath, Hans. 1949. A word geography of the Eastern United Romani 650,000 Armenian 149,694 States. Ann Arbor: University of Michigan Press. Hawaiian 600,000 Navaho 148,530 Kurath, Hans, and Raven I. McDavid. 1961. The pronunciation Creole of English in the Atlantic States. Ann Arbor: University of Korean 626,478 Hungarian 147,902 Michigan Press. Vietnamese 507,069 Hebrew 144,292 Lighter, J.E. 1994-97. Random House historical dictionary of Portuguese 429,860 Dutch 142,684 American slang, vols. 1–2. New York: Random House. Japanese 427,657 Mon-Khmer 127,441 Linguistic Atlas Projects website, http://us.english.uga.edu (Cambodian) Labov, William. 1994 and 2001. Principles of linguistic Greek 388,260 Gujarathi 102,418 change, vols. 1 and 2 (Internal factors, social factors). Oxford: Blackwell. The economic power and cultural influence of the Linguistic Society of America website, http://www.lsadc.org Linn, Michael D. (ed.) 1998. Handbook of dialects and language United States is furthering the rapid spread of English variation, 2nd edition. San Diego: Academic Press. across the globe, with some 700 million fluent speak- Mencken, H.L. 1992. The American language, abridged edition. ers around the world, and English functioning as an New York: Knopf. important language in 75 different countries. Many Mufwene, Salikoko S., John R. Rickford, Guy Bailey, and John countries that formerly looked to a British model for Baugh (eds.) 1998. African American English: structure, history, and use. London: Routledge. the teaching of English as a foreign language are now Preston, Dennis R. (ed.) 1993. American dialect research. adopting American English norms. Thus, it is true that Amsterdam: John Benjamins. the dominant language of the United States has an Pyles, Thomas. 1952. Words and ways of American English. impact far beyond its borders. The importance of New York: Random House. American English is not likely to decline soon, either Trudgill, Peter, and Jean Hannah. 1994. International English, 3rd edition. London: Arnold. within or outside the United States. Valdman, Albert (ed.) 1997. French and Creole in Louisiana. New York: Plenum. Valli, Clayton, and Ceil Lucas. 2001. The linguistics of American References sign language, 3rd edition. Washington: Gallaudet University American Dialect Society website, http://www.americandialect. Press, org Wolfram, Walt, and Natalie Schilling-Estes. 1998. American Anzaldúa, Gloria. 1987. Borderlands/La Frontera. San Francisco: English. Oxford: Blackwell. Aunt Lute. ELLEN JOHNSON Universal Grammar Universal grammar (UG) is a theory both about the generative grammar, and it continued to evolve through- essential nature of human language and about how lan- out his later writings on syntactic theory. A detailed guage is acquired by children. The concept of UG was exposition of UG appears in his 1981 work Lectures on first suggested by Noam Chomsky in his early work on government and binding. This book serves as the basis 1147 UNIVERSAL GRAMMAR for a particular version of UG envisioned as a system of Language may be structured as a system of inter- universal principles interacting with parameters of vari- acting modules or subsystems. These modules are spe- ation. This version of UG is central to the syntactic the- cialized in their function, much like bodily organs, and ory, known as Principles and Parameters Theory, and are usually assumed to include systems devoted to has been the basis for much current research in the vocabulary, grammar, pronunciation, and meaning. fields of syntax and language acquisition. UG is not seen as a list of universally valid rules for The history of UG is intimately related to conflicting the correct formation of language structures. As theories about the nature of language acquisition. In Chomsky points out, lists of individual rules cannot response to the behaviorist view of language acquisition adequately model the nature of language competency as imitation and stimulus–response learning, Chomsky in fluent speakers or how this competency is acquired. argued that language acquisition did not involve learn- On the contrary, learners do not need to acquire any ing, but was instead the result of the activation of a total- organizational principle for the complexities of lan- ly innate capacity for language, namely, UG. guage: the framework is already provided for them by Support for UG comes from observations of how UG, and it is the same for all languages. children’s language acquisition differs from other UG is assumed to contain a set of universal princi- kinds of learning and how it is not merely imitation. ples that are valid for all human languages. These are First, all children acquire language at a time in their thought to include principles about the way language life when they cannot learn any other skill of similar is structured into phrases. For example, a basic tem- complexity. Furthermore, all children achieve approx- plate governing the shape of syntactic structures and a imately the same degree of skill (i.e. native speaker general rule relating different syntactic structures to fluency) regardless of the infinite variations in their each other are both assumed to be principles of UG. environmental stimuli. The language children produce The principles are innate and immediately available to is not merely an imitation of the adult speech they hear the learner, and they are perhaps the most crucial tool around them: they often use innovative structures that in UG in its role as a language acquisition device. they could not possibly have heard from an adult. Because they radically reduce the number of possible Children are not actively taught by adults which lan- analyses that a learner might assign to any one sen- guage structures are grammatical and which ones are tence, the UG principles, in a sense, allow the learner not; they acquire a sense of the difference merely to automatically recognize the correct structure of lan- through exposure, and they generally do not respond guage input during the acquisition process. to corrections when they do receive them. Finally, only Of course, individual human languages show a humans seem to be able to acquire language. Primates great deal of surface variation in terms of the sounds cannot acquire most grammatical features of language, they use, the structure of words, and the order in which even through intensive behavioral training. words are combined. The language learner’s task is to Chomsky concluded from observations such as these acquire the specific characteristics of the language of that language must in some sense be an inborn ability his or her community. UG therefore is said to include universally present in humans. At first, he described the a set of parameters listing which aspects of language innate ability to acquire language by using the metaphor may show variation and defining what the boundaries of a ‘black box’, or device in the brain: the language for such variation may be. For example, one area in acquisition device (LAD). Today, the concept of UG has which languages exhibit systematic variation is their replaced the metaphor of the device and the term LAD. placement of wh- words (what, who, and so on) in the It is UG that is envisioned as a biological ‘language sentence. Some languages, such as English, position organ’, and it is the nature of UG itself that allows wh- words at the beginning of the sentence, whereas humans to acquire language quickly and flawlessly. others, such as Chinese, do not. Languages on the UG contains the basics that all human languages whole seem to pattern either like English or like have in common. Specifically, this includes the modu- Chinese in this respect. The two options are therefore lar organization of the language faculty itself, a set of interpreted as the two possible UG ‘settings’ for a wh- universal principles governing the structures of human placement parameter. A language learner is said to set language, and a set of parameters defining the possible the value for such a parameter through direct experi- limits of language-specific variations. This is the pic- ence with input from the language being learned. Most ture of UG that has arisen within Principles and parameters are thought to be binary in nature, i.e. they Parameters Theory. It provides an account for both the allow only two alternatives, like this one. Parameter fundamental similarities and the vast amount of setting has been compared with a process of setting observable variation among existing human lan- electrical switches or filling in the blanks of a form. guages. It also embodies predictions about what con- Thus, parameter setting is the real work of language stitutes a possible human language. acquisition. UG provides a structured framework and 1148  UTTERANCE-CENTERED LINGUISTICS all the possible choices, and ensures that the choices Cook, Vivian, and Mark Newson. 1997. Chomsky’s universal will be set correctly. grammar: an introduction. Oxford: Blackwell. Herschensohn, Julia. 1998. Universal grammar and the critical age. Behavioral and Brain Sciences 21. Meisel, Jürgen. 1995. Parameters in acquisition. The handbook References of child language, ed. by Paul Fletcher and Brian Atkinson, Martin. 1996. Children’s syntax. Oxford: Blackwell. McWhinney. Oxford: Blackwell. Chomsky, Noam. 1981. Lectures on Government and binding. Pinker, Steven. 1994. The language instinct. New York: Harper. Dordrecht: Foris. SHEILA DOOLEY COLLBERG Utterance-Centered Linguistics When approaching the study of language, a funda- This discourse may be viewed as consisting of three mental issue is to determine the basic unit for analysis. utterances: (1) the comment uttered by A and (2) and (3) This decision will depend on the features of language the response to it uttered by B, with each utterance real- and its use, which are of most interest to a particular izing at least one speech act: viz. (1) reproaching, (2) researcher. In the case of utterance-centered linguis- apologizing, and (3) justifying. In this discourse, also, it tics, utterances rather than, for instance, words or sen- is noteworthy that whereas Sorry. is an incomplete sen- tences are taken as the unit for analysis. tence which therefore might be said to have a defective An utterance is any contextualized sequence of nat- grammatical structure, this is not the only utterance urally occurring language that may be distinguished where the hearer’s understanding cannot be determined from other sequences of language spoken or written by by the grammatical structure of the message alone. For the same person or by other people in the same context. instance, utterance (1)—although grammatically com- Very brief utterances may take the form of far less than plete—is not accurately interpretable as a reproach only a sentence: perhaps just a sigh or a grunt that conveys on the basis of its grammatical structure: grammatical- meaning without words or grammar. Other utterances ly, it is just an assertion. Full understanding depends on may be realized by a single word or a grammatically taking nonlinguistic contextual factors into account as incomplete sentence: for example, Okay. or Got it! Or well: what prior arrangement did the two speakers utterances may take the form of a whole sentence— make? and what grounds might A have for feeling dis- Could you close the window?—or even a number of appointed or angry if B failed to carry through? Clearly, sentences in a row (as in an extended request for assis- A has framed utterance (1) on the assumption that B will tance, with reasons and expressions of anticipatory readily comprehend the reproach through reliance not gratitude). One or more naturally occurring oral or only on linguistic features but also on contextual written utterances constitute a discourse. Utterance- knowledge and inferences. And the same assumption centered linguistics studies discourses: component underlies B’s response. utterances may be analyzed either according to the lin- So utterance-centered linguistics views language as guistic features that mark or create coherence among a socially contextualized rather than a strictly form- them, or in terms of their embodiment of speech acts. based phenomenon; this focus makes utterance-cen- A speech act is an intended meaning conveyed by a tered linguistics particularly effective for the study of speaker or writer. Arguably, the most influential natural language, since actual language users plainly approach to utterance-centered linguistics has focused conceptualize speaking or writing as a process of using on speech acts, particularly in oral language. utterances to perform speech acts with real-life conse- Although utterances have no fixed grammatical quences, not as a process of producing grammatically structure, they do have structural requirements with correct sentences. A key early figure in this tradition regard to speech acts. Each will convey at least one was J.L. Austin, the title of whose influential book, speech act, sometimes a number of them simultane- How to do things with words (1962), reflects his thesis ously. Consider the following short discourse: that language use is a form of social action. Utterances are actions in the sense of being attempts to get others A: It’s twenty minutes past six. to do certain things, to entertain certain thoughts, or to B: Sorry. The boss called an unexpected meeting. experience certain emotions. Thus, Austin’s work drew 1149 UTTERANCE-CENTERED LINGUISTICS attention to the fact that real-life language users would profoundly, its willingness to recognize the multiple be perplexed or even annoyed if their interlocutors and indirect functioning of many utterances allows for interpreted their utterances only in terms of grammar, thought-provoking links with the concept of decon- which in itself commonly gives no explicit indication struction in semiotics, with all its cultural, political, of the intended meaning (speech act). For instance, and aesthetic reverberations. Moreover, utterance-cen- anyone who asked another person in the street, Do you tered linguistics has opened the way to important inno- know where the nearest bus stop is? would doubtless vations in language teaching methodology and course feel insulted by the reply, Yes, I know. To treat an utter- design, made possible by structuring activities and ance simply as a more or less well-formed grammatical curricula around utterances rather than sentences. sentence is willfully to miss the point. In the above example, the lost person’s utterance is evidently intend- ed as a request for information, not as a mere yes–no References question, and so any appropriate response would nec- Austin, J.L. 1962. How to do things with words. Cambridge, essarily include an effort to provide directions. US: Harvard University Press, and Oxford, UK: Oxford Analyzing utterances and the utterance structure of University Press. Cole, P., and J.L. Morgan (eds.) 1975. Syntax and semantics 3: discourses requires linguists to develop principles for speech acts. New York: Academic Press. disentangling separate but sometimes closely similar Grice, H. Paul. 1975. Logic and conversation. In Cole and speech acts, for example promising and threatening. Morgan. John Searle (1969, 1979) made important contribu- ———. 1981. Presupposition and conversational implicature. tions in that respect. Also, he proposed (1975) a useful Radical pragmatics, ed. by P. Cole. New York: Academic Press. distinction between direct and indirect speech acts: Schiffrin, Deborah. 1987. Discourse markers. New York and this contrast has proven to be valuable because it Cambridge, UK: Cambridge University Press. allows analysts to address the way in which, for ———. 1994. Approaches to discourse. Cambridge, US and instance, the above utterance, Do you know where the Oxford, UK: Blackwell. nearest bus stop is? indirectly realizes the speech act Searle, John R. 1969. Speech acts. Cambridge, UK: Cambridge University Press. of requesting information by directly performing a ———. 1975. Indirect speech acts. In Cole and Morgan. quite different speech act, a simple yes–no question. A ———. 1979. Expression and meaning. New York and further significant contribution was made by H. Paul Cambridge, UK: Cambridge University Press. Grice (1975, 1981), who suggested a possible set of Stubbs, Michael. 1983. Discourse analysis: the sociolinguistic maxims (guidelines) which—if shared by all speakers analysis of natural language. Chicago: University of Chicago Press, and Oxford, UK: Blackwell. in a speech community—could explain the remarkable Widdowson, Henry G. 1978. Teaching language as communi- degree of confidence and accuracy with which speak- cation. New York and Oxford, UK: Oxford University Press. ers formulate and hearers interpret utterances embody- JOHN N. SIVELL ing indirect speech acts. The social orientation of utterance-centered linguis- See also Austin, John Langshaw; Grice, H. Paul; tics has had very fruitful implications. Perhaps most Searle, John; Speech Acts 1150  V Variation What is often known as ‘variation theory’ in sociolin- For instance, the phonological environment that fol- guistics refers to the research paradigm developed by lows the cluster might be an important internal fac- William Labov in the 1960s for the quantitative study tor—it is possible that word-final consonant clusters of language variation and change. All variation in are more likely to be reduced when followed by a con- speech that cannot be explained on a purely linguistic sonant (e.g. east coast) than by a vowel (e.g. east end). level was once regarded as random or haphazard. In In addition, there are social or external factors that English, for example, the alternation of -in’ and -ing in may favor or disfavor word-final consonant cluster words such as swimmin’ or swimming is sometimes reduction. Age, gender, social class, and contextual called ‘free variation’, that is, speakers might choose style are some of the external factors that are often to say swimmin’ or swimming randomly. Variation the- taken into consideration in variation studies. ory aims to show that this type of variability is by no Investigators might hypothesize that a speaker from a means random; rather, a statistical and comparative working-class area would be more likely to say eas’ study involving multiple speakers and social contexts instead of east than speakers from other communities. can demonstrate that it is correlated with various Similarly, a speaker might say pac’ for pact in casual social and linguistic dimensions. In other words, the speech but not in formal speech. Finally, multiple- goal of variation theory is to explore the systematic regression analyses—statistical analyses that deter- nature of variability in language. mine the significance of multiple factors—are To conduct a variation study, investigators first need performed. The rule of word-final consonant cluster to select a variable—such as a sound (phonological) reduction is not categorical (that is, not all speakers segment that tends to vary in pronunciation—and from working-class areas reduce consonant clusters in quantify occurrences of variants of this variable in the word-final positions, and no speaker would do so at all speech of different individuals in a given community. times). The goal of the statistical analyses is to deter- Take the example of word-final consonant cluster mine which factors increase or decrease the likelihood reduction in various English dialects. Words such as of consonant cluster reduction. wind, test, and pact are sometimes reduced to win’, With the use of quantification and statistics, investi- tes’, and pac’, respectively. In this case, the two vari- gators can make accurate statements about fine-grained ants of this variable (i.e. word-final consonant cluster differences between groups of speakers in a communi- such as -nd, -st, and -kt) are (1) reduced consonant ty. When examined at the level of the community, what clusters (e.g. tes’) and (2) nonreduced consonant clus- seems to be random phonetic variation is actually sys- ters (e.g. test). Investigators then propose possible fac- tematically patterned and correlated with various tors that might influence word-final consonant cluster social and language-internal factors. In his study of reduction. These factors are of two major types. First, New Yorkers’ speech, for example, Labov (1972) there are factors internal to the language system itself. showed that both higher social class and higher 1151 VARIATION degrees of formality favored conservative (or ‘stan- the notion of social class is still considered a key dard’) pronunciations. Furthermore, an innovation that dimension in sociolinguistic change, John Rickford occurred in working-class speech was likely to appear (1986) and a host of others have argued that social class in the informal speech of all speakers. In addition to needs to be modified for non-Western societies. social class and contextual style, Labov showed that Furthermore, the concept of social network—that is, variation in New Yorkers’ speech was correlated with the sum of relationships that an individual shares with age as well. In the case of syllable-final r, its occur- other people—has received increased attention in rence was not significant for speakers from all social recent years (see Milroy 1987). Others, most notably classes in the two oldest age groups (50–75 and Penelope Eckert (2000), advocate an ethnographic 40–49). In other words, speakers in these age groups approach to the study of language variation and did not usually pronounce the r in words like car and change. Investigators who adopt this approach do not teacher. However, the pronunciation was a prestige decide a priori which social factors should be examined marker for speakers under 40 years old: speakers from in relation to language variation; rather, they participate the upper middle class consistently pronounced r in in the everyday lives of the speakers and understand the syllable-final position more often than those from social dimensions that matter most to members of the other social classes, and a regular increase in the use community. Instead of relying on ad hoc explanations, of r marked the move from casual to formal speech they offer more grounded social explanations for lin- styles. Examining age as one of the social factors guistic behavior, in that they possess a better under- allows investigators to observe language change in standing of the social meanings associated with ‘apparent time’ by listening to the speech of different linguistic variation in the communities they study. generations in a given community. The use of different variants of a variable shows that language change does not take place abruptly, but rather in small gradations. References Nevertheless, generalizations obtained through sta- Eckert, Penelope. 2000. Linguistic variation as social practice. tistical techniques are merely descriptive statements. Oxford and Cambridge, MA: Blackwell. Correlations between variation in language, on the one Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. hand, and linguistic and social dimensions, on the Milroy, Lesley. 1987. Language and social networks. Oxford other, need to be explained with solid reasoning. While and Cambridge, MA: Blackwell. the relationship between variation in speech and lan- Rickford, John. 1986. The need for new approaches to social class guage-internal factors can be accounted for by certain analysis in sociolinguistics. Language and Communication linguistic mechanisms (e.g. word-final consonant clus- 6. 215–21. Wolfram, Walt, and Natalie Schilling-Estes. 1998. American ter reduction can be explained in terms of the simplifi- English. Oxford and Cambridge, MA: Blackwell. cation process), sociolinguists look to sociology and ANDREW WONG anthropology for insights into the relationship between language variation and variation in society. Although See also Labov, William Vietnamese Vietnamese is spoken by approximately 65 million played a secondary role to other languages in Vietnam, of the 80 million inhabitants of the Southeast Asian particularly Chinese, which was the language of country of Vietnam. Nearly two million additional Vietnam’s ruling class from 111 BC to AD 939, and speakers of Vietnamese live outside the country, with French, which served as the country’s official language over a million residing in North and South America, from the late nineteenth century until the mid-twenti- mainly in the United States and Canada, and approxi- eth century. However, Vietnamese is now recognized as mately 400,000 in Europe, particularly in France, Great the official language of Vietnam, becoming the medi- Britain, Germany, and Switzerland. Speakers of um of instruction in schools in 1945. Vietnamese can also be found in Australia, New Because of the similarities and differences Zealand, Hong Kong, and Japan. Vietnamese has often that Vietnamese shares with other Southeast Asian 1152  VIETNAMESE languages, linguists have had difficulty establishing Vietnamese and Chinese share a number of other the genetic affiliation of the language with absolute characteristics leading to the false impression that the certainty. Due in part to the great number of Chinese two languages are closely related. One of the most sig- words in the Vietnamese lexicon, Vietnamese was at nificant of these characteristics is that both languages one time considered to be a member of the Sino- are considered tonal because they recognize tone as Tibetan family. More recently, linguists have attributed phonemic, i.e. tone makes a difference in meaning. For this similarity in word stock to the long domination of example, in the Hanoi dialect of Vietnamese, which Vietnam by the Chinese and have explored similarities recognizes six different tones, the word phonetically between Vietnamese and the members of other lan- realized as [ma] has six different meanings depending guage families. French scholar Henri Maspéro (1912) on which of the six tones (high, low-falling, high-ris- proposed that the genesis of Vietnamese resulted from ing, creaking-rising, dipping-rising, and constricted) is the fusion of a member of the Mon-Khmer family, used. In addition to the explanation of tonal genesis by such as Khmer (formerly Cambodian), and a Tai lan- Haudricourt, linguists have attributed this commonali- guage, such as Thai. Ultimately, Maspéro identified ty to contact between speakers of Vietnamese and Vietnamese as a Tai language because of its use of Chinese, particularly the Chinese dialects spoken in the tone, a characteristic feature of Tai languages that is southern provinces near the border of Vietnam, which uncharacteristic of Mon-Khmer languages. A number tend to have more tones than other Chinese dialects. of other linguists, however, have argued that Like other languages spoken in Southeast Asia, Vietnamese originated as a Mon-Khmer language that Vietnamese is a morphologically isolating language, a adopted a tonal system through contact with speakers typology that Vietnamese also shares with Chinese. of Thai. French botanist–linguist André Haudricourt Characterized by the absence of bound inflections, (1954) argued that the phenomenon of tonal genesis in isolating languages contrast with synthetic languages, Vietnamese was an evolutionary process in which the which use affixes on words to indicate such things as language acquired three tones by the sixth century number, tense, and agreement. For example, by affix- through language contact, and doubled this number by ing the bound morpheme /s/ to a noun or /ed/ to a verb, the twelfth century to the six tones currently used in speakers of English respectively show plurality and the Hanoi dialect of Vietnamese. This account, cou- past tense. In an isolating language like Vietnamese, pled with the observation of the similarity between however, words and morphemes tend to have a one-to- Vietnamese and Muong by Pryzluski (see Meillet and one relationship, so that when Vietnamese speakers Cohen 1924), has convinced many linguists that both want to show plurality, they must use a separate word Vietnamese and Muong belong to the Mon-Khmer or grammatical particle elsewhere in the sentence. As phylum in the Austro-Asiatic family. in Chinese, there tends to be a one-to-one relation- Although linguists have repudiated anything more ship between syllables and morphemes in Vietnamese, than a distant genetic relationship between Vietnamese which has led some scholars to deem Vietnamese as and Chinese, the two languages have a number of monosyllabic. Other scholars, however, have taken similarities. The most apparent among these is the great issue with this characterization and have pointed out number of Chinese words in the Vietnamese lexicon, that the Vietnamese lexicon includes many words that many of which were borrowed during the Chinese dom- are disyllabic, as well as words that have three or four ination of Vietnam from 111 BC to AD 939. During this syllables (Hannas 1997:76–77). The perception that period, Chinese was not only used among many of the Vietnamese is monosyllabic has also been attributed to officials of Vietnam, but also as a medium of educa- the Vietnamese orthographic practice of putting a tion; as a result, the language of law, politics, history, space between every syllable of a word. medicine, science, and technology in Vietnamese Although it lacks inflectional morphology, include a great number of borrowings from Chinese. Vietnamese employs derivational morphology as a While some of these words have been modified to means for creating new words. Vietnamese uses prefix- such an extent by Vietnamese speakers that they can ing and suffixing, which are two processes of affixation hardly be recognized as Chinese anymore, others are used productively in many of the world’s languages, referred to as ‘Sino-Vietnamese’ because they are pro- including English; however, in contrast to many other nounced as they would have been in ancient Chinese. Southeast Asian languages, Vietnamese does not use Vietnam’s proximity to populations of Malay and Thai infixing, a type of affixation in which a morpheme is speakers has resulted in the Vietnamese lexicon inserted into the root of a word. Another very produc- including borrowings from these languages; tive process of deriving new words in Vietnamese is Vietnamese has also borrowed words from French, compounding, not only as a process of creating new English, and Russian as a result of colonization, war, nouns but in the derivation of new verbs and adjectives. and political alliance. Reduplication, which is the process of deriving a word 1153 VIETNAMESE by copying all or part of a word and affixing it to itself, first-person pronouns depending on whether they are is yet another morphological process that is used often addressing their parents, elder siblings, or younger sib- in Vietnamese and other Southeast Asian languages, as lings. In using the second person, children must use well as Chinese. Like many of the world’s languages pronouns reflecting the distinction in gender when that use reduplication productively, Vietnamese com- addressing their parents, a distinction that is preserved monly uses reduplication to pluralize and intensify the when children address elder brothers and sisters. This original semantics of a word, as well as to show reci- distinction is not maintained when parents address chil- procity. Reduplication is also commonly used in ono- dren or when elder siblings address younger siblings. matopoetic expressions in Vietnamese, as is also the As in the use of personal pronouns, the choice of an case in other languages of the world. inappropriate kinship term may have consequences. The basic syntactic structure of Vietnamese is sub- Apparently borrowed from the South Chinese, one cus- ject–verb–object (SVO). Because it is an isolating lan- tom that is commonly used among the lower classes of guage, Vietnamese relies heavily on word order to Vietnam is the use of kin numeratives. A device used show the grammatical roles being played by various for designating grown-up children, a common pattern arguments in a sentence. Vietnamese also uses special is to call the eldest child ‘Number One’, the second markers to indicate such things as the tense of verbs child ‘Number Two’, and so on. An interesting varia- and the number and gender of nouns. In Vietnamese, a tion of this occurs in southern Vietnam, where ‘Number tense marker is placed before the verb that it modifies, Two’ is used to refer to the eldest son or daughter, with and negative markers also appear in preverbal posi- speculation that the Vietnamese word for ‘Number tion. Adjectives come after the nouns they modify in One’ was reserved for the mother because she is con- Vietnamese. Content questions, or questions that sidered the father’s first child (Dinh-Hoa 1980:32). demand an answer besides yes or no, are formed by Three different orthographic systems have been inserting an interrogative word in the position in the used during the history of writing in Vietnamese. The sentence in which the unknown material should be, first was chu nho, which was borrowed directly from leaving the sentence in its basic SVO order. Although the Chinese and was the official written language from SVO is the basic word order in Vietnamese, object– as early as the eleventh century until the nineteenth subject–verb (OSV) order is also commonly used century. The next was chu nom, a writing system based when a speaker moves an object to the beginning of a on the Chinese system but altered to phonetically rep- sentence to give it emphasis over the subject, a process resent spoken Vietnamese. The invention of this sys- known as topicalization. tem is generally attributed to Han Thuyen, a famous The system of personal pronouns in Vietnamese is Vietnamese poet of the thirteenth century. The modern rather complex due to the use of pronouns not only as spelling system is quoc ngu, a Roman alphabet substitutes for the name of a person or group of per- devised by Catholic missionaries from Europe to sons, as they are in English, but also as a way of mark- translate scripture into the language in a way that ing the status of participants and referents in would make sense to the Vietnamese, as well as to the conversations. Based largely on factors such as age and missionaries who read aloud from it. During French familiarity, the use of various pronouns to mark status domination of the country from the late nineteenth requires speakers to evaluate whether they are superior century to the mid-twentieth century, the French used or inferior to those with whom they are speaking, and quoc ngu to introduce the Vietnamese to French lan- this decision affects which first-person pronouns they guage and literature. Although many have praised the use, as well as second-person pronouns. As is the case success of the Vietnamese in replacing the Chinese in other cultures that use languages with similar pro- writing system with a more phonetic alphabet, others noun systems, there are social consequences for choos- have pointed out that the extensive use of diacritics to ing the wrong pronouns in Vietnamese. The pronoun indicate tones in quoc ngu creates a system nearly as system also has two forms for the first-person plural complicated as the one it was intended to replace pronoun: one to be used when speakers wish to include (Hannas 1997:75). their addressees in the group for which they are using Vietnam is characterized by three dialect areas: the pronoun (inclusive) and another if the addressees northern, central, and southern, perhaps reflecting the are not included (exclusive). division of Vietnam into three states by the French dur- Like other Asian languages, Vietnamese also has a ing their domination. The northern dialect area includes complex system of kinship terms, or terms that speak- Hanoi, the capital city of Vietnam. Ho Chi Minh City ers use to address and to refer to members of their own (formerly Saigon) is the largest city in Vietnam and is family, which reflects the extended and patrilineal in the southern dialect region. The differences between nature of the family structure. In speaking to members the three dialects are mainly at the levels of pronuncia- of their immediate family, children must use various tion and vocabulary and are not great enough to cause 1154  VISUAL WORD RECOGNITION much difficulty in communication among speakers of Haudricourt, André-Georges. 1954. De l’origine des tons in different dialects. Although Vietnam does not have a viêtnamien. Journal Asiatique 242. 68–82. Luong, Hy V. 1990. Discursive practices and linguistic mean- spoken standard, with each locale maintaining its own ings: the Vietnamese system of personal reference. dialectal flavor, the dialect spoken in Hanoi is the Amsterdam: John Benjamins. dialect used most often in the media. Maspéro, Henri. 1912. Études sur la phonétique historique de la langue annamite. BEFEO 12. 1–127. Meillet, A., and Marcel Cohen. 1924. Les Langues du Monde. References Paris: Centre national de la recherche scientifique. Dinh-Hoa, Nguyen. 1980. Language in Vietnamese society: Ngo, Binh Nhu. 1999. Elementary Vietnamese. Boston: Tuttle. some articles by Nguyen Dinh-Hoa. Carbondale, IL: Asia Thompson, Laurence C. 1965. A Vietnamese grammar. Seattle: Books. University of Washington. ———. 1997. Vietnamese. Amsterdam: John Benjamins. LAMONT ANTIEAU Gregerson, Kenneth J. 1969. A study of middle Vietnamese phonology. Saigon: Summer Institute of Linguistics. See also Austroasiatic; Chinese (Mandarin); Hannas, Wm. C. 1997. Asia’s orthographic dilemma. Honolulu: Khmer and Mon-Khmer Languages; Southeast University of Hawaii Press. Asia; Thai and Tai Languages Visual Word Recognition Visual word recognition refers to the process by which rules. For example, the grapheme consisting of the first readers identify written words. Recognizing a word two letters of SHEEP is translated as the phoneme that enables a reader to use the word’s phonology to read sounds like ‘sh’. Both of these routes compute a pro- aloud or to use the word’s semantics to determine its nunciation that is represented as a sequence of meaning. Skilled reading is marked by the ability to phonemes. The eventual response is based on combin- recognize words quickly and with relatively little effort. ing the outputs of both routes in a response buffer. There are two major approaches to modeling cog- The lexical route is needed in order to correctly pro- nition in general, and word recognition in particular. nounce words that have an irregular spelling–sound Cognitive neuropsychologists of the nineteenth centu- correspondence such as PINT because the phonology ry assumed that the language processing system was generated by the rule route favors a pronunciation that modular and that component structures and processes would rhyme with the regular words HINT and MINT. could be represented as a box-and-arrow diagram. The rule route is needed in order to model the ability of This approach fell out of favor in the first half of the humans to easily read aloud nonwords like MAVE. In twentieth century, but remains popular in the twenty- this case, the lexical route would have no entry for first century. The revival is bracketed by two seminal MAVE and could not generate a pronunciation other articles in Psychological review: Morton’s (1969) than by analogy to similar real words. This division of logogen model and Coltheart’s (2001) dual route cas- labor allows dual-route models to offer compelling caded (DRC) model. explanations for different types of acquired dyslexias That portion of the DRC concerned with reading that together form a double dissociation. For example, aloud can be used to provide an example of the modu- damage to the lexical route should lead to specific dif- lar approach. The model assumes two principal path- ficulties with irregular words like PINT (surface ways by which a printed word can be read aloud. Both dyslexia), but preserve the ability to read nonwords and routes are initiated by a common process of ortho- regular words. On the other hand, damage restricted to graphic analysis, which leads to the identification of the the rule route should lead to specific difficulties in read- word’s constituent letters. The lexical route uses the let- ing nonwords (phonological dyslexia), but preserve the ter identities to activate an entry in a mental dictionary ability to read both regular and irregular words. called the orthographic input lexicon. In turn, the path- The DRC is also a computational model, meaning way from the orthographic input lexicon to the phono- that it has been implemented as a computer program logical output lexicon activates the word’s phonology. that performs reading tasks by using the same struc- The other major route processes the letters in left-to- tures and processes hypothesized for human readers. right order and applies a set of grapheme–phoneme The reading tasks studied most often are reading aloud 1155 VISUAL WORD RECOGNITION Speech Response buffer Phonological output lexicon Grapheme-phoneme Semantic rule system system Orthographic input lexicon Orthographic analysis (letter identification) Print Figure 1 is also a mediated pathway through a semantic system Hidden that contains semantic representations of each word. units However, the semantic pathway has not been imple- mented in the computational model because developing realistic semantic representations is more difficult than Orthographic Phonological representing orthography and phonology. However, units units powerful new tools like latent semantic analysis and (input) (output) other high-dimensional representations based on the co-occurrence of words in very large corpora of text Figure 2 make it likely that computational models will soon be able to address more adequately the very important role of semantics in word recognition. and the lexical-decision task. In the latter, participants The second major approach used to model word are asked to make speedy responses as to whether a recognition uses learning algorithms in conjunction letter string is a word or a nonword. DRC can simulate with networks that have distributed representations. the speed and accuracy of human readers in response The seminal model of this type was Seidenberg and to many different characteristics of words. These McClelland’s distributed developmental model include spelling–sound regularity, frequency of occur- (DDM, 1989). Using the DDM as an example and rence, and the orthographic similarity of a target word reading aloud as the task of interest, this computation- to other words. al model is based on a network with three layers. The One heavily researched phenomenon in word input units represent orthography, but that representa- recognition that the DRC has not simulated is seman- tion is distributed across a large set of nodes. This tic priming. Semantic priming occurs when the means that individual nodes do not symbolize individ- response to a target word like DOCTOR is faster when ual letters. Rather, a given node is sensitive to many preceded by a related prime like NURSE. In theory, letter sequences and, conversely, a letter in a specific DRC could account for semantic priming because, in input location will partially activate many different addition to the direct pathway from the orthographic input nodes. The third, or output, layer codes phonol- input lexicon to the phonological output lexicon, there ogy with a similar type of distributed representation. 1156  VOICE The middle hidden layer has no symbolic representa- References tions at all. Each node in one layer is connected to every node at the next layer. Each connection has a Besner, Derek, and Glyn Humphreys (eds.) 1991. Basic processes in reading: visual word recognition. Hillsdale, NJ: weight that is initially set at a random value, but dur- Lawrence Erlbaum Associates. ing training the back propagation learning algorithm Coltheart, Max, Kathleen Rastle, Conrad Perry, Johannes progressively adjusts the weights such that each input Ziegler, and Robyn Langdon. 2001. DRC: A dual route cas- (printed word) more and more closely approximates caded model of visual word recognition and reading aloud. the correct output (pronunciation). In the DDM, both Psychological Review 108(1). 204–56. Frost, Ram, and Leonard Katz (eds.) 1992. Orthography, irregular words and nonwords can be read aloud by phonology, morphology, and meaning. Amsterdam: North- using a single pathway from input (orthography) Holland. through the hidden layer to output (phonology). In McClelland, James, and David Rumelhart. 1988. Explorations in contrast to dual-route models, it does so without the parallel distributed processing. Cambridge, MA: MIT Press. explicit representation of grapheme–phoneme rules or Morton, John. 1969. Interaction of information in word recog- nition. Psychological Review 76. 165–78. the explicit representation of words in a mental lexi- Seidenberg, Mark, and James McClelland. A distributed, devel- con. The DDM and its successors have also enjoyed opmental model of word recognition and naming. Psycho- considerable success in simulating the performance logical Review 96. 523–68. obtained with people in reading tasks. KENNETH R. PAAP Voice ‘Voice’ is a technical term for the description of sen- (typically a patient) subject, whereas the agent may or tence structure. It encompasses a range of grammatical may not appear as an object introduced by a preposition constructions that differ in the correlation between the (e.g. Mary was kissed [by John]). The passive has tra- semantic and grammatical function of particular sen- ditionally been analyzed as the result of a grammatical tence elements. The English sentence John kissed Mary, operation (passivization); the active sentence used to be for example, has two noun phrases: John and Mary. The regarded as the basic one, which would somehow need first, the subject, performs the action described by the to be changed to become a passive sentence. Passive sentence and is thus interpreted as the ‘agent’. The sec- sentences were considered to be synonymous para- ond, the object, undergoes the action and is thus inter- phrases of their active counterparts. John kissed Mary preted as the ‘patient’. In an English active sentence and Mary was kissed by John were regarded as essen- such as this one, the grammatical function ‘subject’ thus tially equivalent utterances. Since the 1980s, however, correlates with the semantic role ‘agent’, whereas the several studies have taken into consideration that the ‘object’ fulfills the semantic role of ‘patient’. In lan- passive may have different connotations in context; i.e. guages such as English, ‘active’ is the basic voice, the choice between an active and a passive sentence is because it appears with all verbs and does not require more than just a random stylistic one. Rather, it is seen any specific marking on the words involved (e.g. was as a matter of relative prominence of agent and kissed). These languages typically also have ‘passive’ patient—passive and active differ in terms of informa- and ‘middle’ constructions, which are discussed below. tion packaging or sequencing, or in terms of fore- There are many languages (so-called ergative lan- grounding of information. Because the subject is guages), however, in which sentences where the patient usually the topic of the sentence, a passive sentence is surfaces as subject are the basic ones. A cross-linguis- thus argued to foreground the patient role. Although tically valid definition of voice is thus difficult, and any supported by psycholinguistic experiments, this view of characterization of voice in terms of information passive constructions has been widely challenged, and sequencing or in terms of cognitive processes raises a it has been claimed that the main function of passive number of problems, because the voice systems of dif- sentences is not the foregrounding of the patient but ferent languages differ enormously from one another. rather the defocusing or backgrounding of the agent. A passive construction or sentence is one in which This would account for the fact that in a significant a somehow marked form of the verb has a nonagent number of cases the passive voice does not signal that 1157 VOICE the patient is topical, but simply indicates that the agent agent usually has subject properties; for example, it is not important. A case in point is the so-called imper- agrees with the verb. Talmy Givón (1994) distinguish- sonal passive (e.g. German Es wurde hier getanzt, es between semantic and pragmatic inverse systems. In ‘There was dancing here’). Because in impersonal pas- the latter, the choice between a direct or an inverse sives there is no patient role, patient topicalization can- clause is a matter of which of the two participants is not be considered as the function of these passives. more salient in the specific discourse. However, in Some verbs, such as English sink, appear in pairs semantic inverse systems, the decision between inverse such as The enemy sank the ship vs. The ship sank. The and direct voice depends on more objective relations, object of the first clause corresponds to the subject of such as animacy: if the subject refers to an animate the second. The verb in the second clause would thus be entity (e.g. an animal), while the object refers to an expected to occur with passive marking, which, howev- inanimate one (e.g. a rock), the direct voice would be er, is not the case. These verbs have sometimes been used, because the animate subject is higher on the labeled ‘middle verbs’ because they seem to break out empathy scale than the inanimate object. of the active/passive dichotomy. Other constructions typically referred to as ‘middle’ are reflexive sentences. An example would be the Spanish sentence Juan se ve, References ‘John sees himself’. The primary function of this con- Andersen, Paul Kent. 1991. A new look at the passive. struction is to encode the identity of agent and patient. Frankfurt: Lang Verlag. The argument that this is somewhat in between active Blevins, James P. 2003. Passives and impersonals. Journal of and passive has been extended to other functions or Linguistics 39. meanings of the reflexive marker, such as the encoding Cornelis, Louise H. 1997. Passive and perspective. Amsterdam and Atlanta, GA: Rodopi. of an action performed on the subject’s body, as in the Dixon, R.M.W. 1994. Ergativity. Cambridge, UK, New York: Spanish Juan se peina el pelo, ‘John combs his hair’; Cambridge University Press. the expression of situations that occur entirely within Fox, Barbara, and Paul J. Hopper (eds.) 1994. Voice: form and the subject’s sphere, as in sentences with the Spanish function. Amsterdam and Philadelphia, PA: John verb sentirse, ‘feel’ (literally ‘feel-self’); or even the Benjamins. Givón, Talmy. 1981. Typology and functional domains. Studies expression of passive meaning, as in Spanish aquí se in Language 5. habla Español, ‘Spanish is spoken here’. In some lan- ——— (ed.) 1994. Voice and inversion. Amsterdam and guages, the opposition active/middle is actually encod- Philadelphia, PA: John Benjamins. ed in alternating verb forms. This can be observed in Haspelmath, Martin. 1990. The grammaticization of passive Sanskrit and Greek (both ancient and modern), both morphology. Studies in Language 14. Jaeggli, Osvaldo. 1986. Passive. Linguistic Inquiry 17. members of the Indo-European family, and in lan- Keenan, Edward L. 1985. Passive in the world’s languages. guages such as Tula (a Niger-Congo language) and Language Typology and Syntactic Description, Vol. 1, ed. Tamil (a Dravidian language). In these languages, the by Timothy Shopen. Cambridge: Cambridge University middle voice usually denotes actions that are performed Press. on the body or on a part of it, as well as spontaneous Kemmer, Suzanne E. 1993. The middle voice: a typological and diachronic study. Amsterdam and Philadelphia, PA: John actions, i.e. events presupposing no causal agent. Benjamins. Thus, many languages distinguish between active Kirsner, Robert S. 1976. On the subjectless ‘pseudo-passive’ in and passive voice and may also include a middle voice. standard Dutch and the semantics of background agents. Ergative languages approach the same issue from the Subject and topic, ed. by Charles N. Li. New York: opposite side and seem to focus on the patient rather Academic Press. Klaiman, M.H. 1992. Middle verbs, reflexive middle construc- than the agent. To further complicate the issue, in some tions, and middle voice. Studies in Language 16. languages, the question of whether or not the subject of Myhill, John. 1997. Toward a functional typology of agent the sentence is an agent or not becomes altogether sec- defocusing. Linguistics 35. ondary. These languages, for instance, Philippine lan- Perlmutter, David M., and Paul M. Postal. 1977. Toward a uni- guages or native American Algonquian languages, versal characterization of passivization. BLS 3. Berkeley, California. concentrate instead on ‘empathy’. Empathy is an all- Sansò, Andrea. 2003. Degrees of event elaboration. Passive encompassing complex system involving features such constructions in Italian and Spanish. Milan: Francoangeli. as animacy, (perceptive) salience, (potential) agency, Shibatani, Masayoshi. 1985. Passive and related constructions: whether or not something has been mentioned before, a prototype analysis. Language 61. etc. ‘Inverse voice’ contrasts here with ‘direct voice’. Siewierska, Anna. 1984. The passive. A comparative linguistic analysis. London: Croom Helm. The inverse voice indicates that the agent ranks lower ANDREA SANSÒ on the empathy scale than the other role, whereas the direct voice is used when the agent is higher on the See also Functional Approaches; Generative Gram- empathy scale. In both inverse and direct clauses, the mar; Syntax; Thematic Structure 1158  VOWEL HARMONY Vowel Harmony Vowel harmony is a well-known but not yet well- Tuvan understood phenomenon found in many languages Low vowels (pronounced with less tension, as in native to Eurasia, Africa, and, to a lesser extent, other long) do not obey labial harmony: places. Hungarian, Finnish, and Turkish number xol-da ‘in (the) hand’ among the best-known cases of harmony. The Uralic xöl-de ‘in (the) lake’ and Altaic language families, spoken across a vast High vowels must obey labial harmony: geographic area of Eurasia from Korea to Turkey, xol-u ‘his hand’ exhibit vowel harmony to varying degrees in most of xöl-ü ‘his lake’ their member languages (Hungarian, Finnish, Turkish, Vowel harmony may be thought of as a restriction on Mongolian, Manchu, and so forth). Harmony is not which vowels may occur together within the same found in any familiar Indo-European languages (e.g. word. In Turkic languages, all vowels can be divided Greek, French, or English). into two natural classes: those pronounced with the Simply stated, harmony is when vowels within a tongue in the back part of the oral cavity, and those pro- word are required to resemble one other in terms of nounced with the tongue toward the front. In a Turkic some property. For example, in the Tuvan word language with eight vowels, four of them will belong to tooruksug, meaning ‘smelling like a pine cone’, all the front (palatal) class and four to the back (velar) vowels are ‘labial’ (pronounced with rounded lips). class. For a typical Turkic vowel inventory, see below: Tuvan has an ample supply of vowels that do not use front (palatal) back (velar) lip-rounding, yet the vowels here—taking their cue from the first vowel—are all rounded. High iü ɯu There are several basic types of harmony, each Low eö ao named for the speech organs involved. These include (labial vowels are ü, u, ö, o) palatal/velar (also called ‘backness’) harmony, labial (also called ‘rounding’) harmony, and tongue root Within any single word (including any affixes (also called ATR) harmony. Palatal harmony requires added to that word), vowels from the front and back vowels to be alike in terms of whether they are pro- classes do not mix: nounced with the tongue body toward the front of the Tuvan oral cavity (close to the palate) or toward the back is ‘footprint’ (close to the velum). Labial harmony requires vowels is-ter-ivis-te ‘in our footprints’ to be alike in having lip rounding. Tongue root harmo- aas ‘mouth’ ny requires vowels to be alike in whether the tongue aas-tar-ɯvɯs-ta ‘in our mouths’ body is pushed toward the front of the mouth or pulled toward the back. Harmony applies only within a word. No harmony A language may have more than one harmony sys- system allows harmony to carry across a longer span, tem. Turkish has palatal and labial harmony, whereas such as two distinct words or an entire sentence. At the Classical Manchu had tongue root and labial harmony. level of word structure, harmony systems may be Two separate harmony systems, when present, may divided into two basic types, depending on what ele- operate in an independent or semi-independent manner. ment controls harmony. In ‘root-controlled’ harmony, Not all vowels participate equally in harmony. vowels in a word root determine the quality of vowels Typically, some restrictions are placed on which vow- appearing in affixes. In ‘dominant–recessive’ harmo- els cause harmony and which ones obey it. Such ny, neither roots nor affixes take precedence. Instead, restrictions usually refer to some feature other than the if any vowel belonging to the dominant class appears harmonic one. In Tuvan, for example, any labial vowel anywhere in the root or a suffix, it forces all other vow- can trigger labial harmony (forcing adjacent vowels to els in the word to shift over to that class. A word may also be labial), but only high vowels (vowels that are contain weak (recessive) vowels only if no strong pronounced with more tension in the tongue and vowels are present. mouth, such as in seek or plume) undergo labial har- No harmony system is exceptionless: typically, a mony (vowels that obey harmony are underlined language will have some nonharmonic native words or herein): affixes. A language that has been in long-term contact 1159 VOWEL HARMONY with a nonharmony language (e.g. Russian or Farsi) syllable is palatal, he or she can be fairly certain that may have a large supply of nonharmonic loan-words. each following vowel also belongs to the palatal set. Nonetheless, these disharmonic elements may par- This reduces by one half the range of potential vowels ticipate productively in harmony. Usually, the final he or she must listen for in succeeding syllables, and it vowel of a nonharmonic word determines the harmo- may lessen the cognitive processing required. ny of the following suffix vowels: Many harmony languages have vowels that fail to participate fully or at all. Instead, they may either Classical Manchu (disharmonic words) obstruct the harmony pattern or remain invisible to it. age ‘prince’ So-called opaque vowels block the current harmony age-se ‘princes’ pattern and start a new pattern. The Turkish suffix -gen dahema ‘uncle’ (corresponding to English -gon, as in octagon) con- dahema-sa ‘uncles’ tains a front (palatal) vowel that never alternates to How is harmony related to the mechanics of speak- obey harmony. Vowels following -gen must be palatal, ing? Vowel harmony makes some sense when viewed in even when the prevailing pattern in the word is velar: terms of natural movements of the speech organs. For Turkish example, when you round your lips to pronounce ‘oo’, alt ɯ ‘eight’ this gesture may carry over onto the next vowel. This alt ɯ-gen ‘octagon’ tendency might account for common slips of the tongue alt ɯ-gen-ler ‘octagons’ in English, for example, saying kangaroo moot when you meant to say kangaroo meat. Physical speech So-called transparent or neutral vowels allow har- mechanisms provide a natural foundation for harmony. mony to pass through them, without being affected and But they do not explain why certain languages (and not without blocking it. In Karelian (a Uralic language), others) adopt harmony as a fundamental organizing backness harmony dictates that if the first syllable con- principle of their sound systems. Harmony likely traces tains a front vowel, all the following syllables contain its origins to natural speech mechanics, but it has front vowels. Likewise, if it contains a back vowel, the evolved into a more abstract way of organizing the following syllables have back vowels. Exceptions are sounds of particular languages. Thus, harmony can the front vowels [e] and [i], which remain indifferent to work at cross-purposes with the mechanics of speech— harmony and may appear anywhere. it may impose conditions that do not obviously facilitate Most harmony languages tolerate some disharmony, either speech production or perception. which can result from internal processes or from bor- For example, in Tofa, if a word contains [j] (the rowing foreign words. Internally, consonants may sound written with a ‘y’ in English), all vowels next to interfere with the harmony system, imposing their own [j] must be pronounced in the front of the mouth, as if conditions that override it. Syllable structures may also they belonged to the palatal class. However, for pur- impose limitations on harmony: in Turkmen, for exam- poses of harmony, the speaker may ignore the fronting ple, labial harmony does not affect vowels in word- effect that [j] has on some vowels, considering them to final syllables containing a vowel. In Turkish, the be back vowels despite their palatal quality. Any suffix presence of a velar consonant [g] may force an adjacent the speaker adds will thus contain a velar vowel: vowel to be velar, even though the other vowels in the word are palatal. Modern Turkish also has a high per- Tofa centage of nonharmonic loan-words in its lexicon: taksi xöjen-da ‘in (the) cliff’ ‘taxi’, kitap ‘book’. Some harmony languages alter In this case, harmony clearly does not facilitate pro- borrowed words to render them more harmonic. nunciation, because Tofa speakers are not accustomed The Turkic languages range from almost perfectly to mixing palatal and velar vowels within a word. But harmonic to hardly harmonic at all. These languages it does serve to maintain the integrity (at an abstract illustrate the possibilities of change and evolution in level) of harmony as a system. harmony. The plural suffixes in Old Turkic and two of Besides ease of pronunciation for the speaker, what its daughter languages, Turkish (highly harmonic) and other advantage might harmony offer in language Uzbek (barely harmonic), show two possible paths design? Some linguists have argued that harmony helps along which harmony can proceed (see below). to make clear to the hearer the boundary between one word and the next. But this can be true only if two adja- back suffix front suffix cent words belong to different harmony classes. It has also been suggested that harmony increases the likeli- Old Turkic (harmonic) -lar -ler hood that the hearer will correctly identify certain vow- Turkish (harmonic) -lar -ler Uzbek (not harmonic) -lar -lar els. If a speaker notices that the vowel of the first 1160  VOWEL HARMONY Harmony as a pattern can best be understood by con- Clark, Larry. 1998. A reference grammar of Turkmen. sidering those in which it fails to apply as well as those Wiesbaden: Harrasowitz Verlag. Johannson, Lars, and Eva Csato (eds.) 1999. Turkic languages. in which it applies. Many unanswered questions remain London: Routledge. in understanding vowel harmony. Dozens of languages Kornfilt, Jaklin. 1997. Turkish. London: Routledge. known to have harmony remain largely undocumented, Li, Bing. 1996. Tungusic vowel harmony: analysis and descrip- and new types will certainly be discovered. tion. The Hague: Holland Academic Graphics. Sinor, Denis (ed.) 1988. The Uralic languages. Leiden: E.J. Brill. References Vago, Robert M. (ed.) 1980. Issues in vowel harmony. Anderson, Gregory D.S., and K. David Harrison. 1999. Tyvan Amsterdam: John Benjamins. [Tuvan]. Munich: Lincom Europa. DAVID K. HARRISON 1161  W Wackernagel, Jacob Wilhelm Wackernagel, Jacob’s father, originally from ‘P2’, because it is situated immediately after the first Berlin, moved to Basle in 1833, where he was appoint- word with intonational emphasis (accent). All words in ed professor for German Studies at the local universi- P2 share the characteristic that they are unaccented (or ty. The Wackernagel family developed a very close weakly accented). This tendency is most visible in link with the city of Basle: Jacob’s brother Rudolf, Sanskrit and Homeric Greek, but Wackernagel found its who worked for the city archives, wrote a history of traces in other Indo-European languages as well, and he Basle reaching back to the sixteenth century, and surmised that Proto-Indo-European, the hypothesized Jacob lived in Basle and taught at the University of ancestor language of all Indo-European languages, fol- Basle until his death, with only one interruption, dur- lowed the same pattern. Thus, Wackernagel reconstruct- ing which he taught at the University of Göttingen, in ed Proto-Indo-European as having essentially a similar the years preceding World War I. Jacob’s life was word order to Modern High German, having the tensed entirely devoted to studying and teaching classical verb obligatorily in second position in main clauses and philology and the ancient Indo-European languages. in final position in subordinate clauses. He had his first appointment at the age of 24; his Wackernagel’s hypothesis was received with vary- teacher of Greek was Friedrich Nietzsche. ing degrees of approval by his contemporaries; espe- Wackernagel can be considered a member of the cially the idea that the position of the tensed verb in second generation of the ‘Neogrammarians’, a school Proto-Indo-European was the same as in German did of linguists interested in studying how languages not meet general consent. However, the fact that vari- change over time. Among his contemporaries, he ous types of unaccented item were placed in P2 found occupies a special position because of his interest in a striking confirmation when Hittite was finally deci- sentence structure, something that had been investigat- phered in 1916, for unstressed pronouns and particles ed to a much lesser extent than pronunciation and follow the pattern described by Wackernagel virtually word structure. A survey of the courses he gave in his without exception in this language. years as professor shows that syntax was a recurring In the second half of the twentieth century, progress topic already in the 1880s; in 1904 Wackernagel was in the study of unaccented particles (usually referred offering a course in historical syntax, a topic which to as ‘clitics’) and knowledge of previously unknown can still sound innovative today. non-Indo-European languages has demonstrated that In his 1892 essay Über ein Gesetz der indogermanis- the Wackernagel position is not a unique feature of chen Wortstellung (On a law about Indo-European word Indo-European. Being unstressed, clitics need what is order), Wackernagel pointed out that in the ancient Indo- commonly called a phonological host, i.e. an accented European languages a number of particles, conjunctions, word with which they can form a phonological unit. pronouns, and verbs took a peculiar position in the sen- P2 is one of the possible positions for clitics, found in tence, nowadays known as the ‘Wackernagel position’ or e.g. some Australian and Uto-Aztecan languages. 1163 WACKERNAGEL, JACOB Wackernagel’s other major accomplishment in the appointed in the same university, where he taught from field of syntax is constituted by the two volumes of 1877 to 1902. He was appointed in Göttingen in lectures on syntax (Vorlesungen über Syntax), 1902–1915. Then he returned to Basle and resumed 1926–1928, in which the scholar addressed several his former position, which he held until 1937. He died problems connected with the syntactic behavior of in Basle on May 22, 1938. nouns and verbs in the Indo-European languages (mostly based on Latin, Greek, and Germanic). References Wackernagel’s approach is typical of historical lin- Collinge, N.E. 1985. The laws of Indo-European. Amsterdam guistics in the late nineteenth to early twentieth centu- and Philadelphia: Benjamins. ry, insofar as—from today’s perspective—the word is Eichner, Heiner, and Helmut Rix (eds.) 1990. Sprachwissenschaft still the dominant unit of analysis. In Wackernagel’s und Philologie. Jacob Wackernagel und die Indogermanistik times, linguists were in the process of developing dis- heute. Wiesbaden: Reichert. Hale, Ken. 1983. Warlpiri and the grammar of non-configurational tinctive theories of word and sentence structure, but languages. Natural Language and Linguistic Theory 1(1). word-related issues were still defining and delimiting 5–47. syntactic problems. Although he announced a further Halpern, Aaron, and Arnold Zwicky (eds.) 1996. Approaching volume on sentence structure, Wackernagel never second: second position clitics and related phenomena. wrote it. Since the newly individuated field of syntax Chicago: CSLI Publications. Luraghi, Silvia. 1998. The grammaticalization of the left sen- could not yet rely on a theoretical framework, tence boundary in Hittite. The limits of grammaticalization, Wackernagel’s syntactic work remains descriptive, his ed. by, A. Giacalone Ramat and P. Hopper, 189–210. insights mostly deriving from his deep knowledge of Amsterdam and Philadelphia: Benjamins. languages and texts. Wackernagel, Jacob. 1926–1928. Vorlesungen über Syntax, 2 vols. Basle: Birkhäuser. ———. 1953. Kleine Schriften, 2 vols. Göttingen: Vandenhoek Biography & Ruprecht. Watkins, Calvert. 1963. Preliminaries to a historical and com- Jacob Wackernagel was born in Basle in 1853. He parative analysis of the Old Irish verb. Celtica 6. 194–249. studied Classical Philology in Basle, and was then SILVIA LURAGHI Warao Warao is an Amerindian language spoken in the The estimated 1,000 Warao on the Guyanese side Orinoco Delta and adjoining areas in northeastern are all in close contact with wider Guyanese society so Venezuela and Guyana. Its speakers, the Warao, are that all of them can be assumed to be English speak- the aboriginal population of the Delta, who have been ing: some being bilingual Warao-English or even present for as long as 9,000 years (W. Wilbert trilingual including Spanish. There is however the ten- 1995:336). They have an estimated population of dency to give up Warao (Forte 2000). about 30,000, which is increasing. Although they form the second largest indigenous group in Venezuela, Social and Political Status of the Language given that the total indigenous population of the coun- try comprises only 1.5%, they are clearly a minority. Up to now, Warao is a language with no genuine writ- The latest reliable official census rates 90% of the ing tradition. Nevertheless, recordings mainly of Venezuelan Warao as speaking their indigenous lan- myths have been written down and published by guage and 48% of them as being bilingual in Warao anthropologists (J. Wilbert 1970; Heinen 1988), mis- and Spanish (Venezuela 1993). Especially in the cen- sionaries, and linguists (Osborn 1966a, b, 1967; tral part of the Delta, missionary boarding schools Lavandero 1991, 1992). The diversified oral tradition have had the—albeit unintended—effect of language includes different styles of speech and various genres. loss. There are therefore a growing number of mono- It differentiates for instance between ‘denobo’ (deje lingual Spanish speakers in this area and in the slums nobo: old stories) mythology on the one hand and found in Tucupita and Barrancas. ‘deje jiro’ (new stories) or ‘deje kwamotane abane,’ 1164  WARAO which contain recent stories, gossip, and jokes, on the to have one morpheme (smallest meaningful unit, other hand. In addition, there is special language use which can be a word or a part of a word) per grammat- during ritual as in shamanistic curing sessions and ical function. Larger words are then composed of a wailing for the dead (Briggs 1993). basic word or root and these morphemes. In the case of The large majority of Warao speakers live in com- Warao, most of them—suffixes—occur after the root, munities that are exclusively Warao. Outside these although some of them—prefixes—are attached in front communities there is no social space for the usage of of it. For example, the phrase: ‘ine najoro-turu-ae’, the indigenous language. In Guyana, there has been no structurally ‘I foodalmost-would-havecompleted- bilingual education so far (Forte 2000). In Venezuela, action’, equals the English expression ‘I almost ate.’ indigenous languages were authorized for school edu- (Please note that in these examples of Warao, the cation by a presidential decree in 1979, and official Spanish-based official spelling is used and the letter ‘j’ alphabets for most of them including Warao were therefore corresponds to ‘h’ for English speakers. released subsequently. Nevertheless, due to lack of Hyphens are intended to clarify the structure of a word.) political support these regulations were never put to practice on a larger scale (Villalón 1994). Warao lan- Word Classes guage and culture are subject to strong pressure by the Spanish- and English-speaking majority, urging them The categories of adjective and noun are not clear-cut. to assimilate into the national Criollo or Creole culture A word can function as noun or adjective according to of the respective countries. The awareness among the its place in a sentence and the suffixes it combines Warao concerning their language and culture as being with. Nouns and adjectives do not have gender. There endangered is just starting to rise, but still the future of are furthermore ‘noun–verbs’ (Osborn 1967). their whole cultural heritage has to be regarded as Normally verbs and nouns can be distributionally dif- most uncertain. ferentiated according to the suffixes and prefixes (classed together as affixes) that can occur with them. Dialects Some roots, however, can combine with verb and noun affixes alike. To give an example, the root ‘najoro’ There is no reliable information concerning the num- (food) alone is a noun that can combine, for example, ber of dialects. Some authors claim that there are none with the suffix ‘-noko’ (place/instrument) to form (Romero-Figeroa 1997); others have suggested that ‘najoro-noko’ (place of the food/instrument for eat- there are only slight regional variations (Osborn ing). On the other hand, in the construction ‘najoro-ya’ 1966a:108f.). For their part, Warao speakers claim not (literally, fooddurative aspect), ‘is/are eating’, the being able to understand speakers from certain other same root functions as part of a verb combining with a groups. Due to Delta internal migration, the situation verb suffix. has become regionally complex. Nevertheless, at least four major linguistic and cultural variants have been Tense, Mood, and Time assumed by most researchers (Weisshar 1982; Wilbert 1957). Certainly, the different Warao groups are for In Warao a wide range of such notions as tense, aspect, ecological and historical reasons culturally quite het- and mood are marked on the verb. The above-men- erogeneous (Heinen and Garcia-Castro 2000). tioned morphemes accomplish this task mainly suf- fixed to the basic verb form. In order to clarify this Genealogy point, there is a nonexhaustive list of examples given below. However, as there is no consistent labeling in Because of a lack of historical information, it is impos- the literature for these forms, I will limit myself to sible to prove membership in a language family for give an approximate translation into English: Warao and it has therefore to be considered an isolate (Weisshar 1982). Certain resemblances with other lan- najoro-kitane (to eat) guages have to be attributed to the phenomenon of a i-najoro-kitane (to cause someone to eat/to feed ‘linguistic area’, where languages of a certain region someone) (in this case the Amazon and neighboring areas) show najoro-kitia (going to eat immediately) similarities across the boundaries of different language najoro-te (probably going to eat in the future) families. najoro-ya (is/are eating at that moment/will surely be eating in the future) najoro-ae (ate) Typology najoro-ya-ja (eating) Typologically speaking, Warao shows characteristics of najoro-kuna (could eat) an ‘agglutinating’ language. Languages of this type tend najoro-turu-ae (was/were about to eat) 1165 WARAO najoro-mejerei (in order to have him/her eat) ‘ine warao’ (literally, I Warao) would be the equivalent najoro-moana (has/have not yet eaten but should of the English sentence. have done so) The direct complements of a verb (subject, object) najoro-komoni (cannot eat) do not have case marking or gender and are often najoro! (eat!, singular) omitted, especially when they have been mentioned najoro-kotu! (eat!, plural) previously or are assumed to be known. Besides, there najoro-na-ja (is/are not eating) are stylistic reasons for omission. Nouns can take on najoro-na-tan-ae (has/have not eaten) suffixes (-si/-ma/-to) that specify that they are not the najoro-naka! (don’t eat!, singular) subject of a sentence: dima-si konaria (he takes father najoro-naka ja-kotu! (don’t eat!, plural) along) as opposed to dima konaria (father takes some- thing/someone along). In this last example, ‘jakitane’ (to be/to have) func- A study of the ‘pivot’ in Warao is yet lacking. In tions as an auxiliary verb taking on the imperative plu- languages with nominative and accusative case mark- ral ending ‘-kotu’ because it is impossible to attach it ing, the nominative (case of the subject of an intransi- to the preceding suffix ‘-naka’. tive and a transitive sentence) functions as the pivot. Taken together, ‘jakitane’ with ‘tikitane’ (to say/to Thus, two sentences are connected having a common do) and ‘takitane’ (to be/to do) form an interesting subject: ‘the man ate an apple and smiled’. In ergative group. They can not only be used as auxiliaries and as languages, however, which show the same case for the suffixes but also as clause introducers. As in ‘ta-kore’ subject of an intransitive sentence and for the object of (literally, this being so), then/when. a transitive one, this juncture would be impossible. Analyzing ‘najoro-na-ta-n-ae’ (has/have not eaten) ‘The man ate an apple and smelled good’ would hence from the above list, ‘ta-n-ae’ really is the root of the mean that the apple smelled good, not the man. In verb ‘ta-kitane’, followed by a marker for singularity Warao storytelling, such connections are frequently of action ‘-n-’, followed by the suffix expressing com- made. This gives rise to the question of whether this pleteness of an action: ‘foodnotdosingularity of language is syntactically ergative. actioncompleted action’. Person, Number, and Plurality of Action Discourse Markers Basically, there is no agreement for person and number In addition, there are morphemes that allow the modi- on the verb, although interestingly singularity or plu- fication of verbs, nouns, and adjectives and sometimes rality of the action itself can be marked and thus range over the whole sentence: So, for instance, ‘- emphasized. Compare, for example, ‘boro-te’ (will be yama’ is a citation marker specifying that something is jumping) with ‘boro-bu-te’ (will be jumping repeated- known from hearsay: najoro-ae-yama (it is said/some- ly/a lot of people will be jumping). Here the morpheme one told that he ate). It is prevalent in storytelling. ‘-bu-’ underlines that the action is performed repeated- Other morphemes like ‘-kore’ or ‘-rone’, which ly either by one person or simultaneously by many. mean ‘while/if’ and ‘although’, respectively, play the The case of a morpheme that depicts singularity role of conjunctions connecting sentences or phrases: (‘-n-’) is more complicated as some verbs exist both nojoro-ya-kore dani mi-ae (while I was eating I saw with and without this morpheme and others only occur my mother); najoro-ya-rone dani mi-ae (although I in one of these forms. An example of alternating verbs was eating I saw my mother). For questioning a sen- would be ine oa-e (I grabbed several things) in oppo- tence, the suffix ‘-ra’ is attached to the last word of sition to ‘ine oa-n-ae’ (I grabbed once/one thing). It is that sentence. not clear whether or not ‘-n-’ is still a productive mor- pheme. Sentence Structure With respect to number marking on nouns, there is a suffix ‘-tuma’ that is often regarded as plurality As regards the ‘basic’ or unmarked word order in marker. But it is not obligatory and, especially when Warao, there is no general agreement. Some take a used with people, rather expresses the idea of ‘the- Subject–Object–Verb (SOV) word order to be basic ones- belonging-to’ as in ‘Maria-tuma’ (Maria and her (Osborn 1966b) while others argue in favor of friends/family). Object–Subject–Verb (OSV) (Romero-Figeroa 1997). All that can be stated safely is that Warao is a ‘verb- Possession and Article final’ language. In a number of languages, the copula ‘to be’ (as in The possessive construction has the form ‘possessor an English sentence of the type: ‘I am Warao’) is not possession-markerpossessed-item’: dima a-janoko obligatory. The same holds true for Warao. Thus, (father’s house, literally ‘father his-house’). 1166  WAYAMPI AND TUPI-GUARANI LANGUAGES There is no definite or indefinite article in Warao. But Osborn, Henry. 1966a. Warao I: phonology and morphophone- person markers used in possessive constructions func- mics. IJAL 32. ———. 1966b. Warao II: Nouns, relationals, and demonstra- tion like a definite article. In the following text record- tives. IJAL 32. ed by the author, definiteness is assured in this way: ———. 1967. Warao III: Verbs and suffixes. IJAL 33. Romero-Figeroa, Andrés. 1997. A reference grammar of Warao. Tau-tuma manamo ja. Ta-kore a-raiba a-rajia tane. München, Newcastle: Lincom. Two womenfolk is/has. Then her older sister her Vaquero, P. Antonio. 1965. Idioma Warao. Morfología, sintaxis, literatura (Warao language. Morphology, syntax, literature). younger sister so (literal translation). Caracas: Estudios Venezolanos Indígenas. There were two women: an older and a younger sis- Venezuela. 1993. Oficina Central de Estadística e Informática, ter (free translation). Censo Indígena de Venezuela 1992, Tomo I (Venezuelan Indigenous Census of 1992, Part I). Caracas: OCEI. Villalón, María E. 1994. Educación para indígenas en Future of the Language Venezuela: Una crítica razonada (Education for indigenous Clearly a most interesting language with phenomena people in Venezuela: a reasoned critique). Caracas: CEVIAP. Weisshar, Emmerich. 1982. Die Stellung des Warao und awaiting to be investigated, Warao presents a challenge Yanomama in Beziehung zu den indigenen Sprachen to linguists. Unfortunately, like most amerindian lan- Südamerikas nördlich des Amazonas: Studien zur genetis- guages the future vitality of Warao is not assured. chen und areal-typologischen Klassifikation (The position of Warao and Yanomama in relation to the indigenous lan- guages of South America North of the Amazon: study in References genetic and areal-typological classification). University of Briggs, Charles L. 1993. The pattering of variation in perform- Tübingen: Dissertation. ance. American dialect research, ed. by Dennis R. Preston, Wilbert, Johannes. 1957. Prologo. Diccionario Guarao- 379–431. Amsterdam: Benjamins. Español, Español-Guarao, ed. by Sucre, 7–18. Basilio María Forte, Janette. 2000. Amerindian languages of Guyana. As lín- de Barral. Caracas: Sociedad de Ciencias Naturales la Salle. guas amazonicas hoje, ed. by F. Queixalós and O. Renault- ———. 1970. Folk literature of the Warao Indians. Narrative Lescure. Sao Paulo: Instituto Socioambiental. material and motive content. Los Angeles: Latin American Heinen, Dieter. 1988. Oko Warao: marshland people of the Center University of California. Orinoco-Delta. Münster: Lit. Wilbert, Werner. 1995. Conceptos etnoecológicos Warao. Heinen, Dieter, and Alvaro Garcia-Castro. 2000. The multieth- Naturaleza y Ecología Humana en el Neotrópoco/Nature nic network of the lower Orinoco. Ethnohistory 47. 561–79. and human ecology in the neotropics (Scientia Guaianæ. Lavandero, Julio Perez (ed.) 1991. I Ajotejana, Mitos (Myths No. 5), ed. by Dieter Heinen, José J. San-José and Hortensia from Ajotejana). Caracas: Hermanos Capuchinos. Caballero-Arias, 312–70. Caracas: Instituto Venezolano de ———. 1992. II Ajotejana, Relatos (stories from Ajotejana). Investigaciones Científicas. Caracas: Hermanos Capuchinos. STEFANIE HERRMANN Wayampi and Tupi-Guarani Languages Wayampi is one of the more than 40 closely related lan- was no longer mutually intelligible. There are no writ- guages that comprise the Tupi-Guarani language fami- ten records for this early history. ly. These languages descend from a common ancestor, The first contacts that Europeans had with Tupi- referred to as Proto-Tupi-Guarani, whose speakers Guarani languages were with Tupinambá and Guarani, probably lived in or around the state of Rondonia in at that time spoken along the eastern coast of South western Brazil. Long before the arrival of Europeans to America. As explorers, anthropologists, and linguists the New World, extensive migration of small groups came in contact with other indigenous groups whose contributed toward the development of a large number languages showed a high degree of similarity in vocabu- of closely related languages, scattered throughout the lary and word structure to Tupinambá and Guarani, they Amazon region and beyond. As individual groups referred to these languages as Tupi-Guarani languages. migrated, many of them through the Amazon rain for- The Tupi-Guarani family is one branch of a larger est, they eventually lost contact with each other. The grouping of Tupi languages. Tupi-Guarani languages language spoken by each group gradually changed in have a number of features that define them as a group, unique ways, until the speech of the individual groups separate from the other Tupi languages. For example, 1167 WAYAMPI AND TUPI-GUARANI LANGUAGES they have a set of free personal pronouns that are dialect, there is only one back nasalized vowel, and obvious cognates (having been derived from a single younger speakers are merging the central nasalized source); other Tupi languages do not have this set. vowels as well. Tupi-Guarani languages also share a set of verbal pre- Grammatically, Wayampi has a number of features fixes which refer to the subject: a- ‘I’, oro- ‘we exclu- that are different from Indo-European languages but sive (he/she and I)’, ja- ‘we inclusive (you and I)’, are quite common to Tupi-Guarani languages and, in ere- ‘you (sg)’, pe- ‘you (pl)’, and o- ‘he/she/it/they’. fact, to Amazonian languages in general. Like other They also share a number of basic vocabulary items Tupi-Guarani languages, Wayampi is agglutinative, that are different from forms in other Tupi languages. which means that words can take multiple prefixes and suffixes. This is most visible in verbs. In the following example from the Cuc dialect, the verb stem (kusu Wayampi ‘wash’) is preceded by an incorporated object (po Wayampi (with variant spellings Guayapi, Waiãpi, ‘hand’), which is in turn preceded by three prefixes. ~ Wayãpi, Wajapi, Oiampi, Oyampi, and Oyampik) is the There are also two suffixes following the verb stem: autodenomination of a language group of approximate- n-o-i-po-kusu- t  -i ly 1,800 speakers, living on the two sides of the not-he-self-hand-wash-never-not Brazil–French Guiana border. They are one of the three ‘He never washes his (own) hands.’ (literally, ‘he Tupi-Guarani language groups residing on the north never hand-washes himself’) side of the Amazon River. Originally residing near the mouth of the lower Typologically, one of the areas of greatest interest to Xingu River, the Wayampi migrated north across the linguists studying Wayampi and related languages is Amazon and up the Jari River in the mid-1700s. During the system of person markers which occur on verbs. the first half of the twentieth century, there were four Transitive verbs, those which express an action on an dialect groups of Wayampi: one on the Oiapoque and object, are prefixed by person markers from one of two Camopi rivers in French Guiana, another on the Cuc sets, one referring to the subject and the other referring River in the Brazilian state of Amapá, a third group liv- to the object. For example, there are two prefixes refer- ing east of the Cuc River in the area that they call ring to first-person singular: a- ‘I’ (subject) and e- Kumakary [kumakar], and a fourth group living farther ‘me’(object), as in the following examples: a-pota ‘I south, on small tributaries of the Amapari and Nipuku like (him)’ and e-pota ‘(he) likes me’. There are com- rivers. plicated rules governing the choice of sets, but, in gen- The group on the Cuc River had ongoing contact eral, first person (‘I/we’), whether subject or object, is with the groups in both Oiapoque and Kumakary. considered the most important and must be referred to Around the middle of the twentieth century, the on the verb, as in the examples above. Second person Amapari group came in contact with the group from (‘you’) is second in priority, and third person Kumakary, and the two groups eventually merged. (‘he/she/it/they’) has the lowest priority. This is one Through these contacts the people from the Amapari type of ‘split-ergative’ system. The ‘subject’ prefix is area learned of the group on the Cuc River, with whom also used for intransitive verbs, and the ‘object’ prefix they began to develop a trading relationship. However, is used for a set of stative verbs: a-poreŋeta ‘I speak’ this relationship was interrupted around 1970 when a and e-kara ‘I have a fever’. This type of system is char- dysentery epidemic in the Cuc River area took the acteristic of Tupi-Guarani languages. Variations of split lives of a large number of people. After that, most of ergativity occur in other Amazonian languages as well. the survivors moved to French Guiana, where they When two verbs are perceived by the speaker to be merged with the Oiapoque group. part of a single action, with the same subject, they In spite of their relative isolation from the outside occur together in a single phrase as ‘serial verbs’. If world, the Wayampi were visited by French explorers in the second verb is transitive, as in the second sentence the late nineteenth century, among them Henri Coudreau below, it must take an object prefix rather than a sub- (1892), who published an impressive word list. ject prefix. (In third person, there is no grammatical The general phonological pattern for Wayampi is distinction between masculine, feminine, and neuter.) consonants alternating with vowels. The consonant set o-o o-jau includes p,t,k,kw,ʔ,m,n,ŋ, ŋw,r,s,h, and, in the Amapari he-go he-bathe dialect, β. There are two semivowels, w and j. The set ‘He went to bathe.’ of vowels includes i,e,,a,u,o. There is an oral set and a nasalized set of vowels. When nasalization occurs, it o-nupã pira i-juka affects the entire morpheme and not just a single he(subj)-beat fish it(obj)-kill ~ ~ vowel, such as apasi [ãpãsi] ‘(to) tie’. In the Amapari ‘He beat the fish and killed it.’ 1168  WAYAMPI AND TUPI-GUARANI LANGUAGES Unlike other Tupi-Guarani languages, Wayampi prefixing on the verb or is simply understood from does not take a special suffix to mark the second context. In the sentences above, it is obvious that ture serial verb. ‘flute’ is an object, not a subject, since the verb (it Noun phrases are short and are generally limited to ‘quit’) requires an animate subject. one other constituent besides the noun. They follow Very little has been published on the discourse of one of the following patterns: other Tupi-Guarani languages, so it is not possible to say to what degree Wayampi discourse is characteristic Demonstrative-Noun: of the family as a whole. However, this type of dis- aʔe kuimaʔe ‘that man’ course system is not unusual among American indige- that man nous languages. Number-Noun: moap kuimaʔe ‘three men’ three man Other Tupi-Guarani Languages Noun-Relative Clause: The Tupi-Guarani family and member languages kuimaʔe i-tekorã maʔe~ ‘(the) man that is sick’ have been studied extensively from the perspectives man him-sick RELATIVIZER of historical linguistics and syntactic typology. Genitive-Noun: Reconstructions have been done of Proto-Tupi- papa po ‘father’s hand’ Guarani grammar and sound system. Using the criteria father hand of a shared history of phonological changes from the Noun-Postposition: reconstructed proto-language, Aryon Rodrigues has a pupe ‘in the canoe’ divided the Tupi-Guarani languages into eight sub- canoe in groups. Pronominal possessors are indicated by the ‘object’ Languages in subgroup 1 are Guaranian languages, prefixes: e-po ‘my hand’, ne-po ‘your (sg) hand’. The spoken in southern Brazil, Paraguay, Bolivia, and same is true for the pronominal object of a postposi- northern Argentina. One form of Guarani, now tion: e-pupe ‘inside me’, i-pupe ‘inside it’. referred to as Old Guarani, has been known from the Wayampi is what is referred to as a ‘flexible word- time of the Spanish explorers who first came to the order’ language. Unlike English, which generally area south of São Paulo. This language was docu- puts the subject before the verb and the object after the mented and described by the Jesuit priest Antonio verb, Wayampi uses other criteria to determine what Ruiz de Montoya (1639). In time, Guarani came to be information goes before the verb and what goes after. used extensively among colonists. Gradually, a no For example, new information is introduced before the indigenous variety of the language developed, which verb, but when it is referred to again, it occurs after the is now spoken by 95% of the population of Paraguay. verb. In the following example, the participants of an Languages in subgroup 2 are spoken in Bolivia: all-night festival announce their intentions: Sirionó and Guarayu. Languages in subgroup 3 were spoken along the Atlantic coast and up the Amazon jeʔiwe ture oro-it-ta, river. The most well-known language from this group early.morning flute we(exclusive)-quit-future is Tupinambá, which was documented extensively by eʔi kupa another Jesuit priest, Joseph de Anchieta (1595). said plural.subject Although now extinct, Tupinambá was used extensive- ‘‘Early in the morning we will quit playing flutes’, ly for many years among colonists. Finally, a royal they said.’ decree in the eighteenth century dictated that Portuguese citizens use Portuguese rather than an Both the time and the object (flute) are new infor- indigenous language as their means of communica- mation and come before the verb (underlined). Once tion. Even so, the nonindigenous variety of introduced, they are subsequently referred in postver- Tupinambá, now called Nheengatu, is still spoken bal position, as in the following example: today as the mother tongue of a nonindigenous popu- lation of about 3,000, in the upper Amazon region. o-it jeʔiwe ture kupa Brazilian Portuguese has a large number of words bor- they-quit early.morning flute plural.subject rowed from Tupinambá, especially place names and ‘They quit playing the flutes early in the morning.’ terms for flora and fauna native to Brazil. Even In spite of the flexible word order, there is no gram- English has borrowed words from Tupinambá: jaguar, matical case marking on nouns to distinguish subjects toucan, piranha, tapioca, and manioc. from objects. Often only one noun is actually specified Speakers of languages of subgroup 4 reside at the in the sentence and the other is indicated only by the eastern edge of the Amazon rain forest and beyond, in 1169 WAYAMPI AND TUPI-GUARANI LANGUAGES the Brazilian states of Tocantins, Maranhão, and Pará. References They include Asuriní of Tocantins, Parakanã, Tapirapé, Anchieta, Joseph de. 1595. Arte de gramatica da língua mais Guajajára. usada na costa do Brasil. Coimbra: Antonio de Mariz. Languages of subgroup 5 are west of subgroup 4, in Coudreau, Henri. 1892. Vocabulaires méthodiques des langues Pará and northern Mato Grosso. They include Asuriní Ouyana, Aparai, Oyampi, Emerillon. Paris: Bibliothèque of Xingu and Kayabí. Linguistique Américaine. Derbyshire, Desmond C., and Geoffrey K. Pullum. 1986 and Still further west, in Amazonas, Rondônia, and Acre, 1998. Handbook of Amazonian languages, vols. 1 and 4. are the speakers of subgroup 6: particularly the Berlin: Mouton de Gruyter. Kawahíb dialect cluster. These groups live closest to the Jensen, Allen. 1994. Wayampi. Typological studies in negation, area where their Proto-Tupi-Guarani ancestors probably ed. by Peter Kahrel and René van den Berg. Amsterdam/ lived. Philadelphia: John Benjamins. Jensen, Cheryl. 1999. Tupi-Guarani. The Amazonian lan- Subgroup 7 includes just one language: Kamayurá, guages, ed. by R.M.W. Dixon and Alexandra Y. Aikhenvald. located in the Xingu Park in Mato Grosso. Cambridge: Cambridge University Press. Subgroup 8, of which Wayampi is a member, Navarro, Eduardo de Almeida. 1998. Método Moderno de Tupi includes the three language groups located on the north Antigo: a língua do Brasil dos primeiros séculos. Petrópolis: side of the Amazon, who live in southern French Guiana Editora Vozes. Rodrigues, Aryon Dall’Igna. 1986. Línguas brasileiras: Para o con- or northern Brazil. It also includes several language hecimento das línguas indígenas. São Paulo: Edições Loyola. groups who live on the south side of the Amazon in the Ruiz de Montoya, Antonio. 1639. Tesoro de la lengua Guaraní. states of Pará or Maranhão, among them Guajá and Madrid: Juan Sanchez. Urubú-Kaapor. CHERYL JENSEN Weinreich, Uriel An authority in a number of disciplines, Uriel work together toward a better understanding of lan- Weinreich made significant contributions to several guage. Weinreich also worked on a model of language aspects of linguistics and was particularly interested in change that would be completed and published those facets dealing with the interaction of language and posthumously by his students William Labov and society. His specialties included but were not limited to Marvin Herzog in an article entitled ‘Empirical foun- language change, semantic theory, and the Yiddish lan- dations for a theory of language change’ (1967). In this guage. Born in Lithuania, Weinreich moved to the work, Weinreich et al. argue that a model of language United States in his early teens, served in the US Army that ‘accommodates the facts of variable usage and its during World War II, was naturalized as an American social and stylistic determinants not only leads to more citizen soon after, and worked for a brief time with the adequate descriptions of linguistic competence, but State Department. As a scholar, Weinreich made his also naturally yields a theory of language change that home at Columbia University, where he received his bypasses the fruitless paradoxes with which historical undergraduate and graduate degrees and later taught lin- linguistics has been struggling for over a half century’ guistics and Yiddish until his death at the age of 40. (Weinreich, Labov and Herzog, ‘Empirical founda- One of Weinreich’s major contributions to linguis- tions for a theory of language change’, 1967:99). tics was his work on language change and variation. As part of their proposal, Weinreich et al. downplay Under the guidance of his advisor Andre Martinet, the role of the idiolect in linguistic change in favor of Weinreich wrote both his master’s thesis and his dis- an explanation that takes into account the interrela- sertation on the role of language contact in linguistic tionship of linguistic and social factors. In this respect, change, and this work later served as the foundation Weinreich’s ideas had a profound effect on Labov, for for his book Languages in contact (1953). In his arti- whom Weinreich served as thesis and dissertation cle ‘Is a structural dialectology possible?’ (1954), director, in his pioneering work in the emerging field Weinreich called for scholars working in the seeming- of sociolinguistics. ly adverse frameworks of structural linguistics and Weinreich also made a significant contribution to dialectology to find a common ground on which to the study of semantics. Frustrated with a doctrine of 1170  WEINREICH, URIEL linguistics that chose to ignore the issue of semantics in Israel, France, the United States, and Canada, with a because of the difficulty it created, Weinreich saw questionnaire comprising more than 3,000 questions promise in the generative grammar espoused by and taking some 15 hours to complete. This work Chomsky (1957, 1965) because he believed that this served as the foundation for a comparative study of formal approach could be used to account for mean- Jewish culture throughout Europe before World War II ing, despite Chomsky’s avoidance of semantics in his that culminated in the Language and cultural atlas of own work. Weinreich was, however, critical of many Ashkenazic Jewry, which Uriel directed until just of the approaches to semantics that were in vogue dur- before his death. Now under the direction of Herzog, ing his career. For instance, he asserted that ‘systemat- the first volume of the project was published in 1991. ic semantics is doomed to an isolated place within The Modern English–Yiddish and Yiddish–English dic- linguistics as a whole so long as paradigmatic relations tionary, which Weinreich compiled by applying his are the only form of patterning to which attention is theory of systematic lexicography, was published given. It is undoubtedly important to understand how posthumously in 1968. To honor the man and his work, the meaning of a word in a vocabulary is determined the Uriel Weinreich Program in Yiddish Language, by other words in the same vocabulary; however, an Literature, and Culture was established as a summer account must still be given of the way in which the program in 1968, cosponsored by Columbia University meaning of a sentence is composed out of the mean- and the YIVO Institute for Jewish Research. ings of individual words’ (Weinreich, Explorations in semantic theory, 1972:112). Biography Weinreich also took exception to the theoretical assumption of many working in semantics that seman- Uriel Weinreich was born in 1926 in Vilnius (Russian tics begins where syntax ends; instead, Weinreich Vilna), Lithuania. He moved to the United States with believed that syntax and semantics are so intertwined his family in 1940, when his father, Yiddish scholar that only a theory taking into account the interrelation Max Weinreich, joined the faculty of New York City between them could adequately represent the complex- College. He served as first lieutenant in the US Army ity of language. He pointed out that although genera- from 1944 to 1946 and was naturalized as an tivists postulated syntax and semantics to be separate American citizen in 1945. After his term of service, domains, they had never succeeded in discovering the Weinreich enrolled in Columbia University and boundary between the two. Weinreich was particularly received his BA in 1948, his MA in 1949, and his PhD critical of the Katz–Fodor model of generative seman- in 1951, completing a thesis and dissertation under tics (see Weinreich 1972) on the grounds that their Andre Martinet, work which would eventually culmi- analysis focused on a limited part of semantic compe- nate in his book Languages in contact (1953). He per- tence and concentrated too much on ambiguity. formed fieldwork in Switzerland from 1949 to 1950 Weinreich argued that ‘grammatical theory is not under a research fellowship from the American required to explain how a hearer of such ambiguous Council of Learned Societies. He worked as an editor expressions guesses which of two deep structures is rep- and information specialist with the State Department resented by a given occurrence of a surface structure, from 1951 to 1952, and was Chair of the Department nor is the goal of grammatical theory limited to the cal- of Linguistics at Columbia University. Weinreich culation of such ambiguities’ (Weinreich, Explorations became the first Atran Professor of Yiddish Language, in semantic theory, 1972:18–19). Weinreich was also Literature and Culture at Columbia University in critical of the Katz–Fodor model because he said that it 1952, and was Member, Linguistic Society of America ignored semantic deviations, instead of trying to in 1951, serving as vice president in 1964, and co-edi- account for them. tor of Word from 1953 to 1960. He began work on the Like his father, Yiddish scholar Max Weinreich—a Language and cultural atlas of Ashkenazic Jewry in driving force behind the YIVO Institute for Jewish 1959 and served as its director until his demise. Research and a proponent of the claim that Yiddish was Weinreich continued teaching until two days before a unique creation of Jews and not simply derived from cancer took his life on March 30, 1967. German—Uriel Weinreich was a scholar of Yiddish and worked hard for the preservation of the language, References collecting a great number of Yiddishisms for various books on the subject. His textbook on the subject, Chomsky, Noam. 1957. Syntactic structures. The Hague: College Yiddish, was first published in 1949 and has Mouton. ———. 1965. Aspects of the theory of syntax. Cambridge: since been published in numerous editions. In 1959, MIT Press. Weinreich and his wife, the folklorist Beatrice Labov, William. 1994. Principles of linguistic change, Vol. 1: Silverman Weinreich, began interviewing Jews living internal factors. Oxford: Blackwell. 1171 WEINREICH, URIEL Malkiel, Yakov. 1967. Uriel Weinreich (1926–1967). Language ———. 1980. On semantics, ed. by William Labov and 43. 605–7. Beatrice S. Weinreich. Philadelphia, PA: University of Weinreich, Uriel. 1949. College Yiddish: an introduction to the Pennsylvania Press. Yiddish language and to Jewish life and culture. New York: Weinreich Uriel, William Labov, and Marvin I. Herzog. 1963. YIVO Institute for Jewish Research. On the semantic structure of language. Universals of lan- ———. 1953. Languages in contact: findings and problems. guage, ed. by Joseph H. Greenberg. Cambridge, MA: MIT New York: Linguistic Circle of New York. Press. Empirical foundations for a theory of language ———. 1954. Is a structural dialectology possible?. Word 10. change. Directions for historical linguistics, ed. by Winfred 388–400. P. Lehmann and Yakov Malkiel. Austin: University of Texas ———. 1968. Modern English-Yiddish, Yiddish-English dic- Press. 1968. tionary. New York: YIVO Institute for Jewish Research. LAMONT ANTIEAU ———. 1972. Explorations in semantic theory. The Hague: Mouton. See also Labov, William; Martinet, Andrè Western Caribbean Creole(s) Western Caribbean Creole (WCC, hereafter) belongs gua franca of the majority of residents. In Jamaica, for to the western branch of Atlantic, English-based instance, the official language is English, but the use Creoles. It has features common to the broader lin- of WCC is more widespread. It, therefore, functions guistic designation, West Indian Creole (Winford like the quasi-official language. 1985; Grimes 2001; Roberts 1996). Varieties of WCC Despite its frequent use, WCC continues to be stig- developed in predominantly English-speaking western matized, particularly by its educated speakers (Le Caribbean nations, including Jamaica, Belize, and Page 1966; Nettleford 1966; Allsopp 1996; Spears and Barbados. Believed by some linguists (Hancock 1980; Winford 1997), many of whom use this code as well. Niles, 1980) to be a dialect of English, Barbados WCC may have started out as a pidgin or, in Creole, popularly known as Bajan or ‘Barbadian non- Romaine’s (1994) opinion, as a jargon or a short-lived standard English’ (Fields 1995), ‘has the fewest Creole pidgin. Instances of the earlier form are found in features’ (Roberts 1996:87), and its Creole status is anglophone Caribbean work-songs, and in dancing somewhat controversial. It differs from other varieties and folk songs from 1793 to 1907 such as the anony- of West Indian Creole primarily phonologically mous ‘Freedom a Come Oh!’ (Freedom, Here I (Grimes 2001). Researchers such as Cassidy (1980) Come!, where the ‘Oh’ might very well be a remnant and Fields (1995) maintain that early Bajan was a of the Kwa honorific particle ‘o(h)’ used extensively in Creole. Fields (1995) makes a case for its Creole Yoruba and Nigerian Pidgin), ‘Negro Song at ancestry, arguing that contemporary Bajan is simply a Cornwall’, and ‘Sangaree Kill de Captain’ (Sangaree Creole-in-transition. While Hancock (1980) and Niles Killed the Captain) (see Burnett 1986:3–12). (1980) agree that early Bajan was probably a Creole, Sprinklings of WCC can also be found in veteran they differ in their accounts of precisely when Jamaican writer, Roger Mais’s (1905–1954) The hills Barbados Creole existed. Cassidy (1986) contends that were joyful together, and in Vic S. Reid’s works. WCC Barbados Creole could only have developed after is presently the native or first language of a large per- 1650, when conditions were favorable for its genesis. centage of users; hence the descriptive term, Creole. The largest number of speakers of WCC is found in Creolization is the term used to describe the forma- Jamaica, followed by Belize. In Jamaica, this language tive process it is believed to have undergone. WCC and is known as Jamaican, Jamaican Creole (English), or its eastern Caribbean and West African equivalents, Bongo Talk [Bongo is an Afro-Caribbean dance]. such as Guyanese (see Bickerton 1973; Rickford Linguists tend to use the label West(ern) Caribbean 1987), Krio, or Nigerian Pidgin (Faraclas 1996), Creole. The variety spoken in Belize—slightly differ- respectively, exhibit a Creole continuum, or a range of ent—is referred to as Creola, Kriol, or Belize Creole varieties. At one end is the basilect, a variety with min- (Boland 1986, 1991; Escure 1981, 1997; Greene imal structure and associated functions, and at the other 1999). So frequent is the use of WCC in these coun- end lies the acrolect, a fully expanded variety, used, tries that it is the primary language and preferred lin- like any primary language, for all possible language 1172  WESTERN CARIBBEAN CREOLE(S) functions. The mesolect constitutes the mid-range. It is that constitute the sociolinguistic mosaic of this region more compact than the acrolect, but serves a greater (Burnett 1986:xxv). In the words of well-known range of functions than the basilect. The different lects Jamaican poet and storyteller, Louise Bennett: tend to be socially stratified (Pandey 1997). Mesolectal WCC is a combination of English and Creole words. The Jamaican patois has obviously Standard English as its roots but has been spiced with the tongues of the The acrolect tends to contain few Creole words. Arawak and Carib Indians (the original inhabitants; Examples include: Africans brought over as slaves; Spaniards, the first con- (1) Acrolect: ‘Don’t lie to yuself’, ‘Dis is crazy’ querors; the migrating French, Chinese and East Indians; (Patterson 1964:16, 186). the Colonizing British; and in recent times, visitors from (2) Mesolect: ‘No matter if he is Jesus or him is North America and the Rastafarian religion (1981:i). Jah. Him not gwan like dis one lickle bit’ (Cliff In many ways, then, it is a linguistic chutney of 1996:16). sorts, much like chutney music, a variety of Soka [Translation: It doesn’t matter if he is Jesus or Jah music composed in Trinidad. Bennett draws heavily (Haile Selassie). He will not like this even a little.] on WCC in her works. Exemplary poems include, among others, ‘Jamaica Patois’, ‘Jamaica Obeah’ (3) Basilect: ‘God nuh mus’ be Hinglish. But me (Jamaican Juju Man), ‘Weh Dem Deh?’ (Where are did’ ‘ear once dat Jesus did’ ave bad ‘air. Mus’ They?), ‘Solja Bwoys’ (Soldier Boys), ‘White be one joke ‘pon we’ (Cliff 1996:17–18). Pickney’, ‘Show Yuh Foot’ (Show Your Foot), ‘Him de [Translation: It seems God must be English then. Yah’ (He is Jesus), ‘Me Bredda’ (My Brother), ‘New But once I heard that Jesus had bad hair (reference to Govanah’, ‘Train Leff Miss Hayes’, ‘Pass fe White’, him as African). It must be a joke being played on us.] and ‘De Royal Commotion’. The speech of the poor and less educated Barbadian One of the most widely studied Caribbean Creoles, is very similar to the mesolect found in neighboring the variety used in Jamaica has the longest-standing anglophone nations, leading linguists like Alleyne history of use in Caribbean literature, including folk- (1979) and Fields (1995) to contend that it is in the lore (see McKay 1933; Bennett 1966; Burnett 1986; mesolectal stages. Mordecai and Morris 1980; Senior 1989; Cliff 1996; The term creole continuum was first used by cre- Donnell and Welsh 1996; Patterson 1998; MacCallum olist Derek Bickerton (1981), and has been applied to 1999, among others), music, and other artistic media. a variety of Pacific and Atlantic Creole contexts (colo- The international attention and recognition Jamaican nized coastal territories). It is also termed a post-cre- music, dub poetry and music (see Allen 1990; Breeze ole continuum (Rickford 1987), as it mirrors a process 1991, 1997; Smith 1979), and Rastafarianism have of decreolization or progressive assimilation toward a attracted have helped spread WCC to other parts of the standard language. Bajan might be one such example Caribbean, and have helped familiarize others outside (Cassidy 1980, 1986; Rickford 1992; Rickford and the region with its unique rhythms, sounds, and struc- Handler 1994; Fields 1995). In Cassidy’s (1980) view, ture. Much like rap is to the African American com- Bajan decreolized earlier than other Caribbean munity in the United States, dub poetry and music are Creoles. Decreolization is not necessarily chronologi- signature forms of Black folk-culture in Jamaica and cal. In other words, it is more often the case that the the Caribbean as a whole. different lects coexist, rather than in isolation, chang- Dub poetry refers to ‘Lines meant to be spoken, gen- ing over time in lectal designation. erally to a two-beat rhythm, and dealing mostly with Jamaican Creole is viewed as a paradigm example of the life experiences and/or point of view of Black peo- WCC (Lalla and D’Costa 1990; Mufwene 1993:142). ple in or from the Caribbean’ (Allsopp 206). Dub Many Jamaicans (Bennett 1966, 1981; MacCallum music, derived from Reggae, is also a two-beat rhythm 1999; Donnell and Welsh 1996) refer to it as Patois, the created principally by bass and drums. Both art forms French term for a pidgin. The presence of large numbers employ principally WCC lyrics. When one speaks of of French migrants, and the island’s proximity to Creole, it is no wonder, then, that the layperson gener- French-colonized Caribbean nations might have some- ally thinks of Jamaica(n). The increasing use of ‘I’, a thing to do with this label. The anglicized Patwa is also symbol of self-respect and group solidarity, in the used by some. Like all varieties of WCC, Jamaican Jamaican variety of WCC (see Pollard 1986), as a pop- Creole is English-based, which means that the bulk of ular substitute for the Creole ‘mi’, and in words like ‘I- its vocabulary is drawn from English, while its gram- lect’ (Rasta dialect or Rasta Talk), ‘I-quality’ (equality), matical and discourse structure mirror the influence of ‘Idren’ (brethen), and ‘I’re’ (alright) can be attributed to other (non-European) languages. It contains ingredients Rastafarianism, and are a reminder of the Rastas’s rev- from the different colonizing and Diaspora languages erence for Ras Tafari (Emperor Haile Selassie). 1173 WESTERN CARIBBEAN CREOLE(S) Exemplary proverbs in WCC include ‘Cockroach African languages. According to one view, for no business in fowl-yard’ (i.e. Mind your own busi- instance, ‘Created in West Africa, ... work pidgins ness), ‘De higher monkey clim, de mo im expose’ (i.e. were transported across oceans where they took on promotion has its price—pretentiousness), and ‘Yu lay new roles as lingua francas among the enslaved....’ wid dawg, yu get wid flea’ (i.e. bad company cor- (McWhorter 1997: 240). Dalphinis (1985) observes rupts). WCC is so important a part of Belizean and that ‘[B]oth in the analysis of the Creole languages of Jamaican culture that visitors to these countries are the Caribbean as well as Caribbean culture, it is clear advised to familiarize themselves with the distinctive that Europe has affected Africa but equally Africa has features of WCC (see Porter and Prince 2000; affected Europe; the words are European, but the syn- Brosnahan and Keller 1997). As regards Jamaica, tax is African’ (95). Porter and Prince (2000) offer the following overview Representative features of WCC include (see and advice: Winford 1985, 1993; Roy 1986; Escure 1978, 1983; The unofficial language is a patois.... This archaic and Patrick 1993; LaCharité 1999, among others): simplified structure, coupled with African accents and special intonation, can make the language difficult to (1) The zero copula or ‘be’ verb before adjectives, understand. Some of the most interesting anecdotes and as in ‘Di pikni sik’ (The child is sick) or ‘Di fables are usually told in the patois, so understanding its chile/chilin sik’ and ‘im hongry’ (She or he’s structure can add to your insight into Jamaican cul- hungry) in Kriol, ‘im [Bradda Anancy] always ture.... Proverbs and place-names [e.g. Red Gal Ring] foolin every oda creature’ (MacCallum express some of the vitality [of patois].... The patois has 1999:19), and ‘We making history’ (Cliff been embellished and altered with the growth of 1996:5). Rastafarianism. (196) (2) Nonmarking of the suffix in the third-person In their focus on Belize, Brosnahan and Keller singular present, as in ‘I laik it’ (Kriol for ‘She (1997) describe WCC as Belize’s ‘own colorful Creole or He likes it’), and the use of nonfinite verbs dialect’ (276). Bolland (1977) describes it as ‘a means to convey multiple meanings, e.g. ‘Dat sopoz of expression through which proverbs, sayings, and to daan’ (Kriol for ‘That’s supposed to have folktales’ convey ‘African values and wisdom’ (12), been done’) and ‘De chilin jos ron’ (Kriol for and as a code language used to satirize the British. ‘The children just keep running around’). In Belize, the Kriol Projek, a national organization, (3) Spelling pronunciations; words are pro- continues to play a seminal role in educating citizens nounced the way they are spelled, and vice about the culture and creativity inherent in Kriol. This versa, e.g. ‘dat’ for ‘that’, ‘de’ for ‘the’, ‘dis’ organization, and the present Governor General have for ‘this’, and the use of ‘Dem’ for many compiled and printed folktales in Kriol for use in English words, including plural marker, as in (Belize’s) schools. As few Belizeans are used to seeing (15); and ‘they’, ‘them’, and ‘their(s)’ as in Kriol in writing, these publications are accompanied ‘Mi a sel dem az mi laik’ (Bailey 1966:57), by CDs, so that students can hear the Kriol. meaning ‘I am selling them as I like’. The presence of branches of the University of the (4) Nonmarking of possessive case; possession West Indies in both Barbados and Jamaica helps clarified through noun juxtaposition, as in ‘so ensure that WCC continues to be well researched. many people life’ (MacCallum 1999:20), ‘she Some of the earliest studies of WCC include F.G. sister husban get one job up a Mona’ (from Cassidy’s 1961 text; B.L. Bailey’s (1962) Language Louise Bennett’s ‘Jamaica Patois’), ‘mai li guide to Jamaica, for US Peace Corp volunteers, her breda fren’ (i.e. Kriol for ‘My little brother’s (1953) M.A. thesis, a version of which was published friend’), and ‘person business’ in (5). in 1966; and Cassidy and Le Page (1967). In the (5) Multiple meanings associated with ‘for’, pro- United Kingdom, WCC is the most identifiable and nounced ‘fi’, as in ‘Dem always start wid influential of all Caribbean languages owing to the anancy ... lookin fi smbady... im always wan fi large numbers of Jamaican immigrants. In fact, the push up imself inna every oda person busi- evolution of British Black English has been attributed, ness’ (MacCallum 1999:19). [Gloss: They in large part, to the frequency of use of Jamaican always start with Anancy looking for some- Creole in the United Kingdom (Sutcliffe and Wong body.... He always tries to stick his nose in 1986). WCC is very similar to West African Pidgin other people’s business.] An example from English, leading many creolists (Le Page 1966; Bailey (1966:127) is ‘Im kom ya fo bada mi’, Dalphinis 1985; Faraclas 1996; Roberts 1996) to argue meaning ‘He has come here to bother me’. in favor of the Afrogenesis hypothesis, which traces its (6) Multiple meanings for ‘no’ (i.e. as ‘not’, ‘don’t’, roots to Africa, and highlights the influence of West ‘doesn’t’, and ‘didn’t’); double negatives 1174  WESTERN CARIBBEAN CREOLE(S) and preverbal negation, e.g. ‘Im no know what (14) Consonant reduction, as in ‘im’ vs. ‘him’, ‘an’ mek anancy stay so wutless... and no seem fi for ‘and’, ‘a’ for ‘at’, ‘ow’ vs. ‘how’, ‘wha’ vs. care’ (MacCallum 1996:19) [He didn’t know ‘what’, ‘round’ for ‘around’, ‘ac’ vs. ‘act’, and what makes/made anancy so worthless... and ‘lookin’ vs. ‘looking’. This is sometimes not seem to care]. accompanied by the prevocalic insertion of (7) Absence of auxiliary verbs, including ‘do’ in sounds like aspiration or /h/. Examples from negations, as in ‘I no kom bak’ (Escure Cliff (1996) include: ‘Is hokay wid me, wunna 1980:34), meaning ‘She or he didn’t come know. Dem nuh ‘tink like brute—hall a dem?’ back’, ‘Me no ‘ave not’in fe hide’ (Patterson (10) [Translation: It’s okay with me, you (plu- 1964:62), meaning ‘I don’t have anything to ral) know. Don’t they think like brutes—all of hide’, and ‘People talking around you’ (V.S. them?] Reid, ‘New Day’, p. 198), as well as the use of (15) ‘Cyaan’ or ‘Kyaan’ has context-dependent, the double negative, as in Black English opposite meanings, including ‘can’ and Vernacular. ‘can’t’, as in ‘di biebi-dem kyaan fiid them- (8) Zero tense specifiers or the default use of the selves’ (Bailey 1966:132), meaning ‘The present tense to represent all tenses. Tense is babies can’t feed themselves’ and ‘Lord Jesus, clarified through the conjoining of base verbs hear me nuh, why dem cyaan behave dem- and/or context, e.g. (7), ‘Toder mornin me go selves’ (Cliff 1996:18), meaning ‘Lord Jesus, ask her wat she tink about de war’ (from please hear me. Why can’t they behave them- ‘Jamaica Patois’), and ‘Licklemos’ she draw selves?’ One of dub poet Michael Smith’s me in. But me stop because of lickle monies...’ poems is ‘Mi cyaan believe it.’ (Cliff 1996:19), which translates as ‘A little (16) Emphatic ‘nuh’ as in ‘God nuh say so?’ (Cliff more and she draws/drew me in. But I stopped 1996:18), meaning ‘Didn’t God say so?’; because I didn’t have a lot of money.’ Africanisms like deh and fe; and other culture- (9) Nonmarking of plurals, as in ‘Tu Bwai’ (Kriol) specific terms such as Myal-man, Obeah-man, or ‘Tu bwoy’ (Jamaican Creole), and ‘Ho and Jumby-man; e.g. ‘Hear me, Myal-man. moch man?’ (Kriol for ‘How many men?’) Hear me, Obeah-man. Hear me, Jumby- (10) The use of ‘vex’ to mean ‘distraught’, ‘trou- man.... Suffering nuh mus’ be meant for we.... bled’, ‘disturbed’, and ‘disoriented’, as in How long mus’ we wait to get t’rough?’ (Cliff ‘anancy being anancy was vex’ (MacCallum 1996:16–17), which is essentially an invoca- 1999:20). Given the pronunciation differences tion that translates as ‘Please hear me Myal- between Kriol and Jamaican Creole, this word man (one who performs good miracles). is prounced and written as ‘bex’ in Kriol. This Please hear me Obeah-man (one who works in probably has to do with the Spanish influence the ‘dark’ arts). Please hear me Jumby-man in Belize; /b/ and /v/ tend to be conflated in (one who is like a spirit or ghost). Suffering Spanish. Hence, ‘I bex’ would mean ‘She or must be meant for us. How long must we wait He’s mad/frustrated.’ to get through?’ (11) The use of ‘tief’ as a verb for ‘steal’, ‘stole’, and ‘stolen’, as in ‘afta anancy tief away dat Most of these are characteristic of West African magic calabash’ (MacCallum 1999:19) and Pidgin. WCC is an oral tongue, so it is written in ‘Tiefin de cow’ (MacCallum 1999:22). An accordance with a writer’s perception of how it example from Kriol is ‘Bra Anansi de plan tu sounds. It is believed to echo the orality of African go tiif ya hag’ (Escure 1981:34) or ‘Brother folklore and traditions (see Burnett 1986; Bennett Anansi is planning to go steal a hog.’ 1940, 1966, 1979; Toolan 1992; Brathwaite 1993; (12) A relatively compact vocabulary (in contrast Lalla 1996; Warner-Lewis 1997; Balutansky and to Standard English) in which many words Sourieau 1998). Those who employ it argue in favor of mean more than one thing, depending on the its legitimacy. Like Jamaican poet Claude McKay, context, e.g. ‘me’ for ‘me’, ‘my’, and ‘mine;’ who was the first to employ solely WCC in his first and man hag and uman hag versus the distinct two (1912) books, one of the earliest Caribbean writ- English words, boar and sow. ers to make use of WCC in her writing, despite the (13) Related to (12), two or more meanings are criticism she received for doing so (see Brathwaite assigned to a single verb; for instance, ‘get’ 1993: 282–3), was Louise Bennett. One of the region’s expresses both ‘possession’ and ‘existence’, most influential writers—particularly as regards her as in ‘Im get wan pikni’ (She or He has one use of Patois—Bennett has been described as the ‘only child) and ‘im get good life.’ poet who has really hit the truth about her society 1175 WESTERN CARIBBEAN CREOLE(S) through its own language’ (Espinet 1990:346). She is conventional modes of representation associated with famous for her Anancy poems, many of which can be colonial domination’ (5). In the collection Caribbean found in Jamaican labrish, which means Jamaican creolization, some of the region’s foremost artists Gossip. Anancy is a legendary trickster (spider) in reflect on creolization and their reasons for employ- Jamaican folklore, and references to Anancy tend to be ing Creole. The discussion opens with a quote from more authentic and culturally flavored in WCC than in Edouard Glissant: (Standard) English. An example of an Anancy poem in Creolization’s most manifest symbol is the Creole lan- which the poet switches between WCC and English is guage. Its genius rests on its being always open.... Thus, Shara MacCullum’s poem ‘Jack Mandoora: Me No creolization carries in itself the adventure of multilin- Choose None’. gualism along with the extraordinary explosion of cul- After Bennett paved the way for WCC usage, many tures. But this explosion does not mean their scattering Caribbean writers began to ‘explore ways of working nor their mutual dilution. It is the violent manifestation the rich ore of dialect in literary contexts’ (Burnett of their assented, free sharing (1). 1986:xxv). As a result of the negative associations of the term ‘pidgin’, the term ‘nation language’, coined Part One is titled ‘Creolization and the creative by Barbadian poet, Kamau Brathwaite—another lumi- imagination’, and Part Two, ‘Creolization, literature, nary Caribbean poet—is frequently employed in the and the politics of language’. The consensus among region’s literary circle. This term is believed to aptly the contributors is that Creole (including WCC) is a echo the functional dimensions served by WCC malleable, poetic-license-engendering and liberating (Burnett 1986:xxv). Brathwaite (1993) defines it as indigenous code, free from colonial associations, and ‘Creole English... the kind of English spoken by the best suited to the themes that Caribbeans can relate to. people who were brought to the Caribbean’ (260) and Creolization is defined as ‘a cultural process rooted in ‘not the official English’ which ‘is an imperial lan- long-lasting psychological, spiritual, anthropological, guage’ (259). and linguistic experiences’ (9). Various historical and Louise Bennett, and other Caribbean writers who metaphorical terms are used to describe WCC and its employ WCC in their works have helped both promote formation, including: blendings, cradle-hammock, and sustain this language. In Burnett’s words, ‘The marronage, rhizomatic transcultural exchanges, hybrid tongues which result [in the literature] have an hybridity, métissage, mediation (contexts), linguistic enormous range of nuance and vigor of expression, crossbreeding borne of plantations, and ‘a language of with the limitation that only the locals catch every res- mixed sounds, tonalities, rhythms, and flavors’ (10). onance’ (xxv–xxvi). Noted artists include Orlando Some of these descriptors are constructs associated Patterson, Michelle Cliff, Lorna Goodison, Merle with prominent Caribbean writers. The metaphor of Collins, Erna Brodber, Patricia Powell, Honor Ford- the shipwreck is utilized by some, including Nobel Smith, Olive Senior, and Pauline Crawford (all laureate Derek Walcott (1993:11), to describe cre- Jamaican), and Zee Edgell, Amy Nicholas, Rutheen olization. Taylor, Ushanda Io Elima, and Ziola M. Ellis from A prime example of an instructional text aimed at Belize. Edgell’s Beka lamb, co-recipient of the eliciting respect for WCC or ‘Patois English’ (Bennett Fawcett Society Book prize in 1992, contains some 1981:i) is Jamaica maddah goose. This bilingual col- WCC and was the first Belizean novel to win interna- lection of nursery rhymes has a clear message, stated tional acclaim. in Bennett’s introduction: Any discussion of WCC would be incomplete Standard English is the official language of the Jamaican without reference to the historical context that engen- businessman, educator, professional and mass media. dered WCC, and its popular acceptance. Caribbean But it is the mellifluous patois that is the language of the Creole is believed to define and authenticate the people.... [T]he pictures and, basically, the words are Caribbean experience (Lamming 1960; Glissant authored by Jamaican art students. Hopefully they 1989; Gikanda 1992; Benítez-Rojo 1992; Walcott reflect what all Jamaican children behold around them. 1993; Hamner 1997). It is no wonder the term Creole Our only regret is that you can’t hear these nursery has many meanings in this region; it is used to rhymes spoken. Because patois is a language of melody and rhythm, a language to be heard. It is a vital lan- describe the language, the people, and the culture (see guage, an everchanging language, a language of the Brathwaite 1977, 1993; Edgell 1982). To the street and the scholar. And a language to be appreciated Caribbean writer, Creole and creolization have in its own right (i). acquired vital sociopolitical associations. Creole is conceptualized as ‘a new mode of expression and Exemplary nursery rhymes in WCC include ‘Lickle representation’ (Gikanda 1992:2) and creolization Miss Julie’ (Little Miss Muffet), ‘Ole Maddah Hubad’ as a ‘counter-discourse away from outmoded and (Old Mother Hubbard), and ‘Mary Had Wan Lickle 1176  WHORF, BENJAMIN LEE Lam’. Bennett’s mission—to destigmatize WCC—is Glissant, E. 1989. Caribbean discourse: selected essays, trans- clear in her poem ‘Bans O’ Killing’, which begins: lated by J. Michael Dash, Charlottsville: University of Virginia Press. So yuh a de man, me hear bout! Harney, Stefano. 1996. Nationalism and identity: culture and Ah yuh dem say dah-teck the imagination in a Caribbean diaspora. London and New Whole heap ‘English oat’ sey dat Jersey: Zed Books. Yuh gwine kill dialect! Hart, Richard. 1998. From occupation to independence: a short history of the peoples of the English-speaking Caribbean [Translation: So you’re the man I’ve heard about. region. Barbados and London: Canoe Press and Pluto Press. Ah! You’re the one they say uses piles of words and Holm, John, A. 2000. An introduction to Pidgins and Creoles. Cambridge: Cambridge University Press. (that) vows to kill Creole!] James, Louise (ed.) 1999. Caribbean literature in English. The poem continues with the following message, London and New York: Longman. portending a bright future for WCC: Journal of Pidgin and Creole languages. 1986 to date. Amsterdam and Philadelphia: John Benjamins. Dah language wey yuh proud o’ Mais, Roger. 1974. Brother man. London: Heinemann. Wha yuh honour and respeck McKay, Claude. 1933. Banana bottom, New York: Harper and Po’ Mass Charlie: Yuh noh know sey Brothers. Dat it spring from dialect! McWhorter, John, H. 1997. Towards a new model of Creole genesis. studies in ethnolinguistics, Vol. 3. Berlin, New York [Translation: The language that you are proud of, and Paris: Peter Lang. that you honor and respect, Poor Mr. Charlie; you don’t Patteson, Richard, F. (ed.) 1998. A critical perspective on new (even) know that it came from the dialect you despise!] fiction from the West Indies. Boulder and London: Three Continents Press. Powell, Patricia. 1993. Me dying trial. Ibadan and Gaberone: References Heinemann. Alleyne, Mervyn. 1979. Acculturation and the cultural matrix Senior, Olive. 1989. Arrival of the snake-woman and other short of Creolization. Pidginization and Creolization of lan- stories. Harlow, Essex: Longman. guages, ed. by Dell Hymes, 169–86. Cambridge: Cambridge Walcott, Derek. 1993. The Antilles: fragments of epic memory. University Press. New York: Farrar, Strauss, and Giroux. Bickerton, D. 1981. Roots of language. Ann Arbor: Karoma. ANITA PANDEY Collins, Merle. 1998. Angel, London: Seal Pr Feminist Publishers. Collymore, F.A. 1965. Glossary of Barbadian dialect, 2nd edi- See also Arawak; Discourse Analysis; Hawaiian tion. Bridgetown, Barbados: Advocate Co. Creole English; Nigeria; Pidgins and Creoles; Edgell, Zee. 1982. Beka lamb. London: Heineman. Krio; Saramaccan; Sranan; Tok Pisin Whorf, Benjamin Lee Along with his mentor, Edward Sapir (1884–1939), his reading a French mystic, theosophist, and quasilin- Benjamin Lee Whorf is best known today for the princi- guist, Antoine Fabre d’Olivet (1768–1825), whose The ple of linguistic relativity, also called the Sapir–Whorf Hebrew tongue restored, originally published in hypothesis (see ‘Sapir–Whorf hypothesis’). Basically French in 1815–1816 (translated by Nayan Redfield, self-educated in linguistics and without a Ph.D. (although Whorf’s neighbor, New York: G.P. Putnam’s Sons, he did study for a time with Edward Sapir at Yale 1923), exerted a profound influence on the Methodist University starting in 1931), Whorf published extensive- Episcopal/Theosophist Whorf, about whom he wrote: ly on American Indian languages, especially (Uto- ‘[he] cut loose from dead traditional grammar, antici- Aztecan) Hopi, and also wrote on many other related pated the modern concept of the phoneme…, saw the linguistic topics for a variety of publications. As a result importance of language for logic and what today of his research into the structure of these languages, he would be called psychology’ (Rollins 1980: 49). came to assert that language and culture were intimately Reading another theosophist–linguist helped develop related such that one’s native language, i.e. its structure or Whorf’s ideas even further—Frederick Max Müller grammar and vocabulary, influences the way a speech (1823–1900), who maintained in his The science of community perceives and conceives of its reality. thought (1887:295): ‘Language is not outside the Whorf’s fascination with ciphers, puzzles, and mind, but is the outside of the mind. Language is very codes as a child as well as his interest in religion led to thought as thought is very language.’ 1177 WHORF, BENJAMIN LEE Part of the groundwork for the Sapir–Whorf The milder version of the hypothesis is labeled lin- hypothesis was laid by Whorf’s work as a fire insur- guistic relativity. This claims that the native language ance investigator. During his career, he had the oppor- influences thoughts and perceptions. In fact, it was tunity to analyze many reports as to why fires broke Whorf who coined the phrase ‘linguistic relativity’, out. He found that workers would use extreme caution since he always tried to qualify his assertions. In when around ‘full’ drums of gasoline. Just as one ‘Linguistics as an exact science’, Whorf maintains: might expect, workers were careful not to smoke around ‘full’ drums. Yet, these same workers when …what I have called the ‘linguistic relativity principle’ … means, in informal terms, that users of markedly dif- around ‘empty’ drums of gasoline would sometimes ferent grammars are pointed by their grammars toward toss lit cigarettes nearby. This caused a violent explo- different types of observations and different evaluations sion because an empty drum still contained volatile of externally similar acts of observation, and hence are gasoline vapor. Thus, an ‘empty’ drum was really not equivalent as observers but must arrive at somewhat much more of a threat than a ‘full’ one. Using these different views of the world. (Carroll 1956:221) data, Whorf concluded that the meanings of certain words had an affect on a person’s behavior. One of Whorf’s most famous essays, ‘A linguistic It was the painstaking research of both Sapir and consideration of thinking in primitive communities’, Whorf into the grammatical systems of American asserts that the problem of ‘thinking’ by so-called Indian languages, however, that proved to have the ‘primitive’ peoples is ‘approachable through linguis- greatest impact on Whorf’s subsequent thinking. By tics’ (Carroll 1956:65). Further, as linguists have come predicating their insights into the interrelationships of to fully appreciate only fairly recently, Whorf believed language and culture on what they had learned from that ‘linguistics is essentially the quest of MEANING’ the structures of these languages, the basic idea of lan- (1956:73). Examples are given of things which are rel- guage shaping the perceptions of its speakers and pro- atively easy to say in Hopi but awkward or clumsy to viding for them a vehicle so that their experiences and say in SAE (Standard Average European) languages. emotions could be placed into significant cognitive Whorf concludes that Hopi reveals a ‘higher plane of categories was given its scientific underpinnings. thinking’ (ibid.). Generally, Sapir is credited as giving the problem of In ‘Language, mind, and reality’, published in India establishing the link between language and culture its a year after his death, Whorf explains the ‘obviative’ in initial formulation (continuing in the tradition of Algonkian languages. It is in essence two third persons Johann Gottfried Herder, 1744–1803, and Wilhelm (Carroll 1956:265), one of which we traditionally refer von Humboldt, 1762–1835), while Whorf is honored to as the fourth person. This aids in compact descrip- as the one who took this idea and developed it further tion of complicated situations, for which we should to include grammar in addition to lexis, thereby mak- have to resort to cumbersome phraseology. Let us ing it into a bona fide hypothesis. However, it must be symbolize their third and fourth persons by attaching noted that there are really two different versions of the the numerals 3 and 4 to our written words. The hypothesis. This is understandable when one considers Algonkians might tell the story of William Tell like that Whorf did all of his professional writing in the this: ‘William Tell called his3 son and told him4 to rather short period from 1925 until his untimely death bring him3 his3 bow and arrow…’ He remarked that in 1941, and his ideas, quite naturally, were continu- ‘such a device would greatly help in specifying our ously evolving. The strong version of the hypothesis, complex legal situations, getting rid of the ‘party of which is called linguistic determinism, holds that lan- the first part’…, ‘the party of the second part’, the guage determines thinking. This position is most diffi- ‘aforementioned’ or the ‘aforesaid’ (ibid.). cult to defend primarily because translation between For those wishing to go to the fountainhead of the one language and another is possible, and ‘thinking’ debate over language’s influence on thought, Whorf’s can take place without language at all; e.g. an artist or writings still provide for a satisfying journey. sculptor can and often does think with his fingers. Mirroring Sapir’s thoughts, Whorf notes in his Biography (1940) ‘Science and linguistics’: Benjamin Lee Whorf was born in Winthrop, We dissect nature along lines laid down by our native languages… We cut nature up, organize it into con- Massachusetts on April 24, 1897. He graduated from cepts, and ascribe significances as we do, largely Winthrop High School in 1914, and did his B.S. (1919) because we are parties to an agreement to organize it in in chemical engineering (MIT). In 1920, he began his this way—an agreement that holds throughout our career as a trainee in fire prevention engineering for the speech community and is codified in the patterns of our Hartford Fire Insurance Company. He was appointed as language. (Carroll 1956:213) a Special Agent for the company in 1928. Eventually, 1178  WITTGENSTEIN, LUDWIG he became a recognized expert in industrial fire pre- References vention, writing ‘Blazing Icicles’, which offered a lin- Carroll, John B. (ed.) 1956. Language, thought, and reality: guistic interpretation of the cause of some fires. He was selected writings of Benjamin Lee Whorf. Cambridge, MA: promoted as Assistant Secretary of the company in MIT Press. 1940. Gumperz, John J., and Stephen C. Levinson (eds.) 1997. During the 1930s he regularly visited a Hopi Rethinking linguistic relativity. Cambridge: Cambridge University Press. informant in New York City, and five years later had Hoijer, Harry (ed.) 1954. Language in culture. Chicago, IL: prepared a grammar and dictionary (Unpublished University of Chicago Press. Papers, Yale University). He published much of his Lee, Penny. 1996. The Whorf theory complex. Amsterdam and research in the major recognized outlets, such as Philadelphia: John Benjamins. Language, International Journal of American Rollins, Peter C. 1980. Benjamin Lee Whorf: lost generation, theories of mind, language, and religion. Ann Arbor, MI: Linguistics, and American Anthropologist. Although University Microfilms International. he was offered academic positions, he decided to Whorf, Benjamin Lee. Unpublished papers. Sterling Memorial remain with the Hartford Fire Insurance Company, Library, Yale University. since he believed it afforded him greater opportunity ALAN S. KAYE to do what he wanted. He died of cancer on July 26, 1941. See also Sapir, Edward; Sapir–Whorf Hypothesis Wittgenstein, Ludwig Ludwig Josef Johann Wittgenstein was a charismatic investigations (1954) influenced the so-called ordinary Austrian philosopher whose philosophical career language philosophy (which relinquishes the search developed mainly in England. Wittgenstein’s life was for an ideal logical language). This is the reason why a search for a special kind of spirituality by means of one traditionally speaks of the ‘first’ and the ‘second’ philosophical activity. This spirituality is evidenced by Wittgenstein. However, these ‘two’ Wittgensteins are some idiosyncratic ethical and even metaphysical reconcilable, as shown below. ideas in his Notebooks 1914–1916 and Secret diaries. Wittgenstein wrote the Tractatus logico-philosoph- Wittgenstein’s work was very influential in twentieth- icus during World War I, while in a prison camp, and century philosophy. He also shaped linguistics —espe- sent it to his teacher and colleague Bertrand Russell cially pragmatics— mainly through his Philosophical (1872–1970). The Tractatus was published in England investigations, published in 1953. His theory of lan- in 1922 with the help of Russell, who also wrote the guage-games and his idea that ‘the meaning of a word Prologue. The Tractatus is a collection of numbered lies in its use’ became the basis of certain theoretical statements elaborating on some philosophical prob- frameworks in linguistics. His philosophy of language lems discovered by Frege and Russell. The Tractatus and his scientific epistemology are also reflected in expounds a philosophy of language and mathematics most of the current social sciences. and is a reflection on the nature of life and philosophy. The end of the nineteenth and the beginning of the For Wittgenstein, philosophy is mainly an activity, not twentieth century saw the arrival of the philosophy of a doctrine, and so his philosophy must be understood language. Analytical philosophers like Gottlob Frege through his works. (1848–1925) conceded a central role to language in Wittgenstein’s philosophy, as presented in the philosophical theory, and they begun the construction Tractatus, relates three elements: reality, thought, and of a general theory of meaning, which greatly influ- language. Reality exists, thought is the interpretation of enced Wittgenstein’s thinking. Wittgenstein made two reality, and language is the expression of thought. prima facie contradictory contributions to this linguis- Wittgenstein calls the logical identity between the tically oriented philosophy: his Tractatus logico-philo- structure of reality, the structure of thought, and the sophicus (1922) represents an exercise in analytical structure of language ‘isomorphism’. For him, logic is philosophy (which regards everyday language as an the ideal language, but only a tool for description. impure vehicle of thought), while his Philosophical Language is used to describe facts, although there 1179 WITTGENSTEIN, LUDWIG remain things that cannot be said. Therefore, the Guessing riddles- Tractatus reaches the now famous conclusion: ‘what Making a joke; telling it- we cannot speak about we must pass over in silence’. Solving a problem in practical arithmetic- Thus, Wittgenstein studies the limits of language and Translating from one language into another- correspondingly the limits of oneself: ‘The limits of my Asking, thanking, cursing, greeting, praying.’ language are the limits of my world’ (Tractatus 5.6). Because the game metaphor thus extends to all rule- The Tractatus inspired some meetings of a group of governed uses of language, formal language or the lan- philosophers referred to as the ‘Vienna Circle’, who guage of logic can also be viewed as a language game. developed the so-called logical positivism or neoposi- Hence, the Tractatus is also reflected on in the tivism. The principles of particular appeal to the Philosophical investigations. Wittgenstein’s work is Vienna Circle were the analytical characteristics of thus a language game itself, which highlights the self- logic, the claim that sentences need to be verifiable to reflexive perspective a Wittgensteinian approach has be true and the idea that philosophy clarifies truth to take. instead of discovering it. Understanding the meaning of a word requires a The blue and brown books mark a transition from knowledge of linguistic and extralinguistic facts, and the Tractatus to the Philosophical investigations, that Wittgenstein asserted that the analysis of language use is, from the first to the second Wittgenstein. The is as important as logical analysis. This represented a Philosophical investigations were published posthu- ‘linguistic turn’ in the philosophy of language, and the mously, and are a collection of Wittgenstein’s work resulting requirement of critically reflecting on termi- between 1941 and 1949. The concept of philosophy, nology now forms an important part of the scientific the rules of language, and the theory of language epistemology of many social sciences. Since the games are its main topics. It is precisely the concept of Philosophical investigations with its ‘language games’ ‘language game’ that reconciles the Tractatus and the introduced a rule-based view of discourse, it can be Philosophical investigations as discussed below. viewed as the philosophical basis of pragmatics. The Philosophical investigations shares with the Members of the Oxford School such as G. Ryle, P.F. Tractatus that philosophy is seen as an activity, and Strawson, and notably J.L. Austin, who developed the this activity consists of the analysis of language. theory of ‘speech acts’, were heavily influenced by However, in the Tractatus, Wittgenstein was only Wittgenstein’s ideas. Thus, after philosophy turned to interested in the logical truth or falsity of sentences. In language, linguistics turned to philosophy: ‘this was the Investigations, on the other hand, he was willing to the “pragmatic turn”, which led to a proliferation of look for the meaning of words as expressed in their communicative and social studies of speech in action everyday use. Wittgenstein thus acknowledged that we and action in speech’ (Nerlich and Clarke 1996:6). are primarily users of language and only secondarily analysts of language. For Wittgenstein, the meaning of words and propo- Biography sitions is identifiable in their use in language, and this Born in Vienna, Austria, on April 26, 1889 to a use is shaped in ‘language games’. The concept of lan- wealthy family, Ludwig Wittgenstein was educated in guage games highlights the idea that language use is an artistic and intellectual environment at home until rule-governed, i.e. the game metaphor is a model of he was14. From 1903 to 1906 he studied mathematics the rule-based aspects of language. Wittgenstein uses and natural sciences in Vienna, and from 1906 to 1911 the term ‘game’ broadly: he studied Mechanical Engineering in Berlin and ‘23. […] Review the multiplicity of language- Manchester. He went to Cambridge in 1911 to attend games in the following examples, and in others: lectures given by Bertrand Russell on mathematical Giving orders, and obeying them- logic, but left Cambridge somewhat disappointed with Describing the appearance of an object, or giving the knowledge he gained there and went to Skjolden, its measurements- Norway. He joined the Austrian army from 1914; in Constructing an object from a description (a drawing)- 1918 he became prisoner of the Italians. During these Reporting an event- war years he wrote the Tractatus, which he sent to Speculating about an event- Russell while he was in a prison camp. Released in Forming and testing a hypothesis- 1919, he became primary teacher in 1920 and taught Presenting the results of an experiment in tables and until 1925. After fulfilling various posts, he returned to diagrams- academic life in 1929: he taught at Trinity College. Making up a story; and reading it- That year he submitted the Tractatus as his doctoral Play-acting- thesis. In 1939 he became Professor of Philosophy, Singing catches- and remained at Cambridge except for the World War 1180  WOLOF AND ATLANTIC LANGUAGES II period. In 1947 he retired to Ireland, and in 1949 America, 1780–1930. Amsterdam and Philadelphia: John came back to Oxford because of his illness. He con- Benjamins. Specht, E.K. 1963. The foundations of Wittgenstein’s late phi- tinued working until he died in Cambridge on April losophy. Manchester: Manchester University Press. 29, 1951. A large part of his works was published Winch, Peter (ed.) 1969. Studies in the philosophy of posthumously. Wittgenstein. London: Routledge and Kegan Paul. Wittgenstein, Ludwig. 1961. Notebooks 1914–1916. Oxford: References Blackwell. ———. 1922 and 1961. Tractatus logico-philosophicus. Anscombe, G. Elizabeth. 1971. An introduction to London: Routledge and Kegan Paul, reprinted 1993. Wittgenstein’s Tractatus. Philadelphia: University of ———. 1958. The blue and brown books. Oxford: Blackwell Pennsylvania Press. [Harper Collins Publishers, 1976]. Baum, Wilhem. 1985. Ludwig Wittgenstein. Berlin: Colloquiun ———. 1953. Philosophical investigations. Oxford: Blackwell, Verlag GmbH. reprinted 1996. Kripke, Saul A. 1982. Wittgenstein on rules and private lan- ———. 1973. Letters to C.K. Ogden. Oxford: Blackwell. guage. Cambridge, MA: Harvard University Press. ———. 1986. Philosophical grammar. LA: University of Malcolm, Norman. 1958. Ludwig Wittgenstein. A memoir. California Press. Oxford: Oxford University Press. ELENA BATTANER-MORO Nerlich, Brigitte, and David Clarke. 1996. Language, action and context. The early history of pragmatics in Europe and See also Pragmatics Wolof and Atlantic Languages Wolof is an Atlantic language of the Niger-Congo phy- are for the most part mutually intelligible, although lum, spoken primarily in Senegal and the Gambia on influence from neighboring languages has resulted in the Atlantic coast of West Africa. Wolof serves as a lin- some significant lexical differences between dialects. gua franca in Senegal, where approximately 40% of Contemporary research confirms Koelle’s (1854) the population speak it as a first language and 45% hypothesis that places the Fula-speaking heartland in speak it as a second language. Extrapolating from the the Fouta Toro region of northern Senegal, with subse- 1988 census figures, there are currently at least three quent eastward expansions of Fula-speaking peoples, million native speakers of the language, and at least as many of whom were and still are cattle-herding pas- many who speak it as a second or third language, mak- toralists. Seereer is spoken by approximately seven ing for a total of no less than six million speakers. hundred thousand people concentrated in the western The Atlantic languages are spoken almost exclu- and central areas of Senegal, centered on the historical sively along the West African coast from the Senegal region of Siin (Sine), and in parts of the Gambia. Most river that forms the border between Mauritania and Seereer speakers speak Wolof as a second language. Senegal southward to Sierra Leone and Liberia, gener- Temne is one of the major languages of Sierra Leone, ally abutting languages of the Mande family toward spoken in the central part of the country, mostly to the the east. Atlantic languages with the greatest numbers north of Freetown, and current estimates, although of speakers include Fula, which is spoken across the possibly low, are that it has approximately one million Sahel; Wolof and Seereer, both spoken primarily in two hundred thousand speakers. Senegal; and Temne, spoken in Sierra Leone. Fula Atlantic was recognized as a group of related lan- (also Peul in French from Wolof Pël, and Fulani in var- guages as early as Koelle’s (1854) Polyglotta Africana, ious languages, including English, from Hausa where it goes by the name of North-West Atlantic. Fùláánìì) is known by its speakers as Pulaar (or Pular) Westermann (1927) renamed the language group West in the western dialects and Fulfulde in the eastern Atlantic, a name retained by Greenberg in his classifi- ones. It is spoken by several million people across a cation of African languages published in 1963. The noncontiguous stretch of West Africa. Fula speakers name was subsequently shortened to Atlantic by are found from Mauritania and Senegal in the west to Doneux (1975). Most contemporary classifications of Chad and Sudan in the east, including significant pop- Atlantic consider it to be an early branching of Niger- ulations in Guinea, Mali, Burkina Faso, Nigeria, Congo, concurrent with or subsequent to the Mande Niger, and Cameroon. The different dialects of Fula branching. Until Greenberg’s classification, and largely 1181 WOLOF AND ATLANTIC LANGUAGES as a result of using cultural and ethnic criteria rather alveolar, palatal, and velar. Some languages, including than linguistic ones, Fula had been classified apart from Wolof and Seereer, also have one or more uvular con- the Atlantic languages. Linguistic evidence, however, sonants. Prenasalized stops and implosive consonants showed that Fula was clearly an Atlantic language with are common, although Wolof does not have the latter, a strong resemblance to Seereer. Most contemporary and word-initial voiceless prenasalized stops have all classifications of Atlantic, including Wilson (1989), are but disappeared from the language, being retained only based on or respond to Sapir’s (1971) lexicostatistical in place names (such as Mpal) or by older rural Wolof study of the languages. Comparing 100 core lexical speakers. Seereer is phonetically quite unusual in hav- items across languages, which revealed rather low cor- ing a series of both voiced and voiceless implosive respondences, Sapir nevertheless posits a Northern stops in three places of articulation: bilabial, alveolar, Branch, Southern Branch, and a single isolated lan- and palatal. A basic five-vowel system is augmented guage, Bijago (or Bijogo), spoken on the Bijago Islands with [/ATR] (advanced tongue root) contrasts for off the coast of Guinea Bissau. Sapir’s Northern Branch mid-vowels in many Atlantic languages. These are of Atlantic is comprised of five subgroups: (A) Senegal phonemic in Wolof, as evidenced by pairs such as reer languages (Fula and Seereer, Wolof); (B) Cangin lan- [ATR] ‘dinner’ and réer [ATR] ‘to be lost’. Wolof guages (Non, Ndut, Lehar, Safen, also sometimes also makes an [ATR] contrast between the low vowel called Seereer-Non, Seereer-Ndut, etc.); (C) Bak lan- [a], which is [ATR], and a central schwa-like vowel guages (Diola cluster, Manjaku and Papel, Balanta); written as ë. Many Atlantic languages, including (D) Eastern Senegal — Portuguese Guinea (now Wolof, contrast both consonant and vowel length; Guinea Bissau) languages (Tenda group: Tanda, Basari, Seereer contrasts only the latter. While almost all Bedik, and Konyagi; Biafada and Pajade; Kobiana and Niger-Congo languages, including most Atlantic lan- Banhum; and (E) a nameless group (Nala, Mbulungish, guages, are tonal, the northern Senegal languages, and Baga Mboteni). The Southern Branch is comprised Fula, Seereer, and Wolof, are not, although a high-pitch of three subgroups: (A) Sua; (B) Mel languages (Temne accent is associated with ideophones. and Baga Koba, Sherbro, Mmani, Krim and Kisi, Morphologically conditioned consonant mutation is Gola); and (C) Limba. Recent unpublished work on the found stem initially in verbs and nouns in a great num- isolate, Bijogo, suggests that it may be the most con- ber of Atlantic languages. In noun stems, initial con- servative Atlantic language, especially since its noun sonants may have up to three homorganic variants class markers most closely resemble those of the Bantu conditioned by noun class. The variants are known as languages. The low scores found by Sapir for cognates grades, following Arnott (1970). In Fula, these take the between Atlantic languages are corroborated by form of a continuant, a stop, and a prenasalized stop as Bennett and Sterk (1977), who find that the Atlantic in the singular (Class 1), plural (Class 2), and diminu- languages generally show no more genetic affinity with tive plural (Class 21) words for ‘man’ (gorko, worKe, n each other than they do with the rest of the Niger- goron) and ‘woman’ (debbo, rewKe, ndewon). Seereer Congo phylum. Although grammatical and lexical sim- exhibits a similar type of consonant mutation, as well ilarities are to be found within various subgroups, to as one that shows alternations between a voiced stop, date linguists have not found any evidence of shared voiceless stop, and prenasalized stop. The word for innovation common to the Atlantic languages, leaving ‘man’ in its singular (Class 1), plural (Class 2) and their coherence as a genetically related family a matter diminutive singular (Class 12) forms (okoor, goor, of convention rather than fact. ongoor) illustrates the latter type of mutation, while the Salient characteristics of the Atlantic languages same forms for the word for ‘woman’ (otew, rew, include an SVO word order, extensive noun class sys- ondew) illustrate the former. This type of consonant tems, and verbal extensions, as well as a perfective/ mutation in the Atlantic languages has been analyzed imperfective aspectual distinction in the verbal sys- as the historical result of the erosion of noun class pre- tem, a syntactically encoded focus system, and verb fixes which had a phonological effect on the stem-ini- serialization. These features are not unique to Atlantic tial consonant. Remnants of prefixes are seen in some languages but are, in fact, common to the Niger-Congo Atlantic languages like Seereer, providing incontro- phylum as a whole. The trait most closely associated vertible evidence that Atlantic languages, like other with, although not unique to, the Atlantic languages is Niger-Congo languages, and particularly Bantu lan- morphologically conditioned stem-initial consonant guages, at one time had a complete set of noun class mutation, which is found on nouns, adjectives (for prefixes. Such prefixes eroded phonologically over those languages that have them), and verbs throughout time only to be renewed in many cases, such as Fula, the language group. by suffixes. Wolof exhibits an advanced stage in the The consonantal inventories of Atlantic languages erosion of noun class morphology: for the most part, generally distinguish four places of articulation: labial, nouns are not marked for class so that their singular 1182  WOLOF AND ATLANTIC LANGUAGES and plural forms are identical. A few lexicalized region of southern Senegal. Urban Wolof is substan- exceptions to this generalization, including the words tially different from dialects spoken in the predomi- for ‘eye/s’ (bët/gët) and ‘person/people’ (waa/gaa), nantly rural Wolof heartland. Its main characteristics provide evidence of a once robust inflectional system. are extensive lexical borrowing from French, the for- Consonant mutation is nevertheless quite productive in mer colonial language and current official language of other areas of the grammar, including diminutive for- the country, and the reduction of the noun class system mation, which requires a prenasalized form, and verb toward the default b-class. Urban Wolof has an to noun derivation. ambiguous status. On the one hand, as the language of Wolof has ten noun classes: eight singular and two the capital it holds a certain prestige, especially among plural. Of the singular noun classes, one, the s-class, youth, and its mastery implies a certain urban sophis- serves a second function as a diminutive class. A noun tication. On the other hand, it is recognized as being a class marker appears in the form of a single consonant hybrid language and looked down upon as ‘impure’, on nominal dependents such as determiners and rela- especially by those who are not ethnically Wolof. tive particles. The word for ‘person’ (nit) is in the k- In the domain of discourse, elaborate speech in class, and the word for ‘bird’ (picc) is in the m-class, Wolof is associated with griots or verbal artists who hence the following forms: nit ki ‘the person’/ picc mi are of low social status. Irvine (1978) has documented ‘the bird’: nit kële ‘that person’ / picc mële ‘that bird’; the avoidance of elaborate speech as an attempt to nit ku baax ‘person who is good’/ picc mu baax ‘bird manipulate social status. that is good’. A problematic aspect of Wolof class Finally, although there is an official Wolof orthog- assignment, remarked upon by several scholars, is the raphy in the Roman script, it is not widely used. The tendency for the class marker to repeat the initial con- preferred writing system is an ajami version known as sonant of certain nouns, in a process akin to consonant wolofal, which is a modified Arabic script, made pop- harmony or reduplication. Phonologically conditioned ular through widespread informal Islamic education agreement poses a substantial problem for linguistic and attendance at Qur’anic schools. theory in general and the lexicalist hypothesis in par- ticular, since syntax should not have access to the References phonology. A second noted trend in the evolution of Arnott, D.W. 1970. The nominal and verbal systems of Fula. the Wolof noun class system is a tendency toward Oxford: Clarendon Press. assigning nouns to the default b-class. The origins of Bennett, Patrick R., and Jan P. Sterk. 1977. South Central this tendency are to be found in a combination of mor- Niger-Congo: a reclassification. Studies in African phological and sociolinguistic factors (Irvine 1978; Linguistics 8. 241–73. Mc Laughlin 1997). Childs, G. Tucker. 1995. A Grammar of Kisi: a southern Atlantic language. Berlin and New York: Mouton de Gruyter. African languages in general, including Atlantic Doneux, Jean L. 1975. Hypothèses pour la comparative des languages, have had a profound effect on contemporary langues atlantiques. Africana Linguistica 6. 41–129. linguistic theory. Consonant mutation in the Atlantic Greenberg, Joseph H. 1963. The languages of Africa. languages, and particularly in Fula, has substantially Bloomington: Indiana University Center in Anthropology, informed underspecification theory and, more recently, Folklore and Linguistics, and The Hague: Mouton. Irvine, Judith T. 1978. Wolof noun classification: the social set- theories of featural affixation. The interaction of con- ting of divergent change. Language in Society 7. 37–64. sonant mutation and reduplication in Seereer-Siin has Ka, Omar. 1994. Wolof phonology and morphology. Lanham, had implications for correspondence theory; and the MD: University Press of America. facts of ATR vowel harmony in Wolof, which has vow- Koelle, Sigismund W. 1854. Polyglotta Africana. London: els that are both opaque and neutral to the process, have Church Missionary House. Mc Laughlin, Fiona. 1997. Noun classification in Wolof: when also been instrumental in theoretical advances in that affixes are not renewed. Studies in African Linguistics area. The clause structure of Wolof has also provided 26(1). 1–28. evidence for the projection of a sigma phrase within the Sapir, J. David. 1971. West Atlantic: an inventory of the lan- principles and parameters framework. guages, their noun class systems and consonant alternation. The sociolinguistics of Wolof are of particular Current trends in linguistics: linguistics in sub-Saharan Africa, ed. by Thomas Sebeok. The Hague: Mouton. interest. Wolofization, or the spread of Wolof as a lin- Westermann, Diedrich. 1927. Die westlichen Sudansprachen und gua franca in Senegal, has been increasing steadily ihre Beziehungen zum Bantu [The western Sudan languages since the colonial period when it was used widely in and their relationship to Bantu]. Berlin: Walter de Gruyter. the coastal cities. Wolofization is an urban phenome- Wilson, W.A.A. 1989. Atlantic. The Niger-Congo languages, non, and it is the specifically urban variety of Wolof ed. by John Bendor-Samuel. Lanham, MD: University Press of America. that continues to spread, even to the point where it has FIONA MC LAUGHLIN replaced Portuguese creole as the predominant lan- guage of Ziguinchor, the largest city in the Casamance See also Niger-Congo 1183 WORD Word Although many language users intuitively know what a inherent to the lexeme as a linguistic sign. However, a ‘word’ is, an unequivocal and clear-cut linguistic defi- lexeme actually forms a set of several linguistic signs: nition cannot be given. The concept of ‘word’ is highly some word forms of a lexeme may be linked to spe- ambiguous and refers to different phenomena. On the cific (sub-) meanings of the lexeme and vice versa. one hand, words are language units at the intersection For example, all orthographic word forms of the verb of two linguistic levels of description: morphology YIELD, i.e. yield, yields, yielded, yielding, are linked (which is the arrangement of meaningful sound to the possible meanings ‘give way’ and ‘lead to’. On sequences, or morphemes, in words) and syntax (which the other hand, a third meaning of the verb, ‘produce’, is the arrangement of words in sentences). Accordingly, is strongly associated with the first two aforemen- lexicology is the linguistic branch that deals with tioned word forms. To be able to refer to such subsets words, which thus implies both morphological and of word forms with a specific range of meaning, the syntactic aspects. On the other hand, words are regard- concept of ‘lexical unit’ or ‘lexical linguistic sign’ has ed as basic units in the psychological reality of lan- been introduced. guage acquisition, production, and processing: words Apart from word forms and lexemes (including lex- in the mind. An attempt to produce a system reflecting ical units), the concept of ‘word’ can also be the versatility of the notion of ‘word’ would need to approached from a grammatical perspective. A gram- describe words as both ‘linguistic structures’ (although matical word (i.e. word3) is immediately relevant to of different kinds) and as parts of the mental lexicon. syntax and has specific morphological features. The The linguistic notion of ‘word’ captures three funda- distinction between word forms and grammatical mentally different, but closely related, concepts that will words is important because one word form may repre- be indicated in the following as word1, word2, and word3. sent different grammatical words (a phenomenon To begin with, so-called ‘word forms’ (i.e. word1) refer called ‘syncretism’). For example, the phonological to the physical substance of words in spoken and written word form /bɔiz/ can be regarded as a realization of language: phonological word forms in speech, and the plural form, the singular genitive, and the plural orthographic word forms in writing. For example, /bçI/ genitive. In applying the concept of grammatical and /bçIz/ are two phonological word forms, i.e. spe- words, linguists are particularly interested in the ques- cific combinations of sounds in the phonic medium. On tion of how words, as minimal syntactic units, are the other hand, boy, boy’s, boys, and boys’ represent four arranged in grammatical structures. Irrespective of the corresponding orthographic word forms, or combina- specific grammatical principles at work, grammatical tions of letters (and apostrophes) in the written medium. words have a certain ability to change position in a The six word forms provide the inventory of possible sentence (i.e. ‘positional mobility’). Thus, it is often word-form realizations of one underlying abstract unit, possible to shift a grammatical word, in its entirety, to the lexeme BOY. a different position in a given sentence without jeop- The lexeme (word2), usually given in capitalized ardizing the grammatical integrity of that sentence. It letters, is what all the actual word forms have in com- goes without saying that in English, which has a com- mon at an abstract level. It is this abstract unit of a lex- paratively fixed word order, the overall degree of posi- eme that is at the basis of a dictionary since all word tional mobility is lower than, for example, in German, forms of one and the same lexeme are subsumed under which allows for much more word-order variation. In the same dictionary entry. terms of their internal structure, however, grammatical While word forms represent merely physical forms words are stable, in that the sequence of morphemes in terms of sounds or letters, lexemes—in a first is, in principle, not subject to variation: for instance, approximation—are examples of linguistic signs. That boy-ish is a possible adjective in English, but the is to say, a lexeme connects potential forms (i.e. the reversed sequence, i.e. *ish-boy, is not permitted. signifier) to an abstract meaning encoded by these Words also play an important role in the mental forms (i.e. the signified), such as the phonological processes that underlie language use. It is reasonable to word forms /teibl/and /teiblz/(i.e. the signifier) to the assume that words are stored in the mind in a highly meaning of ‘piece of furniture consisting of a flat top structured way. This orderliness of the complex mental supported on one or more legs’ (i.e. the signified). lexicon permits the amazing speed of lexical retrieval This arbitrary, but fixed, form-meaning relation is in natural speech. It is widely accepted that the mental 1184  WORD lexicon presumably does not list each and every word attempts have been made to establish criteria for a reli- in a random order. Rather, a list of lexical structures able identification of words in the continuous stream (words, word groups, word components) is supposed to of speech sounds. For example, words are said to be be complemented with a set of rules and principles on demarcated by positions at which pausing is, at least, how to combine these structures. This concept avoids, possible. Another suggestion is to look for those for instance, a listing of different words ending in the stretches of sounds that could stand meaningfully on same suffix as separate entries: for example, boyish and their own. Although these two approaches and others girlish are derived from applying the same combinato- may be useful for the majority of words, other cases rial rule to the base entries boy and girl, respectively. cannot be clarified, such as compounds, which might Such systematic combinations presuppose a mental be considered as one word or as several words. capacity for analyzing complex words in speech pro- In all languages, linguists set out to group words into cessing, which is subsumed under the notion of ‘mor- word classes. Generally, one can distinguish between phological parsing’. In the aforementioned example, ‘open’ and ‘closed’ word classes. Open word classes morphological parsing would refer to speakers’ ability can always have new words freely added to them; in to strip the suffix -ish and to identify a general rule that most languages, these classes include nouns, verbs, makes it possible to derive an adjective from a noun by adjectives, and adverbs. In contrast, usually no new adding this suffix to the noun. members can be added to closed word classes. In The importance of words for the linguistic compe- English, for example, prepositions (e.g. under), con- tence in the mind and the highly structured design of junctions (e.g. because), pronouns (e.g. mine), deter- the mental lexicon are also corroborated by research miners (e.g. the), and numerals (e.g. first) are the into language acquisition. The surprisingly rapid primary closed word classes. The basis for the catego- development of child language is to a large extent rization of words into classes is provided by their mor- based on the acquisition of thousands of words in a phosyntactic behavior. Nouns, for instance, are almost comparatively short time. At a more or less specific always able to take a plural form (disregarding excep- stage of this development, children tend to overgener- tions). As with word identification, word class catego- alize morphological rules. For instance, they add the rization is not straightforward in all cases. An important regular past tense morpheme (-ed in English) to all example in English is, perhaps, the word class of regular verbs (learn becomes learned) and irregular adverbs. This class contains members that are extreme- verbs (go becomes *goed). This phenomenon reflects ly dissimilar in structural terms, such as the prototypical the tendency for a child’s mental lexicon to comprise adverb rapidly and the negative particle not. both basic elements (learn, go, -ed) and so-called rules Words have an internal structure themselves and of morphological parsing to combine these elements are at the same time the basic units of larger struc- and analyze the combinations (verb_present  -ed  tures. The former aspect refers to the domain of verb_past). ‘derivational morphology’, i.e. the linguistic descrip- Lexical disorders, too, shed light on the structure of tion of processes and phenomena of word formation. the mental lexicon. Aphasia—a condition of speech- The latter aspect captures the fact that words play a lessness that stroke patients often display when specif- key role in syntactic structures, which they enter in a ic areas of the brain are affected—is a compelling principle-guided fashion. Moreover, a given word in a example of such a disorder. Broca’s aphasia, for specific position tends to predict, to a large extent, the example, leads to ‘telegraphic speech’, or the omission range of words possible both before and after it. If the of function words (e.g. articles like the, prepositions probability of two words occurring next to each other like of), whereas patients suffering from Wernicke’s is significantly high, this phenomenon is called ‘col- aphasia show difficulty in lexical selection (i.e. they location’ (from co-location). very often do not produce the words they want but With collocations, corpus linguistics has definitely other words with different meanings, which often ren- broken new ground. In searching very large computer ders their speech unintelligible). collections of text (so-called corpora) for co-occur- In the light of their structural and mental signifi- rences, it is now possible to base the concept of collo- cance, it is reasonable to assume that words are the cation on an empirical foundation. By applying such most basic linguistic units. However, the overall corpus linguistic methods, it can be statistically deter- importance of words for language to function is large- mined to what extent a given word predicts another ly at odds with the yet unresolved problem of word word in its immediate neighborhood. For example, identification. What is relatively simple in written lan- strong and argument co-occur in significant frequency guage, where words are usually separated by spaces in authentic English language use. Their combination from each other (though not the case in all languages), is thus not of a random nature but constitutes a collo- is far more complicated in spoken language. Several cation, i.e. a lexical co-selection. Furthermore, specific 1185 WORD words tend to occur in a restricted range of grammati- but do not correspond to the specific vocabulary that cal patterns (a phenomenon called ‘colligation’, from an individual speaker has at his or her disposal. co-ligation). For example, naked eye almost exclusive- The object of inquiry for lexicographers and lexi- ly occurs in the grammatical pattern ‘verb/adjective  cologists is in constant flux. Productivity and imagina- preposition  definite article  naked eye’ as in visible tion in human languages are perhaps most obvious at to the naked eye. Collocations and colligations reveal the level of words. that lexical choices and grammatical choices are not independent of each other but that lexis and grammar are intricately intertwined. One could argue that the tra- References ditionally established distinction between lexis and Aitchison, Jean. 1994. Words in the mind: an introduction to the grammar should eventually be overcome altogether mental lexicon. Oxford: Blackwell. Bauer, Laurie. 1983. English word-formation. Cambridge: and be replaced with an integrated lexicogrammar (or a Cambridge University Press. ‘pattern grammar’). Coates, Richard. 1999. Word structure. London: Routledge. The linguistic level of words, i.e. lexicology, is of Cruse, D. Alan. 1986. Lexical semantics. Cambridge: Cambridge particular interest for dictionary makers or lexicogra- University Press. phers. They take an inventory of the words of a lan- Esser, Jürgen. 2000. Corpus linguistics and the linguistic sign. Corpus linguistics and linguistic theory: Proceedings of the guage but, by definition, are bound to fail to give an 20th ICAME conference, ed. by Christian Mair and exhaustive account of all words in existence at a given Marianne Hundt. Amsterdam: Rodopi. time. Although it is impossible to list all words in a Fischer, Roswitha. 1998. Lexical change in present-day English: given language, the use of dictionaries has led to a a corpus-based study of the motivation, institutionalization, high degree of standardization in orthographic and and productivity of creative neologisms. Tübingen: Narr. Hunston, Susan, and Gill Francis. 2000. Pattern grammar: a phonological word forms; they are, of course, a help- corpus-driven approach to the lexical grammar of English. ful tool for language users in general and language Amsterdam: John Benjamins. learners in particular. Dictionaries should include Katamba, Francis. 1994. English words. London: Routledge. orthographic, phonological, grammatical, and seman- Lipka, Leonhard. 1990. An outline of English lexicology: lexical tic information about words, thereby strongly resem- structure, word semantics, and word-formation. Tübingen: Niemeyer.; 2nd edition, 1992. bling the composition of the mental lexicon as Marchand, Hans. 1969. The categories and types of present-day hypothesized by psycholinguists. One ought to bear in English word-formation: a synchronic–diachronic approach. mind, however, that the mental lexicon is based on a München: Beck. list of entries and rules to combine them, whereas dic- Stubbs, Michael. 2001. Words and phrases: corpus studies of tionaries usually contain only an alphabetical list of lexical semantics. Oxford: Blackwell. Wierzbicka, Anna. 1985. Lexicography and conceptual analy- entries. Furthermore, large dictionaries represent enor- sis. Ann Arbor: Karoma. mous and idealized word lists of which no single JOYBRATO MUKHERJEE native speaker has command: they approximate the so- called ‘magnavocabulary’ of a language in its entirety, See also Aphasia; Corpus Linguistics Word Order Sentences consist of words that are combined in cer- an asterisk indicates that an order is ungrammatical: tain patterned ways to convey meaning. One important dimension of this system of combination is linear (1) A dog bit the man. order: the order in which words are combined to form (2) The man bit a dog. phrases and sentences is fundamental to the syntactic (3) *Bit a the man dog. structure of any language. The importance of word order in English, for exam- Only the word order change between (1) and (2) sig- ple, can be seen in the following sentences, in which nals a crucial meaning change, and (3) shows that words 1186  WORD ORDER cannot be combined haphazardly, but must be ordered noun–adjective order (house white), whereas SOV lan- by rules. Word order is just as important to the structure guages follow adjective–noun order (white house). Such of separate phrases as it is to a whole sentence. In correlations also support the OV vs. VO typology, English, a long nominal phrase such as these three new although there are languages that contain exceptions. white cotton sheets is grammatical, but it becomes unin- Languages differ in the degree to which they allow telligible or odd if the ordering of even one word is or exploit variations from the basic SVO, SOV, or VSO altered: *new these three white cotton sheets. word order. Variations are typically used to convey a The rules governing word order are specific to each certain style or focus, emphasize a particular word or language and may show considerable variation. phrase, or signal that a sentence has a special discourse Nevertheless, research has identified certain universal function, such as a question or command. It is not word order patterns. Typological studies of hundreds uncommon to find that one word order is standard in of languages show that each one may be classified into finite (tensed) main clauses and that another is the one of three basic word order types according to how norm in infinitive (untensed) clauses. Adverbs (such as the three main syntactic constituents—subject (S), luckily), which are optional elements, are particularly verb (V), and object (O)—are ordered within a basic flexible in their possible placements. In terms of lan- declarative sentence. According to studies by guage processing, the very first and last positions in Greenberg (1966) and Tomlin (1986), the most com- each sentence are significantly prominent positions. mon word order type is SOV, which is found in Often a phrase will be ordered first or last to empha- 37–45% of the languages studied. This order is basic size and focus attention on it. Very long phrases, such in languages such as Japanese, Turkish, and Chinese. as the three young men who washed the car for me last The SVO order of languages such as English, French, Saturday, are often positioned last or at least moved and Spanish is nearly as common, occurring in rightward, possibly because this makes them easier to 35–43%. The VSO order found in languages such as process. The first and last positions are also the most Irish, Maori, and Berber is the rarest, with estimates common ones for special question-marking particles varying from 9% to 20%. in those languages that use such particles. Logically, there are at least three other possible The second position in a sentence (‘Wackernagel word order types—VOS, OSV, and OVS. However, position’) also seems to have special significance. One these orders are virtually nonexistent. Their absence is group of languages normally classified as SVO shows often interpreted as a universal preference for ordering a basic word order in which the first position may subjects before objects. One possible functional expla- actually be occupied by any type of phrase (not just the nation for this preference may be that it reflects the way subject), as long as the verb consistently appears in in which humans arrange information cognitively and second position. This word order pattern is known as that subjects are often more salient cognitively than verb-second or V2 order and has been studied exten- objects. The relative rarity of the VSO order suggests sively in research on the Scandinavian languages, that there is a preference for keeping the verb and German, Dutch, and Yiddish. Second position also object together as a unit, as they are in the SVO and seems to be a preferred position in many languages for SOV orders. the placement of clitics, i.e. elements that cannot stand Some researchers, for example Lehmann, instead alone but must attach to an independent ‘host’ word. simplify the three basic word order types into two, mak- Finally, some languages, such as Polish, may be ing VSO a variation of an underlying SVO structure. In classified as showing ‘free word order’ at the sentence this view, all languages use either a basic VO or OV level. That is, it is possible to combine phrases in many order. Hawkins and Vennemann have also expressed different ways and still convey essentially the same this as a general head-dependent (i.e. VO) vs. depend- meaning. Also, some languages allow for discontinu- ent-head (OV) order, which can be seen to govern the ous constituents, in which the words of one phrase ordering of all types of phrases within one language. may be separated rather than appearing as one contin- Phrase-internal orderings among word categories uous string of words. These ordering phenomena pres- such as nouns (house, dog, etc.), determiners (the, a, ent interesting challenges to existing theories of both etc.), adjectives (green, dark, etc.), possessives linear and hierarchical syntactic structure. (my, your, etc.), and adpositions (in, for, of, etc.) show In summary, it is most common for languages to interesting correlations with the three basic sentential identify one basic word order for the subject, object, and word orders. For example, VSO languages tend to verb in each sentence, but to allow for some stylistic use prepositions (in in the phrase in the house), where- variations. In either case, word order is rule governed as SOV languages use postpositions (literally translat- and shows some universal tendencies that may reflect ed, the house in). Similarly, VSO languages follow possible constraints on how humans process language. 1187 WORD ORDER References Hawkins, John A. 1983. Word order universals. New York: Academic Press. Baker, Mark C. 2001. The natures of nonconfigurationality. The Lehmann, Winfred. 1973. A structural principle of language handbook of contemporary syntactic theory, ed. by Mark and its implications. Language 49. 47–66. Baltin and Chris Collins. Oxford: Blackwell. Siewierska, Anna. 1988. Word order rules. London: Croom Greenberg, J.H. 1996. Some universals of grammar with partic- Helm. ular reference to the order of meaningful elements. Tomlin, Russell S. 1986. Basic word order: functional princi- Universals of language, 2nd edition, ed. by J.H. Greenberg. ples. London: Croom Helm. Cambridge, MA: MIT Press. Vennemann, Theo. 1974. Analogy in generative grammar: the Haider, Hubert, and Martin Prinzhorn (eds.) 1986. Verb second origin of word order. Proceedings of the Eleventh phenomena in Germanic languages. Dordrecht: Foris. International Congress of Linguists, ed. by L. Heilmann. Hale, Ken. 1992. Basic word order in two ‘Free Word Order’ Bologna: Il Mulino. languages. Pragmatics of word order flexibility, ed. by D. Payne. Amsterdam: John Benjamins. SHEILA DOOLEY COLLBERG Halpern, Aaron, and Arnold Zwicky (eds.) 1996. Approaching second: second position clitics and related phenomena. See also Greenberg, Joseph Harold; Syntactic Stanford: CSLI Lecture Notes. Typology; Wackernagel, Jacob Word Sense Disambiguation A word is semantically ambiguous if it has more than (3) Idiom lists are kept. If an entire idiomatic one sense. Word sense disambiguation (WSD) is the phrase is found in the text being analyzed, it process of deciding which sense is correct in a given is assumed that the idiomatic meaning of each context. The process of WSD is illustrated by word in the phrase is more likely than indi- Rayson and Wilson’s SEMSTAT (Thomas and Wilson vidual interpretations of the words. 1996), a semantic tagger that reads in a text and (4) The domain of discourse can be an indicator. assigns a code number standing for a particular word For example, if the topic of discussion is sense to each word in that text. For each word, a lexi- footwear, then ‘boot’ is unlikely to refer to the con is first checked to see what senses that word can boot of a car. take. Many words are unambiguous, but if more than (5) Special rules have been developed for the one sense is possible for a given word, WSD tech- auxiliary verbs ‘be’ and ‘have’. niques come into play, making use of the following (6) Collocations are pairs or groups of words that types of information: frequently appear in the same context. Are any collocates of the word, suggesting a par- (1) The part of speech (POS) tag assigned by the ticular interpretation, found in its immediate CLAWS POS tagger (Garside et al. 1987). vicinity? We would, for example, prefer the For example, if ‘spring’ is a verb, we know it dental sense of ‘bridge’ if the context contains must mean ‘jump’. Wilks and Stevenson collocates such as ‘dentist’ or ‘cavity’. This (1996) have shown that POS tagging greatly technique is also called proximity disam- assists in the problem of word sense disam- biguation. The amount of text on either side biguation. of a word in which we look for collocates is (2) The general likelihood of a word taking a par- called the window. One statistical measure of ticular meaning, as found in certain frequency collocation strength is mutual information dictionaries. For example, if a corpus has (Church and Hanks 1990). 5,651 occurrences of ‘bridge’ in the sense of a bridge over a river, and only 194 occurrences SENSEVAL is an open evaluation exercise for of ‘bridge’ in the sense of a dental bridge, WSD programs, first taking place in 1998 with WSD then the simplest technique is to assume that tasks for English, French, and Italian (Kilgarriff and the more common sense of ‘bridge over a Palmer 2000). river’ is always the correct one. According to A corpus manually annotated with the correct Allen (1995), this simple technique is about sense of each word was used as a ‘gold standard’, 70% accurate over English as a whole. against which the output of each of the programs was 1188  WORD SENSE DISAMBIGUATION concerned. The SENSEVAL systems could make use (7) Use of a thesaurus to overcome data sparse- of the rich HECTOR dictionary rather than a simple ness. A recurring problem with WSD, com- lexicon. In HECTOR, for each homograph, there is a pared with POS tagging, is that there are separate entry for each sense distinction, including more word senses than syntactic categories, fields for word sense definition, POS information, and meaning a much larger amount of training examples of usage. This enabled the development of data is required. Use of a thesaurus helps the following techniques: overcome this problem, as frequencies of word classes are studied rather than those (1) Collocation filters. These were used by vari- of individual words. ous authors, the main variations being in (a) In the next SENSEVAL evaluation, the WordNet the window length and (b) whether to meas- hierarchy will be used rather than the HECTOR dic- ure word–word, word–phrase, or noun–argu- tionary. Other interesting WSD methods described in ment collocations. In order to calculate the literature are as follows: collocation statistics for verbs and the nouns they are associated with, shallow parsing is (1) Gale et al. (1992) used machine-readable required beforehand. texts and their translations, noting, for exam- (2) Manually created idiom extraction patterns, ple, that the sense of ‘drugs’ which translates such as ‘shake in \w* (shoes|boots|seat)’ into French as ‘mediacaments’ collocates would cover any of the (variable form) with ‘prescription’, ‘patent’, and ‘generic’, idioms ‘shake in your boots’, ‘shake in one’s while the sense which translates as ‘drogues’ seat’, etc. collocates with ‘abuse’, ‘paraphernalia’, and (3) Matching dictionary example sentences with ‘illicit’. a window of input text. A score is given for (2) Biber (1993) used the multivariate statistical the number of identical words occurring in technique of factor analysis to discover the the dictionary example and the text window. four basic senses of the word ‘right’ in a cor- The sense with the highest scoring example is pus, according to their various collocates. chosen. Similarly, one can count matches that involve semantically related words (such as References words with the same Roget’s thesaurus cate- gories) in the matching score, rather than Allen, James. 1995. Natural language understanding. Redwood insisting on exact word matches. Various City, CA: Benjamin Cummings. Biber, Douglas. 1993. Co-occurrence patterns among colloca- measures of similarity between test sequences tions: a tool for corpus-based lexical knowledge acquisition. and dictionary examples (glosses) have been Computational Linguistics 19(3). 531–8. suggested. Church, Kenneth Ward, and Patrick Hanks. 1990. Word associ- (4) Conceptual density. All possible senses ation norms, mutual information and lexicography. of all content words in the input sentence Computational Linguistics 16(1). 22–9. Fellbaum, Christiane. 1998. WordNet, an electronic lexical are marked in a hierarchy such as Word- database. Cambridge/London: MIT Press. Net (Fellbaum 1998). The portion of the Gale, William, Kenneth Ward Church, and David Yarowsky. hierarchy with the greatest concentration 1992. A method for disambiguating word senses in a large of marked nodes (including one for the corpus. Computers and the Humanities 26(5–6). 415–39. word being tested) will reflect the sense of Garside, Roger, Geoffrey Leech, and Geoffrey Sampson (eds.) 1987. The computational analysis of English, a corpus- the test word. based approach. London: Longman. (5) Development of decision trees. According to Hug, Marc. 2000. Partial disambiguation for very ambiguous the nature (word classes or individual words) grammatical words. Journal of Quantitative Linguistics 7(3). of the neighbors of the test word, decisions 217–26. are taken in an order designed to yield maxi- Kilgarriff, Adam, and Martha Palmer (Guest Editors of the Special Issue on SENSEVAL). 2000. Computers and the mum information at each branch point, and to Humanities 34. 127–34. determine the sense of the test word in as few Thomas, Jenny, and Andrew Wilson. 1996. Methodologies for steps as possible. A similar technique for studying a corpus of doctor–patient interaction Using cor- French grammatical words is described by pora for language research, ed. by Jenny Thomas and Mick Hug (2000). Short, 92–109. Harlow: Longman. Wilks, Yorick, and Mark Stevenson. 1996. The grammar of (6) A number of systems use machine learning sense: is word-sense tagging much more than part-of-speech techniques for finding the combination of fea- tagging? Technical Report CS-96-05, University of tures (e.g. collocates, POS of words in the Sheffield. window) most likely to reveal word sense. MICHAEL P. OAKES 1189 WORKING MEMORY Working Memory The term ‘working memory’ (WM) refers to the The visuospatial sketch pad (or scratch pad) has a assumption that some form of temporary storage and visual component, which is concerned with the analy- manipulation of information within the time range of a sis of color and shape, and a spatial component, which few seconds is necessary for performing virtually all is concerned with location. Besides the important role cognitive operations. Thus, understanding and analyz- of the prefrontal cortex for both components, the visu- ing a sentence, performing a mental arithmetic prob- al component has its neuronal correlates in bilateral lem, or recalling a telephone number requires the occipitotemporal brain regions whereas the spatial activation of the working memory system. component has its neuronal substrate in right occipi- A multicomponent WM system was proposed toparietal regions. as an alternative conceptualization of the so-called The phonological loop, which also comprises two short-term memory, considering the unitary system components, is responsible for the storage and manip- and experimental results, which led to its abandon- ulation of sounds. The first component is a phonologi- ment as a concept. The most commonly accepted WM cal store, which is capable of holding traces of acoustic model was proposed by Alan Baddeley and col- or speech-based material for 1–2 seconds. The phono- leagues during the 1970s and consists of three main logical store receives directly and unavoidably any components: an attentional-controlling system (cen- information auditorily presented and stores it in terms tral executive) aided by two slave systems responsible of a sound-based code. It has its neuronal correlates in for the temporary storage and manipulation of either the left posterior parietal cortex and is also able to visual/spatial material (visuospatial sketch pad) or receive visually presented items, but these must first be acoustic/speech-based material (phonological loop) converted into an articulatory form before gaining (Figure 1). access to the store. The second component is an articu- The central executive, the most important yet least latory control process, which refreshes memory traces well-understood component of the WM, is a limited- by subvocal rehearsal of phonological information. The capacity processor that is responsible for providing the neuronal substrate for subvocal rehearsal lies in left link between the slave systems and the so-called long- hemispheric speech areas, including Broca’s area, the term memory. It is assumed to be an attentional con- premotor cortex, and the supplementary motor area trol system and is responsible for online data storage (SMA). Prevention of subvocal rehearsal results in very and the selection, initiation, and termination of pro- rapid forgetting. cessing routines (e.g. encoding, storing, and retriev- Convincing evidence for the existence of such dif- ing). Recent findings point to distinct areas of the ferent components included in the WM model comes prefrontal cortex of the brain as neuronal substrate for the central executive (Figure 2), which controls the Parietal activity of the more posterior sensory areas that also Frontal 3 lobe contribute to WM. The prefrontal cortex appears to lobe 2 hold relevant information online as well as to perform 4 Occipital lobe complex processing functions. 1 6 7 5 VISUOSPATIAL PHONOLOGICAL SKETCH PAD LOOP Temporal Visuospatial Phonological lobe CENTRAL EXECUTIVE Store Store Figure 2. Left-hemispheric scheme of the brain lobes. In the frontal lobe, schematic positions of left-hemispheric brain areas Visuospatial Articulatory Control Processes Processes are indicated, which are important concerning WM: Broca’s area (1); premotor cortex (2); supplementary motor area (3); Figure 1. Scheme of the multicomponent WM model after dorsolateral prefrontal cortex (4); orbital prefrontal cortex (5); Baddeley (1986). primary auditory cortex (6); primary visual cortex (7). 1190  WORKING MEMORY from studies of brain-damaged patients with specific as in (3b), which are even more difficult to comprehend memory impairments. Some patients show selective due to greater demands on WM. Subjects with low deficits of auditory verbal WM, which was attributed WM span have substantially more problems in under- either to selective impairment of the phonological standing the more difficult sentences, and furthermore store or of rehearsal processes. Other patients have a need more processing time. specific impairment of the visuospatial sketch pad or of the central executive. (1) a. The dog chased the child. b. The child was chased by the dog. (2) a. The child chased the dog that jumped from WM and Language Processing the table. Concerning the relationship between WM and lan- b. The dog that jumped from the table chased guage processing, the phonological loop and the cen- the child. tral executive are the most important components of (3) a. The dog that chased the child jumped from the model. In fact, the visuospatial sketch pad might the table. also play a role in some aspects of language process- b. The dog that the child chased jumped from ing, for example, in text comprehension in as much the table. as mental imagery might be involved. The functional WM is studied in humans mainly by the use of role of the phonological loop is probably to facilitate behavioral paradigms, most recently in conjunction long-term phonological learning, as is needed to with brain-imaging techniques. It has been shown that acquire both native and second language. Indeed, WM relies on cooperation among distributed areas of phonological learning is probably a primary determi- the brain, with the precise regions depending on nant for successful first-language acquisition. It is whether tasks entail remembering objects, locations, necessary for the acquisition of new vocabulary and or words. The prefrontal cortex is apparently working language comprehension as a whole. The function of as the coordinator of the activity of these various the central executive is the inhibition of irrelevant regions. information as well as activation and maintenance of information relevant during language comprehen- sion. While listening to human speech, WM is used to References hold the segments of sentences ‘on-line’ millisecond by millisecond. WM is necessary to carry forward, in Baddeley, Alan D. 1986. Working memory. Oxford: Oxford University Press. real time, the subject of a sentence and associate ———. 1999. Essentials of human memory. Hove: Psychology it with verbs and objects in order to comprehend the Press. sense and meaning of sentences. Some of the best ———. 2000. Short-term and working memory. The Oxford evidence for WM playing a substantial role in lan- handbook of memory, ed. by Endel Tulving and Fergus I.M. guage processing comes from studies that show Craik. New York: Oxford University Press. Baddeley, Alan D., and Graham J. Hitch. 1974. Working mem- substantial correlation between the measure of a per- ory. The psychology of learning and motivation, Vol. 8, ed. son’s verbal working memory capacity and the by Gordon A. Bower. New York: Academic Press. person’s performance on language-understanding Caplan, David, and Gloria S. Waters. 1999. Verbal working tasks. Therefore, subjects are divided into those with memory and sentence comprehension. Behavioral and either high or low WM span, with the latter being more Brain Sciences 22. 77–126. Daneman, Meredyth, and Patricia A. Carpenter. 1980. likely to be misled by inappropriate context. Individual differences in working memory and reading. Generally, persons with a high working memory span Journal of Verbal Learning and Verbal Behavior 19. 450–66. seem to be better at language comprehension and syn- Gathercole, Susan E., and Alan D. Baddeley. 1993. Working tactic analysis. memory and language. Hove, Sussex: Erlbaum. To test the influence of WM load on comprehension, Gazzaniga, Michael S. (ed.) 2000. Cognitive neuroscience: a reader. Malden: Blackwell. some common paradigms in language research are Just, Marcel, and Patricia A. Carpenter. 1992. A capacity theory used such as the comparison of simple subject– of comprehension: individual differences in working memo- verb–object constructions (1a) with passive construc- ry. Psychological Review 99. 122–49. tions (1b), which are harder to understand. Similarly, Miyake, Akira, and Priti Shah (eds.) 1999. Models of working right-branching structures (2a) are easier to understand memory: mechanisms of active maintenance and executive control. New York: Cambridge University Press. than center-embedded sentences (2b), since the main Roberts, Angela C., Trevor W. Robbins, and Lawrence clause subject has to be kept in mind till the end of the Weiskrantz (eds.) 1998. The prefrontal cortex: executive and embedded phrase. Furthermore, sentences with embed- cognitive functions. New York: Oxford University Press. ded relative clauses are used (3a) or constructions such SABINE WEISS 1191 WRITING SYSTEMS Writing Systems Writing is distinguished from other forms of pictorial and Cyrillic alphabets (the last used, for instance, for representation in that it relates directly to linguistic Russian) are all at least in principle alphabetic writing expression. Thus, a picture of a man feeding a dog, no systems, so that in regularly spelled English words like matter how eloquent, is not writing. Writing as a ran, run, bun, but, each orthographic symbol (a, b, r, n, means of representing linguistic expression probably t, u) corresponds to a single phoneme. The earliest first arose in Mesopotamia at the end of the fourth mil- alphabetic writing systems departed from this pattern lennium BC, although some recent discoveries suggest in an interesting and systematic way: they wrote only that the earliest writing in Egypt might be contempo- consonants. This consonantal writing system survives, raneous or even earlier. with certain modifications, in the modern Arabic and From a linguistic point of view, writing systems can Hebrew writing systems. The Arabic word as written be classified in terms of the kinds of linguistic ele- in (g) represents only the consonants [k], [t], and [b], ments that form the basic units of the system, with a in that order (although read from right to left in basic division into logographic, syllabic, and alphabet- Arabic). The word would most probably be read as ic types, although there are also intermediate types kataba ‘he wrote’, although it could in principle also (such as alphasyllabic) and writing systems that com- be read as kutiba ‘it was written’. It is possible to add bine more than one type. The scale logographic–syl- diacritics to indicate the vowels, as in (h) for kataba labic–alphabetic also corresponds approximately to and (i) for kutiba, but it is not usual to do so, since the the historical development of writing systems. correct interpretation can usually be derived from the In a logographic system, there is a separate symbol context by someone with native-like fluency in read- for each morpheme of the language in question. The ing the language. The first alphabet to use both conso- closest modern writing system to a pure logographic nant and vowel symbols consistently was the Greek system is that of Chinese. In (Mandarin) Chinese, the alphabet. words for ‘lake’ and ‘pot’ are both pronounced alike, hú, but have written forms that are completely differ- ent, as in examples (a) and (b), respectively. Although (g) (h) (i) the term logographic might suggest one symbol per word, this is not strictly accurate, since a word con- An interesting halfway stage between alphabetic sisting of more than one morpheme, such as wo E-men and syllabic writing is found in the so-called alphasyl- ‘we’, literally I-PLURAL, is written with two symbols, labic writing system, found in most of the indigenous one for each morpheme, as in (c). scripts of South Asia (e.g. Devanagari, Bengali, Tamil), South-East Asia (e.g. Burmese, Thai), and Ethiopia and Eritrea (e.g. Amharic). Here, the basic (a) (b) (c) symbols of the writing system represent consonants, but modifications are added to indicate vowels, thus In a syllabic system, there is a separate symbol for giving rise to complex symbols that represent sylla- each syllable. Japanese uses two syllabaries (sets of bles. Examples (j), (k), (l), and (m) are from the syllabic symbols), illustrated here with the hiragana Devanagari script, used for writing modern Hindi, and syllabary. The symbols in (d), (e), and (f) are, respec- represent, respectively, [ke], [pe], [ku], and [pu]. The tively, the syllables [ta], [to], and [no]. Note that there vowel [e] is represented by a kind of hook above the is no similarity between the symbols for [ta] and [to] consonant symbol, and the vowel [u] by a kind of hook corresponding to the shared initial [t], and no similar- below. ity between the symbols for [to] and [no] correspon- ding to the shared vowel [o]. (j) (k) (l) (m) (d) (e) (f) Occasionally, one finds orthographic representation of distinctive features, i.e. the individual phonetic In an alphabetic system, there is a separate symbol parameters that make up a phoneme, although only in for each phoneme of the language. The Latin, Greek, particular parts of the system. In Czech, for instance, 1192  WRITING SYSTEMS which uses the Latin alphabet, palatalization is indi- English is notorious for its departures from the reg- cated by means of a so-called hook placed above the ular principles of alphabetic writing, so that we find nonpalatalized correspondent, so that alongside s, z, words with the same pronunciation but different and c (the last like the initial consonant of tsetse), one spellings (e.g. bear, bare), words with the same has S (like the initial consonant of ship),  (like the spelling but different pronunciations (e.g. lead as in I final consonant of rouge), and J (like the initial conso- will lead you to the restaurant; lead as in lead piping), nant of chip). as well as words whose spelling seems to bear no rela- Few writing systems correspond exactly to one or tion to their pronunciation (e.g. eye). As Bernard Shaw other of the ideal types. The Chinese writing system, once observed, from the spelling of the [f] of enough, although basically logographic, makes some conces- the [I] of women, and the [S] of nation, one might sions to pronunciation, so that like-sounding words deduce that fish ought to be spelled ghoti! This con- sometimes incorporate an identical element. In trasts with some other languages using the Latin Mandarin Chinese, the pronunciation huáng corre- alphabet, where the phonemic principle is strictly sponds to the three meanings (among others) ‘emper- observed, as in Finnish, or at least more strictly or’, ‘locust’, and ‘afraid’, which use the symbols observed, as in Spanish. shown in (n), (o), and (p), respectively. It will be seen It is interesting to ask why such discrepancies that the symbols for ‘locust’ and ‘afraid’ include as between spelling and pronunciation exist. Occasionally, their right-hand component the symbol for ‘emperor’. they simply reflect random decisions by scribes that In addition, symbols can be used in writing foreign happen to have gained acceptance into the language, as words purely for their phonetic value; thus, the in the spelling delight, where the gh has no justification, Chinese for Berlin is bólín, written as in (q), where the either synchronically or diachronically. In many two symbols normally represent, respectively, instances, however, irregular spellings reflect earlier ‘cypress’ and ‘grove’. This is essentially the same as pronunciations, i.e. the spelling was originally justified the rebus principle often used in children’s picture in terms of the pronunciation at that time, and in the writing puzzles, where for instance a picture of a bee meantime the pronunciation has changed but the might represent the like-sounded verb ‘be’. spelling has remained conservative. The gh of such words as light and bright was pronounced like the ch sound in loch until medieval times in English, and (n) (o) (p) (q) indeed it is still so pronounced in some Scottish dialects; similarly, until medieval times the initial con- Although vowels are not universally written in sonants of knee and gnaw were pronounced. Where Arabic, long vowels nearly always are, a departure another language has or had prestige in a community, away from a purely consonantal writing system, so loan words may retain, in whole or in part, the spelling that the word kaatib ‘secretary’, as in (r), does differ in of the original language, which accounts for the initial p writing from item (g). in psychology (from Greek, the source of much English scientific terminology). One recurrent deviation that is found from phone- mic writing in many languages of the world is the (r) application of the so-called morphophonemic princi- ple, whereby the same morpheme is written in the Probably the most complex mixed script in use same way even when its pronunciation changes. In today is that of Japanese, which uses (simplifying English, for instance, the regular plural suffix is writ- somewhat) logographic symbols of Chinese origin for ten -s even though it is sometimes pronounced [s] roots (e.g. hito ‘man’, mi- ‘see’), syllabic symbols (after voiceless consonants, e.g. cats), sometimes [z] from the hiragana syllabary for particles and suffixes (elsewhere, e.g. dogs, zebras). Likewise, unstressed (e.g. nominative ga, accusative o, past tense -ta), and vowels in English are often reduced in pronunciation syllabic symbols from the katakana syllabary for loan- to schwa, but the same spelling is retained as when the words from other languages (e.g. doa ‘door’), as illus- vowel is stressed; compare the spelling of both stem trated in (s). vowels in torrent, stressed on the first syllable, and torrential, stressed on the second syllable. Another example of morphophonemic spelling in English is the hito ga do-a o mi-ta. invariant spelling of the root in the adjective divine and man NOMINATIVE door ACCUSATIVE see-PAST the derived noun divinity, even though the stressed ‘The man saw the door.’ vowel is pronounced [ai] in the former, but [I] in the (s) latter. 1193 WRITING SYSTEMS It is often difficult to tell whether certain spellings even before the advent of audio and video recording, are really historical or morphophonemic, since mor- the preservation of texts for future generations and their phophonemic alternations often reflect sound changes transmission to distant locations. The decipherment of that have differentially affected an originally identical such ancient writing systems as Egyptian hieroglyphs sound in different environments. Thus, the spelling of in the nineteenth century and Mayan glyphs in the the Russian word for ‘wave’ has an o in the first twentieth century is not only an intellectual achieve- unstressed syllable in the singular, as in (t) (corre- ment on the part of the decoders, but has also provided sponding to volna in Latin script), even though this us with new insights into societies of the past. unstressed vowel is pronounced [a]. This can be explained both historically and morphophonemically. References Historically, because this was an earlier pronunciation, still preserved in northern dialects, whereas other Daniels, Peter T., and William Bright (eds.) 1996. The world’s dialects have merged earlier unstressed [o] and [a]. writing systems. New York: Oxford University Press. Hooker, J.T. 1990. Reading the past: ancient writing from Morphophonemically, because this vowel is stressed cuneiform to the alphabet. Berkeley: University of and pronounced [o] in the plural, as in (u) (correspon- California Press and London: British Museum. ding to volny in Latin script). Pope, Maurice. 1999. The story of decipherment: from Egyptian hieroglyphs to Maya script, revised edition. London: Thames and Hudson. Robinson, Andrew. 1995. The story of writing: alphabets, hiero- glyphs and pictograms. London: Thames and Hudson. Sampson, Geoffrey. 1985. Writing systems: a linguistic intro- The development of writing must be seen as one of duction. London: Hutchinson. the greatest achievements of humanity, permitting, BERNARD COMRIE 1194  Y Yémba and the Grassfields Bantu Languages The Grassfields Bantu Languages Yémba The Grassfields languages form the principal sub- Yémba is a Bamiléké language spoken by over 300,000 group of Southern Bantoid, within the Niger-Congo people in the region around the town of Dschang [t∫aŋ], phylum that dominates sub-Saharan Africa. Some 65 and the language is most widely known by this latter Grassfields languages are spoken in the mountainous name. About ten dialects have been identified, the most plateau of the western and northwestern provinces of significant being those of Bafou to the north and Foréké- Cameroon (see Figure 1), the region that is thought to Dschang to the south. A Roman-based orthography was be the origin of the original Bantu expansion. These established in the early 1930s and literacy was promot- languages exhibit predominantly subject–verb–object ed first at a school in the paramount chief’s compound, word order, a noun class system (typically prefixes), and later in mission and government schools. Around noun class agreement between words in the noun the time of independence in 1960, education in local phrase, and complex tone systems. Many of these fea- languages was forceably halted because it was viewed tures are illustrated below for Yémba. The segmental as tribalism and an obstacle to nation building. By 1980 inventories of Grassfields languages typically include the political climate had changed and the development (prenasalized) stops, fricatives, affricates, liquids, of local languages was linked to the construction of a nasals and glides with labial, labiodental, alveolar, new noncolonial pan-African identity. A national alpha- palatal, velar and glottal places of articulation. Some bet based on the International Phonetic Alphabet (IPA) languages also exhibit labiovelars and implosives. and derived from the Africa script was introduced, and Key distinguishing features of Grassfields languages the Yémba orthography was duly modified. The follow- with respect to narrow Bantu are the simplified verbal ing text illustrates the orthography, where acute accent morphology and the smaller number of noun classes, marks high tone and macron marks mid-tone. the loss of final vowels from Proto-Bantu roots, the   ‘floating’ tones that interact in highly complex ways Kaŋ pɔ´ mbhJ ¯ é lelá’ ŋ´gɔ¯ méso  , m´bú n´z ŋε´ ta’ enɔ. Pɔ´ with their surroundings, and the addition of a high cen-  n´naŋ tε eshJ lelá’ ¯ ’ amɔ  ¯’ ál ’í, m´ bε´ á ápa, n´dɔk ŋgJ ¯ ɔ´ á ŋ´ka’ ŋiŋ njJJ´ a apuma. Pɔ´ le ge¯ é to¯ á m´ba, ŋ´gɔ´ ɔ mbɔŋ. tral vowel to the five-vowel system of Proto-Bantu. Pɔ´ le gJ ¯ ɔ´ tε ŋkó éwú, kaŋ á le mε´ mbhJ¯ ŋ´gε Esó, pá’ Within Grassfields, three main subfamilies are dis- meŋ ŋkJɔ´k m´ biŋ nzéŋ n´za¯ŋné la¯, meŋ ε¯ kɔ´’á áthJ ¯, tinguished: Mbam-Nkam (34 languages), Ring (16 é ka¯p, o gɔ´ á és¯ n´nɔŋ mbε´ tε´ n´néŋ á n´te¯ á ápa. languages), and Momo (eight languages). Within Mbam-Nkam four subgroups are distinguished, the Yémba lacks the labiovelars and implosives found largest being the Bamiléké languages (the name in some of the other Grassfields languages, and it has ‘Bamileke’ is a corruption of mbə ləkJɔ, literally peo- augmented the Proto-Grassfields six-vowel system ple of the ravine). with a series of mid-vowels. Diachronic processes, 1195 YÉMBA AND THE GRASSFIELDS BANTU LANGUAGES Cameroon Language Map Legend Afro-Asiatic Nilo-Saharan Adamawa-Ubangian West-Atlantic Benue-Congo Grassfields Bantu 0 100 200 300 400 Km Figure 1. Cameroon language map showing principal families. such as the loss of final consonants and the devoicing The investigation of Grassfields tone systems has of high vowels, are still manifested synchronically in stimulated important theoretical advances in autoseg- some complex morphophonological processes. mental phonology. Notable consequences of these processes are highly complex tonal patterns, and heavily aspirated sylla- bles. This aspiration is most striking after voiced References obstruents (e.g. /ghíε´ / [γiε´] fly!). Bird, S. Orthography and identity in Cameroon. Written Like the other Grassfields languages, Yémba Language and Literacy 4. 131–62. exhibits lexical and grammatical tone. The following ———. 1999. Dschang syllable structure. ed. by H. van der nouns have a lə´ - class 5 prefix, followed by the təŋ syl- Hulst and N. Ritter, The syllable: views and facts, Studies in lable that carries H, HL, LH, and L tones (from Proto- Generative Grammar, 447–76. Berlin: Mouton de Gruyter. ↓ Bird, S., and Tadadjeu, M. 1997. Petit Dictionnaire Bantu HH, HL, LH, and LL): lə` tɔ Eŋ [-] feather, lə` t ɔ´ŋ Yémba–Français (Dschang–French Dictionary). Cameroon: [--] reading, lə` tɔ` ŋ [--] navel, and lə` tɔ` ŋ [- ] finishing. \ ANACLAC. These examples demonstrate that tone is a lexical prop- Hyman, L.M. 1972. A phonological study of Fe?fe?-Bamileke. erty of words. However, tone in Yémba also has a Studies in African Linguistics (Supplement 4). grammatical function. The following examples illus- ———. 1980. Relative time reference in the Bamileke tense system. Studies in African Linguistics 11. 227–37. trate how tonal distinctions alone can convey tense dis- ———. 1981. Noni grammatical structure. Occasional papers tinctions. The words used in the examples are: ə`-fɔ` in linguistics, Vol. 9. University of Southern California. CL1-chief, kə` mtè bury, and mə` m-bh↓J´ CL10-dogs (ə`- ———.1985. Word domains and downstep in Bamileke- and mə` m- are class 1 and 10 noun class prefixes). The Dschang. Phonology Yearbook 2. 45–83. words ɔ´ and ɔ` are known as concord markers, serving Hyman, L.M., and M. Tadadjeu. 1976. Floating tones in Mbam- Nkam. Studies in Bantu tonology, ed. by L.M. Hyman, no purpose other than to link the subject and verb. 57–112. University of Southern California. Occasional (1) ə` fɔ` ɔ´ ↓kə` mtè ↓mə´ mbhJ´ [-|--|---|-] ‘the chief Papers in Linguistics, Vol. 3. Voorhoeve, J. 1971. Tonology of the Bamileke noun. Journal of buried dogs (immediate past)’ African Languages 10. 44–53. (2) ə` fɔ` ɔ` kə` mtè mə` mbh↓J´ [-|--|---|-] ‘the chief Watters, J.R., and Leroy, J. 1989. Southern Bantoid. The Niger- buries dogs (simple present)’ Congo languages, ed. by J. Bendor-Samuel, 430–49. (3) ə` fɔ` ɔ` ↓kə` mtè ↓mə` mbhJ´ [-|--|--|--] ‘the chief Lanham, MD: University Press of America. will bury dogs (immediate future)’ STEVEN BIRD 1196  YIDDISH Yiddish Yiddish is the historical language of the North in the ‘Jewish Autonomous Region’ of Eastern Siberia European ‘Ashkenazic’ Jews and their descendants (Russia) formed after the Revolution, and in Soviet worldwide (Ashkenaz denoted an Iranian people in the Belarus’ and Ukraine (until the early 1930s), it has Bible, which in the Middle Ages was used to refer to been the object of organized standardization efforts in Slavic, and then, by about 1100, to Germans and Europe and America since 1908. German Jews). Since the 1600s, the language has been In addition to being Indo-European, Yiddish is a called jidiš ‘Jewish’ (earlier native names are member of a genetically mixed ‘family’ of some two unknown), and more recently mame-lošn (‘mother dozen languages adapted by Jews from non-Jewish tongue’) and D argon (‘jargon’, often contemptuous in linguistic stock, such as Arabic, Berber, Chinese, languages other than Slavic). Greek, Iranian, Romance, and Slavic, and it therefore Yiddish is traditionally regarded as a derivative of owes its raison d’être to common processes of forma- High German, which first formed in southern tion and (Hebrew) enrichment. Of the Jewish lan- Germany c. 1000 CE. In the thirteenth century, guages, Yiddish has by far the largest number of Yiddish speakers began migrating to Poland; dialects speakers, and the richest literature. Yiddish was taught spoken there and in Belarus’, Ukraine, Russia, the in the early 1700s in German universities, and since Baltic lands, and Romania are called ‘Eastern the twentieth century has stood at the cutting edge of Yiddish’, which is used by most contemporary speak- modern linguistic theory, particularly since top theo- ers of the language. Speakers in Holland, Hungary, the reticians such as Edward Sapir, Roman Jakobson, Max German-speaking and Bohemian-Moravian lands, and Uriel Weinreich, and Edward Stankiewicz made it however, use what is known as ‘Western Yiddish’. their occasional or sole area of interest. Yiddish was Although the two spoken Yiddishes were always dis- the first Jewish and Soviet language to be the subject tinct, the Eastern Yiddish literary language was large- of a linguistic atlas (see Vilenkin 1931, and now ly based on Western norms until the early nineteenth Herzog et al. 1992–2000). century. A new view holds that Eastern and eastern Taking Western and Eastern Yiddish as a whole, Western Yiddish (in the originally bilingual eastern Yiddish enjoys the broadest expanse of any European German and Slavic lands) are instead a multilayered language, with the exception of Russian and perhaps Slavic language with a predominantly German lexicon Romani (Gypsy); in the sixteenth century Yiddish was formed between the ninth and twelfth centuries, while spoken from Holland to Ukraine, and from the Baltic western Western Yiddish (best called ‘Judeo-German’) Sea to Northern Italy and in Ottoman Palestine. While is Germanic, having been formed by French Jews in Yiddish was obsolescent in its German homeland and the Rhineland between the tenth and late fourteenth western Hungary from about 1750 (being displaced by centuries (and now extinct). Both Romance and Slavic standard German, after passing optionally through a Jewish settlements in Germany (the latter of local and stage of slightly Yiddishized German until the early Balkan stock) date to the ninth to tenth centuries. twentieth century), it reached Russia and America in the Hence, Yiddish is the only living Indo-European lan- late eighteenth to early nineteenth century, and by the guage whose genetic assignment is still being debated. late nineteenth century had arrived in the other inhabit- The mass migration of speakers to Palestine and the ed continents, including remote Harbin (Manchuria) Americas, beginning in the 1880s, resulted in most and Shanghai. In an unusual sociolinguistic develop- migrants eventually ‘losing’ their language; they ment, Yiddish is now obsolescent for the most part stopped speaking their first language and switched to except among ultra-orthodox Jews. the predominant language of their new homeland. Yiddish is largely written phonetically in the Language shift from Yiddish to Russian was also rapid Hebrew (Aramaic) alphabet, although Hebraisms and in the Soviet Union, without outward migration. Aramaisms are spelled etymologically (except in Nevertheless, in 1939, Yiddish speakers probably Soviet Yiddish, which uses an entirely phonetic numbered ten million worldwide. Accelerated by the spelling). Eastern Yiddish generally comprises approx- Nazi German extermination of the European Jews, the imately 75% Germanisms, 15% Hebraisms, and 10% number of speakers today, found primarily in Israel Slavisms; in German-speaking milieus, Yiddish often and the Americas, has dropped to around one million. acquires a much higher Hebrew component so as not to Although Yiddish never enjoyed official status except be easily understood by German speakers (some of 1197 YIDDISH these Hebraisms entered German beginning with the meanings were assigned by the original Slavic vocab- 1400s, then—through German—into other European ulary they replaced (see Horvath and Wexler 1997). In slang registers). There are also small unique Romance, the kind of vocabulary replacement known as relexifi- Greek, Turkic, and Iranian components. cation, the substratal Slavic semantic and derivational The origin of Yiddish has recently become a key parameters are kept, so that one can predict with con- topic in Yiddish linguistic discussion. In the ‘tradition- siderable accuracy which superstratal German compo- al’ model of Yiddish genesis, proposed by Max nents can be accepted—unlike straightforward Weinreich (1973), the language’s unique Romanisms bilingual interference, which is always unpredictable suggest the founders of Yiddish came from France and and usually involves vocabulary exclusively. Parts of Italy. Weinreich envisioned Yiddish born in the the German lexicon incompatible with Sorbian are Rhineland and Regensburg, then largely reshaped in blocked in Yiddish, to be replaced by genuine and Bavaria-Franconia, prior to spreading into monolin- many newly coined Hebraisms, unrelexified Slavisms, gual Slavic lands. Seven hundred years of contact with or other acceptable Germanisms. For example, Slavic, beginning with Sorbian and Polabian in the German uses a common root for Gewitter (‘storm’) ninth century and culminating with Russian in the and Wetter (‘weather’). Wexler expects relexifiers to eighteenth century, rendered parts of Western and all accept at best one Germanism because Upper Sorbian of its Eastern Yiddish offspring heavily Slavicized. denotes the two differently; see hrimanje and wjedro, Examples of Slavic features are the possessive pro- respectively. Not surprisingly, Yiddish has only veter nouns inflected only for number; the leveling out/addi- ‘weather’. The use of separate words in Slavic makes tion of morphophonemic alternations in Germanisms; German Sturm ‘storm’ possible for Yiddish (as the distribution (and often indeclinability) of the šturem). Only relexification can account for the mass reflexive pronoun zix; two degrees of diminutive; the of Hebraisms as well as the very reduced German lex- double negative; analytic expression of comparative icon in Yiddish (and almost total impoverishment of and superlative adjectives; recalibration of German German synonyms), compared with Central and verbal prefixes; psycho-ostensive expressions; the Bavarian German dialects that conceivably were the resumptive pronoun; no verb-final position in subordi- relexifier sources for Yiddish. Following the relexifi- nate clauses; exclusive use of the present perfect to cation process that created it, Yiddish came to borrow express past time; initial s and x (in non-German com- additional German, Hebrew, and Slavic lexicon wihout ponents); the absence of [ç]; and final voiced conso- reference to Slavic grammar. nants (the last feature was probably original, and not Yiddish speakers have an enormous Hebrew- restored under Slavic influence). While some of these Aramaic lexicon (in all semantic domains) and a mor- features are found in German dialects (often spoken in phology used mainly with Semitisms, unmatched by former Slavic areas), Eastern Yiddish almost never has any other Jewish language. This is because the blockage ‘German’ features unknown to Slavic. of so much German vocabulary obliged Yiddish speak- Since the late 1980s, the obvious similarity of ers to invent many Hebrew forms and/or meanings as Yiddish to Bavarian dialects and the lack of features replacements, such as Yiddish nadn (‘dowry’) and from southwestern German led linguists to favor a xmime (‘extreme heat’—an invention of Yiddish speak- (south)eastern birthplace. The Danube rather than the ers, based on Slavic lexicons which, unlike German, Rhine venue is further supported by the likelihood that have a word for ‘extreme heat’ that differs from the the few unique old Romance elements all derive from usual word for ‘hot’) vs. Hebrew na¯da¯n (‘sheath’) and Balkan Romance languages, Friulan, Rhaeto- am ‘hot’. The Yiddish pronunciation of Hebrew pre- Romance, or Italian and hardly ever from French, dates the emergence of Yiddish, but the origins remain except in Germanic Western Yiddish. The origin, lan- unclear. Yiddish Bibles, first attested in Germany in the guage, and fate of the Jews in fourth-century Roman 1400s, reveal many atypical Germanisms, and slavishly Rhineland are unknown. follow Hebrew syntax and derivational patterns; hence, Between 1991 and 1993, while accepting an eastern they are really ‘relexifications’ of Biblical Hebrew into German venue for the birth of Yiddish, Wexler pro- Yiddish lexicon, i.e. they utilize the original Old posed that Sorbian Jews in Eastern Germany—com- Hebrew grammar but replace almost all of the Old prising mainly local and Balkan proselytes and some Hebrew lexicon by German-origin words. The latter are ethnic Jews—‘relexified’ their West Slavic language to often used in ways that are unintelligible to German High German by the twelfth century. Relexification speakers. ‘Yiddish’ Bible language is best defined as a means that the Jews retained Sorbian grammar, lexically deviant variant of Old Hebrew. phonology, and phonotactics while replacing most of A Slavic affiliation for Yiddish is supported by its the original lexicon with German words—whose Slavic-type syntax, phonology, and phonotactics, as 1198  YIDDISH well as the role of the derivational morphology; only gation is suggested by its geography within Yiddish (it the lexicon (including the inflectional and most of the is much less productive in Western Yiddish and alto- derivational morphology) is predominantly German. gether unknown in German slang), and by its use in A specifically Sorbian substratum for Yiddish is sup- Judeo-Belarusian (attested in the seventeenth century) ported by the shape and geography of some unrelexi- and in the Turkic and Iranian languages spoken by fied Yiddish Slavisms and the striking parallels in the Muslims (for Arabic loans). distribution of Hebraisms in Yiddish vs. Germanisms Many scholars speak of a Western–Eastern Yiddish in Sorbian, which points to a common speech commu- dialect continuum, with Western providing the input nity. Hence, Eastern Yiddish is probably an unusual for Eastern Yiddish (Herzog et al. 2000). Yet it is also dialect of Sorbian rather than a Slavicized form of commonly assumed that Eastern Yiddish dialects German. developed in situ, rather than being imported from There is agreement that Yiddish speakers reached Western Yiddish ‘readymade’. A multilayered the Belarusian and Ukrainian lands in the 1400s, West–East Slavic affiliation for Yiddish requires where they encountered Slavic-speaking Jews. The exploring, in future, to what extent Eastern Yiddish ‘Yiddish-is-Germanic’ school believes that the latter, dialects match differences among the dialects of coter- few in number, shifted to Yiddish. In the ‘Yiddish-is- ritorial Slavic languages. Slavic’ model, the local East Slavic Jews allegedly car- Regardless of its origin, Yiddish holds an abiding ried out the process of relexification a second time, as interest for students of Jewish linguistics and history, is borne out by the existence of uniquely East Slavic bilingualism, minority ethnolects, creole linguistics, grammatical features in Yiddish, such as gender corre- German, Slavic, unspoken Ashkenazic Hebrew, spondences that differ from those found in German, Modern Hebrew, and Esperanto (the last three are and the original use of the German plural suffix -(e)n demonstrably Slavic Yiddish—the first two relexified as a dual (significantly, the Yiddish dual matches the to Old Hebrew and the third to Latinoid roots). This is distribution of the Belarusian-Ukrainian ‘pseudo-dual’ because Yiddish provides clues to Jewish ethnogenesis ending used after 2–3–4; the dual is a category that has and migrational patterns that are not recoverable from never been acquired through language contact). The the notoriously sparse historical and archeological evi- attraction to Yiddish by East Slavic-speaking Jews dence; Yiddish is an ideal laboratory for the study of attests to their high regard for Ashkenazic culture and relexification, since the dual processes of relexification desire to differentiate themselves more sharply from in that language are considerably older than the relexi- the coterritorial non-Jews; their ability to reform fication found in numerous creole languages (e.g. imported Sorbian Yiddish attests to their numerical Haitian Creole); finally, repeated exposure over a mil- superiority over the Ashkenazic immigrants. lennium to the same languages (German, Slavic, Indigenous East Slavic Jews could only be descended Hebrew) makes Yiddish a unique laboratory for stu- largely from the Turko-Iranian converts to Judaism in dents of bilingual interference. the Khazar Empire. A Khazar component in Yiddish is also suggested by a few unique Iranian and Turkic ele- ments: e.g. Yiddish šabaš ‘tip given to musicians at a References wedding by guests who join in the dancing’; šibeš Birnbaum, Salomon A. 1979. Yiddish: a survey and a grammar. ‘small coin; trifle’ stems from Persian ša¯ba¯š (in Slavic Toronto-Buffalo: University of Toronto Press; includes a and other European slang vocabularies, the word is comprehensive bibliography. formally and/or semantically distorted); nar, plural Herzog, Marvin, et al. (eds.) 1992–2000. Language and culture atlas of Ashkenazic Jewry, 3 vols. Tübingen-New York: Max naronim ‘fool’, derives from Persian ner ‘man’ rather Niemeyer and YIVO; ongoing publication. than from German Narr ‘fool’
Old Ukrainian nar’- Horvath, Julia, and Paul Wexler (eds.) 1997. Relexification in ci, nor’ci ‘Ossetian epic heroes’ (Yiddish -on- from Creole and non-Creole languages (with special attention to Old Hebrew and Old Ukrainian -ci both express an Haitian Creole, Modern Hebrew, Romani and Rumanian). agentive noun); pejsex ‘Passover’ comes from Hebrew, Wiesbaden: Otto Harrassowitz. Mark, Yudel, et al. (eds.) 1961–1980. Grojser verterbux fun der but its use as a male name for children born during the jidišer šprax, 4 vols. New York-Jerusalem: Yiddish holiday mirrors Turkic practice (the word as a name is Dictionary Committee; incomplete. first attested in tenth-century Khazar Hebrew); and a Šapiro, Mojsej Al’teroviJ, et al. (eds.) 1984. Russko-evrejskij periphrastic conjugation that combines Hebrew mas- slovar’. Moscow: Russkij jazyk. culine singular participles, now indeclinable, with a Vilenkin, Lazar. 1931. Jidišer špraxatlas fun sovetn farband afn grunt fun di dialektologiše materjaln, vos zajnen cuno- Yiddish auxiliary (e.g. Hebrew bo¯deq ‘inspecting’ jfgezamlt gevorn durx der špraxkomisje fun jidišn sekter fun became Yiddish bojdek zajn ‘inspect’, with zajn ‘be’ der vajsrusišer akademje unter M. Vejngers onfirung. Minsk: from German). A Turko-Iranian origin for this conju- Vajsrusiše visnšaft-akademje, jidišer sekter. 1199 YIDDISH Weinreich, Max. 1973. Gešixte fun der jidišer šprax 4 vols. ———. 1992. The Balkan substratum of Yiddish (a reassess- New York: YIVO; partial English translation, The history of ment of the unique Romance and Greek components). the Yiddish language. Chicago: University of Chicago Press, Wiesbaden: Otto Harrassowitz. 1980. ———. 1993. The Ashkenazic Jews. A Slavo-Turkic people in Weinreich, Uriel. 1968. Modern English–Yiddish Yiddish– search of a Jewish identity. Columbus, Ohio: Slavica. English dictionary. New York: YIVO-McGraw-Hill. ———. 2002. Two-tiered relexification in Yiddish. Jews, Wexler, Paul. 1990. References and bibliography of Yiddish lin- Sorbs, Khazars and the Kiev-Polessian dialect. Berlin-New guistics, 1979–1988. Studies in Yiddish linguistics, ed. by York: Mouton de Gruyter. Paul Wexler. Tübingen: Max Niemeyer. PAUL WEXLER ———. 1991. Yiddish—the fifteenth Slavic language. International Journal of the Sociology of Language 91. See also German; Hebrew: Biblical; Hebrew: Mod- (focus issue). 1–150, 215–25. ern; Old Church Slavonic Yoruba and Yoruboid Languages Yoruboid languages (YRB)—Yoruba, Igala, and Education, which requires high school students to Itsεkiri—are classified as the Defoid group, a branch learn a major Nigerian language in addition to their of Benue-Congo, which is a subgroup of Niger-Congo mother tongue. Furthermore, Yoruba is the official lan- languages. Yoruba, the most widely spoken of the guage in southwestern Nigeria and is used for govern- group, has over 20 million speakers in southwestern ment notices, radio, television, and newspapers. Igala Nigeria, southeastern Benin, and central and northern is spoken by approximately 800,000 people in central Togo. In Sierra Leone, the influence of Yoruba is evi- Nigeria and Itsεkiri is spoken by about 500,000 people dent in Krio loanwords and personal names. Outside in midwestern Nigeria. Igala and Itsεkiri are taught in West Africa, Yoruba is used in religious contexts in primary schools and are used in radio and television Brazil and Cuba, as well as in cultural revitalization broadcasting. The distribution of Yoruboid languages movements such as the ɔyɔtu& Ôi village, a Black is shown in Map 1. Nationalist community in South Carolina. Yoruba has over 20 distinct dialects. In Nigeria, Standard Yoruba is the literary usage and is taught in schools up to the Phonology (Sound Patterns) university level. It is also spoken by younger genera- The sound inventory of YRB includes the consonants tions as a second or third language because of the shown in Table 1. In the Stop series, voiced (sounds implementation of the 1981 National Policy on produced with the vibration of the vocal cords) and Map 1. Yoruba and Yoruboid lan- guages. 1200  YORUBA AND YORUBOID LANGUAGES unvoiced (sounds produced without vocal cord vibra- standing the phonology and syntax (sentence structure) tion) consonants are attested. However, /b/ and /7/ have of YRB languages. For example, tone can distinguish no unvoiced counterparts. The velar stops—kw and words: bí ‘give birth’, bì ‘vomit’, b ‘ask’, arò ‘lamen- gw—are found in Igala, Itsεkiri, and southeastern tation’, aro ‘cymbal’, aró ‘indigo’, àró ‘a granary’. dialects of Yoruba such as Ondo and, but they are real- Tone may also serve a grammatical function. For ized as /k/ and /g/ in Standard Yoruba and northwest- example, within a statement, a high tone must occur ern dialects. Fricatives (sounds produced with audible between the subject and the verb: Dajɔ` (a name); Dayɔ` friction) are generally unvoiced. The only voiced frica- ɔ´ ra òrom ` bó ‘Dayo bought (some) oranges’. tive, /γ/, is attested in Itsεkiri and southeastern Yoruba Tone is used in other contexts. For instance, it is dialects, and it is the equivalent of /w/ in Igala and used in whistled Yoruba, a language used to communi- northwestern Yoruba dialects. Yoruba /s/ may corre- cate over long distances on farms. As speakers talk and spond to Igala [r], [l], or [h], depending on the context. whistle simultaneously, the language is transformed: An /s/ occurring before an oral vowel (vowel pro- consonants are devoiced (produced without vocal cord nounced through the mouth) is an [r] in Igala (Yoruba vibration) or turned to [h] and all vowels are changed /εsε` / is realized as /ε´rε` / ‘foot’); when /s/ precedes a to [u] because whistling involves lip puckering, the nasal vowel (vowel pronounced by allowing air to pass articulatory gesture required for the production of [u]. out through the nose), it is realized as [l] (Yoruba /ε`sɔ´ & / However, all tones are retained without any alteration. is Igala /ε`lá / ‘nine’); an /s/ occurring before /i/ is an The retention of tones enables speakers to understand [h] (Yoruba /èsì/ is Igala /òhì/ ‘answer’). the meaning of whistled language. For instance, a Besides consonants, there are seven oral vowels: [i farmer just arriving on the farm may greet farmers in e ε a ɔ o u]; in addition, [I υ] occur in Igala and some nearby farms using any of the whistled forms in (2) Yoruba dialects. Nasal vowels are also attested. As and (3). As shown below, although the consonants and shown in Table 2, dialects of Yoruba have four to six vowels of these whistled forms are altered, they have nasal vowels, Itsεkiri has three nasal vowels, but Igala the same meaning as the nonwhistled form in (1): has no nasal vowels. (1) Nonwhistled Yoruba: ara iwájú, ε kú iʃε´ o ‘peo- Oral vowels exhibit co-occurrence restrictions. For ple who arrived before me, greetings at work’ instance, /ε/ and /ɔ/ do not co-occur with /e/ and /o/ in (2) Whistled form with devoiced consonants: ur#ú nouns: ɔ` bε ‘knife’, εbɔ ‘sacrifice’, ekpo ‘palm-oil’, uw#úÔú u kú uʃú u òkpó ‘pillar’. Consonant–vowel co-occurrence restric- (3) Whistled form with [h] replacement: uhú tions are also attested. For example, /r/, /y/, /w/, and /h/ uhúhú, u hú uhú u are nasalized when nasal vowels occur after them: yɔ´& ‘pounded yam’, ir& ‘iron’, ìwɔ´& ‘measurement’, ah& ɔ´& The Yoruba talking drum, a pressure drum, which ‘tongue’. accompanies singing during festivals and important There are three tones (pitches): High (´), Low ( `), ceremonies, also uses tone. This drum ‘speaks’ by and Mid (unmarked). These tones are crucial for under- reproducing the tones of vowels. For example, a TABLE 1 Consonants Labial Alveolar Palatal Velar Labial–Velar Glottal Stop Voiceless t kw kp Voiced b d 7 gw gb Fricative . h Voiceless f s  h Voiced  Nasal m n Liquid 1r Glide y w TABLE 2 Nasal Vowels Yoruba Itsεkiri Igala I7ea: & i& ε& O& U& u& & O& u& No nasal vowels Ila7ε: & ε& a& O& u& Standard Yoruba: & ε& O& u& 1201 YORUBA AND YORUBOID LANGUAGES notable citizen who has just returned from a trip may Consonant–Vowel syllable may bear a high tone in be greeted with these words using drums: Igala, only Consonant–Vowel syllables may have a high tone in Yoruba and Its εkiri . Hence, vowel-initial (4) Words: a ti n´ retí ìrε; káábɔ`, ʃé dáadaa lo dé words with high tones in Igala surface with mid tones ‘we have been expecting you; welcome, did in Yoruba: you arrive well?’ Talking drum: Mid Mid High Mid High Low Mid, (9) Yoruba Igala Gloss High High Low High High Mid High Mid Mid High imú ímɔ nose Because the talking drum produces only tones, its lan- etí étí ear guage could be ambiguous. For instance, the signature fújε´ fújε´fújε´ to be light tune of Radio Nigeria Ibadan, given in (5), could be tí &rí& fíílí narrow, thin assigned the interpretations in (6): (5) Talking drum: High High Low Low High Morphology (Word Structure) High Mid High High Mid Low (6) Possible interpretations Prefixation, reduplication, truncation, compounding, and desentencing are the principal word formation a. rédó nàìÔíríá la tí n´ fɔhù& devices. Although inflection does not play a role in ‘we are broadcasting from Radio Nigeria’ Yoruba and Igala, Itsεkiri distinguishes between sin- b. tólúbàdɔ`‡ bá kú ta ní ó Ôoyè gular and plural nouns, as shown below: ‘if the king of Ibadan dies, who will succeed (10) Yoruba Igala Itsεkiri Gloss him?’ ob‡r‡ ónobùlε obirε& woman c. ní &nú& kòkò dúdú la ti n´ sebε`, ob‡r‡ ónobùlε ebirε& women ‘we cook soup in a black pot’ Prefixation (the process of attaching an affix to a Syllables may be expressed as Consonant–Vowel (lɔ word to derive a new word) is a device used to derive ‘go’), Vowel (ilé ‘house’), or N, a syllabic nasal (òn`tε` nouns from existing verbs. For example, a vowel may ‘stamp’). Syllables with consonant clusters and those be prefixed to a verb or verb phrase to form a noun, as ending in consonants are unattested. Hence, when follows: English words with such unacceptable sequences are borrowed, they are restructured through consonant (11) àlɔ ‘going’ <àlɔ ‘prefix deletion and vowel insertion, as follows: go’ ɔ` ‘singer’ <ɔ` ‘prefix (7) English YRB kɔr& kɔr& sing’ kettle kε´tù buckle bɔ´kù Reduplication (process of repetition) is used to travel tíráfù form words denoting intensity; it is also used to form brick bíríkì distributive nouns, profession-based names, and names denoting endearment: The Consonant–Vowel syllable plays an important role in Yoruba phonology. For instance, εnɔ`&, a Yoruba (12) Base Reduplicated Form secret language game, disguises words by adding a Intensity: Consonant–Vowel syllable to the end of every syllable. kíá ‘quick’ kíákíá ‘quickly’ As shown in (8), the consonant of the added syllable is kpú ‘much’ kpúkpɔ` ‘very much’ fixed—a/g/; however, the vowel and tone of this sylla- kpɔ´ kpúkpɔ` ble vary, based on the vowel and tone of the base form. Another property of this game is that the final Distributive: Consonant–Vowel tag is always preceded by a syllab- où ‘month’ oooú ‘every ic nasal: month’ àg ‘elder’ àgbà ‘every elder’ (8) Nondisguised form: olú kò lɔsílé bà àgbà ‘Olu did not go home’ Olu NEG. go to house Disguised form: ogolúgú kògò lɔgɔ sígílé"´gè Profession agentive: ‘Olu did not go home’ wolé ‘examine woléwolé ‘sanitary house’ inspector’ Word-initial high tones are also sensitive to syllable mɔ& ‘build mɔ&lé ‘builder’ shapes. For example, whereas a word-initial Vowel or lé house’ mɔ&lé 1202  YORUBA AND YORUBOID LANGUAGES Diminutives: olú tu& n´ sáré ‘Olu is running adé ‘a name’ ádéadè ‘diminutive again’ form’ Olu repeat is run olú ‘a name’ ólúolù ‘diminutive form’ The basic word order is S(ubject–)V(erb–)O(bject). Truncation (shortening or abbreviation) is the device Within phrases, the head of the phrase occurs in the ini- adopted in creating shortened names, which are com- tial position. Hence, in a noun phrase, the noun occurs monly used in informal contexts: first, followed by adjectives and relative clauses. In verb and prepositional phrases, the verb and preposi- (13) Full Name Shortened Form tion appear first and their objects occur after them: adéwɔlé adé or wɔlé olúkɔ´lá olú or kɔ´lá (17) a. Word olú ra ‘Olu bought ɔlábɔ`dé ɔlá or bɔ`dé order: ìwé a book’ modúkpε´ modú or dúkpε´ Olu buy book Compounds (words consisting of two or more words) b. Noun ɔmɔ& ‘black child’ are formed from two or more nouns, as follows: phrase: dúdú (14) ìdí ‘motor ì ‘base/ ɔkɔ` ‘vehi- child black kɔ` park’ dí bottom’ cle’ c. Verb ra ìwé ‘buy a book’ εrɔ& ‘ani- εrɔ& ‘meat’ oko ‘farm’ phrase: buy book ko mal’ d. Prepositional sí oko ‘to the farm’ ɔmɔ& ‘girl’ ɔmɔ& ‘child’ ob& ‘wo- phrase: to farm b&r& r& man’ The basic SVO word order may be altered to form ɔmɔ& ‘toe’ ɔmɔ& ‘child’ ì ‘fin- focus sentences, sentences in which a given word or kas ka ger’ phrase is fronted for emphatic purposes. In some ε` εsε` Yoruba dialects, a focused word or phrase is usually ‘leg’ followed by the focus marker ni; in other dialects, the focus marker appears at the end of the sentence; in Its A rather different but common morphological process Ekiri, only fronting applies, and the focus marker is in YRB and Benue-Congo languages is desentencing, not used. The following examples illustrate subject a process used to form names. Desentencing involves fronting, derived from a basic sentence such as (17a): turning a full sentence into a name, as follows: (18) Standard (15) Sentence Name Gloss Yoruba ɔlá wɔ ilé ɔláwɔlé honor enters Subject οlú ni ó ra ìwé ‘It is Olu the house Focus: who bought adé yí mi ká adéyí"ká crown sur- a book’ rounds me Olu focus akí& dé ilé akí &déilé valor arrives 3SG buy in the house book olú dá mi sí olúdámisí the Lord kept me (19) Ukarε olú ó ra ìwéa~¯ ‘It is Olu alive Yoruba: who bought a book’ Subject Olu 3SG Syntax (Sentence Structure) Focus: buy book Nouns, verbs, and prepositions are the major word cat- focus egories. Some words, which function as adjectives and 2 3 4 adverbs in English, are expressed as verbs. For exam- ple, the adjective ‘red’ and the adverb ‘again’ are used (20) Itsεkiri as verbs in the following Yoruba sentences: Subject olú òu& ‘It is Olu Focus: é ra ìwé who bought (16) olú pukpa ‘Olu is red: Olu is a book’ fair-skinned’ Olu 3SG Olu red buy book 1203 YORUBA AND YORUBOID LANGUAGES As shown in (18) through (20), a pronoun (3SG) Long pronouns are also called independent pronouns replaces a fronted subject. In contrast, a fronted object because they behave like nouns in terms of fronting does not produce the same effect: and coordination, as shown by these examples: (21) Standard (25) Basic ìwɔ´ ra ‘you bought Yoruba Sentence: àgbádá a flowing Object ìwé ni olú rà ‘It is a book gown’ Focus: that Olu Subject ìwɔ ni o ‘it is you bought’ Focus: ra àgbáda who bought book focus a flowing Olu buy gown’ Serial-Verb-Construction (SVC) is another princi- Coordination: ìwɔ àti o ‘you and pal point of syntactic interest. SVCs have sequences of lú ra à Olu bought verbs, which share the same subject and object. They gbáda a flowing are different from coordinate clauses in that the verbs gown’ use one tense and aspect, and allow only one negative: Unlike long pronouns, short pronouns can neither be (22) a. olú ra àgbàdo .ε ‘Olu will buy fronted nor coordinated with a noun. yóó some corn and eat them’ Olu FUT Politeness buy corn eat Finally, politeness is an essential aspect of Yoruba, b. olú ra àgbàdo .ε ‘Olu did not Itsεkiri, and Igala culture, and language is used to con- kò buy and eat vey respect, especially for older people and those in some corn’ authority. Politeness is expressed by the choice of pro- Olu NEG nouns in reference and address. For instance, the plu- buy corn eat ral forms of the second and third person pronouns (ε`y¸& As shown below, it is ungrammatical to assign sepa- ‘you pl.,’ àwO& ‘they’) are considered respectful and rate tense or negative markers to a serial verb con- appropriate whereas singular forms are considered struction: impolite (ìwO ‘you sg.,’ òu& ‘he/she’). (23) *olú yóó ‘Olu will buy some ra àgbàdo yóó 7ε corn and eat them’ References Abraham, Roy C. 1958. Dictionary of Modern Yoruba. London: Olu FUT buy corn University of London Press. FUT eat Awobuluyi, Ó ladele. 1978. Essentials of Yoruba grammar. Ibadan: Oxford University Press. *olú kò, ‘Olu did not buy and eat Bamgbose, Ayó’. 1967. A short grammar of Yoruba. Ibadan: ra àgbàdo some corn’ Heinemann Publishers. kò 7ε Capo, Hounkpati B.C. 1989. Defoid. The Niger-Congo lan- Olu NEG buy corn guages, ed. by J. Bendor-Samuel. Lanham: University Press NEG eat of America. Fresco, Edward. 1970. Topics in Yoruba dialect phonology. YRB languages typically divide pronouns into two Studies in African linguistics, Supplement to Vol. 1. Los sets, long and short pronouns: Angeles: University of California Press. Rowlands, Evan C. 1969. Teach yourself Yoruba. London: (24) Yoruba Igala Gloss English Universities Press. Long Short Long Short Schleicher, Yetunde. 1993. Je. ka so. Yoruba (let us communicate in Yoruba). New Haven: Yale University Press. èmi mo, m omi ù, n` I Williamson, Kay, and Roger Blench, Niger-Congo. 2000. ìwɔ o ùwE, ε` You African languages: an introduction, ed. by Bernd Heine, and òu& ó òù& ì He/ Derek Nurse. Cambridge: Cambridge University Press. she/it O . LANIKE. O . LA ORIE 1204  Z Zipf, George Kingsley A philologist who studied and taught German and lin- abbreviations among speakers followed this same guistics at Harvard University for most of his academic trend. Zipf’s explanation for the correlation between career, George Kingsley Zipf was one of the first schol- the two was that ‘High frequency is the cause of small ars to use statistics in the study of language. Although magnitude’ (1935:29) and he attributed this correlation he is perhaps best known today for the statistical for- to a Principle of Least Effort, which he said ‘means … mula bearing his name, Zipf did not actually create the that a person … will strive to solve his problems in formula; it was used earlier in the writings of such such a way as to minimize the total work that he must scholars as J.B. Estoup and Vilfredo Pareto. However, expend in solving both his immediate problems and Zipf served as its most ardent proponent, applying the his probable future problems…’ (1949:1). formula to language and a variety of other social phe- To explain, Zipf used the analogy of an artisan’s nomena in his argument for a human ecology. workbench. According to this analogy, the tools of an Zipf began his work on the relationship between artisan that are used more often will be placed nearer frequency and language as a graduate student at the artisan’s chair so that they may be retrieved with Harvard, where he successfully defended a disserta- the least possible effort on the artisan’s part, while the tion on the effect of relative frequency on phonetic tools that are rarely used will be placed farther away change. In this work, Zipf argued for a Principle of from the artisan. Relative Frequency, which said that Following the same line of reasoning, Zipf also worked on the relationship between rank and frequency the accent or degree of conspicuousness of any word, of words in a language. Zipf found that the frequency of syllable, or sound is inversely proportionate to the rela- a work in a language is inversely proportional to the rank tive frequency of that word, syllable, or sound, among of that word in the language, so that if a word ranks first its fellow words, syllables, or sounds in the stream of spoken language. As any element’s usage becomes more in the language, it will be used far more frequently than frequent, its form tends to become less accented, or those words ranked lower in the language. Conversely, a more easily pronounceable, and vice versa. (1932:1) word that ranks 100th in the language will be used far more rarely than words that rank closer to one in the lan- Using the Chinese, French, and English languages, guage. In the form of an equation, this takes the form of as well as a number of writings in Latin, Zipf Rank times Frequency equals Constant. addressed the correlation between the size of a word Zipf later applied his formula to the populations and and its frequency in a language, finding that the size of ranks of city with some success. In the introduction to words in the language is inversely proportional to its his National Unity and Disunity (1941), Zipf explains frequency in the language, so that short words will that if readers were to look at the US Census in any given occur very often in the language and long words will year they would find the city with the greatest population occur infrequently. He also argued that the use of to rank first, the second-ranking city would have half the 1205 ZIPF, GEORGE KINGSLEY population of the first, the third-ranking city would have (1930) in Comparative Philology from Harvard one third the population, and so on. Zipf added that the College, completing a dissertation on relative frequen- size and rank of incomes would also follow this simple cy of use as a cause of phonetic change in language mathematical formula. Zipf inferred from this that com- evolution. He was Instructor of German at Harvard munities—cities as well as nations—worked by the until 1936, Assistant Professor of German until 1939, same principles as other organisms, so that we should and University Lecturer until 1950. He spent the year think of a population as a ‘natural biosocial entity’. after his graduation from Harvard studying at the From the beginning, Zipf’s work has been the sub- University of Bonn and the University of Berlin. He ject of much criticism. Perhaps greatest among these was Member of Linguistic Society of America (LSA), concerns is Zipf’s attempt to explain his findings with 1931, and served on the Committee on Quantitative the Principle of Least Effort, due to the significant prob- Linguistics, created by the Sixth International lem of identifying articulatory features that are univer- Congress of Linguists, in 1948. He died in Newton, sally ‘difficult’ or ‘easy’. Another criticism that many Massachusetts, on September 25, 1950. people have made—and one that is supported by Zipf’s evidence of the formula working in fields besides lin- References guistics—is that Zipf’s Law does not succeed so much at describing how language works as it does at describ- Mandlebrot, Benoit, B. 1953. An information theory of the sta- ing how mathematics works. Finally, although Zipf’s tistical structure of language. Communication theory, ed. by W. Jackson, 486–502. London: Butterworths. Law typically works well with respect to objects that are ——— 1982. The fractal geometry of nature. San Francisco: ranked in the middle of lists, it is less dependable with W.H. Freeman and Co. objects ranked very low or very high, a problem that has Miller, George. 1965. Introduction. The psycho-biology of lan- led some, such as mathematician Benoit Mandelbrot in guages: an introduction to dynamic philology, ed. by George his work on fractal geometry (1982), to make some Zipf, v–x. Boston: MIT. Rapoport, Anatol. 1982. Zipf’s law revisited. Studies on Zipf’s modifications when using the formula. law, ed. by H. Guiter and M.V. Arapov. Bochum: Although there are problems with Zipf’s Law, and Studienverelag Brockmeyer. certainly with the inferences that Zipf made from his Walsh, J.L. 1949. Another contribution to the rapidly growing formulations, Zipf should be recognized as one of the literature of mathematics and human behavior. Scientific pioneers of the use of statistical analyses in linguistics American, August. 181(2). 56–8. Zipf, George Kingsley. 1929. Relative frequency as a determi- in general and the current subfield of corpus linguis- nant of phonetic change. Harvard Studies in Classical tics in particular. Zipf’s Law is also commonly used in Philology, XL. mathematics, physics, library science, and memetic ———. 1932. Selected studies of the principle of relative fre- science, which is the quantitative analysis of cultural quencies in language. Havard University Press. transfer. Zipf died on September 25, 1950 in Newton, ———. 1935. The psycho-biology of language: an introduction to dynamic philology. Boston: Houghton Mifflin. Massachusetts, after a long illness. ———. 1941. National unity and disunity: the nation as a bio- social organism. Bloomington, IN: Principia Press. Biography ———. 1949. Human behaviour and the principle of least effort: an introduction to human ecology. Cambridge: Born in Freeport, Illinois, on January 7, 1902, George Addison-Wesley. Kingsley Zipf received his B.A. (1924) and his Ph.D. LAMONT ANTIEAU Zulu and Southern Bantu Languages The Sintu (Bantu) languages are the most widespread southern third of Africa, from the equatorial rain for- and best-known subgroup of Niger-Congo (the form est of Gabon and southern Cameroon to the highlands Sintu reflects the Nguni prefix isi-, denoting language of Kenya and from the Ubangi River in the Central and culture). According to Greenberg’s (1963) classi- African Republic nearly to the Cape of Good Hope in fication, the Niger-Congo is a subfamily of the Niger- South Africa. Some of the well-known Sintu lan- Kordofanian family, one of the four language families guages include Zulu, Kongo, Lingala, Shona, Bemba, in Africa. The Sintu languages cover most of the vast Ganda, and Gikuyu. 1206  ZULU AND SOUTHERN BANTU LANGUAGES Despite their vast extent today, Sintu languages high degree of structural unity among most Sintu lan- belong to just one subgroup of the Benue-Congo guages, the wide area of contact among them, and sig- branch of Niger-Congo (Williamson and Blench 2000). nificant mutual influence among Sintu languages Sintu languages spread eastward and southward from make classification difficult. West Africa (modern-day eastern Nigeria and The most typical feature of Sintu morphology is the Cameroon) in the early centuries of the first millenni- class prefix system (Campbell 1991). Proto-Bantu had um AD. They only spread over the regions they now 19 noun classes, while Sotho, for example, has only occupy during the last 4,000 years. The ancestral Sintu seven. Sintu languages have long been appreciated by language, proto-Bantu, was spoken somewhere proba- scholars for their distinctive morphology, being highly bly in eastern Nigeria, West Africa. Then, for reasons agglutinative, and allowing great structural complexity unknown, early Sintu communities began expanding to nominal and even more so to verbal forms. There are into new territories, first through the equatorial rain for- very few root adjectives in Sintu; some examples are est belt and along its margins, and then, between about –kulu (large in isiZulu) and –bi (bad in isiZulu). The 500 BC and AD 300, eastward and southward into east- conjunctive pronouns, subjective and objective, are ern and southern Africa. It remains unknown what lan- remarkably homogeneous over most of the Sintu area. guages preceded the Sintu in the equatorial forest and Most primary verb roots are disyllables, and derived the adjoining savannas, but in southern Africa the ear- stems are formed by suffixation. For example, bonana lier languages often belonged to the Khoisan family. means ‘to see each other’ in isiZulu. Some Sintu lan- As Sintu languages spread throughout Africa, their guages have a neutral passive of state ub –Vka(la); for features remained remarkably stable over great dis- example, in isiZulu inkanyezi ya.bona.ka.la (‘the star tances and long periods of time. Their main features was visible’). Moods are generally marked by suffix. are the systems of noun classes and agglutinative verb Most Sintu languages have seven moods: infinitive, morphology. Guthrie (1948) divided Bantu (Sintu) indicative, imperative, subjunctive, perfect, continua- languages into six areas: northwest Central Africa, tive, and relative. In terms of syntax, the typical Sintu west and southwest Central Africa, east central, north- verbal complex consists of prefix (subject concord east central, southeast, and south. Some of the most marker)–tense marker–object marker–stem–modal/ important representative members of the southeast voice marker. area include Shona, Tsonga, Ronga, Makua, and Yao; Current thinking is that the Nguni were one of three those of the south area include Swazi, Tswana, and large African migrant groups whose tradition of horti- Zulu-Xhosa. In Voegelin and Voegelin’s (1997) similar culture and cattle breeding combine the major cultural classification of the world’s languages, ‘Bantu proper’ attributes of West, Central and North East Africa, from is divided into seven major areas: central eastern, cen- where they moved along separate routes to southern tral western, Kari, northeastern, northwestern, south- Africa. The Nguni followed an inland course via the eastern, and southwestern areas. Languages of the headwaters of the Zambesi, where contact with San southeastern area are divided into two groups: Makua, hunters produced the ‘click’ sounds that characterize Matumbi, and Yao, with speakers in Tanzania, their languages today. They moved southward to the Mozambique, and Malawi; and Chopi, Nguni, Shona, most northerly bend of the Limpopo River, which Sotho-Tswana, Tswa-Ronga, and Venda, with speak- marks the boundary between present-day South Africa ers in Zimbabwe, Botswana, South Africa, Swaziland, and Zimbabwe, sometime between the middle of the and Lesotho. The Nguni include isiSwati, isiXhosa, fifteenth and the third century AD. The Nguni suppos- Gaika, Gcaleka, Mpondo, Mpondomse, Ndlambe, and edly split into separate migrations, moving in different isiZulu. The southwestern area contains two subdivi- stages into what is now called KwaZulu-Natal and the sions: Chokwe-Luchazi, Lozi, Luyana, Subiya, with Eastern Province (formerly known as the Transkei). speakers in Botswana, Zambia, Angola, and Zaire; and They fell broadly into four regional divisions: the Herero, Ndonga, Umbundu, and Yeye reside in Mhlubi, Mtetwa, Ndwandwe, and Ngwane. Angola, Namibia, and Botswana. Some of those who settled in northern KwaZulu- The languages that Guthrie (1948) recognized as Natal doubled back into what is now Swaziland, while Bantu are now referred to as ‘Narrow Bantu’ those who first entered the Transkei were the forebears (Williamson and Blench 2000). All major modern of the Pondo. The last to leave the Limpopo settled for attempts to clarify Narrow Bantu have not produced an a while in what is now the southeastern region of the agreed overall scheme (Nurse 1996). The most wide- Mpumalanga province; and then they moved on in spread agreement is that there is a northwest Bantu easy stages into central KwaZulu-Natal. Finding the (Sintu); some authors see a division between east and northeast and northwest already occupied, two small- west Bantu (Sintu). The southern Sintu languages er groups moved on. One of these, finding the coastal include Xhosa, Zulu, Sotho, Tsonga, and Tswana. The regions of the south settled by the Pondo, kept to the 1207 ZULU AND SOUTHERN BANTU LANGUAGES inland high ground, to become the Xhosa. The other of dental, palatal, and lateral click, each containing four the two smaller groups found a home as the coastal phonemes—a surd, its aspirate, its voiced allophone, neighbors of the Pondo to become the Thembu of and its nasal (Campbell 1991:163). today. Thus, the Nguni divided into two large seg- IsiZulu is now the best-known language in South ments, North and South. The Xhosa, Pondo, and Africa, besides English. In 1995, there were 9,140,000 Thembu of the Eastern Cape (formerly Transkei) and Zulu people in southern Africa. They consisted of areas further south are major representatives of the 8,778,000 Zulu people in South Africa (about 18.8% South Nguni, while the Zulu, the Swazi of Swaziland, of the population), 37,480 in Malawi, 76,000 in and the Ndbele (in the present provinces of Gauteng Swaziland, and 248,000 in Lesotho (Grimes 1996). and Mpumalanga) are of the Northern Nguni. Now about 22.4% of the South African population use The final Nguni migration populated the heart of it as their home language, while others speak it as a KwaZulu-Natal where the small and unimportant Zulu second language. clan was later to succeed the Ndwandwe and Mthetwa IsiZulu was first studied by foreign linguists in the empires, respectively, in the northwest and northeast. nineteenth century. Natal’s first Anglican bishop, John Under their famous chief, Shaka kaSenzangakhona William Colenso, and his assistant, William Ngidi, (c. 1787–1828), the first kind of this unified group of translated and studied the Zulu language (isiZulu) in Nguni clans and chiefdoms, they became the rulers of the 1850s and early 1860s (Guy 1997). IsiZulu is a the northern half of KwaZulu-Natal from the Tugela tonal language, closely related to Swazi and Xhosa. Its River in the south to the border of Mozambique in the two dialects are Lala and Qwabe. The standardization north. A long line of descendants link these historic of IsiZulu is now controlled from Kwa-Zulu. figures with the current royal house headed by King There are also several slang languages that are vari- Goodwill Zwelethini. ations of isiZulu. For example, Tsotsitaal, Iscamtho, Prior to the Zulu king Shaka, Zulu referred strictly and Fanagolo are slang languages that are based on to a relatively small clan of people living in the middle Afrikaans, English, and Zulu. Fanagolo is a second reaches of the White Mfolozi River. When Shaka language only, with no mother tongue speakers, wide- founded the Zulu kingdom in the early nineteenth cen- ly used in towns and gold, diamond, coal, and copper tury, he extended the use of Zulu to the many clans he mining areas, while Tsotsitaal and Iscamtho are spo- incorporated into the Zulu state. European anthropolo- ken in black urban townships in South Africa (Slabbert gists and linguists, who saw the Zulu language as cen- and Myers Scotton 1997). tral in defining identity, categorized people of Natal and Zululand as Zulus (Harries 1993). The Nguni languages are closely related, and now References spoken by most people in South Africa. On purely lin- Bendor-Samuel, John (ed.) 1989. The Niger-Congo languages: guistic grounds, there are two main languages in south- a classification and description of Africa’s largest language ern Africa that have severally mutually intelligible family. New York: University Press of America. dialects: Nguni (with varieties like Ndebele, Swazi, Bryant, A. T. 1929. Olden times in Zululand and Natal. London: Xhosa, Zulu) and Sotho (northern Sotho, southern Longmans; reprint, Cape Town: C. Struik, 1965. Campbell, C., G. Mare, and C. Walker. 1995. Evidence for an Sotho, Tswana) (Wolff 2000). In particular, isiZulu is ethnic-identity in the life-histories of Zulu-speaking Durban closely related to Swazi and Xhosa. However, mem- township residents. Journal of Southern African Studies bers of these speech communities, for several reasons, 21(2). 287–301. consider themselves to have separate languages, not Campbell, George L. 1991. Compendium of the world’s lan- just dialects, and unique historical, ethnic, and social guages. London: Routledge. De Kadt, E. 1998. The concept of face and its applicability to backgrounds (Campbell et al. 1995). Language is the Zulu language. Journal of Pragmatics 29(2). 173–91. closely linked with ethnic identity in South Africa. Greenberg, Joseph H. (ed.) 1963. Universals of language. Outside Khoisan, clicks are known to occur only in Cambridge, MA: MIT Press. South African Sintu languages such as Zulu, southern Grimes, Barbara F. (ed.) 1996. Ethnologue. Summer Institute of Sotho, Yeyi, and Xhosa, and in the Cushitic language Linguistics. Guthrie, M. 1948. Classification of the Bantu languages. Dahalo. A click is a multiply articulated sound pro- Oxford: Oxford University Press. duced by forming one closure in the front of the mouth Guy, J. 1997. Class, imperialism, and literary criticism: William with the lips of the tongue in front and another in the Ngidi, John Colenso and Matthew Arnold. Journal of back of the mouth with the tongue dorsum. The clicks Southern African Studies 23(2). 219–41. are a result of borrowing from Khoisan by way of long- Harries, Patrick. 1993. Imagery, symbolism and tradition in a South African Bantustan: Mangosuthu Buthelezi, Inkatha, enduring and intensive contacts, rather than a genuine and Zulu history. History and Theory 32(4). 105–26. part of the sound inventories in Sintu languages them- Nurse, Derek. 1996. ‘Historical’ classifications of the Bantu selves. For example, the Zulu click sounds consist of a languages. The growth of farming communities in Africa 1208  ZUNI from the Equator southwards, ed. by J.E.G. Sutton. Nairobi: Williamson, Kay, and Roger Blench. 2000. Niger-Congo. British Institute in Eastern Africa. African languages: an introduction, ed. by Bernd Heine and Slabbert, Sarah, and Carol Myers-Scotton. 1997. The structure Derek Nurse. Cambridge: Cambridge University Press. of Tsotsitaal and Iscamtho: code switching and in-group Wolff, H. Ekkehard. 2000. Language and society. African lan- identity in South African Townships. Linguistics: An guages: an introduction, ed. by Bernd Heine and Derek Interdisciplinary Journal of the Language Sciences 35(2). Nurse. Cambridge: Cambridge University Press. 317–43. CAMILLA M. COCKERTON Voegelin, C.F., and F.M. Voegelin. 1997. Classification and index of the world’s languages. New York: Elsevier. See also Niger-Congo Zuni The Zuni language spoken in the Pueblo region of the tact on Zuni grammatical structure, however, is less southwestern United States is a linguistic isolate as yet apparent. Influence on grammatical structure as a result unproven to be related to any other known language. of contact with Uto-Aztecan speakers to the west is According to suggestive archeological records, Zuni more easily detected. For instance, the system of clause land claims, and references to sites in oral tradition, the linkage indicating continued vs. changed events along territory in which Zuni was originally spoken extended with a continued vs. change in sentence subject is con- over much of what is now the four-corner region of structed in similar fashion to that of Hopi, currently New Mexico, Arizona, Utah, and Colorado. Today, spoken in Arizona several hundred miles away. Zuni is spoken at Zuni Pueblo and its associated vil- While a great deal of archeological and anthropo- lages and farming communities (Nutria, Ojo Caliente, logical work has been done at Zuni over the past cen- Blackrock, Pescado) in northwest New Mexico, tury, the Zuni language has received less attention than approximately 150 miles west of Albuquerque. other languages of Native North America, possibly The Pueblo region is recognized as an important due to its isolate status. Early documentation includes area of shared culture and history encompassing four word lists collected by Catholic missionaries. other language families native to North America in Transcriptions of letters, stories and prayers from the addition to Zuni (Keresan, Kiowa-Tanoan, Uto- nineteenth century have also been found in archives. Aztecan, and more recently Athapaskan). Despite sev- Brief grammatical sketches were published in the eral attempts, genetic linguistic affiliation has not 1930s and 1960s, as well as a short lexicon of words successfully been demonstrated between Zuni and any and grammatical endings. An orthography was con- other language. One such attempt proposed a link to structed in the 1970s that is currently in use at the the Penutian languages of California. Subsequent Pueblo for citation of Zuni words and names; howev- investigation of the evidence presented has discredited er, Zuni in daily life largely remains an unwritten lan- this proposal, although the mistaken categorization of guage. The orthography does not strictly follow Zuni as Penutian has persisted. standard linguistic practices for transcription; for Zuni differs from its closest geographical linguistic example, it transcribes the voiceless stops [p, t] as ‘b’ neighbors in many typological respects, for example, in and ‘d’, respectively, and it reduces some geminate lacking much of the complex verbal morphology typi- (doubled) consonants to single consonants. cal of languages of this area as well as their associated Among the interesting phonetic properties of Zuni free word order. Zuni has a basic word order of are the glottalized series of consonants and the correla- S(ubject)–O(bject)–V(erb), but special conditions on tion between word stress and vowel tensing. (Word the position of pronouns or on words containing con- stress falls on the initial syllable, and the further a vowel trastive focus frequently result in OSV order. Evidence occurs from this stress the laxer its pronunciation.) The of loanwords in Zuni from Keresan (e.g. Zuni k’yašši- phonological structure allows only a single consonant at ta, Acoma Keresan sk’a’šu ‘fish’; Zuni pu:la, Acoma the beginning of a word, only two consonants contigu- ° Keresan pú:r’ai’ka ‘butterfly’; Zuni ’uwakya, ‘great- ous in the middle of a word, and only a subset of con- ° grandson’ Acoma Keresan ’úwa’ka ‘baby’) suggests a sonants at the end of a word. The language contains ° history of contact between Zuni and the Pueblos further both long vowels and consonants (geminates), and adja- to the east along the Rio Grande. The effect of this con- cent consonants frequently assimilate to result in a 1209 ZUNI geminate, both historically as well as in active pronun- uses this finely detailed system of sentence initial par- ciation. Similar to English, Zuni has a nominative/accu- ticles to convey attitude. sative case system and inflects only pronouns for case. The population of Zuni Pueblo is currently around Unlike pronominal elements of other languages in the 10,000, and it is estimated that 70–80% of speakers geographical area, Zuni pronouns are obligatory inde- use Zuni as a first language. Many of the oldest gener- pendent forms and not inflectional morphology on the ation of speakers are bilingual in Spanish, while verb. Interestingly, while Zuni does not allow for pro- English has been increasing as second and first lan- nouns to be dropped, it has no third person pronouns. guage among younger generations. Consequently, a verb appearing on its own receives third person anaphoric interpretation for its subject (and References object if it takes one), cf. ’awakya, find-past, ‘he/she found him/her/it’. Zuni noun stems are marked with dif- Bunzel, Ruth. 1932a. Zuni origin myths. Report of the Bureau ferent inflectional endings in the singular and in the plu- of American Ethnology. Smithsonian Institution, Washington, DC. Reprinted in Zuni Ceremonialism. 1992. ral (te’le ‘pot’ vs. tewe ‘pots’), and nouns have endings Albuquerque: University of New Mexico Press. that classify the noun based partly on semantic classes ———. 1932b. Zuni ritual poetry. Report of the Bureau of and partly on the number of syllables of the stem (no’le American Ethnology. Smithsonian Institution, Washington, ‘bean’ vs. no’linne ‘nose’). DC. Reprinted in Zuni Ceremonialism. 1992. Albuquerque: Syntactic properties of broader typological interest University of New Mexico Press. ———. 1933. Zuni texts. Publications of the American include noun incorporation, which takes the special Ethnological Society, Vol. 15. American Ethnological So- phonological form of a noun reduced to its initial con- ciety. Washington D.C. sonant and vowel and attaches this reduced form as an ———. 1933. Zuni. Handbook of American Indian languages, affix to the verb. Furthermore, Zuni makes abundant Part 3, ed. by F. Boas, 389–515. New York: J.J. Augustin. use of sentence initial particles to indicate a range of Newman, Stanley. 1958. Zuni dictionary. International Journal of American Linguistics 24(1) (pt. 2). meaning from aspect to modality. Zuni lacks the abil- ———. 1965. Zuni grammar. University of New Mexico ity to form finite subordinate clause structures of the Publications in Anthropology, No. 14. Albuquerque: Univer- kind selected by verbs of attitude evaluation, such as sity of New Mexico Press. the equivalent of ‘I think that I will go’ and instead LYNN NICHOLS 1210  Index Entries containing suffix ‘f’ and ‘t’ along with page numbers indicate figure and table, respectively. A question of affix or process, 13 AAVE (African American vernacular English), 14–16, 301, 423–425, types of, 12 1140–1141 Affixes, See also Evidentials Abkhaz, 757, 1004, 1132 Igbo and Igboid languages, 498 Aboriginal English, 102 inflectional and derivational, 529–530 Abrupt metathesis, 679 Javanese, 560t–561t Absolute tense system, Coptic Egyptian, 243t productivity of, 715–716 Accents, ethnic, 300–301 Affix stripping, 816 Accommodation, linguisitic, in India, 504 Affricates, 649 Achievement testing, 876 assimilation and, 95–96 Acoustic phonetics, 1–3, 350 Africa Acoustics, speech, 838–839 French spoken in, 356 Acoustic signals, 1011 tone languages, 1118–1119 Acquired dyslexia, 276 African American vernacular English (AAVE), 14–16, 301, 423–425, Acquisition 1140–1141 bilingual, 136–138 African languages language, 3–7 classification work by Greenberg, 413 theories of, 7–9 logophoric pronouns, 1059 planning goals, 605 Saramaccan words derived from, 929 what it means in second language acquisition, 5 spoken in Brazil, 992 word formation in, studied by Clark, 208–209 Afrikaans, 275–276, 514 Activities–accomplishments, differences between, 94 Afroasiatic Adamawa-Ubangi, 739–740 Chadic branch, 17 Adaptation, See also Preadaptations origin of language family, 19 grammatical, 621 protolanguage, 18–19 Adjacent metathesis, 679 relatedness of languages, 17–18 Adjectives Age effects on language, 19–21 Burushaski, 165 Agentive, 1101 Hindi-Urdu, 457 Agglutinating languages Italian, 544 Mongolian, 703 Latin, 610 Turkish, 713–714 Sindhi, 967 Agglutinative theory (Bopp), 149 Adults, older, aging effects on language, 19–21 Agnéby languages, 28 Adverbial modification, 900 Agrammatic aphasia, 1027 Adyghey, 757, 1004 Agreement Aerodynamics of vocal tract, 10–11 and case, 22 Affixation ergative/absolutive, 22 Carib, 175 evolution of, 24–25 delimitation issues, 12–13 exotic forms of, 23–24 headedness, 14 gender, 22 Mande languages, 122 Guaymí and Chibchan languages, 422 number marking and, 760 and inflection, 23 1211 Index Agreement (continued) Alveolar fricatives, 859 reduction of, 23 Alzheimer’s disease rich, 21–22 effect on language production and comprehension, 20–21 syntactic typology and, 1067 impairment of word finding, 1027 theory of, 25 Amazigh, See Berbers verb Ambiguity in signed languages, 964 distinguished from vagueness, 40–41 in Sindhi, 969 lexical and structural, 40 Ainu structural, 1069 culture and language, 27–28 Ambonese Malay, 647 genetic and linguistic history, 25–26 American English, 1037 relationship to Altaic, 37 history of, 1140–1144 Yukar (epic poems), 26 American Indian languages, See also Native American languages Akan languages Dakota and Siouan, 253–255 subfamilies, 30 studied by tone terracing, 30 Chafe, 184–185 Twi dialects, 29 Sapir, 925 vowel harmony types, 30 American Indian mythology, studied by Lévi-Strauss, 619 Akkadian, 113–114, 953–954 American Sign Language (ASL), 1146–1147 Aktionsart, 94 chimpanzees exposed to, 63 Albanian grammatical evolution of, 41–42 attributive articles in, 33 nonmanual signs, 42 loanwords in, 32 vs. oral method of education, 43 origins of, 31 structure of, 963 Tosk and Gheg dialects, 31 Amharic verb system in, 32 sentence construction, 46 Aleut dialects, 297–298 spoken throughout Ethiopia, 44 Algeria written, 45 Algerian Arabic vs. Modern Standard Arabic, Analogical change, 46–48, 599 35–36 proportional analogy, 47 Berber dialects, 1125 theoretical approaches to, 48 linguistic independence, 34 Anansi tori, 1033 Algonquian languages, 144 Anaphora, 48–50, 336 Ojibwe and, 764–767 Anaphoric reference, 217 Aliyah, 541 Anatolian, as Indo-European language, 511–512 Alliteration, 336 Anatolian hypothesis, archaeological, 85 Allomorphs, 760 Anatomy Allophones, 829–830, 836, 840 articulatory system, 50–53 Alphabet auditory system, 53–54 Albanian, 31 Ancient Egyptian, 54–57 ancient Egyptian, 55 Ancient Greek, 58–62 Arabic, 74–75 Anglo-Irish, 407 Armenian, 87 Angloromani, as bilingual mixed language, 139 Burmese, 161 Angolar, 250 Copto-Greek, 241–242 Animal communication and human language Dania, invented by Jespersen, 562–563 dolphins, 64–65 Devanagari, 922 overview, 62–64 Gothic, 394 parrots, 66–67 IPA, 831–832 primates, 67–70 Italian, 543 Animal sign systems, 142 Latin, 610 Animate hierarchy, Cape Verde Creole, 173 Mongolian, 704 Annamese-Muong, 108 Old Church Slavonic, 774 Anomic aphasia, 71 Roman, in Soviet Union, 997 Anthropological linguistics, 875 Slav, 521–522 Anticipatory assimilation, 97 Alphabetic writing system, 1192 Antilles, Papiamento creole language, Alphasyllabic writing system, 1192 808–809 Alsatian, 355 Antimodularity principle (Givón), 392 Altaic Apache, 728 Manchu-Tungus languages, Ape language research (ALR), 67–70 1003 Aphasia Mongolian, 702–704, 1003 anomic and conduction types, 71–72 presence in China, 193 Broca’s, 71, 214, 1026 putative cognates of, 39 global, 72 subfamilies of, 37–38 prognosis, 72 Turkic languages, 1003 psychosocial consequences, 72 1212  Index resulting from hemispheric damage, 70 Artificial intelligence, 88–90 Wernicke’s, 71 main components of generic system of, 89–90 Applied linguistics Turing test, 89 subfield of second language teaching and learning, 73 Artificial languages as transdiscipline, 72 Esperanto, 91–93 Approximants, 649 historical overview, 90–91 Arabic interlinguistics, 91 Algerian, vs. Modern Standard Arabic, 35–36 Ashkenazim, 540 alphabet, 74–75 Asia Juba and Nubi pidgin-creoles, 567–568 Southeast, 993–995 lexicon of Farsi and, 325 tone languages of, 1118 Old and New, 75–77 ASL, See American Sign Language (ASL) Arabic traditional grammar, 78–79 Aspect, 93–94 Aragonese, 1007 Igbo and Igboid languages, 498 Aramaic markers, in Javanese, 562 nomenclature of, 79–80 Mayan languages, 663 used for legal purposes, 539–540 Navajo and Athabaskan-Eyak languages, 730 writing, 955 predicational, 94 Aranese, 1007 Aspect marking, tense and, 1088–1089 Arapaho kinship terms, 576 Assimilation Araucanian, 989 and coarticulation, 94–96 Arawak and dissimilation, 96–97 Carib/Arawak pidgin, 81–82 manner of, 95 complicated nature of, 82–84 voicing and, 95 endangered, 82 Assiniboine, 254 Arbitrariness, opposed to iconicity, 487 Asturian, 1007 Archaeology Asymmetrical discourse, in medical encounter, 672–674 Anatolian hypothesis, 85 Ataylaic languages, 193 continuity hypothesis, 84–85 Athabaskan-Eyak language, Navajo and, 728–731 Kurgan model, 85 Atkan, 297 role in determining language family origins, 19 Atlantic languages, 740, 1181–1183 of Taklamakan Desert region, 1113 Atlases, linguistic, 267–268 Areal linguistics, 1030 Attic-Ionic, 58 Argentina Attié languages, 28 indigenous languages Attributive articles, in Albanian, 33 Araucanian, 989 Audiolingual method of teaching, 1085 Chon, 989–990 Auditory pathway, brain organization and, 150–153 Guaycuruan, 989 Auditory phonetics, 350 Lule-Vilela, 989 Auditory system anatomy, 53–54 Matacoan, 989 Augmented transition networks, 814–815 Quechua, 988 Austin, John Langshaw, 98–100, 1016–1017 Tupi-Guarani, 988–989 Australia Yahgan, 990 comprehensive language policy (since 1980s), 102–103 other languages, 990 genre school, 270 Argobba immigrant population, 101 divergence from Amharic, 46 Australian languages EES language, 44 contact languages, 101–102 Armenian, 1002 Kriol (Roper River Creole), 585–587 alphabet, 87 Murrinh-Patha and Daly languages, 717–720 as cover term, 85 Pama-Nyungan languages, 804–807 literary form (nineteenth century), 87 and Pintupi, 852–855 Lord’s Prayer, 87–88 studied by Hale, 433 shared features, 86 Tangkic languages, 569–571 Armenian gypsies, 1132 tied to social units, 100 Articles Austria attributive, in Albanian, 33 ethnic minority languages, 104–105 definite and indefinite, 262 pluricentrism, 103–104 Warao language, 1166–1167 Austroasiatic Articulation in India, 499 double, developed by Martinet, 656–657 Mon-Khmer branch, 106–108 manner of, 649–650 Munda branch, 106 place of, 858–860 Austronesian and manner of, 95 comprising 1200 languages, 108 stage in speech production, 1025 distinction between possession types, 111 Articulatory system anatomy, 50–53 Indonesia and Philippines, 108–109 Artificial, discrete-combinatorial systems (ADCs), dolphins, 64–65 New Guinea, 735 1213 Index Austronesian (continued) Bidirectional assimilation, 97 Oceanic subgroup, 110, 459 Bilingual acquisition, 136–138 Southeast Asia, 994 Bilingual education, in Australia, 854–855 speakers in China, 193 Bilingualism, See also Code-switching; Multilingualism Autocorrelation, 985 balanced and compound bilinguals, 140 Automatization, 1044 Canada, 169 Auxiliaries, 111–112 Egyptian–Greek, 242 Auxiliary language, international, 925 in Ethiopian region, 44–45 Avaro-Andi-Tsez languages, 382 first language acquisition and, 4 Avikam–Alladian languages, 28 individual, 601 Azeri, 1003 mental flexibility and, 140–141 Bilingual mixed languages, 138–139 B Bilinguals Babylonian traditional grammar, 113–114 balanced and compound, 140 Backformation, 47 transfer by, 601 Bakhtin Circle, 220 virtual, 825 Balkans, 114–118 Binding domain, for anaphor, 49–50 geography, 115–116 Biosemiotics, 141–142 history, 116–117 Birmingham model of discourse analysis, 270 linguistic landscape, 115 Bislama, 803, 1115–1116 linguistic union, 117–118 Blending, 47 Baltic languages, 118–121 Bloomfield, Leonard, 143–145, 1041–1042 Latvian, 120–121, 1002 Boas, Franz, 145–147 Lithuanian, 119–120, 1002 Body language, See Paralanguage Slavic languages related to, 522 Bonaire, Papiamento creole language, 808–809 Bambara language, 123–124 Bonobos Banjarese, 646 duetting behavior, 314 Bantu symbol use by, 69 Fanakalo and, 323–324 Boolean approach to information retrieval, 531–532 Grassfields Bantu languages, 1195 Bootstrapping, semantic and syntactic, 8 Southern Bantu languages, 1206–1209 Bopp, Franz, 148–150, 221 Bar-Hillel, Yehoshua, 124–126, 206 Borrowing Barlavento (Cape Verde), 172 in American English, 1140 Basic vocabulary, 1051 folk etymology and, 306 Basque, 126–127, 354, 1006 Japanese words from Chinese, 553–554 Baudouin de Courtenay, Jan Ignacy NiecisBaw, 128–130, Khmer from outside Southeast Asia, 572 830–831 as language contact type, 848 Belarusian, 916–917, 1002 large-scale, 378 Belfast English, 407 lexical, 279, 462, 620–622 Belgium linguistic, Egyptian–Greek, 242 Flemish (Vlaams), 274 role in JSL, 557–558 history, 130–131 by Slavic languages, 522 Bemba, class marking, 370 Bosnian, 1132 Bengali, ethnic identity, 302 Bottom-up parsing, 813–814 Bennett, Louise, 1175–1177 Bound morphemes, 716 Benue-Congo language family, 740 Brain organization and auditory pathway, 150–153 Benveniste, Emile, 132–134 Brand names, 726 Berber languages Brazil, 990–993 evolution of, 1125–1126 African languages, 992 linguistic features, 1125 creoles, 992 in Morocco, 1124–1125 illiteracy, 992–993 Tuareg, 1123–1124 indigenous languages, 991 Berbers, 34–36 other European languages, 992 Bever, Thomas, 134–136 Portuguese, 991–992 Bia languages, 29 Bresnan, Joan, 153–155 Biblical Hebrew Breton, 183–184, 354 basic characteristics of, 451 British Sign Language (BSL) classification, 450 grammar of, 155–156 diglossia, 451 historical records of, 155 emphatics, 452 as minority language, 156 loanwords, 452 regional dialects, 155 morphology, 452 social situation of, 156 orthography, 451 time lines in, 964 phonology, 452 Broca’s aphasia, 71, 214, 1026 stages of, 450–451 Broca’s area, 630–631, 882 syntax, 452 Brugmann, Karl, 157–158 1214  Index BSL, See British Sign Language (BSL) Northeast, 1004 Bühler, Karl, 158–160 Northwest, 383, 756–758, 1004 Bulgarian, 958, 1132 in Turkey, 1132 Buli, 711 Causation, 180–182 Burmese Causative construction, 1068 dictionaries and grammars, 161 Cayuga, 700 grammar, 162–163 Cellular sign systems, 141 historical overview, 160–161 Celtiberian, 1005 literary form and colloquial language, 161 Celtic languages phonology, 161–162 historical overview, 182 significance of, 161 Welsh, 405–406 Burushaski, 163–167 Census (1991) India, problems with, 500 morphology, 164–166 Central executive component, working memory, 1190 phonology, 164 Central German dialects, 783–784 sentence structure, 166–167 Central nervous system, 150–151 Bzhedukh West Circassian, 383 Central Semitic, 954 Central Tibetan dialect, 1104–1105 C Cepstral analysis, 985, 1015–1016 Cacaopera, 693–694 Chabacano, 822–823 Cajun French, 1145–1146 Chadic languages, 17 Canada Hausa and, 445–447 bilingualism, 169 Chafe, Wallace, 184–186 English presence in, 171 Chaha, EES language, 44 French speakers in, 356 Chain shift, 408–410 history of settlement, 170 Chamorro/Austronesian, reduction of agreement, 23 Québec Francophones, 170–171 Change Cape Verdean Creole, 172–173 analogical, 46–48 Cape Verdean Crioulo, 250–251 between Classical and Modern Armenian, 86–87 Carib language, See Language change morphology, 175 studied by historical linguistics, 462–463 phonology, 175 Chao Yuen Ren, 187–189 syntax, 175–176 Chart parsers, 814 two grammars, 174 Chechen, 1004 Cariban languages, 173–176 Chiasmus, 336 morphology, 175 Chibchan languages phonology, 175 connection with Misumalpan, 695 poor documentation of, 174–175 and Guaymí, 419–423 syntax, 175–176 Chicano English, 301–302 Carib/Arawak pidgin, 81–82 Chicano linguistic competence, 1145 Caribbean Chickasaw, 1059 French and Dutch creoles, 178 Children historical overview, 176–177 acquisition of phonology by, 549 Western Caribbean creoles, 1172–1177 language acquisition, studied by Case Jespersen, 564 agreement and, 22 Slobin, 974 Germanic, in Gothic, 394 meaning acquisition, studied by Clark, 209 lacking in Farsi, 326 Chimpanzees Tamil, 1078 exposed to ASL, 63 Case assigner, 179–180 lexical and syntactic ability, 310 Case grammar, 398–399 nursing poke, 318 Case-marking signing by Nim, 68–69 and grammatical relations, 22 China patterns of, 1067 Altaic languages spoken in, 193 Case/number marking, Sindhi, 967 bilingual speakers in, 193–194 Case theories, 179–180, 808 ethnic Tibetans, 1104 Cassubian, 864 linguistic absorption, 194 Castilian, 1006, 1008 living languages, 189 Catalan, 354, 519, 1006, 1010 minority languages, 192t Cataphoric reference, 217 Sinitic languages, linguistic tree, 191f Catawba, 254–255 speakers of Tai languages, 1099 Categorial grammars, 398 tone languages, 1118 Categorical perception, 1019 typological comparison of major language groups, 190t Categorization in linguistics, 1065 Chinese Caucasian languages characteristics shared with Vietnamese, 1153 Georgian, 380–382, 1004 dialects, studied by Chao, 187–188 Nakh-Daghestanian languages, 382–383 Japanese words borrowed from, 553–554 1215 Index Chinese (Mandarin) prepragmatic capacities, 313–314 characters and sounds, 195 presemantic capacities, 312–313 distribution of, 190–191, 194 presyntactic capacity, 312 morphology, 196–197 Cognitive theory of metaphor, 677–678 phonology, 196 Cognitive theory of metonymy, 680–681 syntax, 197–198 Coherence Chinese Pidgin English, 200–201 in discourse, 216–218 Chinese Pidgin Russian, 201–203 text, 1097 Chinese traditional grammar, 198–200 Cohesion, grammar- and vocabulary-driven, 216–217 Chinook Jargon, 203–205 Collective nouns, 862 Chippewa, See Ojibwe Collocation, 1185–1186 Chiwere languages, 253–254 studied by Firth, 345 Chöke (Classical Tibetan), 1104 Colloquial Burmese, 161 Chomsky, Noam, 205–208, 256, 283–284 Color terms, 218–219 Chon, 989–990 Common Norse, 792 Chukchi language, 714 Communication Chuvash, 38–39 animal, and human language, 62–70 Circumfixation, 12 of emotions, 281–282 Clark, Eve V., 208–210 ethnography of, 303–304 Classical Arabic script, 1124 modes of, in India, 499–500 Classical Armenian, Lord’s Prayer in, 87–88 organon model (Bühler), 158–159 Classical French, 358 theory of, 219–221 Classical phonetics, 327–328 Communicative competence, 979 Classical Sanskrit, 923 Communicative language teaching, 944, 1083, 1085 Classical Tibetan, 1104 Comparative and historical linguistics, 466 Classification codes, 532 Comparative method, 221–223 Classificatory verbs, Navajo and Athabaskan-Eyak languages, 730 employed by Humboldt, 479 Classifiers introduced by Darwin, 310 in ASL, 42 Comparison, long-range, 633–635 Chinese (Mandarin), 197 Competition Model, learning acquisition, 9 noun, Hmong-Mien languages, 474 Complement clauses, 211 numeral, 280 Complementizer phrase, 210–211 in signed languages, 963 Complex concept formation, 313 Thai and Tai languages, 1100 Complex sentences, Burushaski, 166–167 Class marking, gender, 370–371 Compositional semantics, 224 Clause Compounding, 13, 224–226 chaining, in Eskimo-Aleut, 297 Chinese (Mandarin), 196 Japanese, 554–555 Compounds as modifier, 698 representation in lexicon, 624 in traditional grammar, 210–211 Yoruboid languages, 1203 Clause-type indicators, 211–213 Comprehension Clefting, 236 effects of Alzheimer’s disease, 20–21 Click experiments by Bever, 134 sentence, 310 Clicks (velaric ingressive), 11, 650 Computational linguistics, 125 Zulu, 1208 Computer-assisted language learning (CALL), 229–231 Clinical linguistics, 213–214, 731–732 Computer programming, stylistic analysis applied to, 351 Coalescent assimilation, 97 Comrie, Bernard, 231–232 Coarticulation, 838 Concatenative synthesis, 1029 distinguished from assimilation, 95 Concept formation Cochlea, 54 basic, 312–313 Cockney, 405 complex, 313 Coded meanings, 669 Conceptualization stage of speech production, 1025 Code-switching, 137, 140, 214–216, See also Bilingualism Conceptual metaphor theory (Lakoff), 671 in India, 505 Concordancing programs, 246 purposeful, 601 Conduction aphasia, 72, 1026 studied by Gumperz, 426–427 Configurationality, 232–234 Cognitive development Conjugation system, Coptic Egyptian, 243–244 first language acquisition and, 3–4 Connectionism, 234–235 language development and, 9 Consecutive mode in interpreting, 534–535 Cognitive grammar, 400 Consonants developed by Langacker, 595–597 Afroasiatic languages, 17–18 Cognitive linguistics (Lakoff), 594 Arabic, 76 Cognitive plausibility principle (Givón), 392–393 Atlantic languages, 1182–1183 Cognitive preadaptations, 311–315 Burushaski, 164 elementary symbolic capacity, 314–315 epenthesis, 294–295 prephonetic capacity, 312 Gbe, 320t 1216  Index Gur languages, 709–710 Corpus planning goals, 605 Hindi-Urdu, 457t Corsican, 354–355 Javanese, 559t Cortex, electrical stimulation of, 733 Korean, 581t Count nouns, 861–862 loss of clusters in AAVE, 15 Courtroom discourse, 247–249, 352 Mohawk and Iroquoian languages, 700 Creoles, See also English-based creoles; French creoles; Pidgins; Old Norse, 793 Spanish-based creoles reduction, Western Caribbean creoles, 1175 Brazil, 992 retroflex, 859 Cape Verdean, 172–173 Semitic languages, 953t Caribbean, 177–178 spectral properties, 2 comparison with pidgins, 847–850 Telugu, 1087 Gulf of Guinea Crioulo, 249–250 Consonant Shift, Old High German, 783–784 Haitian, 178, 431–432 Constative language, 1016–1017 Hawaiian Creole English, 448–450 Constatives, distinguished from performatives, 98–99 Indian Ocean, 506–507 Constituency Juba and Nubi Arabic, 567–568 and phrase structure, 235–236, 843–844 Krio, 584–585 relevance of, 236 language contact and, 601, 848 Constraints, rules vs., 907–910 Louisiana, 635–636 Contact New Guinea, 736–737 Balkans, linguistic changes induced by, 117–118 Pacific region, 802–803 languages in, 600–602 Papiamento, 808–810 migration and, 692 Philippine Spanish, 822–824 restructuring through, 424 Saramaccan, 928–930 Contact languages in Southeast Asia, 994–995 Australia, 101–102 Upper Guinea Crioulo, 250–252 Chinese Pidgin Russian, 201–203 Western Caribbean, 1172–1177 Kituba, 577–579 Crioulo Melanesian Pidgin, 1114–1116 Cape Verdean, 250–251 in Oto-Manguean language history, 799 Gulf of Guinea, 249–250 Contamination, in context of analogy, 47 Upper Guinea, 250–252 Context, 237–238 Critical period, first language acquisition and, 4 phonetic, 829 Criticism, textual, 825 retrospective and prospective, 239 Croatian, 105 Context-free parsers, 813 Crow, 254 Contextualization cues, studied by Gumperz, 426 Cultural sign systems, 142 Continuity hypothesis, archaeological, 84–85 Culture Contrastive function of phonemes, 829 Ainu, impacted by Japanese, 27–28 Converbs, Tungusic, 1128 of Deaf people, 43 Conversational implicature, 1097 link with language, 927–928 Conversation analysis, 238–240, 270, 871 Turkish, Balkans influenced by, 117 Conversion, in English, 293 Cuneiform, 825–826, 1046 Cooperation, as prepragmatic capacity, 313 Curaçao, Papiamento creole language, 808–809 Cooperative principle (Grice), 416–417 Cushidic, 17 Coordinate Structure Constraint, 240–241 Cutting, reverse, 199 Coordination, 240–241 Cymraeg (Welsh), 405–406 Coordination test, 236 Cypriot Greek, 410–411 Coordinative compounds, 225 Cyrillic, 956–958 Coptic Egyptian Czech, 864 bilingualism and linguistic borrowing, 242 spoken in Austria, 105 conjugation system, 243–244 writing system, 1192–1193 Copto-Greek alphabet, 241–242 dialect variation, 242 D noun morphology, 242–243 Dagaare, 710 phonology, 242 Dagbane, 710 syntax, 244 Dakota languages, 253–254 verb morphology, 243 Daly languages, Murrinh-Patha and, 717–720 Copto-Greek, alphabet, 241–242 Damin, 570 Corachol, 722 Danish, 513, 1055 Coreference, 217 dialect alphabet invented by Jespersen, 562–563 Coronal sounds, 859 vs. North Scandinavian languages, 1054 Corpus linguistics spoken in Schleswig, 389 analysis tools in, 246 Dardic, 517 corpus-based investigations of language use, Darwin, Charles, 309–310 245–246 Data collection, in field situation, 333–334 potential limitations of, 246–247 Davaoeno, 995 1217 Index Decipherment, 825–826 Chinese, studied by Chao, 187–188 Tocharian, 1111 Dutch, 274–275 Declarative sentences, 212 Estonian, 343 Declensions, Old Church Slavonic, 776 ethnic Decoding AAVE, 14–16 phonological, 895 in Austria, 104–105 text, 1096 English, 300–302 Deconstruction, structuralism and, 670 Finnish, 342–343 Deep structure, 255–257 German, 389–390 Definite clause grammar, 400 Greek, 58–59 Definite descriptions, theory of, 911–912 Italian, 542–543, 546–548 Definiteness, 258–260 Japanese, 552–553 Deictic gestures, 812 Latvian, 120 Deixis, 260–262 Lithuanian, 119 Delimitation issues in affixation, 12–13 Malagasy, 641 Dementia, Alzheimer’s, 20 Marshallese, 655 Demographic profile Middle English, 688 Mexican languages, 682 modern Greek, 411–412 Nilo-Saharan, 752–753 Old English, 777 Demonstratives, 262 Old High German, 784 Demotic regional characteristics of, 57 BSL, 155 present state of, 411 England, 405 writing system, 55 Tibetan, 1104–1105 Denmark (Danmark), 933 variation, Coptic Egyptian, 242 Dental sounds, 859 Warao language, 1165 Dependency Yiddish, Western-Eastern continuum, 1199 long-distance, 631–633 Dialogism, and intersubjective meaning, 670 in syntactical relationships, 1070–1071 Dialogue management, 1022 Dependency grammar, 399 Digitalization, 984 Depictive secondary predication, 873 Diglossia Derivation Biblical Hebrew, 451 explained in terms of affixation, 760 Burmese, 161 inflection and, 529–530 modern Greek, 411–412 in Vietnamese, 1153–1154 presupposed by standard language, 1036 Descriptive approach to reference, 896–897 situations of, 268–269 Descriptive linguistics (Firth), 345 sociolects and, 975–976 Desemanticization, 403 in Switzerland, 1060 Desentencing, Yoruboid languages, 1203 Tamil, 1077–1078 Determiner, 262–263, 886 Dik, Simon, 364–365 Determinism, linguistic, 927 Diphthongs Devanagari script, 456, 921 in AAVE, 15 Developmental dyslexia, 276 and Great Vowel Shift, 408–410 Developmental stages of language modern Greek, 412 in infancy, 264–265 Diplomatics, 825 in preschool years, 265–266 Direct access theories of irony, 537–538 in school years, 266 Direct reference theories, 897–898 Dhegiha languages, 253–254 Disambiguation, word sense, 1188–1189 Diachronic approach to language history, Discontinuist view, animal and human language, 143 62–63 Diachronic linguistics, 647–648 Discourse Diachronic metathesis, 679 coherence in, 216–218 Diachronic pragmatics, 871 courtroom, 247–249, 352 Diagrammatic iconicity, 487–488 deixis, 261 Dialect geography, 462 forensic, 351 Belarusian, 917 function topic, 234 eastern United States, 1141 in medical encounter, 672–674 Korean, 580 related features in JSL, 558 Rusian, 913–914 relevance in, 901–903 Russian, 915 Discourse analysis, 269–271, 442 Ukrainian, 916 Discourse markers, Warao language, 1166 Vietnam, 1154–1155 Discourse strategies, 271–272 Dialectology, 266–268, 978 studied by Gumperz, 426 Dialects Discourse typology, 945–947 Ainu, 27 Disharmonic words, Classical Manchu, Berber, 1124–1125 1160 1218  Index Dispersal morphosyntax, 56–57 archaeological models of, 84–85 phonology and pronunciation, 55–56 Austronesian languages, 109 Egyptian, Coptic Pama-Nyungan languages, 806–807 bilingualism and linguistic borrowing, 242 Displacement conjugation system, 243–244 in animal communication, 63 Copto-Greek alphabet, 241–242 by dolphins, 65 dialect variation, 242 Displacement tests, 236 noun morphology, 242–243 Dissimilation, and assimilation, 96–97 phonology, 242 Distinctive features, studied by syntax, 244 Jakobson, 549 verb morphology, 243 Ladefoged, 592 Ejectives (pulmonic egressive), 10 Distinctive feature theory Elderspeak, 20 in phonology, 328–329, 840–841 Electroencephalography, 732 redundancy as example of use of, 329–331 Elementary symbolic capacity, 314–315 Distributed developmental model, 1156–1157 Embedded, See Subordinate clause Distributed Morphology, 436 Emblems, 812 Distribution of reading impairment, 895–896 Embodied contexts, 238 Distribution pattern of phonemes, 830 Emeneau, Murray Barnson, 279–281 Document examination, forensic, 352 Emotion and language, 281–282 Dogon, 740 Emphatics, Biblical Hebrew, 452 Dolphins, artificial, discrete-combinatorial systems (ADCs), 64–65 Empiricism, English, 465 Domain concept (Fishman), 347, 978–979 Empty categories, 282–284 Doric, 58–59 Empty morphemes, 285–286 Dorso-velar sounds, 859 Enclitics, Igbo and Igboid languages, 498 Double articulation, developed by Martinet, 656–657 Endangered languages Dowty’s model, 1103 Abkhaz, 757 Dravaeno-Polabian, 865–866 Arawak, 82 Dravidian, 272–273, 499–502 Athabaskan, 728–729 studied by Emeneau, 279–280 causes of, 287 Dual route cascaded model, 1155–1156 Celtic, 183–184 Duetting behavior by bonobos, 314 in China, 194 Dutch, 514 Dakota, 254 Afrikaans, 275–276 Hawaiian, 109 dialects, 274–275 in North America, 286–287 history of, 274 Pama-Nyungan, 804–805 influence in Papiamento, 809 and Pintupi, 852–855 loanwords, 275 Wakashan, 754 Pennsylvania, 387 Endocentric compounds, 225 Dutch creoles, spoken in Caribbean, 178 England Dyslexia, 276–277, 894–895 Cockney, 405 English, 404 E regional dialects, 405 Ear, anatomy of, 53–54 Standard English and Received Pronunciation, 405 Eastern Mon-Khmer, 107 English East Germanic, 513 Aboriginal, 102 East Oto-Manguean, 798 African American vernacular, 14–16 East Slavic languages, 912–917 American, 1037, 1140–1144 Belarusian, 1002 auxiliaries, 112 Russian, 1001 Chinese Pidgin English, 200–201 Ukranian, 1001–1002 configurationality, 232–234 Ebonics, 15 contact in Australia, 101 Echo, as form of reduplication, 797 creole-speaking communities in Caribbean, 178 Education definiteness, 258–259 bilingual, in Australia, 854–855 external sandhi, 919–920 and government policies in India, 503–504 as a foreign language (EFL), 5 sociolects and, 976–977 formation of “Englishes,” 293–294 sociolinguistics and, 979–980 Hawaiian Creole English, 448–450 Tok Pisin language in, 1115 history, 288–290 Western, in Nigeria, 747–748 importance of word order, 1186–1187 EES (Ethiopian and Eritrean Semitic languages), 44–46 in India, 503 Ega, as Kwa language, 28 mixing with, 505 Egressives, pulmonic and glottalic, 10 Nigerian, 742–745, 747–748 Egyptian, ancient, 54–57 Northern Irish English, 407 alphabet, 55 presence in Canada, 171 Demotic, 57 role in South Africa, 987 1219 Index English (continued) Ewe and Gbe languages Scottish English, 407 history, 319 as a second language (ESL), 5 ideophones, 322 spelling-pronuciation discrepancies, 1193 morphology, 320 spoken in England, 404–405 noun phrase, 322 spoken worldwide, 514 phonology, 320 structure, 290–293 preposition and postpositions, 322 taboo words, 1073–1074 syllable structure, 321 Welsh English, 405 syntax, 320–321 English-based creoles, 293–294 tone, 320 Sranan, 1031–1033 Ewenki, 1127 Western Caribbean, 1172–1177 Exocentric compounds, 225 Epenthesis, 294–295 Expansionism, Scandinavian, 935–936 Epiglottal sounds, 860 Experience-dependent learning, 7–8 Epigraphy, 824 Experience-expectant learning, 7–8 Equipment in field situation, 334 Experiential hypothesis, and meaning and metaphor, Equivalence, concept in translation studies, 1120–1121 670–671 Ergative/absolutive agreement, 22 Experimental neurolinguistics, 732–733 Ergative systems, Guaymí and Chibchan languages, Experimental phonetics, 592 422–423 Expert systems, 227 Ergativity Expressives, Hmong-Mien languages, 474 Cariban languages, 175–176 External sandhi rule, 919–920 Mayan languages, 664 Eye movements, faulty, 895 syntactic typology and, 1068 Ezafe construction, Farsi, 326 Erosion (phonetic reduction), 403 Errors F evolving definitions of, 5–6 Face, concept of, 866–867 speech, studied by Fromkin, 361 Facial expressions Eskimo-Aleut, 296–299 in IPSL, 527 Esperanto, 91–93 in signed languages, 964–965 Estonian, 1003–1004 Facilitation, morphological, 716–717 dialects and written, 343 Family tree model, 378 Ethics Sino-Tibetan languages, 971f contributions of Grice, 417 Fanagolo, 1208 in field situation, 334–335 Fanakalo, 323–324 Ethiopian and Eritrean Semitic languages (EES), 44–46 Faroese, 513–514 Ethnic dialects Farsi, 324–327 AAVE, 14–16 Modern Persian, 325 in Austria, 104–105 Old and Middle Persian, 325 English, 300–302 properties of, 325–327 Ethnicity Feature inventory, organization of, 727 and language, Soviet Union, 996 Feature theory, 327–331 link with language, 299–303 distinctive Ethnography of communication, 303–304 in phonology, 328–329 Ethnography of speaking (Hymes), 484 redundancy as example of use of, 329–331 Ethnomethodology, 239 feature sets, 329 Etymology, 304–306 Feedback folk, 47, 599 relevance, 532 Euphemism, 306–307 speech motor control regulated by, 838 European colonization, Pacific islands, 801–802 Femininity, See Gender European traditional grammar, 307–309 Ferguson, Charles Albert, 268–269, 331–333 Euskara (Basque), 126–127 Field methods Evaluation metrics, speech recognition, 1022–1023 data collection, 333–334 Evidentials equipment, 334 Arawak, 83 ethics, 334–335 mood and, 708 transcription techniques, 334 Tuyuca, 212 Figurative speech, 335–337 Evolution Filler-gap dependencies, 1070–1071 of agreement, 24–25 Fillmore, Charles John, 337–339 Berber, 1125–1126 Filtering, 983 Esperanto, through communicative use, 92 Fingerspelling, in ASL, 42–43 Evolution of language Finite-state morphology, 339–340 cognitive preadaptations, 311–315 Finland (Suomi), 933–934 overview, 309–311 Finnish and Finnic languages, 341–344 physical preadaptations, 315–317 Finno-Ugrian languages, 342 social preadaptations, 317–319 Finno-Ugric languages, 1003–1004 1220  Index First language acquisition Function words, 362–364, 846 and bilingualism, 4 Fusional language, 714 and cognitive development, 3–4 and critical period, 4 G and language acquisition device, 3 Ga–Dangme subfamily, 30 and social development, 4 Gaelic Firth, John Rupert, 345–347 Irish, 407 Fishman, Joshua A., 347–348 Scottish, 183, 406 Five Nations languages, 699–700 Gafat, EES language, 44 Flemish, 130–131, 355–356 Galician, 1006, 1009–1010 in Belgium, 274 as romance language, 519 Flexibility, mental, bilingualism and, 140–141 Gallaudet University, 41, 43 Focality, 348–349 Gallego, 1009–1010 Focus system, Tagalog, 1076 Games, language, 1180 Folk etymology, 47–48, 306, 599 Gbe languages, See Ewe and Gbe languages Foregrounding, 1044 Ge’ez, EES language, 44–46 Foreign language teaching, 144, 1044–1045 Gender Foreign workers, Germany, 390–391 agreement, 22 Forensic linguistics, 350–353 Chadic languages, 446 Formants, 1014 class marking, 370–371 Form-classes, structuralist, 1064 German nouns, 385–386 Formosan, 109–110 as grammatical category, 367–368 Formulation of message, 883, 1025 identity and language and, 489–490 Fossil words, 495 noun paradigms in Sindhi, 968t Fourier analysis, periodic waveforms and, 1012–1014 Semitic languages, 954t Fourier’s theorem, 1 Slavic languages, 523 Fragments, method introduced by Montague, 705–706 as social concept, 368–369 Fraktur style, 387 Gender distinction, Arawak, 83 Frame sematics (Fillmore), 338 Genderlect vs. genrelect, Sumerian, 1050 France Genealogy, Warao language, 1165 colonial expansion and la Francophonie, 355–356 Generalized Quantifier Theory, 885–886 French language, 353–354 Generation, problems unique to, 372–373 regional languages of, 354–355 Generative grammar, 373–376, 1101–1102 Free word order languages, 631 pro-form in, 878 French Generative semantics (Lakoff), 593 in Belgium, 130–132 Genericity, 376–377 classical and modern, 358 Genetic affiliation, Vietnamese, 1153 CODOFIL program in Louisiana, 1145–1146 Genetic classification, American Indian languages, 925 evolution of, 353–354 Genetic relationship, 377–378 gender agreement, 22 Genetive possessive, 868 in Italy, 547 Genre la Francophonie, 355–356, 359 endangered definition of, 380 Old and Middle, 358 stable and flexible, 379 protecting the language, 358–359 Geographical continuum, Gullah, 424 rich agreement, 21 Geographical dialect continua, 267 roles in linguistic repertoire of users, 357 Geography as romance language, 359–360, 519 Balkans, 115–116 in Southeast Asia, 994 of dialects, 462 spoken French, current changes in, 360 Georgian, 380–382, 1004 in Switzerland, 1059–1061 German, 514–515 unofficial official language of Algeria, 34–36 agreement and past tense, 23 French creoles Austrian, 103–104 Indian Ocean, 506–507 characteristics of, 385–387 Lesser Antillean, 615–617 class marking, 370 spoken in Caribbean, 178, 356 history of language, 384–385 Tayo, 803 as language of instruction in United States, 1144–1145 French Sign Language (LSF), 41–42 spoken in Fricatives, 649 Italy, 547 alveolar, 859 Switzerland, 1060 Frisian, 514 studied by North and East, 389 Grimm, 418–419 Friulian, 547 Hermann, 817–818 Fromkin, Victoria Alexandra, 360–362 Germanic languages Fulani, 746–747 Afrikaans, 514 Functional approaches to grammar, 364–365 Danish, 513 Functionalist structuralism, studied by Martinet, 656–657 Dutch, 514 1221 Index Germanic languages (continued) phrase-structure, 397–398 East Germanic, 513 generalized, 399–400 English, 514 relational grammar, 398 Faroese, 513–514 role of lexicon, 623 Frisian, 514 Grammatical aspect, 93–94 Icelandic, 514 Grammatical differentiation, 137 North Germanic, 513 Grammatical function, 401–402 Norwegian, 513 Grammaticalization Pennsylvania German, 387, 515 in ASL, 42 studied by Sweet, 1056–1057 assumptions underlying, 403 Swedish, 513 defined, 402 West Germanic, 514 Grammatical relations Yiddish, 514–515 case-marking and, 22 Germany Guaymí and Chibchan languages, 422 adherence to stem principle, 388 Grassfields Bantu languages, 1195 dialects, 389–390 Grassman’s Law, 511 Early Modern period, 387–388 Great Britain guest workers, 390–391 England, 404–405 orthographical regulation, 388 Northern Ireland, 407 pronunciation standardization, 388–389 Scotland, 406–407 regional spoken language forms, 387 Wales, 405–406 spelling reform, 388 Great Vowel Shift, 408–410 Gestures, 811–812 Greco-Roman world, 464–465 Gheg dialect, 31 Greek, ancient Givón, Talmy, 392–393 dialects, 58–59 Glossa language, 91 and European traditional grammar, 307–309 Glossematics, 471–472 and Indo-European, 62 Glottalic egressive (ejectives), 10 morphology, 60–61 Glottalic ingressive (implosives), 11 Mycenaean, 58 Glottalization, in Nootka, 755 phonology, 60 Gothic, 393–395 syntax, 61–62 Government-Binding model (Chomsky), 206–207 written sources, 59 Government policies of India, 503–504 Greek, modern Gradual metathesis, 679 geographical spread, 410–411 Grammar influence of Classical Greek, 411 ancient Greek, 60–62 linguistic structure, 412 BSL, 155–156 regional dialects, 411–412 Burmese, 162–163 Greenberg, Joseph Harold, 413–415 cohesion driven by, 216–217 Grice, H. Paul, 415–418 functional, 365 Grimm, Jacob, 418–419 generative, 373–376, 878, 1101–1102 Grimm’s Law, 222, 511 Korean, 582–583 Guang subfamily, 30 of Panini, 508–509, 808 Guarani, 1169 Romanian, 907 Guaycuruan, 989 story, 1038–1039 Guaymí and Chibchan languages, 419–423 studied by Jakobson, 550 active/nonactive systems, 423 systemic, 364–365 basic word order, 420–421 transformational generative (Chomsky), 205–206 ergative systems, 422–423 universal, 8, 908, 940–943, 1147–1149 nouns and nominal morphology, 421–422 Grammar, traditional verbs and verbal morphology, 422 Arabic, 78–79 word classes and parts of speech, 421 Babylonian, 113–114 Guest Worker German, 390–391 Chinese, 198–200 Guinea-Bissau Crioulo, 250–251 clause, 210–211 Gulf of Guinea Crioulo, 249–250 European, 307–309 Gullah, 14–15, 423–426 history of, 395–396 Gumperz, John Joseph, 426–427 Indian, 507–509 Gurenne, 710–711 Japanese, 198–200 Gur languages, 708–711, 740 Grammar theories Bresnan, 153–154 H case grammar, 398–399 Haas, Mary Rosamond, 429–431 categorial grammars, 398 Hair cells, inner ear, 54 cognitive grammar, 400, 595–597 Haitian Creole, 178, 431–432 definite clause grammar, 400 Hakka, 191–192 dependency grammar, 399 Hale, Kenneth, 432–435 lexical-functional grammar, 399 Halle, Morris, 435–437 1222  Index Halliday, Michael Alexander Kirkwood, 364–365, 437–439 Hixkaryana, two grammars, 174 Hamito-Semitic, See Afroasiatic Hjelmslev, Louis, 470–472 Handedness, lateralization and, 607–609 Hmong-Mien languages, 193, 472–476 Hand gestures, 812 typological characteristics, 473–475 Handwriting writing, 475–476 anatomy of, 439–440 Hochunk, 254 influenced by linguistic factors, 440–441 Hockett, Charles F., 476–478 models of, 440 Hohe, See Assiniboine Hanty, 482–483 Hokkaido Aborigine Protection Act, 27 Hanyu, 189–190 Homeric poems, 59 Harari (Adare), EES language, 44 Homonymy, 40 Harmonic clusters, Georgian, 381 Homophony, Sumerian, 1046 Harris, Zellig Sabbetai, 441–443 Honorifics, 553 Haugen, Einar, 443–445 Tibetan, 1105–1106 Hausa Hopi, 721, 928 Chadic languages and, 445–447 How to Do Things with Words (Austin), 98 spoken in Nigeria, 746 Human vs. nonhuman distinction, Tamil verbal system, 1079 Hawaiian, endangered language, 109 Humboldt, Wilhelm von, 478–480 Hawaiian Creole English, 448–450 Hungarian Headedness, 14 in Burgenland, Austria, 105 Hebrew, biblical and Ugric languages, 480–483 basic characteristics of, 451 Hymes, Dell Hathaway, 484–485, 979 classification, 450 Hypercorrection, 47 diglossia, 451 Hyponyms, 225 emphatics, 452 Hyponymy, 624 loanwords, 452 Hypotaxis, ancient Greek, 61 morphology, 452 orthography, 451 I phonology, 452 Iban, 646 stages of, 450–451 Iberian, 1005 syntax, 452 Iberoromance languages, 1007–1010 Hebrew, modern Icelandic, 514 noun morphology, 454 Iceland (Island), 934 phonology, 455 Iconic gestures, 812 revival in Israel, 541–542 Iconicity, 487–488 verb morphology, 453 role in IPSL, 527 word order, 454–455 Identity Hebrew–Aramaic lexicon, Yiddish, 1198 ethnic, preserving, 302 Helsinki speech, 343 language and, 489–490 Hiberno-English, 407 multiple, and stable multilingualism in India, 504–505 High German, new and modern, 384–385 performance of, 369 Himalayish languages, 192 Ideology Hindi-Urdu, 455–459, 499 language and, 490–492 morphology, 457–458 meaning and social context and, 669–670 phonology, 456–457 Ideophones, Ewe and Gbe languages, 322 syntax, 458 Idiomaticity, 492–493 Hiragana, 551–552 Idioms Hiri Motu, 459–461 influenced by social and cultural heritage, 495 Historical linguistics, 461–464 semantic opaqueness of, 494 Kiowa-Tanoan languages, 1092 words used figuratively in, 494–495 studied by Ido project, derived from Esperanto, 91 Hermann, 818 Igala, 1200–1204 Martinet, 657 Igbo and Igboid languages Historical records, BSL, 155 affixes and verbal phrases, 498 Historic Sinological Reconstruction, Old Chinese, 772 aspect, 498 History of linguistics orthography, 496 Greco-Roman world, 464–465 phonology, 497 Middle Ages, 465 spoken in Nigeria, 746 Renaissance and modern times, 465–466 syllabic structure, 497 19th and 20th centuries, 466–467 syntax, 497–498 Hittite, 467–470 tone system, 497 and Indo-European, 469–470 word order and noun phrases, 498 member of Anatolian, 511–512 Ijoid language family, 740 morphology, 468–469 Ill-formed representations, 908–910 phonology, 468 Illiteracy in Brazil, 992–993 syntax, 469 Illocutionary acts, 937 1223 Index Illocutionary force, 708, 1017 spoken in Imagic iconicity, 487 Balkans, 115–117 Immigrant population Turkey, 1131–1132 America, 300–302 Tocharian, 1112–1113 Australia, 101 Indo-Iranian languages Canada, 171 Dardic, 517 foreign workers in Germany, 390–391 Middle and New Iranian, 516 United States, and language in schools, Modern Indic, 516–517 1144–1147 Old and Middle Indic, 516 Implicational universals, 414, 1134 Old Iranian, 515–516 Implicature, contributions of Grice, 416 Indonesia Implosives (glottalic ingressive), 11 Austronesian languages, 108–109 Incorporating languages, 758–759 Indonesian, 994 India Indo-Pakistani Sign Language (IPSL) education and government policies, 503–504 social and political factors, 528–529 English, 503 structural features languages and modes of communication, 499–500 different from other sign languages, 528 multilingualism, 504–505 shared with other sign languages, 526–528 Persian, 502–503 word classes and sentence structure, 524–526 rural vs. urban divide, 504 Indus Valley civilization, 273 Sanskrit, 501–502 Infancy, language development in, 264–265 scripts, 505 Inferred meanings, 669 Telugu, 1086–1088 Infinitude of language, 1068–1069 Indian Ocean Creoles, 506–507 Infixation, 12 Indian traditional grammar, 507–509 Infixes, dimensional, 1048 Indic, 516–517 Inflection Indigenous languages agreement and, 23 Americas, tone languages, 1119 Cape Verde creole, 173 Argentina, 988–990 derivation and, 529–530 Brazil, 991 explained in terms of affixation, 760 Nigeria, 746–747 Slavic languages, 523 Pacific islands, 802 Inflectional morphology, English, 290–291 Pama-Nyungan languages, 804–807 Inflectional phrase, 210 and Pintupi, 852–855 Information-processing models of handwriting, 440 United States, 1139–1140 Information retrieval, 531–534 Indigenous populations Ingressives, glottalic and velaric, 11 Algonquian speakers, 765 In-group vs. out-group theory, 427 Australia, 101 Inner ear, 53–54 Berbers, 34–35 Inor, EES language, 44 and language death, 602 Input, affecting second language acquisition, 6 Michoacán, Mexico, 1080–1082 Institutionalization of idiom, 493 Pacific Northwest, 203–205 Insular Celtic, 786 Indo-Aryan, 499–500 Intelligence, artificial, 88–90 Dravidianization of, 502 Intentionality Sindhi, 965–970 studied by Searle, 937–938 studied by Emeneau, 279–280 work by Grice, 416 Indo-European Interactional sociolinguistics, 239, 489, 979 Anatolian cuneiform, 511–512 Interference Armenian, 85–88, 1002 shift-induced, 848 Baltic languages, 118–121, 1002 structural, 1039–1040 comparative method, 221–222 Interlinguistics, 91 dialectical divisions, 511 International Auxiliary Language Association, 925 East Slavic languages, 1001–1002 International Phonetic Alphabet (IPA), 831–832, 835 Germanic languages, 512–515 International planned languages, 90–91 grammatical features, 509–510 Internet access to language-teaching resources, 230–231 Greek and, 62 Interpreting, 352 Hittite and, 469–470 in consecutive mode, 534–535 homeland location, role of archaeology, liaison, 536 84–85 physical setting in, 535–536 Indo-Iranian languages, 515–517 relay, 536–537 Iranian languages, 1002 simultaneous, 535 laws describing reconstruction steps, 511 Interrogative generalized quantifiers, 887 major branches, 510 Interrogative sentences, 212–213 romance languages, 517–521, 1002 Udmurt, 1138 Sanskrit and, 923 Intersubjective meaning, and dialogism and reader response, Slavic languages, 521–524 670 1224  Index Intonation Jinyu, 192 meanings of, 811 Johor Malay, 1036 as prosodic feature, 880 Joint attention, as prepragmatic capacity, 314 sentence, 316 Jones, Daniel, 831 studied by Pike, 851 Jones, Sir William, 565–567 Intracarotid sodium amytal test (Wada test), 733 JSL, See Japanese Sign Language (JSL) Intransitive constructions, Okanagan and Salishan languages, 770 Juba, Arabic-based pidgin-creole, 567–568 Inuit Judeo-Spanish, 1009, 1133 dialect groups, 298 Jury instructions, 248 split from Yupik, 296 Inuktitut (Eskimo), portmanteau, 24 K Invariance Kabardian, 757 lack of, 1019 Kabardyn, 1004 studied by Jakobson, 549 Kadugli-Krongo, 751 Iowa-Oto, 254 Kanji, 551–552 IPA (International Phonetic Alphabet), 831–832, 835 Kansa, 254 IPSL, See Indo-Pakistani Sign Language (IPSL) Karelian, 343–344 Iranian languages, 515–516, 1132 Kartvelian, 380–381 Osetin and Kurd, 1002 Kasem, 711 Tajik, 1002 Katakana, 551–552 Ireland, Northern, 407 Katharevusa, 411–412 Irish Kayardild, 569–571 characteristics of English pronunciation, 301 Kazakh, 1003 sound system, 183 Kelantan-Patani Malay, 646 Irish Gaelic, 407 Kesab (Syria) dialect of Armenian, 88 Irony, 537–538 Kestane (Christian), See Soddo Iroquoian languages, Mohawk and, 699–702 Kham Tibetan, 1104–1105 Irrealis mood, agreement and, 23 Khasi, 573 IsiZulu, 1207–1208 Khmer, 571–572 Isle de France creoles, 506–507 Khoisan languages, 574–576, 986 Isoglosses, 267 Kikongo-Kituba, 577–579 Isolating languages, 713 Kinship Israel, 538–542 categories, in Thai, 669 Italian, 518, 542–545, 1060 studied by Lévi-Strauss, 618–619 Italy, 545–548 terms, 576–577 Item & arrangement approach to affixation, 13 in Vietnamese, 1154 Item & process approach to affixation, 13 Ki-Nubi, Arabic-based pidgin-creole, 567–568 Itsεkiri, 1200–1204 Kiowa-Tanoan languages contemporary speech communities, 1092–1093 J historical linguistics, 1092 Jakarta Indonesian, 647 language and culture, 1093 Jakobson, Roman, 549–551 structural features, 1092 Jamaican Creole, 1173–1174 Tewa and, 1091–1094 Japanese Kituba, 577–579 case-marking and grammatical relations, 22 Kiva speech, 1093 clause structure, 554–555 Komi-Zuryanskiy, 1137 dialectical variation, 552–553 Konkomba, 711 honorifics, 553 Kordofanian language family, 741 impact on Ainu language, 27–28 Korean kanji, 551–552 area of use and variants, 579–580 link to Altaic, 37 dialect, 580 loanwords, 553–554 grammar, 582–583 long-distance scrambling, 632 long-distance scrambling, 632 Middle (Classical), 684–687 phonetics and phonology, 580–582 script, 1193 as tone language, 1118 sharing of grammatical characteristics, 552 vocabulary, 583–584 Tokyo, 1118 writing system, 37–38, 580 verb classification, 554 Krama, level of Javanese, 559, 561 Japanese-Okinawan subfamily of Altaic, 37 Krio, 584–585 Japanese Sign Language (JSL), 555–558 Kriol (Roper River Creole), 585–587 Japanese traditional grammar, 198–200 Kru language family, 741 Jargon, See also Creoles; Pidgins Kryz language, 382 Chinook, 203–205 Kurdish, 80, 1002 Pacific Jargon English, 857 Kurgan model, archaeological, 85 Javanese, 558–562 Kusaal, 711 Jespersen, Otto, 562–565 Kwa, 616, 741 1225 Index Kwakiutlan, 754 Language contact, 600–602 Kyakhta Pidgin Russian, 201–203 pidgins and creoles resulting from, 848 Kyrgyz, 1000, 1003 in Scandinavian/Nordic past and present, 934–935 Sranan as product of, 1031 L Language death, 602–603 Labeling, by parrots, 66–67 Language families Labial harmony, 1159 former Soviet Union republics, 1000–1004 Labial sounds, 858–859 migration and, 692 Labov, William, 589–591, 1151–1152 Language games, 1180 Ladefoged, Peter, 591–593 Language-in-education planning, 605 Ladino, 540–542, 546–547, 1133 Language mixing Lakoff, George, 593–595, 671 by adult bilinguals, 137 Lakota, 254 theory of, 128 Langacker, Ronald, 595–597 Language planning, 603–605 Language, See also Artificial languages; Endangered languages; studied by Haugen, 443 Paralanguage Language policy, Soviet Union, 996–1000 aging effects, 19–21 Language shift, 602–603 animal and human Language socialization, 606–607 dolphins, 64–65 Language structure overview, 62–64 English, 290–293 parrots, 66–67 Philippine Spanish Creoles, 823 primates, 67–70 signed languages, 962–965 developmental stages, 264–266 as system of interacting modules, 1148 direction, in interpreting, 536 Language use distribution, in South Africa, 986–987 corpus-based investigations of, 245–246 emotion and, 281–282 mental processes underlying, 1184–1185 evolution of, 309–319 in Swiss communities, 1060 functions of, studied by Jakobson, 549–550 Langue des signes Française (LSF), 41–42 genetic relationship, 377–378 Langue–parole distinction, 1094 legal, 248, 352 Lao, 994 link with ethnicity, 299–303 Lardil, 569–570 mass media and, 658–661 Laryngeal Hypothesis (de Saussure), 222, 512 medicine and, 672–675 Laryngeal sounds, 860 migration and, 691–693 Laryngeal voice quality settings, 810 neural systems for, 965 Larynx neural theory of (Lakoff), 594 preadaptation in, 316 personality and, 821–822 role in speech production, 50–51 philosophy of, 827–828 Late Common Slavic, 774–775 processing, working memory and, 1191 Late Old Chinese, 773 progress in, 563 Lateralization, and handedness, 607–609 of Qur’an, 75–77 Latin in schools, and US immigrant languages, 1144–1147 adjectives and adverbs, 610–611 vitality and attrition, 287 and European traditional grammar, 307–309 Language acquisition history of, 609–610 by children introduction to Spain, 1005–1006 studied by Jespersen, 564 in Italy, 546 studied by Slobin, 974 in linguistic history of Balkans, 116 first language, 3–4 spoken vernacular, 611 second language, 4–7 word structure, 610 theories, 7–9 Latvian, 120–121, 1002 Language acquisition device, 1148 Law of the Palatals, 511 first language acquisition and, 3 Learner language, in second language acquisition, 5–6 Principles and Parameters approach, 8 Learning Language change computer-assisted language (CALL), 229–231 external and internal motivations, 597–598 experience-dependent and experience-expectant, 7–8 lexical change, 600 as natural property of connectionist networks, morphological change, 598–599 234 phonological change, 598 second language, 73, 940–943 semantic change, 599–600 Legal language, 248, 352 studied by Lehmann, Winfred Philipp, 612–613 Greenberg, 414 Lengthening, compensatory, 1063 Weinreich, 1170 Leskien, August, 613–615 syntactic change, 599 Lesser Antillean French Creole, 615–617 work by Brugmann, 157 Leveling Language classification, Old Tibetan, analogical, 47 794 as morphological change, 599 1226  Index Lévi-Strauss, Claude, 617–620 cross-cultural studies in, 628 Lexemes, 1064, 1184 relativistic perspective on, 627 Lexical access, 624–625 Literary languages Lexical ambiguity, 40 Adyghe and Kabardian, 757 Lexical borrowing, 279, 462, 620–622 Burmese, 161 Lexical causatives, 180–181 Georgian, 381 Lexical creativity, English, 292–293 Nakh-Daghestanian languages, 382 Lexical fields, emotion words forming, 282 Nigerian Pidgin English, 744–745 Lexical-functional grammar, 153–154, 399 Sanskrit, 508 Lexicalization, 622–623 Sri Lanka Portuguese, 1035–1036 Lexical statistics, 661 Lithuanian, 119–120, 1002 Lexical tone studied by Leskien, 613–614 Hmong-Mien languages, 473 Liturgical language, Aramaic, 80 Yémba, 1196 Loanwords Lexical words, 362–363, 846 Albanian, 32 Lexicon Biblical Hebrew, 452 Arabic and Farsi, 325 Dutch, 275 compounds in, 624 Japanese, 553–554 JSL, 557–558 rejected by Tewa, 1093 Middle (Classical) Japanese, 686 Localization morphemes, 623–624 of linguistic information, 629–631 Old Japanese, 789 role in IPSL, 527–528 organization of, 624–625 Logicism, advocated by Russell, 911 role in grammatical theories, 623 Logographic writing system, 1192 Saramaccan, 930 Logophoric pronouns, 1058–1059 Sranan, 1032 Lojban language, 91 Lexicon Project at MIT, 434 Long-distance dependency, 631–633 Lezghian languages, 382 Long-range comparison, 633–635 Lezginian language, 1004 Lord’s Prayer, in different varieties of Armenian, 87–88 Liaison interpreting, 536 Louisiana, CODOFIL program, 1145–1146 Likpakpalnli, 711 Louisiana Creole, 635–636 Linear prediction, 981–983 Lower articulators, role in speech production, 52 Linear predictive coding (LPC), 1015–1016 Lower Sorbian, 865 Lingua Franca, history of, 625–626 Low German, 385 Lingua Romana, 611 Lule-Vilela, 989 Linguistic borrowing, Egyptian–Greek, 242 Lungs, role in speech production, 50 Linguistic determinism, 927 Lyons, John, 636–638 Linguistic diversity Soviet Union, 996 M Switzerland, 1061 Mabia languages, 709 Linguistic history Macedonian, 958 Balkans, 116–117 Machine translation, 124–125, 226, 639–640 English, 288–290 Macropragmatics, 871 Linguistic information, localization of, 629–631 Magnetic resonance imaging (MRI), 732–733 Linguistic relativity, 393, 927–928 Magnetoencephalography, 732 Linguistic rights in United States, 1145 Malagasy, 109 Linguistics dialects, 641 applied, 72–74 history, 641 clinical, 213–214 morphology, 642–643 cognitive (Lakoff), 594 phonology, 641–642 computational, 125, 226–229 syntax, 643 diachronic, 647–648 Malay, 109, 994 forensic, 350–353 onomatopoeic words, 796t historical, 461–464 Malay-Indonesian history of, 464–467 and Malayic languages, classification, 644–645 Jakobson’s legacy in, 550–551 Mayalic isolects, 646–647 mathematical, 661–662 Standard modern, 696–697 history, 645 rhetoric and, 903–906 typical features, 645–646 text, 1094–1095 writing systems, 646 utterance-centered, 1149–1150 Malayo-Polynesian, See Austronesian Linguistic sign systems, 142 Malkiel, Yakov, 647–649 Linguists, professions for, 874–875 Mampruli, 711 Literacy Manado Malay, 647 acquisition of, 628–629 Manchu language, 1127–1128 campaign in Soviet Union, 997 disharmonic words, 1160 1227 Index Manchuria Pidgin Russian, 201–203 precognitive modern theories of, 677 Manchu-Tungus languages, 1003 studied by Jakobson, 550 Mandan, 255 Metathesis, 678–680 Mandarin Metonymy, 335, 622–623 characters and sounds, 195 cognitive theory of, 680–681 distribution of, 190–191, 194 studied by Jakobson, 550 morphology, 196–197 Metrical theory, studied by Halle, 436 phonology, 196 Mexicano, See Nahuatl syntax, 197–198 Mexico Mande languages, 121–123, 741 demographic and geographic profile of languages in, 682 Mandenkan (Manding) languages, 123–124 development and characteristics of Mexican Spanish, 682 Mani-Bandama languages, 121–122 distinctions of Spanish in, 683 Manila Bay Creole, 822–823 surviving native languages in, 683–684 Manipulation, as prepragmatic capacity, 313 Tarascan language, 1080–1082 Mansi, 481–482 Miao-Yao, 472–476, 994 Maori and Polynesian languages, 651–654 Micronesian languages Markedness, 909–910 Marshallese and, 654–656 Maroon creoles, 928–929 Nuclear, 802 Marshallese and Micronesian languages, 654–656 Microplanning, component of generation, 372 Martinet, André, 656–658 Middle Ages, 465 Masculinity, See Gender Middle (Classical) Japanese, 684–687 Mass media and language, 658–661 lexicon, 686 Mass nouns, 862 morphology, 686 Matacoan, 989 nominal morphology, 686–687 Matagalpa, 693–694 phonology, 685–686 Mathematical linguistics, 661–662 sources, 685 Mathematical logic, studied by Montague, 705–706 verb morphology, 687 Mau, 507 writing system, 685 Mayan languages Middle ear, 53 aspect, 663 Middle English ergative/absolutive agreement, 22 dialects, 688 Maya and, 662–666 historical background, 688 numerical system and numerical classifiers, 664 orthography, 688–689 phonology, 665 sound system, 689–690 pronominal system and ergativity, 664 vocabulary and word formation, 690–691 spoken in Mexico, 684 word and sentence structure, 690 McCawley, James David, 666–668 Middle French, 358 Meaning Middle German, 384 acquisition by children, studied by Clark, 209 Middle Indic, 516 change in, 599–600 Middle Iranian, 516 coded and inferred, 669 Middle Old Chinese, 773 contributions of Grice, 416 Middle Persian, 325 and experiential hypothesis, 670–671 Middle voice, 1158 hidden, mass media and, 659 Migration intersubjective, and dialogism, 670 ethnic, Soviet Union, 999–1000 of intonation, 811 Germanic tribes, 782–783 naming and, 725 language and, 691–693 noun phrase, 886 Nguni, 1207–1208 semantic and pragmatic, 668–669 Mikeyir, See Shabo and social context and ideology, 669–670 Minangkabau, 646 structuralism and deconstruction, 670 Mind reading, 313 theories on (Meillet), 676 Minimalism, 256 Media Lengua, as bilingual mixed language, 138 Min Nan, 191 Medicine and language, 672–675 Minority languages Meillet, Antoine, 675–677 BSL as, 156 Melanesian Pidgin, 1114–1116 Celtic, 184 Memory, See Working memory in China, 193–194 Mental calculation, 313 Minority/majority group relations, sociolects and, Mental lexicon, 1185 976 Mention theory approach to irony, 537 Miskito and Misumalpan languages, 693–695 Meroitic, 751 Mississippi Valley languages, 253–254 Meronymy, 624 Missouri Valley languages, 254 Mesqan, EES language, 44 Modality effects, signed languages, 962–965 Metaphor, 335, 622, See also Idioms Model/rival training, of parrots, 66–67 cognitive theory of, 677–678 Modern Greek, 410–413 and meaning and experiential hypothesis, 670–671 Modern Hebrew, 453–455 1228  Index Modern Indic, 516–517 study of, 715–717 Modern linguistics, 696–697 Tivoid languages, 1108–1109 Modern Persian, 325 Yoruboid languages, 1202–1203 Modern Standard Arabic, 34–36, 75–77 Morphophonemic principle, 1193–1194 Modification, 697–699 Morphophonology, Sindhi, 966 adverbial, 900 Morphosyntactic properties, loss of, 403 number marking by, 760–761 Morphosyntax Modularity, 361 ancient Egyptian, 56–57 Mohawk and Iroquoian languages, 699–702 Lesser Antillean French Creole, 616–617 Moldavian/Romanian, 1002 Okanagan and Salishan languages, 769 Monemes, studied by Martinet, 657 Old English, 778–779 Mongolian, 702–705, 1003 Sumerian, 1049–1050 Mongolic languages, 38 Morris, Charles W., 950–951 spoken in China, 193 Motivation Mon-Khmer languages, 571–573, 993–994 in iconicity, 487 branch of Austroasiatic, 106–108 in idiomatic expressions, 494–495 Mon language, 572–573 for language change, 597–598 Monosyllabic word structure Motu, 459 Hmong-Mien languages, 473–474 Mouth, role in speech production, 52 Thai and Tai languages, 1099–1100 Mouthing, role in IPSL, 528 Montague, Richard, 705–707 Movement trace, 284 Mood, 707–708 Mozarabic, 1005–1006 Sintu languages, 1207 Multilingualism, See also Bilingualism Warao language, 1165–1166 in India, 504–505 Moore language, 710 language shift and, 602 Moraic theory of syllable, 1062–1063 in Mosquito Coast, 694–695 Morocco, Berber languages, 1124–1125 New Guinea, 736 Morphemes Nigeria, 742–743 concept of, 712–713 sociolinguistics inspired by, 978 empty, 285–286 Multiple negation, in AAVE, 15 handshape and movement, 963 Multistratal theories, 233 multiple, 623–624 Munda branch of Austroasiatic, 106 and study of morphology, 715–717 Murrinh-Patha and Daly languages Morphological change, 598–599 areal features, 718 Morphological typology, 713–715, 1135 classification, 717–718 Semitic languages, 952 free pronouns, 719 Morphology nominal classification, 719 ancient Greek, 60–61 phonology, 720 Arabic, 76 pronominal indexing, 718 Biblical Hebrew, 452 verbal classification, 718–719 Burushaski, 164–166 Muscovite dialect, 915 Carib and Cariban languages, 175 Mutations Chinese (Mandarin), 196–197 consonant, Atlantic languages, 1182–1183 distributed, 436 initial, Celtic languages, 183 Ewe and Gbe languages, 320 palatal, Common Norse, 792 Farsi, 326 Myanmar (Burmese), 160–163 finite-state, 339–340 Mycenaean Greek, 58 Hindi-Urdu, 457–458 Mythology, studied by Lévi-Strauss, 619 Hittite, 468–469 inflectional, English, 290–291 N Kriol (Roper River Creole), 586–587 Nahuatl, 684 Malagasy, 642–643 and Uto-Aztecan languages, 721–724 Marshallese, 655 Nakh-Daghestanian languages, 382–383 Middle (Classical) Japanese, 686 Nakota, 254 Navajo and Athabaskan-Eyak language, Naming, 724–726 729–730 Nasalization, Spanish and English, 96 Nilo-Saharan, 752 Nationality, identity and language and, 490 Nootka, 755 National languages, See State languages Old Japanese, 789 Native American Language Act, 1146 Old Norse, 793 Native American languages, See also American Indian Panini’s treatment of, 508–509 languages Sanskrit, 921, 923 in Caribbean, 176–177 Saramaccan, 930 Chinook Jargon, 203–205 Sindhi, 966–969 grammars of, 309 Siouan-Catawban languages, 255 kinship terms, 576–577 Sranan, 1032 Ojibwe and Algonquian, 764–767 1229 Index Native American languages (continued) Nigeria studied by ethnicity and linguistic identities, 745–746 Boas, 145–146 indigenous languages, 746–747 Haas, 430 linguistic landscape, 742–743 Native languages, Mexico, 683–684 literary utilization of NPE, 744–745 Natural classes, 726–728 water-borne languages of, 743 Naturalism–schematism continuum, 91 Western education in, 747–748 Natural language Nigerian Pidgin English (NPE), 743–745, 747 determiners, 886 Niger-Kordofanian languages, 986 generation, 372 Nilo-Saharan languages, 749–753 interaction, 226–228 demography, 752–753 studied by Montague, 705 external boundaries, 750–751 Natural selection, 310 internal subgrouping, 751–752 Navajo and Athabaskan-Eyak language, Node branching, 845 728–731 Nominal agreement, 24 classificatory verbs, 730 Nominal arguments, Okanagan and Salishan languages, 770 future of, 730 Nominal classification, Murrinh-Patha and Daly languages, 719 morphology and verb structure, 729–730 Nominal morphology phonology, 729 Middle (Classical) Japanese, 686–687 Negative constructions Northwest Caucasian languages, 757 Chadic languages, 447 Old Japanese, 789–790 incompatible with definiteness, 258 Sumerian, 1046–1047 Negative position, universals of, 1067 Nominals, anaphoric dependencies between, 49 Negerhollands, 178 Nonadjacent metathesis, 679 Neogrammarians, 222, 267, 466 Nonlinguistic gestures, 812 Bloomfield, 143–144 Nonmanual signs, in ASL, 42 Brugmann, 157 Nonverbal communication, 220–221 Hermann, 817–818 Nootka and Wakashan languages, 753–756 Leskien, 614 Northeastern Caucasian languages, 1004 Wackernagel, 1163 Northern Athabaskan, 728 Networks, connectionist, 234–235 Northern Ireland, 407 Neural system Northern Irish English, 407 for language, 965 Northern Mon-Khmer, 107 regulating speech production, 317 North Germanic, 513 Neural theory of language (Lakoff), 594 Northwest Caucasian languages, 383, 756–758, 1004 Neurobiology, 1023–1024 Norway (Norge, Noreg), 933 Neurolinguistics Norwegian, 513, 1055–1066 clinical, 731–732 studied by Haugen, 443 experimental, 732–733 Nostratic superstock, 19 profession of, 875 Noun classifiers, Hmong-Mien languages, 474 simulative, 733 Noun class system, Atlantic languages, 1183 New Arabic, 75–77 Noun incorporation, 758–759 New Benue-Congo language family, 740 Noun morphology, Modern Hebrew, 454 New Guinea Noun phrase Austronesian and Papuan language groupings, Chadic languages, 447 735 Ewe and Gbe languages, 322 multilingualism, 736 Igbo and Igboid languages, 498 pidgins and creoles, 736–737 meanings of, 886 political divisions, 734 Wayampi, 1169 New Guinea Pidgin, See Tok Pisin Nouns New Iranian, 516 Burushaski, 164–165 Ngoko, level of Javanese, 559, 561 collective, 862 Nguni, 986, 1207–1208 Coptic Egyptian, 242–243 Nida, Eugene Albert, 737–739 count, 861–862 Niger-Congo German, 385–386 Adamawa-Ubangi, 739–740 Guaymí and Chibchan languages, 421–422 Atlantic languages, 740 Italian, 544 Dogon, 740 Latin, 610 Gur, 740 mass, 862 Ijoid language family, 740 Mohawk and Iroquoian languages, 700 Kordofanian language family, 741 naming and, 724 Kru language family, 741 Ojibwe, 766 Kwa, 741 Old Church Slavonic, 776 Mande languages, 741 Old Irish, 787 (New) Benue-Congo, 740 proper, 878–879 relationship to Nilo-Saharan, 751 Sindhi, 966–967 1230  Index Sri Lanka Portuguese, 1035 Organ of Corti, 54, 151 Tibetan, 1105 Organon model (Bühler), 158–159 Nubi, Arabic-based pidgin-creole, 567–568 Orthography Nucleus–coda division of syllable, 1062 Biblical Hebrew, 451 Number marking, 760–761 German, strict regulation of, 388 Warao language, 1166 Igbo and Igboid languages, 496 Numeral classifiers, 280 Middle English, 688–689 Mayan languages, 664 Old English, 777 Numerical system, Old Japanese, 790 Osage, 254 Numic, 721 Oscillogram, 980–981 Nung language, 970 Osetin, 1002 Nyo, daughter languages in, 28–29 Ossetic, 1132 Ostrogothic, 394 O Oto-Manguean languages, 693, 797–800 Oaths of Strassburg, 785 Outer ear, 53 Obstruents, and voicing, 11 Output, affecting second language acquisition, 6 Occitan, 355 Oxymoron, 336 as romance language, 519 Oceanic subgroup of Austronesian, 110, 735–736 P Motu, 459 Pacific Official languages, See also State languages European languages in, 802 selection of, 763–764 indigenous languages of, 802 Singapore, 1037 islands of, 801–802 Udmurt, 1138 pidgins and creoles in, 802–803 Off-target verbosity, 20 Pacific Coast Athabaskan, 728 Ogihara, T., 1090–1091 Pacific Jargon English, 857 Ohio Valley languages, 254 Pahawh Hmong, 475 Oirat-Khalka branch of Mongolian, 38 Paiwanic languages, 193 Ojibwe and Algonquian languages, 764–767 Palatal harmony, 1159 Okanagan and Salishan languages, 767–771 Palatals Okinawan languages, 37 articulation and, 859 Old Arabic, 75–77 Law of, 511 Old Bulgarian, studied by Leskien, 614 mutations in, Common Norse, 792 Old Chinese, 771–774 Old Indo-Aryan, 280 Old Church Slavonic, 774–776 Palauan, 23 Old English Palaung-Wa, 107 dialects, 777 Paleoasiatic languages, 1004 morphosyntax, 778–779 Paleography, 824–825 orthography, 777 Pama-Nyungan languages phonology, 777–778 current state of, 804–805 possessives in, 869 origin and dispersal of, 806–807 vocabulary and word formation, 779–780 and Pintupi, 852–855 Older adults, aging effects on language, 19–21 status in relation to non-Pama-Nyungan, 805 Old French, 358, 780–782 subgrouping, 806 Old German, 384 Panini Old High German, 782–785 grammar of, 808 Old Indic, 516 language described by, 807 Old Iranian, 515–516 treatment of morphology, 508–509 Old Irish, 183, 786–788 Papiamento, 808–810 Old Japanese, 788–791 Papua, area of language contact, 460 Old Norse, 791–793 Papua New Guinea, Tok Pisin, 1114–1116 Old Persian, 325 Papuan languages, 735–737, 802 Old Prussian, 119 Paralanguage Old Tibetan, 793–795 gestures, 811–812 language classification, 794 intonation, 811 verbal system, 795 voice quality, 810–811 vocabulary, 794 Paralinguistic meaning of intonation, 811 writing system and phonology, 794–795 Parametric coding, 1028–1029 Omaha-Ponca, 254 Paraphrasing, 1097 Omotic, 17 Parison, 336 Onomatopoeia, 796–797 Parrots, communication with humans, 66–67 Onondaga, 700 Parsing, 883 Onset of syllable, 1062–1063 affix stripping, 816 Opaqueness, semantic, of idioms, 494–495 augmented transition networks, 814–815 Optimality theory, 154, 842, 908–909 chart parsers, 814 Oral cavity, role in speech production, 52 context-free parsers, 813 1231 Index Parsing (continued) Korean, 580–582 probabilistic, 815 phoneme and phonetic variation, 835–836 semantic, 816 prosody, 836–837 tagging, 815–816 speech acoustics and speech perception, 838–839 top-down and bottom-up, 813–814 speech motor control and physiology, 837–838 Passive construction, 1157–1158 studied by Sanskrit, 923 Ladefoged, 591–592 Past tense, agreement and, 23 Sweet, 1056 Paul, Hermann, 817–818 Trubetzkoy, 1122 Peirce, Charles Sanders, 819–820, 950 Phonetic transcription, 831–833 Pejoration, euphemism and, 306–307 broad and narrow, 832–833 Pennsylvania German, 387, 515 history of, 832 Perception, speech, 838–839, 1018–1021 IPA charts, 832 Perceptual strategies (Bever), 135 Phonetic writing, Yiddish, 1197–1198 Performance of identity, 369 Phonological awareness, 895 Performatives, 1017 Phonological change, 598 distinguished from constatives, 98–99 Phonological differences Persian, 325 Algerian Arabic and Modern Standard Arabic, 35–36 influence on Indian languages, 502–503 Hawaiian Creole English and mainland English, Persio-Arabic script, 456 449–450 Personality and language, 821–822 Phonological loop, working memory, 1190 Person deixis, 260–261 Phonological opposition, 830 Personification, 336 Phonological statistics, 661 Person markers Phonology Guaymí and Chibchan languages, 422 Ainu, 26 Okanagan and Salishan languages, 769–770 ancient Egyptian, 55–56 Warao language, 1166 ancient Greek, 60 Pharyngeal consonants, 859–860 Biblical Hebrew, 452 Pharynx, role in speech production, 51–52 Burmese, 161–162 Philippine languages, 1074–1075 Burushaski, 164 Philippines, Austronesian languages, 109 Carib and Cariban languages, 175 Philippine Spanish Creoles, 822–824 Chinese (Mandarin), 196 Philology Coptic Egyptian, 242 decipherment, 825–826 distinctive features in, 328–329, 840–841 diplomatics, 825 EES languages, 46 epigraphy, 824 epenthesis and syncope and, 294–295 paleography, 824–825 Ewe and Gbe languages, 320 textual criticism, 825 Farsi, 325 uses of, 826 Hindi-Urdu, 456–457 Philosophy of language, 827–828 Hittite, 468 Phoenician, 955 Igbo and Igboid languages, 497 Phonematic unit, 346 Korean, 580–582 Phonemes Kriol (Roper River Creole), 586 allophones, 829–830 Malagasy, 641–642 ancient Egyptian, 56 Mayan languages, 665 conceptual development by Sapir, 924–925 Middle (Classical) Japanese, 685–686 contrastive function, 829 Modern Hebrew, 455 distribution, 830 Murrinh-Patha and Daly languages, 720 major early works on, 830–831 Navajo and Athabaskan-Eyak language, 729 Nootka, 755 Okanagan and Salishan languages, 768–769 Papiamento, 809 Old English, 777–778 phonetic context, 829 Old Japanese, 789 and phonetic variation, 835–836 Old Tibetan, 794–795 phonological opposition, 830 optimality theory, 842 studied by phonemes and allophones, 840 Bloomfield, 143–144 rules, 841–842 Swadesh, 1051 Sanskrit, 921 Trubetzkoy, 1122–1123 Saramaccan, 930 in study of phonology, 840 Sranan, 1032 term introduced by Baudouin de Courtenay, 128 structural, studied by Jakobson, 549 Phonetic context, 829 studied by Halle, 435–436 Phoneticians, 874 Sumerian, 1046 Phonetics suprasegmentals, 842 acoustic, 1–2 Tarascan, 1082 basic classificatory principles, 833–835 theoretical (Chao), 188 classical, 327–328 Tivoid languages, 1108 1232  Index Tocharian, 1113 Pontic, 411 Yoruboid languages, 1200–1202 Portmanteau, 24 Phonotactics, Sumerian, 1046 in context of analogy, 47 Phrasal coordination, 240–241 Mande languages, 122 Phrasal modification, 698 morphs, 760 Phrase structure Portuguese and constituency, 235–236 inflected infinitive, 1009 distinction between lexical and function words, 846 as romance language, 518–519 maximally binary branching nodes, 845 spoken in Brazil, 991–992 phrasal constituents, 843–844 Sri Lanka, 1033–1036 X bar theory, 844–847 Portuguese-based creoles, in Southeast Asia, 994–995 Phrase-structure grammar, 397–398 Portuguese Creole, 250–251 generalized, 399–400 Positron emission tomography (PET), 732 Physical preadaptations, 315–317 Possession Physical setting, in interpreting, 535–536 in Austronesian, 111 Pidgins, See also Creoles Warao language, 1166–1167 Chinook Jargon, 203–205 Possessive determiners, 263 comparison with creoles, 847–850 Possessives, 867–869 English-based, 293–294 Postalveolar sounds, 859 West African, 585 Postpositions, Ewe and Gbe languages, 322 Fanakalo, 323–324 Potawatomi (Algonquian), 24 Hiri Motu, 459–461 Potou–Tano languages, 28–29 Juba and Nubi Arabic, 567–568 Pragmalinguistics, 870 language contact and, 601, 848 Pragmatic meaning, 668–669 Lingua Franca as, 625–626 Pragmatics, 270, 869–872 in New Guinea, 736–737 Pragmatism, studied by Peirce, 819–820 Nigerian Pidgin English, 743–745, 747 Preadaptations Pacific region, 802–803 cognitive, 311–315 in Southeast Asia, 994–995 physical, 315–317 Tok Pisin, 1114–1116 social, 317–319 Pike, Kenneth Lee, 850–852 Predication, 872–874 Pimic branch of Uto-Aztecan, 721–722 Predicational aspect, 94 Pintupi and Pama-Nyungan languages, 852–855 Prefabricated phrases, Chinese (Mandarin), 197–198 Pipil (Nahuatl variety), 722 Prefixation, 12 Pitcairnese, 855–857 Yoruboid languages, 1202 Pitcairn-Norfolk, 855–857 Prefix conjugation, Semitic languages, 953t Pitch-accent languages, 1117 Prephonetic capacity, 312 Place deixis, 261 Preposing, 236 Place names, 725–726 Prepositional possessive, 868 Planning, language, 603–605 Prepositions, Ewe and Gbe languages, 322 Plantation creoles, in Suriname, 928–929 Prepragmatic capacities Plato, 307–308 cooperation, 313 Plurality, 860–862 joint attention, 314 collective nouns, 862 mind reading and manipulation, 313 count nouns, 861–862 ritualized action, 314 Hmong-Mien languages, 474 Preschool years, language development in, 265–266 mass nouns, 862 Presemantic capacities Warao language, 1166 basic concept formation, 312–313 Pluricentrism, in Austria, 103–104 complex concept formation, 313 Poems mental calculation, 313 Ainu, 26 Prestige planning goals, 605 Homeric, 59 Presyntactic capacity, 312 Pointing, in signed languages, 964 Preverbs, Ojibwe, 766 Polabian, 865–866 Primates Polish, and Cassubian, 863–864 communication with humans, 67–70 Politeness, 488 joint attention, 314 contributions to pragmatics, 870 make-believe play, 318 conveyance of, 866–867 Principle of relative frequency (Zipf), 1205 Yoruboid languages, 1204 Principles and Parameters model (Chomsky), 8, 207, 1148 Politics, and ideology-language intersection, 491 Probabilistic parsing, 815 Polynesian languages, 109, 802 Productive causatives, 181 Maori and, 651–654 Professions for linguists, 874–875 Polysemy, 40 Proficiency testing, 875–877 Polysynthetic languages, 714 Pro-forms, See also Pro phenomenon Ojibwe, 765–766 in generative grammars, 878 Polysystemicity, 346 pronouns, 877 1233 Index Prognosis, aphasia, 72 Quapaw, 254 Progressive assimilation, 97 Québec Francophones, 170–171 Pronominal arguments, Okanagan and Salishan languages, 770 Québecois ethnicity, 302 Pronominal indexing, Murrinh-Patha and Daly languages, 718 Québecois French, 356 Pronominal prefixes, Kiowa-Tanoan languages, 1092 Quechua, 988 Pronouns Question answering, 1097 agreement and, 24–25 in information retrieval, 533 Albanian, 32–33 Questions, as elicits or requests, 888–889 ancient Egyptian, 56–57 Quine, Willard van Orman, 827–828, 889–892 Burushaski, 165 Quotative construction, Sri Lanka Portuguese, 1035 Castilian leísmo, 1009 Qur’an, language of, 75–77 Dutch, 275 free, Murrinh-Patha and Daly languages, 719 R German, 386 Rabaul Creole German, 803 Mayan languages, 664 Racism, linguistic, 491 Mongolian, 704 Ralik dialect of Marshallese, 655 Ojibwe, 766–767 Ranks, introduced by Jespersen, 563 Old Irish, 787 Ratak dialect of Marshallese, 655 Papuan languages, 736 Rationality, substantialist view of, 417 personal Reader response, and intersubjective meaning, 670 analysis by Benveniste, 133 Reading Javanese, 562 definitions of, 893 Vietnamese, 1154 learning of, 894 pro-drop languages, 1068 what is being read, 893–894 as pro-form, 877 Reading impairment Sindhi, 967 distribution of, 895–896 Tagalog, 1076 nature of, 894–895 Tamil, 1078–1079 Reanalysis, in context of analogy, 47 Thai and Tai languages, 1100 Received Pronuciation, Standard English and, 405 Zuni, 1210 Reconciliation Process in Australia, 854 Pronunciation Reconstruction ancient Egyptian, 55–56 Indo-European, 510–512 changes across United States, 1143–1144 modern work of, 222–223 German, standardization, 388–389 Old Chinese, 772 spelling, 1174 phonological and internal, 462 discrepancies in English, 1193 Proto-Indo-European, 221 studied by Labov, 589–590 studied by Greenberg, 413–414 Proper nouns, 878–879 Rectification of names, Chinese debate on, 198 Pro phenomenon, 283–284, See also Pro-forms Recursion, by dolphins, 65 Proportional analogy, 47–48 Redundancy, distinctive feature theory in, 329–331 Prosody, 346, 836–837 Reduplication intonation, 880 Chinese (Mandarin), 196–197 quantity, 881 echo as form of, 797 stress, 880 Gothic strong verbs, 395 Proto-Afroasiatic, 18–19 number marking by, 761 Proto-Gothic, 394 Sotavento, 172 Proto-Indo-European Yoruboid languages, 1202–1203 kinship terms, 577 Reference, 896–898, 948 studied by Reflexive dependencies, 1070 Lehmann, 612–613 Register, 898–899 Wackernagel, 1163 Regressive assimilation, 97 Proto-Uto-Aztecan, 722–723 Regular metathesis, 679 Psycholinguistics, 352, 881–884 Reichenbach, Hans, 899–901 Psychosocial consequences, aphasia, 72 Relational grammar, 398 Pulmonic egressives, 10 Relational structure, studied by Jakobson, 549 Punctuation, Italian, 543 Relative clauses, 211 Purépecha, 1080 structures of, 1067 Tagalog, 1076 Q Tamil, 1080 Q-Celtic, 786 Relative tenses, Coptic Egyptian, 244t Qiangic languages, 192 Relativity, linguistic, 393, 927–928 Quantification, 885–887 Relativization, Tivoid languages, 1110 Quantifiers, 262–263, 886–887 Relay interpreting, 536–537 Quantity, as prosodic feature, 881 Relevance feedback, 532 Quantity iconicity, 487 Relevance in discourse, 901–903 Quantization, in soundwave analysis, 984–985 Relexification, Yiddish and, 1198 1234  Index Religious studies, importance of Aramaic, Sandhi, 919–920 80 San Juan Copala dialect of Trique, 1119 Renaissance, 465 Sanskrit Replacement test, 236 history of, 920–921 Representational matching procedure, 908 and Indo-European, 923 Repubblica italiana, 545–548 influence on Indic languages, 501–502 Restructuring, Gullah, 424 in light of Indo-European reconstruction, 510 Resultative secondary predication, 873 literary language of Hindu India, 508 Retroflex consonants, 859 morphology, 921, 923 Réunionnais, 506 phonology and script, 921 Reverse cutting, 199 sandhi rules, 919 Revitalization programs, 287 studied by Bopp, 148–149 Rhaeto-Roman, 519–520 syntax, 923 Rhetoric variety of, described by Panini, 807 contrastive, 270 Santee, See Dakota languages and linguistics, 903–906 Sapir, Edward, 924–926 Rhyme dictionaries, Chinese, 195–196 Sapir–Whorf hypothesis, 927–928, 1178 Rhyme of syllable, 1061–1062 Saramaccan, 928–930 Riau-Johor Malay, 646 Sardinian, 518, 546–547 Rich agreement, 21–22 Saussure, Ferdinand de, 222, 831, Risorgimento, 546 930–932 Rites of initiation, 318 Scaffolding, 9 Ritualized action, as prepragmatic capacity, 314 Scandinavia Romance languages, 359–360 Denmark (Danmark), 933 Catalan, 519 expansionism and retreat, 935–936 French, 519 Finland (Suomi), 933–934 Galician, 519 Iceland (Island), 934 Italian, 518 language contact types, 934–935 Moldavian/Romanian, 1002 Norway (Norge, Noreg), 933 Occitan, 519 Sweden (Sverige), 932–933 Portuguese, 518–519 Scandinavian languages, 1053–1056 Rhaeto-Roman, 519–520 Danish, 1053–1054 Romanian, 520 Norwegian, 1055–1056 Sardinian, 518 studied by Haugen, 444 Spanish, 518 Swedish, 1054 Roman Empire, Balkans, 116–117 Scerba, Lev, 830–831 Romanian, 520, 906–907 Scheduled languages of India, 501t Romansh, 547, 1060–1061 Schools, language in, and US immigrant languages, 1144–1147 Romany, spoken in Austria, 105 School years, language development in, 266 Root-mean-square amplitude, 981 Scope ambiguity, 40 Roper River Creole (Kriol), 585–587 Scotland Rules Scots, 406–407 vs. constraints, 907–910 Scottish English, 407 governing word order, 1187–1188 Scottish Gaelic, 406 phonological, 841–842 Scots, 406–407 sandhi, 919–920 Ulster, 407 Rural vs. urban divide in India, 504 Scottish English, 407 Rusian, 913–915 Scrambling, long-distance, 632–633 Rusian Church Slavonic, 913 Scripts, See also Writing systems Russell, Bertrand Arthur William, 827, 910–912 cuneiform, 825–826 Russenorsk trade pidgin, 202 Devanagari, 456, 921 Russian, 1001 Indic, 505 Chinese Pidgin Russian, 201–203 Japanese, 1193 and East Slavic languages, 912–917 Mongolian, 704 Russian Empire, 998–999 Tibetan, 1104 Rus’ warrior-traders, 913–915 Tifinagh, 1123–1124 Ryukyuan, 37, 552–553 Search terms, in information retrieval, 531–532 Searle, John R., 936–938 S Secondary predication, 873 Saamáka, See Saramaccan Second language Saami languages, 341 learning of, 73, 940–943 Salishan languages, Okanagan and, 767–771 teaching of, 73, 943–945 Samoyedic languages, 1004 what it means in second language acquisition, 4–5 Sampling Second language acquisition, 938–940 in soundwave analysis, 984 factors in, 6–7 typological, 1134 identity and, 490 1235 Index Second language acquisition (continued) Shijing, 195 knowledge vs. skill, 939 Shqipe, See Albanian meaning of acquisition in, 5 Sibe language, 1128 meaning of second language in, 4–5 Sign languages research, learner language in, 5–6 American Sign Language, 41–43, 63, 1146–1147 Sefardí, 1009 British Sign Language, 155–156 Segmentation of speech stream, 1018 distribution, 962 Semantic bleaching, 403 French Sign Language, 41–42 Semantic bootstrapping, 8 history of, 961–962 Semantic change, 599–600 Indo-Pakistani, 524–529 Semantic feature hypothesis (Clark), 208 Japanese, 555–558 Semantic meaning, 668–669 neural systems for language, 965 Semantic nets, 227–229 universals and modality effects, 962–965 Semantic opaqueness of idioms, 494–495 Sign relation, studied by Peirce, 819–820 Semantic parsing, 816 Sign relationships, 960–961 Semantic representations, 227–228 Sign systems, 141–142 Semantics, 947–949 Silte, EES language, 44 compositional, 224 Similarity/contiguity, studied by Jakobson, 550 forensic, 350 Simile, 336 generative (Lakoff), 593 Simplification, historical linguistics study of, 462 studied by Simulative neurolinguistics, 733 Jakobson, 550 Simultaneous constructions, in signed languages, 963–964 Lyons, 637 Simultaneous interpreting, 535 Weinreich, 1170–1171 Sindhi Semantic typology, 945–947 adjectives, nouns, and pronouns, 966–967 Semiotics, 669, 949–952 historical development and sociopolitical factors, 969–970 studied by Peirce, 819 morphophonology, 966 Semiotic system (Halliday), 437–438 and related Indo-Arayan languages, 965–966 Semitic languages sound inventory, 966 Akkadian, 955 syntax, 968–969 Arabic, 74–77 verbs, 968 Aramaic, 79–81 Sine waves, 980 classification, 953–954 Singapore official languages, 1037 EES, 44–46 Sinitic languages, 189, 194–195, 971–972 general characteristics, 952 linguistic tree, 191f Hebrew, 539 Sino-Tibetan languages, 970–973 semitic writing, 955 Sintu, See Bantu South Arabian languages, 955 Sinusoids, 1011–1012 studied by Harris, 441 Siouan-Catawban languages, 253–255 in Turkey, 1132–1133 Sirenikski, 298 Semito-Hamitic, See Afroasiatic Sisaali, 711 Seneca, 700 Situational irony, 538 Sense, in context of semantics, 948–949 Slave trade, Cape Verde role in, 172 SENSEVAL evaluation, 1188–1189 Slavic languages, See also West Slavic languages Sentence comprehension, 310, 883 affiliation for Yiddish, 1198–1199 Sentence structure borrowings from non-Slavic languages, 522 Amharic, 46 East Slavic languages, 912–917 Burushaski, 166–167 influence on Romanian, 906 IPSL, 525–526 Old Church Slavonic, 774–776 JSL, 557 related to Baltic languages, 522 Middle English, 690 sound structures, 522–523 naming and, 725 South Slavic languages, 956–958 Old French, 781–782 verbs, 523 relationship to syntactic category, 1065 word structure, 523 Warao language, 1166 writing, 521–522 Yoruboid languages, 1203–1204 Slav invasions of Balkans, 117 Sentential coordination, 240–241 Slobin, Dan Isaac, 973–975 Sephardim, 540 Slovak, 864–865 Serbo-Croatian, 547 Slovene, 957–958 and South Slavic languages, 956–958 Slovenian, 104–105 studied by Leskien, 614 spoken in Italy, 547 Serial verb construction, 959–960 Social context, meaning and ideology and, 669–670 Sexist language, 367–368, 491 Social deixis, 261 Sey, 507 Social development, first language acquisition and, 4 Shabo, 751 Social factors, affecting second language acquisition, 6 Shift-induced interference, 848 Socialization, language, 606–607 1236  Index Social preadaptations, 317–319 Source-oriented translation, 1121 Social situation of BSL, 156 South Africa Sociolect English language role in, 987 diglossia, 975–976 Fanakalo, 323–324 education, 976–977 Khoisan and Niger-Kordofanian languages, 986 minority/majority group relations, 976 language distribution in, 986–987 stratification, 976 Nguni, 986 Sociolinguistics official languages of, 987 approach to identity and language, 489 population mix, 985 concept of domain, 978–979 South America contemporary, 977 Argentina, 988–990 dialectology, 978 Brazil, 990–993 education and, 979–980 South Arabian languages, 955 inspired by multi- and bilingualism, 978 South Caucasian languages, 1004 interactional, 239, 979 Southeast Asia, 993–995 JSL, 555–556 Southern Bantu languages, 1206–1209 questioning what language is, 979 Southern Mindanao Creole, 822 Wolof, 1183 Soviet Union, 995–998 Sociopolitical role of mass media, 659 Armenian, 1002 Sociopolitical status, Warao language, 1164–1165 Baltic languages, 1002 Sociopragmatics, 870 Caucasian family languages, 1004 Soddo, EES language, 44 East Slavic languages, 1001–1002 Songhay cluster, Nilo-Saharan languages, Finno-Ugric languages, 1003–1004 750–751 Iranian languages, 1002 Sorbian language and ethnicity, 996 spoken in Lusatia, 390 language families, 1000–1001 Upper and Lower, 865 linguistic diversity, 996 Sotavento Manchu-Tungus languages, 1003 Cape Verde, 172 Mongolian languages, 1003 distant from Portuguese, 251 paleoasiatic languages, 1004 Sound change post-Soviet era, 1000 diffusion, 598 romance languages, 1002 interaction with analogical change, 47 Russian Empire, 998–999 Sound inventory Samoyedic languages, 1004 Sindhi, 966 Soviet language policy, 996–1000 Tai languages, 1099 Turkic languages, 1003 Sound pattern Spain Japanese, 552 historical (extinct) languages, 1005 Yoruboid languages, 1200–1202 introduction of Latin, 1005–1006 Sound structure modern languages, 1006–1007 interaction with affixation, 530 Spanish Tocharian, 1113 Castilian, 1008–1009 Sound symbolism, 796 class marking, 371 Sound system Mexican, 682–683 Georgian, 381 as romance language, 518 Irish, 183 Spanish-based creoles Italian, 543 in Philippines, 822–824 Middle English, 689–690 in Southeast Asia, 995 Oto-Manguean, 799 Speaking, ethnography of (Hymes), 484 Slavic languages, 522–523 SPEAKING mnemonic, 304, 979 Standard Malay-Indonesian, 645 Specific Language Impairment (SLI), 213–214 Soundwave analysis Spectral analysis, 981, 1010–1016 autocorrelation, 985 acoustic signal, 1011 cepstral analysis, 985 cepstrally and LPC smoothed spectra, 1015–1016 digitalization, sampling, and quantization, formants, 1014 984–985 periodic waveforms and Fourier analysis, 1012–1014 filtering, 983 sinusoids, 1011–1012 linear prediction, 981–983 source–filter theory of speech production, 1014 oscillogram, 980–981 spectrograms, 1014–1015 root-mean-square amplitude, 981 Spectrograms, 1014–1015 sine waves, 980 Spectrographic analyses, 983–984 spectral analysis, 981 Spectrum of sound, 1–2 spectrographic anayses, 983–984 Speech act Sound waves, 1 approaches to irony, 537 Source–filter theory, 1–2 constative language, 1016–1017 of speech production, 1014 illocutionary force, 1017 1237 Index Speech act (continued) State languages, See also Official languages performatives, 1017 Croatian, Serbian, and Bosnian, 957 studied by Searle, 936–938 India, 501t theory of, 870 linguistic recognition of, 979 Speech community, 303 post-Soviet era, 1000 Kiowa-Tanoan languages, 1092–1093 Switzerland, 1059–1061 Speech decorum, Javanese, 559 Status planning goals, 605 Speech motor control, and physiology, 837–838 Stoic philosophical school, 308, 464 Speech patterns, United States, 1144 Stoney, 254 Speech perception, 838–839, 1018–1021 Stops, 649 Speech processing Mayan languages, 665 back end, 1022 Stored speech, 1028 dialogue management, 1022 Story grammar, 1038–1039 evaluation, 1022–1023 Storytelling neurobiology, 1023–1024 Kiowa-Tanoan languages, 1093 speech recognition, 1021–1022 Saramaccan, 930 text-to-speech generation, 1022 Sranan, 1033 Speech production, 1024–1026, See also Articulatory system Strategies anatomy learner, 6–7 neurobiology, 1026–1028 perceptual (Bever), 135 preadaptation and, 315 Stratification, sociolects and, 976 source–filter theory of, 1014 Stress Speech recognition, 1021–1022 Hindi-Urdu, 456–457 Speech registers, elderspeak, 20 as prosodic feature, 880 Speech sounds, 1–2 Zuni, 1209 and acoustic signal, 838–839 Structural ambiguity, 40 alternating, studied by Boas, 145 Structural interference, 1039–1040 antalphabetic system for describing, 562–563 Structuralism basic phonetic classificatory principles, 833–835 and deconstruction, 670 transcribed broadly or narrowly, 832–833 developed by Lévi-Strauss, 617–618 Speech styles, assimilation in, 97 functionalist, studied by Martinet, 656–657 Speech synthesis studied by application of technique, 1029 Bloomfield, 1041–1042 concatenative synthesis, 1029 Chao, 188 parametric coding, 1028–1029 Jakobson, 549 stored speech, 1028 Sturtevant’s Paradox, 47 Spelling pronunciation Stylistics, 1043–1045 discrepancies in English, 1193 forensic, 351 Western Caribbean creoles, 1174 Stylistic statistics, 662 Spelling reform, in Germany, 388 Subgrouping Split agreement, 24 internal, Nilo-Saharan languages, 751–752 Sporadic metathesis, 679 Pama-Nyungan, 806 Sprachbund, 1029–1030 Subordinate clause, 210–211 Balkan, 117–118 Substantialist rationality, 417 described, 268 Substitution test, 236 Sprachtheorie (Bühler), 158–159 Suffixation, 12–13 Sranan, 928–930 Cape Verde Creole, 173 lexicon, 1032 Igbo and Igboid languages, 498 morphology, 1032 Old Japanese, 790 phonology, 1032 Suffixes as product of language contact, 1031 inflectional and derivational, 529–530 syntax, 1032–1033 Italian, 545 verbal arts, 1033 pronominal, Sumerian, 1049 Sri Lanka Portuguese, 1033–1036 Sumerian, 113–114, 1045–1051 Standard Eastern Armenian, Lord’s Prayer in, 88 genderlect or genrelect, 1050 Standard English morphosyntax, 1050 American, 1037 nominal morphology, 1046–1047 and Received Pronunciation, 405 phonology and writing interface, 1046 Standard language, 1036–1037 verbal system, 1047–1049 Standard Malay, 1036–1037 Summarization, 1097 Standard Malay-Indonesian single- and multi-document, 533 history, 645 Summer Institute of Linguistics (SIL), 851–852 typical features, 645–646 Sumo, 694–695 writing systems, 646 Suppletion, number marking by, 761 Standard Western Armenian, Lord’s Prayer in, Supralaryngeal voice quality settings, 810–811 88 Suprasegmentals, 842 1238  Index Surface realization, component of generation, T 372 Taboo words, 1073–1074 Swadesh, Morris, 1051–1053 Tagalog, 1075–1077 Swahili, class marking, 370–371 Tagging, 815–816 Sweden (Sverige), 932–933 Tagmemics, 851 Swedish, 513, 935, 1054 Tahgan, 990 Sweet, Henry, 1056–1058 Tahitian, 856–857 Switch reference Tai-Kadai, 192 markers, 1058 Tai language family, 994, 1098–1101 other reference tracking systems, 1059 Taiwan, Austronesian, 193 Papuan languages, 736 Tajik, 1002 typical systems of, 1058 Takic branch of Uto-Aztecan, 721 uncommon variants in, 1059 Talk, context-bound and -free, 239 variations in, 1058–1059 Talking drum, Yoruba, 1201–1202 Switzerland, 1059–1061 Tamazight language, Berber, 35–36 Syllabic writing system, 1192 Tamil, 1077–1080 Syllable structure, 1062–1063 Tangkic languages, 569–571 Ewe and Gbe languages, 320 Tano languages, 29 Hmong-Mien languages, 473 Taps, 650 Igbo and Igboid languages, 497 Taracahitic branch of Uto-Aztecan, 722 Late Common Slavic, 775 Tarascan, 1080–1082 Syllable-tone languages, 1117 Target-oriented translation, 1121 Synchronic metathesis, 679 Tay Boy, 995 Synchronous communication, 230 Taymyr Pidgin Russian, 201–203 Syncope, 294–295 Tayo, 803 Synecdoche, 335 Teachers of English to Speakers of other Languages Synonymy, 624 (TESOL), 5 Syntactic analogy, 48 Teaching Syntactic anaphor, 49 foreign language, 144, 1044–1045 Syntactic bootstrapping, 8 methods of, 1084–1086 Syntactic categories, 1063–1066 second language, 73, 943–945 Syntactic change, 599 Teaching curricula Syntacticians, 874 communicative approach, 1083 Syntactic statistics, 661 needs analysis, 1084 Syntactic typology, 1066–1068 negotiated syllabus, 1083–1084 Syntax oral approach, 1083 ancient Greek, 61–62 Teeth baring, 314 in animal communication, 63 Tekhne Grammatike, 308 Biblical Hebrew, 452 Telecommunication, 220 Carib and Cariban languages, 175–176 Telugu, 1086–1088 Chinese (Mandarin), 197–198 Tense Coptic Egyptian, 244 aspect marking and, 1088–1089 Ewe and Gbe languages, 320–321 syntax, 1089–1091 fundamental properties of languages, time and, 1106–1107 1068–1071 Warao language, 1165–1166 Hindi-Urdu, 458 Tense-mood-aspect (TMA) system, 112 Hittite, 469 Tense system Igbo and Igboid languages, 497–498 absolute, Coptic Egyptian, 243t Malagasy, 643 English, 291–292 Nootka, 755 Hmong-Mien languages, 474 Old Japanese, 790 Sanskrit, 921, 923 Sanskrit, 923 studied by Reichenbach, 900 Saramaccan, 930 Tepiman, 721–722 Sindhi, 968–969 Teton, See Lakota Sranan, 1032–1033 Tewa, and Kiowa-Tanoan languages, standard theory of (Chomsky), 206 1091–1094 studied by Jespersen, 563 Text linguistics, 1094–1095 tense, 1089–1091 Text mining, 533 Tivoid languages, 1109–1110 Text processing, aging effects, 20 Vietnamese, 1154 Texts work by Brugmann, 157 Old Irish, 786 Yoruboid languages, 1203–1204 Sumerian, word lists, 114 Synthetic compounds, 225 Tocharian, 1111 Synthetic languages, 712–714 Udmurt, 1137 Syriac, 80–81 Text-to-speech generation, 1022 Systemic grammar, 364–365 Textual criticism, 825 1239 Index Text understanding, 1096–1098 in legal context, 352 Thai languages, 994, 1098–1101 machine, 124–125, 226, 639–640 studied by Haas, 429 studies in, 1120–1121 Thematic structure, 1101–1103 work by Nida, 738 Theoretical principles of analogical change, 48 Trills Theory of agreement, 25 uvular, 649–650 Theta Theory, 872–873 voicing and, 11 Tibetan, 1103–1106 Trubetzkoy, Nikolai Sergeyevich, 1122–1123 Tibeto-Burman languages, 192, 499, 970–972 Truncation, Yoruboid languages, 1203 Tifinagh writing system, 1123–1124 Tsakonian, 411 Tigre, EES language, 44 Tsouic languages, 193 Tigrinya, EES language, 44–45 Tuareg, 1123–1124 Time Tübatulabal branch of Uto-Aztecan, 721 deixis, 261 Tungusic languages, 38, 1126–1129 and tense, 1106–1107 spoken in China, 193 in Warao language, 1165–1166 Tunisia, Berber dialects, 1125 Time lines, in signed languages, 964 Tupi-Guarani languages, 463, 988–989, 1167–1170 Time reference, in AAVE, 15 Tupinambá, 1169 Tiv and Tivoid languages, 1107–1111 Turing, Alan, 1129–1131 Tlapanecan-Manguean, 798–799 Turing test, 89 Tocharian, 511, 1111–1114 Turkey Tofa, vowel harmony, 1160 culture, Balkans influenced by, 117 Tok Pisin, 460–461, 803, 1114–1116 Indo-European languages in, 1131–1132 Tone Semitic languages in, 1132–1133 lexical, Hmong-Mien languages, 473 Turkic languages, 38–39 terracing, Akan languages, 30 Azeri, 1003 Yoruboid languages, 1201 Kazakh, 1003 Tone languages Kyrgyz, 1003 of Africa, 1118–1119 spoken in China, 193 of Asia, 1118 Turkmen, 1003 Ewe and Gbe, 320 Uzbek, 1003 Igbo and Igboid, 497 vowel harmony, 1159 indigenous to Americas, 1119 Turkish, 1131 Nilo-Saharan, 751–752 as agglutinating language, 712–714 pitch-accent, 1117 influence on Romanian, 906–907 syllable-tone, 1117 words, in modern Greek, 412 word-tone, 1117 Turkmen, 1003 Tongan, 652 Turn-taking, 871 Tongue Tuyuca, evidential markers, 212 anatomy of, 834 Twi dialects (Asante–Akyem–Kwawu), 29 role in speech production, 52 Two-audience dilemma, in courtroom discourse, Tongue root harmony, 1159 248 Top-down parsing, 813 Two-step approach to irony, 537–538 Topicalization, 236 Typological linguistics, 231–232 Topic-comment structure of ASL, 42 Typological statistics, 662 Tosk dialect, 31 Typology Totemism, studied by Lévi-Strauss, 619 Dravidian, 273 Trace, in empty categories, 284 morphological, 713–715, 1135 Traditional grammar sampling, 1134 Arabic, 78–79 semantic and discourse, 945–947 Babylonian, 113–114 Sino-Tibetan languages, 971 Chinese, 198–200 studied by Greenberg, 414 clause in, 210–211 syntactic, 1066–1068 European, 307–309 universals, 1133–1136 history of, 395–396 Warao language, 1165 Indian, 507–509 Wayampi, 1168–1169 Japanese, 198–200 word-order, 1134–1135 Traditional literature, Maori, 653 Transcription U phonetic, 831–833 Udi (Caucasian), pronouns and agreement, 24 techniques in the field, 334 Udmurt, 1137–1139 Transfer, by bilinguals, 601 Ugric languages, Hungarian and, 480–483 Transformational generative grammar (Chomsky), 205–206 Ukrainian, 915–916, 1001–1002 Translation Ulster Scots, 407 of idioms, 495 Underspecified representations, 727 indeterminacy of, 890 Unification, in computational linguistics, 228–229 1240  Index United States Verb stems American English, history of, 1140–1144 Semitic languages, 953t immigrant languages, and languages in schools, Tivoid languages, 1109 1144–1147 Verb system immigrant populations, 300–302 Albanian, 32 indigenous languages, 1139–1140 Cape Verde Creole, 172–173 Universal grammar, 8, 908, 940–943, 1147–1149 Chadic languages, 446 Universals Classical Egyptian, 57 color terms and, 218–219 Farsi, 326 cultural, politeness as, 867 German, 386 explanations for, 1135–1136 Hmong-Mien languages, 474 implicational, studied by Greenberg, 414 Italian, 544–545 of negative position, 1067 Japanese, 554 signed languages, 962–965 Khoisan languages, 575 typology, 1133–1134 Latin, 611 Upper German dialects, 783–784 Mohawk and Iroquoian languages, 700–701 Upper Guinea Crioulo, 250–252 Ojibwe, 765–766 Upper Sorbian, 865 Old French, 781 Uralic group Old Irish, 787 Finno-Ugric languages, 1003–1004 Old Tibetan, 795 Samoyedic languages, 1004 Sindhi, 968 Ussuri Pidgin Russian, 201–203 Slavic languages, 523 Uto-Aztecan languages, Nahuatl and, 721–724 Sri Lanka Portuguese, 1035 Utterance-centered linguistics, 1149–1150 Sumerian, 1047–1049 Utterance-level prosody, 837 Tagalog, 1076 Uvular consonants, 859 Tamil, 1079 Uzbek, 1003 Verner’s Law, 511 Viet-Muong, 107–108 V Vietnamese, 573, 713, 1152–1155 Vagueness, ambiguity distinguished from, 40 Viking period, 935 Vai Nakh languages, 382 Visigothic, 394 Valencian, 1006 Visual cue structure, JSL, 556–557 Vanuatu, Bislama language, 1115–1116 Visual word recognition, 1155–1157 Variation, studied by Labov, 1151–1152 Visuospatial sketch pad, working memory, 1190 Variationist sociolinguistics, 489, 589–591 Vocabulary Varro, 308, 464 Ainu, 26–27 Vector space model, 532 American English, 1141–1142 Vedic Sanskrit, 923 animal communication and human language, 63 Vegetative sign systems, 142 basic, 1051 Velaric ingressive (clicks), 11, 650 Basque, 127 Velum (soft palate), role in speech production, 52–53 cohesion driven by, 216–217 Vendler classes, 94 Hawaiian Creole English, 449 Veps, 343–344 Italian, 543–544 Verb agreement, in signed languages, 964 Korean, 582–583–584 Verbal arts Middle English, 690–691 Saramaccan, 930 Old Church Slavonic, 776 Sranan, 1033 Old English, 779–780 Verbal chain, Sumerian, 1048t Old French, 782 Verbal classification, Murrinh-Patha and Daly languages, Old Irish, 786 718–719 Old Tibetan, 794 Verbal phrase, Igbo and Igboid languages, 498 Tamil, 1078 Verbal syntax, English, 291–292 Vocal folds Verb complex, Navajo and Athabaskan-Eyak language, 729–730 photographed apart, 834f Verb endings, Mongolian, 704 role in speech production, 51 Verb morphology Vocal tract ancient Greek, 60–61 aerodynamics of, 10–11 ASL, 42 cross-sectional view, 834f Burushaski, 165–166 supralaryngeal, 315–317 Coptic Egyptian, 243 Voice Guaymí and Chibchan languages, 422 middle, 1158 Hindi-Urdu, 457–458 passive construction, 1157–1158 Middle (Classical) Japanese, 687 Voice quality, 810–811 Modern Hebrew, 453 Voicing Old Japanese, 790 assimilation and, 95, 97 Verbosity, off-target, 20 speech sounds and, 841 Verb serialization, 959–960 and trills and obstruents, 11 1241 Index Volapük, 91 West Oto-Manguean, 798 Volta–Comoé, See Tano languages West Slavic languages, 862–866 Volta–Niger languages, 122 Czech and Slovak, 864–865 Von Baer’s law, 310–311 Polabian, 865–866 Votyak (Udmurt), 1137–1139 Polish and Cassubian, 863–864 Vowel harmony Sorbian (Upper and Lower), 865 Akan languages, 30 Whistled speech, 799 Altaic languages, 38–39 Whorf, Benjamin Lee, 1177–1179 Finnic and Finno-Ugrian languages, 342 Whorfian hypothesis, 927 Mongolian, 703 Williams syndrome, 214 Old Norse, 793 Winnebago, 254 as pattern, 1160–1161 Witnesses, lay, 247–248 relation to speaking mechanics, 1160 Wittgenstein, Ludwig, 1179–1181 Sumerian, 1046 Wolof, and Atlantic languages, 1181–1183 Telugu, 1087 Word Tungusic, 1128 collocations, 1185–1186 Turkic languages, 1159 linguistic notion of, 1184 Vowels role in mental processes underlying language use, 1184–1185 Burushaski, 164 structural and mental significance, 1185 coarticulation, 96 Word-class endings, in Esperanto, 92 Dravidian, 273 Word classes epenthesis, 294–295 IPSL, 524–525 formant pattern, 2 Okanagan and Salishan languages, 770 Gbe, 321t open and closed, 1185 Korean, 581t Standard Malay-Indonesian, 646 Late Common Slavic, 774–775 Warao language, 1165 Old Japanese, 789 Word formation physiological production of, 835 in acquisition, studied by Clark, 208–209 English, 293 W Hmong-Mien languages, 473–474 Wackernagel, Jacob, 1163–1164 Middle English, 690–691 Wada test (intracarotid sodium amytal test), 733 Ojibwe, 766 Wakashan languages, Nootka and, 753–756 Old English, 779–780 Wales, Welsh and Welsh English, 405–406 Standard Malay-Indonesian, 645 Walloon, 130–132 Word lists, Sumerian texts, 114 Wan Nan, 192 Word order Warao language Atlantic languages, 1182 dialects, 1165 Burushaski, 166 discourse markers, 1166 Farsi, 326–327 future of, 1167 flexible, Wayampi, 1169 genealogy, 1165 Guaymí and Chibchan languages, 420–421 person, number, and plurality, 1166 Igbo and Igboid languages, 498 possession and article, 1166–1167 importance in English, 1186–1187 sentence structure, 1166 Modern Hebrew, 454–455 social and political status of, 1164–1165 Navajo and Athabaskan-Eyak language, 730 tense, mood, and time, 1165–1166 Old Church Slavonic, 776 typology, 1165 phrase-internal, 1187 word classes, 1165 rules governing, 1187–1188 Warlpiri, 1058 Sino-Tibetan languages, 971–972 Waveforms Standard Malay-Indonesian, 646 periodic, and Fourier analysis, 1012–1014 Tibetan, 1105 vowels and consonants presented as, 350 typology, 1134–1135 Wave model of language relation, 378 Yoruboid languages, 1203–1204 Wayampi, 1168–1169 Word recognition, visual, 882–883 Weinreich, Uriel, 1170–1172 Word sense disambiguation, 1188–1189 Welsh, 183 Word-stress, as part of word-level prosody, 836–837 and Welsh English, 405–406 Word structure Wernicke–Geschwind model, 630–631 JSL, 557 Wernicke’s aphasia, 71, 1027 Latin, 610 Wernicke’s area, 630–631 Middle English, 690 West African Pidgin English, 585 Old French, 781 Western Caribbean creoles, 1172–1177 Yoruboid languages, 1202–1203 West Germanic, 514 Word-tone languages, 1117 West Indies Working memory four distinct groups in, 176 central executive component, 1190 historical partition of, 177 and language processing, 1191 1242  Index in older adults, 20 X phonological loop, 1190 X bar theory, 844–847 visuospatial sketch pad, 1190 Xiang, 191 Writing systems, See also Scripts alphabetic, 1192 Y alphasyllabic, 1192 Yankton, See Nakota Amharic, 45 Yémba, 1195–1196 ancient Egyptian, 55–56 Yi-Burmese, 192 Czech, 1192–1193 Yiddish, 515, 540–542, 1197–1200 discrepancies between spelling and pronunciation, 1193 Yoruba, 746 dyslexia and, 277 and Yoruboid languages, 1200–1204 Hindi, 456 Yoruboid languages Hmong-Mien languages, 475–476 morphology, 1202–1203 Khmer and Mon, 573 phonology, 1200–1202 Korean, 580 politeness, 1204 logographic, 1192 syntax, 1203–1204 Middle (Classical) Japanese, 685 Yoruboid languagesYuezhi tribes, morphophonemic principle, 1193–1194 1112 Old Chinese, 771–772 Yukar (epic poems), Ainu, 26 Old Japanese, 788–789 Yupik Old Tibetan, 794–795 distribution and characteristics of, 298 Semitic, 955 split from Inuit, 296 Sino-Tibetan languages, 971 Southeast Asia, 995 Z Standard Malay-Indonesian, 646 Zamboangueño, 822–823 syllabic, 1192 Zapoteco, 684 Tai languages, 1099 Zay, EES language, 44 Tifinagh, of Tuareg, 1123–1124 Zero derivation, 293 Vietnamese, 1154 Zeugma, 336 Written language history, Russian, 915 Zeytun dialect of Armenian, 88 Written sources Zhongyuan Yinyun, 195–196 ancient Greek, 59 Zipf, George Kingsley, 1205–1206 Aramaic, 80 Zulu Middle (Classical) Japanese, 685 Fanakalo pidgin with, 323–324 Old High German, 785 and Southern Bantu languages, Old Japanese, 788 1206–1209 Russian, 913 Zuni language Tungusic language, 1128 phonetic properties, 1209–1210 Wu, 191 syntactic properties, 1210 1243