Adaptive Reuse: Aspects of Creativity in South Asian Cultural History

 ABHANDLUNGEN FÜR DIE KUNDE DES MORGENLANDES Im Auftrag der Deutschen Morgenländischen Gesellschaft herausgegeben von Florian C. Reiter Band 101 Board of Advisers: Christian Bauer (Berlin) Desmond Durkin-Meisterernst (Berlin) Lutz Edzard (Oslo/Erlangen) Jürgen Hanneder (Marburg) Herrmann Jungraithmayr (Marburg) Karénina Kollmar-Paulenz (Bern) Jens Peter Laut (Göttingen) Joachim Friedrich Quack (Heidelberg) Florian C. Reiter (Berlin) Michael Streck (Leipzig) 2017 Harrassowitz Verlag . Wiesbaden  Adaptive Reuse Aspects of Creativity in South Asian Cultural History Edited by Elisa Freschi and Philipp A. Maas 2017 Harrassowitz Verlag . Wiesbaden Published with the support of Austrian Science Fund (FWF): PUB 403-G24 Open access: Wo nicht anders festgehalten, ist diese Publikation lizensiert unter der Creative-commons-Lizenz Namensnennung 4.0. Open access: Except where otherwise noted, this work is licensed under a Creative commons Atribution 4.0 Unported License. To view a copy of this license, visit http:// creativecommons.org/licenses/by/4.0/. Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.dnb.de abrufbar. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the internet at http://dnb.dnb.de. For further information about our publishing program consult our website http://www.harrassowitz-verlag.de © Deutsche Morgenländische Gesellschaft 2017 Printed on permanent/durable paper. Printing and binding: Hubert & Co., Göttingen Printed in Germany ISSN 0567-4980 ISBN 978-3-447-10707-5 e-ISBN PDF 978-3-447-19586-7  Contents Elisa Freschi and Philipp A. Maas Introduction: Conceptual Reflections on Adaptive Reuse ............................. 11 1 The dialectics of originality and reuse................................................ 11 2 The background .................................................................................. 12 3 Some basic conceptual tools .............................................................. 13 3.1 Simple re-use versus different grades of adaptive reuse ....... 13 4 Adaptive reuse: Aspects of creativity ................................................. 17 5 “Adaptive reuse” and related terms .................................................... 20 5.1 Adaptive reuse, intertextuality and adaptation studies .......... 20 6 On the present volume........................................................................ 21 References ................................................................................................ 24 Section 1: Adaptive Reuse of Indian Philosophy and Other Systems of Knowledge Philipp A. Maas From Theory to Poetry: The Reuse of Patañjali’s Yogaśāstra in Māgha’s Śiśupālavadha................................................................................. 29 1 The Pātañjalayogaśāstra ................................................................... 30 2 Māgha’s Śiśupālavadha ..................................................................... 31 3 The Śiśupālavadha and Sāṅkhya Yoga in academic research ............ 34 4 Pātañjala Yoga in the Śiśupālavadha ................................................. 36 4.1 The stanza Śiśupālavadha 4.55 ............................................. 36 4.1.1 The reuse of the Pātañjalayogaśāstra in Śiśupālavadha 4.55 ............................................................... 37 4.1.2 Śiśupālavadha 4.55 in context .............................................. 41 4.2 The stanza Śiśupālavadha 14.62 ........................................... 46 4.2.1 The reuse of the Pātañjalayogaśāstra in Śiśupālavadha 14.62 ............................................................. 47 4.2.2 Śiśupālavadha 14.62 in context ............................................ 49 4.3 The passage Śiśupālavadha 1.31–33 .................................... 51 4.4 The reception of Māgha’s reuse in Vallabhadeva’s Antidote ....................................................... 53 5 Conclusions ........................................................................................ 55 6 Contens References................................................................................................ 57 Himal Trikha Creativity within Limits: Different Usages of a Single Argument from Dharmakīrti’s Vādanyāya in Vidyānandin’s Works ...................................... 63 1 A passage from the Vādanyāya and an overview of corresponding textual material ................................................................................... 65 1.1 The background of the argument .......................................... 66 1.2 Overview of corresponding passages.................................... 68 1.3 Groups of correlating elements ............................................. 71 2 The succession of transmission for the adaptions in Vācaspati’s and Aśoka’s works.......................................................... 73 2.1 Basic types of the succession of transmission ...................... 73 2.2 The adaption in the Nyāyavārttikatātparyaṭīkā .................... 75 2.3 The adaption in the Sāmānyadūṣaṇa .................................... 79 3 Vidyānandin’s use of the argument .................................................... 82 3.1 The adaptions in the Tattvārthaślokavārttikālaṅkāra ........... 82 3.2 The adaptions in the Aṣṭasahasrī .......................................... 87 3.3 The adaptions in the Satyaśāsanaparīkṣā ............................. 95 4 Conclusion........................................................................................ 101 References.............................................................................................. 102 Ivan Andrijanić Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra ................ 109 1 Introduction ...................................................................................... 109 2 Material marked by Śaṅkara or by sub-commentators as being reused from other authors ................................................................. 113 2.1 Indefinite pronouns as markers of reuse ............................. 113 2.2 Identifications of reuse by the sub-commentators .............. 115 2.2.1 Reuse of the views of the Vṛttikāra..................................... 116 3 Different interpretations of the same sūtras ..................................... 118 4 Examples of reuse ............................................................................ 119 4.1 The case of ānandamaya in Brahmasūtrabhāṣya 1.1.12–1.1.19 ..................................... 119 4.1.1 The introduction of the adhikaraṇa .................................... 120 4.1.2 Brahmasūtrabhāṣya 1.1.12 ................................................. 121 4.1.3 Brahmasūtrabhāṣya 1.1.13–17 ........................................... 121 4.1.4 Brahmasūtrabhāṣya 1.1.17 ................................................. 122 4.1.5 Brahmasūtrabhāṣya 1.1.19 ................................................. 123 4.2 The “bridge” (setu) from BS(Bh) 1.3.1 and MU(Bh) 2.2.5 ..................................................................... 126  Contens 7 5 Conclusions and outlook for further research................................... 129 References .............................................................................................. 130 Yasutaka Muroya On Parallel Passages in the Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara ......................................................................................... 135 1 Bhaṭṭa Vāgīśvara’s Nyāyasūtratātparyadīpikā................................. 136 2 Parallel passages in the Nyāyasūtratātparyadīpikā, the Nyāyabhāṣya and the Nyāyavārttika ................................................ 138 3 Parallel passages in the Nyāyasūtratātparyadīpikā and the Nyāyavārttikatātpāryaṭīkā ................................................................ 138 3.1 Vāgīśvara and Vācaspati on Nyāyasūtra 1.1.1 ................... 139 3.1.1 Udayana’s theory of categories........................................... 142 3.2 Vāgīśvara and Vācaspati on Nyāyasūtra *5.2.15(16) ......... 143 3.2.1 Dharmakīrti’s discussion of ananubhāṣaṇa ........................ 145 3.2.2 Vāgīśvara’s and Vācaspati’s references to Dharmakīrti ..... 147 4 On the relative chronology of Vāgīśvara and Vācaspati .................. 148 References .............................................................................................. 150 Malhar Kulkarni Adaptive Reuse of the Descriptive Technique of Pāṇini in Non-Pāṇinian Grammatical Traditions with Special Reference to the Derivation of the Declension of the 1st and 2nd Person Pronouns ............................................ 155 References .............................................................................................. 166 Section 2: Adaptive Reuse of Tropes Elena Mucciarelli The Steadiness of a Non-steady Place: Re-adaptations of the Imagery of the Chariot................................................................................. 169 Premise .................................................................................................. 169 1 The Ṛgvedic ratha: The chariot as a living prismatic metaphor ...... 171 1.1 ratha and swiftness ............................................................. 171 1.1.1 ratha as a means for crossing fields .................................... 173 1.2 The godly character of the ratha ......................................... 173 1.3 ratha and conquest .............................................................. 174 1.4 ratha in the ritual context.................................................... 174 1.5 ratha and poetry.................................................................. 175 1.6 ratha and generative power ................................................ 176 1.7 Summing up: The many semantic values of the ratha in the Ṛgveda Saṃhitā ............................................... 178 8 Contens 1.8 The medieval adaptive reuse of the ratha compared to its Vedic use ................................................... 178 2 The linear re-use of the ratha in the middle Vedic period: The symbolic chariot ............................................................................... 179 2.1 The socio-political context of the re-use ............................. 179 2.2 The chariot in the middle Vedic sacrifices ......................... 180 2.2.1 The chariot in non-royal sacrifices ..................................... 181 2.2.2 The chariot in the royal sacrifices ....................................... 182 2.2.3 The chariot and the evocation of fertility ............................ 187 2.3 Shrinking of meanings in middle Vedic reuse .................... 188 3 Conclusion ....................................................................................... 188 References.............................................................................................. 189 Cristina Bignami Chariot Festivals: The Reuse of the Chariot as Space in Movement ........... 195 1 Introduction ...................................................................................... 195 2 The origins of chariot processions in the Vedic period .................... 197 3 Faxian’s record of chariot festivals ................................................. 198 4 A record of the chariot festival in the southern kingdom ................. 200 5 The modern ritual of rathotsava at the Cennakeśava Temple of Belur, Karnataka ....................................... 202 6 The modern ritual of rathayātrā at Puri, Orissa ............................... 204 7 Applying the concept of reuse: The chariot in the diaspora ............. 205 8 Conclusions ...................................................................................... 209 Figures ................................................................................................... 210 References.............................................................................................. 212 Section 3: Adaptive Reuse of Untraced and Virtual Texts Daniele Cuneo “This is Not a Quote”: Quotation Emplotment, Quotational Hoaxes and Other Unusual Cases of Textual Reuse in Sanskrit Poetics-cum-Dramaturgy ...... 219 1 Introduction: Reuse, novelty, and tradition ...................................... 220 2 Śāstra as an ideological apparatus.................................................... 221 3 The worldly śāstra, its fuzzy boundaries, and the derivation of rasas............................................................................ 224 4 Quotation emplotment and the teleology of commentarial thought....................................................................... 232 5 Quotational hoaxes and novelty under siege .................................... 236 6 Unabashed repetition and authorial sleight of hand ......................... 237 7 Conclusions: The alternate fortunes of the two paradigms of textual authoritativeness ............................................. 239  Contens 9 Appendix: Four translations of Abhinavagupta’s intermezzo ................ 246 References .............................................................................................. 247 Kiyokazu Okita Quotation, Quarrel and Controversy in Early Modern South Asia: Appayya Dīkṣita and Jīva Gosvāmī on Madhva’s Untraceable Citations ... 255 Introduction ............................................................................................ 255 1 The modern controversy: Mesquita vs. Sharma ............................... 256 2 Untraceable quotes and Purāṇic studies ........................................... 257 3 Untraceable quotes and Vedānta as Hindu theology ........................ 259 4 Early modern controversy: Appayya Dīkṣita vs. Jīva Gosvāmī ....... 260 4.1 Appayya Dīkṣita ................................................................. 260 4.2 Jīva Gosvāmī ...................................................................... 267 Conclusion ............................................................................................. 274 References .............................................................................................. 275 Elisa Freschi Reusing, Adapting, Distorting? Veṅkaṭanātha’s Reuse of Rāmānuja, Yāmuna (and the Vṛttikāra) in his Commentary ad Pūrvamīmāṃsāsūtra 1.1.1 ........................................................................... 281 1 Early Vaiṣṇava synthesizing philosophies ....................................... 281 2 Veṅkaṭanātha as a continuator of Rāmānuja (and of Yāmuna) ........ 283 3 The Śrībhāṣya and the Seśvaramīmāṃsā: Shared textual material .. 285 3.1 Examples ............................................................................ 285 3.1.1 The beginning of the commentary ...................................... 285 3.1.2 Commentary on jijñāsā ....................................................... 287 3.1.3 vyatireka cases .................................................................... 288 3.1.3.1 Śaṅkara’s commentary on the same sūtra........................... 288 3.1.3.2 Bhāskara’s commentary on the same sūtra......................... 290 3.2 Conclusions on the commentaries ad Brahmasūtra / Pūrvamīmāṃsāsūtra 1.1.1 .................................................. 293 4 The Śrībhāṣya and the Seśvaramīmāṃsā: A shared agenda concerning aikaśāstrya ..................................................................... 294 4.1 Similarities between the treatment of aikaśāstrya in the Seśvaramīmāṃsā and the Śrībhāṣya ................................... 295 4.2 The Saṅkarṣakāṇḍa ............................................................ 297 4.2.1 The extant Saṅkarṣakāṇḍa .................................................. 299 4.2.2 The Saṅkarṣakāṇḍa-devatākāṇḍa ....................................... 303 4.2.3 Quotations from the Saṅkarṣakāṇḍa ................................... 304 4.2.4 The Saṅkarṣakāṇḍa and Advaita Vedānta .......................... 306 4.2.5 The Saṅkarṣakāṇḍa and the Pāñcarātra .............................. 309 10 Contens 4.2.6 Conclusions on the Saṅkarṣakāṇḍa .................................... 310 4.2.7 The authorship of the Saṅkarṣakāṇḍa................................. 312 5 Yāmuna and the Seśvaramīmāṃsā: Shared textual material ............ 316 6 Conclusions ...................................................................................... 319 References.............................................................................................. 320 Cezary Galewicz If You Don’t Know the Source, Call it a yāmala: Quotations and Ghost Titles in the Ṛgvedakalpadruma .............................. 327 1 The Ṛgvedakalpadruma ................................................................... 329 2 The concept of the daśagrantha ....................................................... 330 2.1 Keśava Māṭe’s interpretation of the daśagrantha ............... 331 2.2 The sūtra within Keśava’s daśagrantha ............................. 332 3 The Rudrayāmala as quoted in the Ṛgvedakalpadruma ................... 336 4 The Rudrayāmala and the yāmalas .................................................. 338 5 Textual identity reconsidered ........................................................... 340 6 What does the name Rudrayāmala stand for? .................................. 341 7 Tantricized Veda or Vedicized Tantra?............................................ 342 8 Quotations and loci of ascription...................................................... 343 9 Spatial topography of ideas .............................................................. 345 References.............................................................................................. 346 Section 4: Reuse from the Perspective of the Digital Humanities Sven Sellmer Methodological and Practical Remarks on the Question of Reuse in Epic Texts .................................................................................................... 355 Introduction............................................................................................ 355 1 Epic reuse. ......................................................................................... 357 1.1 Internal reuse ....................................................................... 358 1.1.1 Repetitions .......................................................................... 358 1.1.2 Fixed formulas .................................................................... 359 1.1.3 Formulaic expressions ........................................................ 360 1.1.4 Flexible patterns .................................................................. 360 1.2 External reuse and its detection ........................................... 361 1.2.1 Unusual vocabulary ............................................................ 363 1.2.2 Exceptional heterotopes ...................................................... 365 1 .2 .3 Specific metrical patterns ................................................... 368 Conclusion ............................................................................................. 369 References.............................................................................................. 370  Introduction: Conceptual Reflections on Adaptive Reuse* Elisa Freschi and Philipp A. Maas 1 The dialectics of originality and reuse The purpose of the present volume is to explore a specific aspect of creativity in South Asian systems of knowledge, literature and rituals. Under the head- ing “adaptive reuse,” it addresses the relationship between innovation and the perpetuation of earlier forms and contents of knowledge and aesthetic expres- sions within the process of creating new works. This relation, although it has rarely been the topic of explicit reflections in South Asian intellectual tra- ditions, can be investigated by taking a closer look at the treatment of earlier materials by later authors. With this in mind, the chapters of this book discuss, for example, the following questions: What is an “original” con- tribution of an author? How can instances of adaptive reuse of older textual materials be detected?1 What are the motives of and purposes for adaptive reuse? Why does an author recur to something already available instead of inventing something new? What did it mean to be an “author,” to be “orig- inal,” or to be “creative” during South Asian cultural history? By dealing * Work on this volume has been generously supported by the FWF in the context of pro- ject No. V-400 (EF), by the Institute for South Asian, Tibetan and Buddhist Studies of the University of Vienna, and by the DFG in the context of the project “Digitale kritische Edition des Nyāyabhāṣya.” We are grateful to the Deutsche Morgenländische Gesellschaft for accepting the volume in their series. All costs related to the publication were covered by the FWF (PUB 403-G24). We would also like to thank Cynthia Peck- Kubaczek for her careful copy-editing and Dania Huber for checking the bibliographi- cal entries. 1 The terms “textual” and “text” should be conceived in this introduction in a very broad way. In accordance with Hanks, we believe that “text can be taken (heuristically) to designate any configuration of signs that is coherently interpretable by some com- munity of users” (Hanks 1989: 95, emphasis by Hanks). It thus includes also works of visual and performative arts. On the presence of an underlying “text” also in oral performances, see Barber 2005. 12 Elisa Freschi and Philipp A. Maas with these and related questions this volume moreover addresses the follow- ing two topical complexes: – detecting specific patterns and practices of adaptive reuse in South Asian cultural history, and – reframing concepts such as “originality” and “authorship” in South Asia by means of a closer investigation of instances of adaptive reuse. 2 The background When we started to conceptualize the present volume, there had only been a few studies on the topic of textual reuse we could build on. – Ernst Prets’ database of Nyāya fragments (available here: http://nyaya. oeaw.ac.at/cgi-bin/wr/listaut.pl) mainly focused on the retrieval of frag- ments rather than the reasons for their reuse. – The conference entitled “Transmission and Tradition: The Meaning and the Role of Fragments in Indian Philosophy” that was organized by Prets and Hiroshi Marui in Matsumoto in 2012 expanded on this first purpose of detecting and identifying fragments of mostly lost works by adding the evaluation of the formative role of early Indian philosophy as it can be re- constructed through such fragments (abstracts and program of the confe- rence can be found here: http://nyaya.oeaw.ac.at/cgi-bin/conf/adv.pl). – The book edited by Julia Hegewald and Subrata Mitra (Hegewald and Mitra 2012b) mainly focused on the political value of the reuse of artistic elements. – The book edited by Elisa Freschi on the form of quotations and references in South Asian śāstras (Freschi 2015b) established a basis for the present project, insofar as the book deals with the various forms of reusing textual materials. The present volume builds on the above work by scrutinizing different pur- poses of adaptive reuse. The editors had the pleasure to discuss these topics in person with the authors of the various chapters in the context of the the- matic panel “Adaptive Reuse of Texts, Ideas and Images” at the 32nd Deut- scher Orientalistentag held in Münster in September 2013.2 An expanded and 2 For the titles and abstracts of the individual contributions to this panel, see http:// tinyurl.com/paefcq3.  Introduction 13 revised version of the presentation of Gianni Pellegrini was published sepa- rately in the Journal of Indian Philosophy (Pellegrini 2016). 3 Some basic conceptual tools The concept of reuse comprises four main aspects, viz. (1.) the involvement of at least one consciously acting agent, who, (2.) in order to achieve a certain purpose, (3.) resumes the usage (4.) of a clearly identifiable object after an interruption in its being used. The attribute “adaptive” presupposes that the reusing person pursues a specific purpose by adapting something already existent to his or her specific needs. The reused object has to be identifiable as being reused, because otherwise the adaptation is not an instance of reuse, but of recycling (see below, section 5). In the fields of city planning and architecture, the theoretical concept of “adaptive reuse” has been influential for at least the last thirty years.3 With “adaptive reuse” scholars in these fields describe a phenomenon that lies at the basis of each re-actualization of an architectural element. The concept of adaptive reuse is thus as old as architecture itself (see Plevoets and van Cleempoel 2013 for a historical survey). In city planning and architecture, adaptive reuse applies to the use of a building (often partially reconstructed) for a new function that differs from the purpose for which the building was originally erected. Adaptive reuse is an alternative to demolition and is em- ployed for a wide range of aims, such as saving material resources, prevent- ing urban sprawl, or preserving, at least to some degree, the appearance of townscapes. Thus, agency, finality and creativity are key elements in adaptive reuse. An additional important factor in the process of reuse is the interruption of a previous use, which leads to questions concerning the many historical, religious, philosophical, social and/or political causes that result in the use of a certain architectural element or, in our case, of a text or concept being inter- rupted. 3.1 Simple re-use versus different grades of adaptive reuse In the context of the present volume we shall differentiate between two ideal types of re(-)use, i.e., simple re-use and adaptive reuse. Simple re-use is the resumption of the previous use of an item without a strong change of pur- 3 See the discussion of the history and prehistory of this concept in Plevoets and van Cleempoel 2011. 14 Elisa Freschi and Philipp A. Maas pose(s). An item is employed again because it is readily available and can be easily used. Usually the re-user does not want the re-used element to be spe- cifically recognized as having been re-used. To elaborate, simple re-use is the act of “again using” something that had been used earlier. Typically, simple re-use implies no change in purpose. This is the case when, for example, a pillar from an ancient monument is re- used to support an architectural element in a new building. Simple re-use is also characterized by the fact that the re-used item is readily available. For instance, re-using a pillar from an old building for a new one constitutes a case of simple re-use, if the re-use is the easiest and cheapest solution for erecting that colonnade. Moreover, in simple re-use, re-used objects are not marked as being re-used, because the audience4 is not supposed to recognize the re-use at all. In contrast to simple re-use, adaptive reuse is not merely the repetition of a previous use; it implies more than an item just being used again.5 In adap- tive reuse, the reuser expects his or her audience to recognize the reused ele- ments in order to achieve a well-defined purpose, as for example adding pres- tige, credibility, etc. to the newly created item. Adaptive reuse may involve a more substantial change in the usage. Moreover, it is not motivated (primary- ly) by economic reasons. Reusing a pillar from an ancient monument consti- tutes a case of adaptive reuse if it is reused, for instance, to show continuity with the past, etc. However, simple and adaptive re(-)use do not mutually exclude each other. In general, different degrees of adaptation characterize individual cases of re(-)use. On the side of simple re-use, economic reasons are more relevant, whereas on the side of adaptive reuse, changes of purpose (“resemantiza- tion”), and authorial expectations concerning the audience’s recognition of the reuse are more dominant (see Fig. 1). simple re-use adaptive reuse Figure 1: The spectrum from simple re-use to adaptive reuse: there is no sharp line – simple re-use and adaptive reuse blend into each other. 4 This term will be used throughout this introduction to indicate all possible targets of a text, work of art or performance, i.e., readers, listeners, viewers, spectators, etc. 5 In order to highlight our differentiation of simple re-use from adaptive reuse, we have decided to refer to the former concept with the word “re-use” (with a hyphen) and to the later, with “reuse” (without a hyphen).  Introduction 15 The grade of adaptation in any case of adaptive reuse lies to a considerable degree in the eye of the beholder, whose ability to determine the adaptation may vary over time and in different historically, culturally or socially deter- mined contexts. Therefore, a shift in the time, place, context or social posi- tion of the audience may lead to varying interpretations of a given instance of re(-)use as being more or less adaptive or simple. For instance, Elisa Freschi (2015c) discussed unmarked passages within a late Mīmāṃsā text derived from previous authoritative sources that a well-informed audience of the time would probably have recognized immediately, although contemporary readers may fail to do so. Accordingly, what today may seem an instance of simple re-use was intended as adaptive in its original context. Figure 2: Minerveo obelisk by Gian Lorenzo Bernini (1667), Rome, Italy. File:SantaMariaSopraMinerva-Pulcin03-SteO153.JPG, detail, (CC BY-SA 3.0) https://commons.wikimedia.org/w/index.php?curid=27583105. 16 Elisa Freschi and Philipp A. Maas Within the visual arts, building stones procured from a monument and re-used for the same purpose as before without any implication of their artis- tic value may be an extreme case of simple re-use. In contrast, adaptive reuse would be materials from a previous monument being reused with the con- scious aim of making the audience aware, for instance, of a foreign civiliza- tion’s subjugation (such as incorporation of monuments of African origin in European monuments during the Colonial period) or of a foreign culture’s dissolution into one’s own culture, such as the Minerveo Obelisk [Rome, 1667], a sculpture by Gian Lorenzo Bernini that combines an ancient Egyp- tian obelisk with Christian elements and other symbols, such as an elephant from the Hypnerotomachia Poliphili (see Fig. 2, above). There are many examples of this type of reuse, such as the reuse – and consequent resemantization – of an image of the Jina at the Doddappa temple in Adargunchi. This temple originally belonged to the Jaina community and was later appropriated by Vīra Śaivas. The new owners did not destroy the Jina image, but rather reused it. As Hegewald and Mitra (2012b, pp. 61–64) have explained, the Vīra Śaivas applied horizontal lines and ashes to the Jina sculpture as signs of the image’s “conversion” to Śaivism. The adaptive reuse thus presumably communicated to the audience that Vīra Śaivism had be- come the new dominant religion (see Fig. 3). Figure 3: A statue of a Tīrthaṅkara adaptively reused as a Śaiva image. Source: Plate 3.5 in Hegewald and Mitra 2012b, with kind permission of Julia Hegewald.  Introduction 17 In the case of textual reuse, adaptive reuse highlights the fact that the textual material has been reused. Its reuse emphasizes the text and its connotations. For example, it possibly adds prestige to the newly created text or situates that text within a continuous and illustrious tradition. In this way, the reused text mediates the new text to its audience. At the same time an explicitly marked quotation also highlights the quoted text as an extraneous element. It thus establishes a distance, putting the reused material in a showcase, so to speak. The quoted text is perceived as alien to the new context, since it comes from the past or some other remote context.6 This may be the reason why authors of Mīmāṃsā texts7 generally did not explicitly indicate quoted passages from authors of their own school; these were silently embedded in the texts. However they highlighted quotations from the works of authors belonging to other schools. In this way, the reuser endorsed his own tradition and explicitly distanced himself from other schools of thought. 4 Adaptive reuse: Aspects of creativity Texts are reused in different historical and intellectual contexts and for different purposes. For example Jīva Gosvāmin and Jayanta both reuse the following stanza from Bhartṛhari’s Vākyapādīya: yatnenānumito yo ’rthaḥ kuśalair anumātṛbhiḥ | abhiyuktatarair anyair anyathaivopapādyate || (Vākyapādīya 1.34) Even a matter that was inferred with effort by skilled experts in infe- rence is later established in a completely different way by those even more competent. Bhartṛhari’s original intention was to point out the unreliability of inference. Any inference leads to results that are only provisionally valid, because pre- vious results can always be superseded by later inferences. Bhartṛhari’s stanza became so well known that it was reused by Jīva Gosvāmin, the 16th c. systematizer of Gauḍīya Vaiṣṇavism, who employed it to voice a more gen- eral criticism of the truth-claims of logic applied independently of the Vaiṣ- ṇava sacred texts.8 6 It is interesting to note that quotation marks are accordingly used both to quote texts and to express distance from certain words or expressions. On the double nature of ci- tations, see Nakassis 2013. 7 And perhaps also of other śāstras; see Freschi 2015c and Freschi 2015a. 8 Note, in this regard, the use of āpāditaḥ “obtained, proved” instead of anumitaḥ “in- ferred.” 18 Elisa Freschi and Philipp A. Maas tathā prācīnair apy uktam. yatnenāpādito ’py arthaḥ kuśalair anumātṛbhiḥ | abhiyuktatarair anyair anyathaivopapādyate || iti. (Jīva Gosvāmin, Bhaktirasāmṛtasindhu, sāmānyabhakti, 1.46) Accordingly, also the ancient ones said: “Even a matter that was pro- ven with effort by skilled logicians is later established in a completely different way by those even more competent.” Whereas Bhartṛhari had aimed at establishing his view that knowledge is al- ways linguistic in nature (because nothing exists apart from the śabdabrah- man and all our cognitions are permeated by language), Jīva aimed at estab- lishing – on the strength of the authority of the “ancients” (a concept that he does not further specify, although probably Bhartṛhari would not have been the favorite reference for Jīva’s audience) – the dominant role of the Vaiṣ- ṇava sacred texts. Thus, its new context and frame gives the Vākyapādīya passage a new meaning.9 A stronger example for the adaptiveness of reuse is Jayanta’s parodical reuse of the same passage, which occurs in the Nyāyamañjarī quite close to a quotation of Bhartṛhari’s original verse (the latter is found on p. 316 of the Mysore edition): yatnenānumito yo ’rthaḥ kuśalair anumātṛbhiḥ | abhiyogaśatenāpi so ’nyathā nopapadyate || (Nyāyamañjarī 2, p. 326 of the Mysore edition) A matter that was inferred with effort by skilled experts in inference cannot be explained differently, even with one hundred attempts! Things become more complicated when studying the reuse of concepts be- cause such cases are often more difficult to identify than those of textual reuse (examples are discussed in the chapter by Philipp Maas in the present volume). However, the possibility of encountering the reuse of a concept should be kept in mind for any historically oriented investigation so that non- literal reuses are not overlooked and the degree of innovation in a new ideol- 9 Karin Barber explains how the same applies to oral texts: “The power of the concept of quotation is that it captures simultaneously the process of detachment and the process of recontextualization. A quotation is only a quotation when it is inserted into a new context. Thus, in the very act of recognizing a stretch of discourse as having an inde- pendent existence, the quoter is re-embedding it. This, I suggest, helps us to understand how ‘text’ (the detachable, decontextualized stretch of discourse) and ‘performance’ (the act of assembling and mobilizing discursive elements) are two sides of a coin, in- separable and mutually constitutive” (Barber 2005: 275).  Introduction 19 ogy is correctly estimated.10 Examples of the reuse of concepts include M. K. Gandhi’s employment of the idea of non-harming (ahiṃsā) in the political context of non-violence, as well as Neo-Hinduism’s reinterpretation of the concept of dharma (see Halbfass 1988). As for the reuse of images, it is interesting to observe how reused images tend to acquire new meanings in their new contexts; suffice it to remember the regular reuse of the image of Mona Lisa, or Andy Warhol’s provocative reuse of images of political icons or sex symbols. Warhol’s works are also an interesting case study regarding the concept of authorship, given that a few of his most famous oeuvres, such as his Che Guevara image reproduced in Fig. 4, were actually fakes by Gerard Malanga that Andy Warhol later “authenti- cated” when developing his own style, whereby he treated these works as if they were his own creation. Figure 4: Andy Warhol’s Che Guevara, 1968. Reproduced from Ziff 2006: 79. The shared element in all of these cases is the fact that an author recurs to something already available instead of inventing something new, which leads to the question of why an author makes this particular choice. 10 Cf. the investigation of the concept of “interlanguage” in Freschi 2015a. The concept of “interlanguage” has in fact been devised in order to deal with the case of ideas spreading between intellectuals without having any specific linguistic form. 20 Elisa Freschi and Philipp A. Maas 5 “Adaptive reuse” and related terms As described above (section 3.1), the present volume explores the phenome- non of adaptive reuse in contradistinction to simple re-use. There are, how- ever, several other concepts similar to these two types of “reuse.” In order to clarify the focus of the present volume, it may be appropriate to explain why these concepts are not relevant to the investigation at hand. For example, “recycling” refers to the re-use of raw materials, such as the sand or lime from a former temple being re-used in the foundations of a new one. Since we are interested in the reuse of distinguishable and identifiable elements, this re-use would be relevant to us only if it were accompanied by a sense of appropriation, superiority, etc., that is, if it were intentional. The concept of “reproduction” refers to the recreation of something that ideally is as similar as possible to the original (e.g., a Xerox copy), so similar that the audience might not even be able to detect a difference between the two. In contrast, we want to focus on reuse as a specific expressive modality, one that needs the audience to be more or less aware of the adaptive nature of the reuse.11 5.1 Adaptive reuse, intertextuality and adaptation studies When trying to estimate the potential of the concept of adaptive reuse as a hermeneutical tool in literary and art historical studies, it may seem that quite a number of phenomena that we place under the heading of adaptive reuse ‒ the creation of new meaning in changed contexts through allusions, refer- ences, quotes, etc., or in the form of citation, plagiarism, parody, creative censorship,12 etc. ‒ have already been extensively researched in the field of literary theory under the heading of intertextuality. In fact, since being invented by Julia Kristeva in 1966, the concept of in- tertextuality has had its own history of adaptive reuse.13 In this process, while various authors have created a large number of theories, no consensus exists on what exactly “intertextuality” means in literary studies.14 As a result, there exist nearly as many definitions of intertextuality as there are academic au- 11 For a discussion of other related terms, see the introduction in Hegewald and Mitra 2012b. 12 On this topic, see Maas 2013–2014. 13 Kristeva wrote the essay “Le mot, le dialogue et le roman” in 1966, although it was only published in 1969. For a short history of the concept of intertextuality and the term “intertextuality” itself, see Martinez Alfaro 1996. 14 See Vögel 1998.  Introduction 21 thors writing on the subject. This makes intertextuality a rather cumbersome hermeneutical tool. Moreover, a common feature of many theories on intertextuality in the post-modernist discourse is that they view literary and artistic works from a synchronic perspective in which the historical contexts of authors and reusers as well as their specific intentions do not play a role. In contrast to this, at- tention to historical processes and contexts is central to our understanding of the concept of adaptive reuse; we apply it to concrete individual cases in their specific cultural and historical settings. This does not, however, rule out the possibility of a fruitful interaction with scholars focusing on intertextuality in the field of the reuse of Indian philosophical texts, as Himal Trikha has shown (Trikha forthcoming). An additional field of scholarship with an excellent prospect for fruitful future exchanges is adaptation studies. Having emerged in the last two de- cades as a novel trend in cultural studies, adaptation studies have now turned into a field of research in their own right. In the course of this development, adaptation studies widened their focus from the almost exclusive study of transformations of literary sources into movies to research in adaptations of a large variety of cultural phenomena across different media in different, main- ly European and North American, historical contexts.15 This extension of the objects of research went along with a deepening of methodological and the- oretical reflections. Especially noteworthy in the context of the present vol- ume are attempts to create a comprehensive theory of adaptation.16 This newly emerging theoretical background will be as useful for future research in adaptive reuse in the context of South Asian cultural history as the con- sideration of the peculiarities of South Asian cultures will be relevant for the creation of any intercultural theory of adaptation. 6 On the present volume The chapters of this volume explore the issues outlined above from various viewpoints, but with similar methodologies and – as far as possible – using similar terminology. The first section, entitled “Adaptive Reuse of Indian Philosophy and Other Systems of Knowledge,” consists of five chapters dealing with the adaptive reuse of traceable texts, that is, texts that are identi- fiable as real works existing or having existed independently from the reusing 15 See Bruhn et al. 2013. 16 See Hutcheon and O’Flynn 2013. 22 Elisa Freschi and Philipp A. Maas works. In some cases these are lost works. For example, the reused text dis- cussed by Ivan Andrijanić has no independent attestation; the work is only preserved within the reusing text. The instances discussed by Maas possibly also entail references to lost texts, although Maas shows that the main reused text is the extant Pātañjalayogaśāstra. The degree of independent attestation increases gradually through the chapters by Yasutaka Muroya, Himal Trikha and Malhar Kulkarni. These five chapters all elaborate on the adaptive reuse of texts: śāstric passages within other śāstric texts (Kulkarni), and philosophi- cal passages found, respectively, in philosophical texts (Trikha, Andrijanić and Muroya) and in a work of poetry (Maas). A common characteristic of all these adaptive reuse cases is that the reuse is not a pedantic repetition of something already known, but a means by which authors acted creatively within (or across) given traditions. This happened with various gradations. The reuse of the Pātañjalayogaśāstra discussed by Maas reproduces both the form and the content of the reused text. It is thus employed to enhance the prestige of both the reusing and reused text through a process of reciprocal legitimation. In contrast, the Grammars reusing the techniques of Pāṇini dis- cussed by Kulkarni deviate in part from their model but achieve only a li- mited degree of real improvement, probably because they were influenced by Pāṇini’s authoritative system to such a degree that they did not dare to intro- duce any real innovation. Following the same line, Muroya and Trikha show how a clever dose of adaptation may lead to final results that differ considera- bly from the reused text; the reusing author can go beyond the intentions of the author being reused. Lastly, Andrijanić shows that an adaptive reuse can be so adaptive that it even supersedes the original premises of the reused text. In section 2, entitled “Adaptive Reuse of Tropes,” the two chapters by Elena Mucciarelli and Cristina Bignami focus on the reuse of the motif of the chariot in late Vedic as well as medieval Indian texts and rituals. Through these case studies, the two authors show that the terminology formulated in this introduction can be successfully applied to a methodologically sound analysis of a given trope under changing cultural and historical circum- stances. The framework of adaptive reuse allows, in fact, meaningful ques- tions to be asked regarding the involved agency and the agenda of the actors adaptively reusing a motif, in this case, that of the chariot. Section 3 consists of four chapters (by Daniele Cuneo, Kiyokazu Okita, Elisa Freschi and Cezary Galewicz ), which like the chapters in section 1 deal with philosophical or śāstric texts. However, the chapters of section 3 focus on untraceable reused texts, that is, texts whose significance is based on the texts in which they are found. It is certainly possible that this lack of inde- pendent testimony is the result of historical contingencies, that is, the loss of  Introduction 23 the original texts. But it is not inconceivable that the lack of a traceable source could be a feature designed by an author to camouflage the introduc- tion of an innovation into his tradition by ostensibly reusing ghost texts to legitimize the reusing text. The prestige that reusing texts gain in such a process increases when such ghost texts continue to be reused. In fact, re- gardless of whether such ghost texts actually had a former historical exis- tence, they apparently lived an independent life even when their textual basis could not be (or could no longer be) identified. This is clearly the case for the Saṅkarṣakāṇḍa, as discussed by Freschi, which was quoted and discussed at length although the original text (if it ever existed) was no longer accessible. The fragments that Madhva seemingly quotes but which remain untraceable other than in his work are similar. Okita shows how they continued to be used and referred to in Gauḍīya Vaiṣṇava texts. Galewicz and Cuneo discuss the phenomenon of ideas or concepts being attributed to an earlier authority. Their chapters suggest the existence of a common tendency to attribute ideas to other authors or texts. Not only Madhva, if Roque Mesquita’s recon- struction is correct, needed to attribute the most innovative traits of his sys- tem to untraceable texts, but also Abhinavagupta saw it necessary to attribute to an alleged authority the various prima facie views that he, according to Cuneo, was about to defeat. Galewicz shows that the same tendency is also attested in modern and contemporary India. In this sense, the tendency high- lighted in these chapters counterbalances the act, discussed in the chapters by Muroya and Andrijanić, of silently reusing older material. Thus the picture of the role and significance of adaptive reuse is more complex than it may have seemed at first glance. In the cultural history of South Asia, it is possible to find opposite tendencies: on one hand, silent appropriation, and on the other, the appeal to authority, which appears most notably in the case of texts and concepts outside an author’s own school.17 Silent reuse is clearly much more difficult to detect, precisely because it is neither acknowledged nor identified as reuse. The chapter by Sven Sellmer in section 4 of the present volume (“Reuse from the Perspective of the Digital Humanities”) proposes the implementation of an IT tool to detect alterations in the uniform texture of a given work to discover unacknowledged reuses. 17 Freschi 2015a argues that the former tendency prevailed in earlier phases of Indian philosophy, especially with regard to texts and concepts reused within one and the same school of thought. 24 Elisa Freschi and Philipp A. Maas References Barber 2005 Barber, Karin. “Text and Performance in Africa.” Oral Tradition 20.2 (2005): 264–277. Bruhn et al. 2013 Bruhn, Jørgen, Anne Gjelsvik, and Eirik Frisvold Hanssen. “‘There and Back Again’: New Challenges and New Di- rections in Adaptation Studies.” Adaptation Studies: New Challenges, New Directions. Ed. Bruhn, Jørgen, Anne Gjelsvik, and Eirik Frisvold Hanssen. London, New York: Bloomsbury Academic, 2013. 1–16. Freschi 2015a Freschi, Elisa. “The Reuse of Texts in Indian Philosophy. Introduction.” In: Freschi 2015b: 85–108. Freschi 2015b —, ed. The Re-use of Texts in Indian Philosophy. Special issues of the Journal of Indian Philosophy 43.2–3 and 4–5 (2015). Freschi 2015c — “Quotations, References, etc. A Glance on the Writing Habits of a Late Mīmāṃsaka.” In: Freschi 2015b: 219– 255. Halbfass 1988 Halbfass, Wilhelm. “Reinterpretations of Dharma in Mod- ern Hinduism.” India and Europe: An Essay in Understan- ding. Ed. Wilhelm Halbfass. New York: State University of New York Press, 1988. 334–348. Hanks 1989 Hanks, W. F. “Text and Textuality.” Annual Review of An- thropology 18 (1989): 95–127. Hegewald and Mitra 2012a = Hegewald, Julia A. B. and Subrata K. Mitra. “The Past in the Present. Temple Conversions in Karnataka and Ap- propriation and Re-use in Orissa.” In: Hegewald and Mitra 2012b: 55–85. Hegewald and Mitra 2012b = —, eds. Re-use: The Art and Politics of Integration and Anxiety. New Delhi; Thousand Oaks, Calif.: SAGE, 2012. Hutcheon and O’Flynn 2013 = Hutcheon, Linda, and Siobhan O’Flynn. A Theory of Adaptation. 2nd ed. London, New York: Routledge, 2013. Kristeva 1969 Kristeva, Julia. “Le mot, le dialogue et le roman.” Sēmei- ōtikē. Recherches pour une sémanalyse. Paris: Éditions du Seuil, 1969. 82–112. Maas 2013–2014 Maas, Philipp André. “On Discourses of Dharma and the Pañcatantra.” Wiener Zeitschrift für die Kunde Südasiens 55 (2013–2014): 5–31. Martinez Alfaro 1996 Martinez Alfaro, María Jesús. “Intertextuality: Origins and Development of the Concept.” Atlantis 18.1–2 (1996): 268–285. Nakassis 2013 Nakassis, Constantine V. “Citation and Citationality.” Signs and Society 1.1 (2013): 51–78.  Introduction 25 Nyāyamañjarī Nyāyamañjarī of Jayanta Bhaṭṭa with Ṭippanī Nyāyasau- rabha by the Editor. Ed. K. S. Varadacharya. Vol. 2. Mysore: Oriental Research Institute, 1983. Pellegrini 2016 Pellegrini, Gianni. “On the Alleged Indebtedness of the Vedānta Paribhāṣā towards the Vedānta Kaumudī: Some Considerations on an Almost Forgotten Vivaraṇa Text (Studies in Vedānta Kaumudī I).” Journal of Indian Phi- losophy 44.3 (2016): 485–505. DOI 10.1007/s10781-014- 9271-2. Plevoets and van Cleempoel 2011 = Plevoets, Bie and Koenraad van Cleempoel. “Adaptive Reuse as a Strategy towards Conservation of Cultural Heritage: A Literature Review.” Structural Stu- dies, Repairs and Maintenance of Heritage Architecture 12. Ed. C. Brebbia and L. Binda. Chianciano Terme: WIT Press, 2011. 155–164. Plevoets and van Cleempoel 2013 = — “Adaptive Reuse as an Emerging Discipline: An Historic Survey.” Reinventing Architecture and Interi- ors: A Socio-political View on Building Adaptation. Ed. G. Cairns. London: Libri Publishers, 2013. 13–32. Trikha forthcoming Trikha, Himal. “Facets of a Fragment: Evaluation and Classification of Intertextual Elements in a Philosophical Jaina Sanskrit Work.” Transmission and Reflection. The Meaning and the Role of ‘Fragments’ in Indian Philo- sophy. Proceedings of a Symposium on Quotations and Paraphrases from and Allusions to Ancient Texts on Indian Philosophy. Ed. Ernst Prets and Hiroshi Marui. Wien: Verlag der Österreichischen Akademie der Wissen- schaften, forthcoming. Vākyapadīya Rau, Wilhelm, ed. Vākyapadīya by Bhartṛhari. Wiesba- den: Steiner, 1977. Abhandlungen für die Kunde des Mor- genlandes 42.4. Vögel 1998 Vögel, Bertlinde. “‘Intertextualität’ – Entstehung und Kontext eines problematischen Begriffs.” MA thesis. Uni- versity of Vienna, 1998. 24 October 2016. <https:// germanistik.univie.ac.at/fileadmin/user_upload/inst_germa nistik/voegel.rtf.> Ziff 2006 Ziff, Trisha. Che Guevara: Revolutionary and Icon. Lon- don: Victoria and Albert Publications, 2006.  Section 1: Adaptive Reuse of Indian Philosophy and Other Systems of Knowledge  From Theory to Poetry: The Reuse of Patañjali’s Yogaśāstra in Māgha’s Śiśupālavadha* Philipp A. Maas The present chapter discusses two cases of adaptive reuse of religio-philo- sophical ideas and text passages from the Pātañjalayogaśāstra (“The Au- thoritative Exposition of Yoga by Patañjali,” PYŚ), as well as a single case of a reference to the same, in a work of high-class poetry, namely, Māgha’s epic poem Śiśupālavadha (“The Slaying of Śiśupāla,” ŚPV). The reuse occurs in the two stanzas 4.55 and 14.62, the reference in the three stanzas 1.31–33. After a brief introduction to the two quite different literary works that serve as the respective source and target of reuse (in sections 1 and 2), the chapter outlines the history of research on the ŚPV and its relationship to Sāṅkhya and Yoga philosophy in section 3. The fact that Māgha alluded to Sāṅkhya and Yoga concepts has been known by scholars of Indology for more than a hundred years, but the exact nature of these references has never been inves- tigated in detail. To address this, the first part of section 4 interprets stanzas 4.55 and 14.62, and the passage 1.31–33, highlights the reused text passages and the concepts of classical Yoga, analyses the specific contexts in which the reuse occurs, and suggests possible answers to the question of what au- thorial intentions may have been behind Māgha’s reuse of Patañjali’s work. The final part of section 4 investigates the reception of Māgha’s reuse by the 10th-century Kashmiri commentator Vallabhadeva. In conclusion, section 5 examines the primary historical result of this investigation, namely that the PYŚ was widely known as a unitary authoritative work of Yoga theory and practice in different parts of South Asia at least from the 8th to 10th century. It was this appraisal of the work in educated circles that may have suggested it to Māgha as a source of reuse. By this, he achieved – irrespective of his actual intentions – two interrelated effects: On one hand, his reuse contri- buted to strengthening and maintaining the authority of the śāstra as a ∗ Many thanks to Elisa Freschi, Dominic Goodall, Petra Kieffer-Pülz, Andrey Klebanov, James Mallinson, Chettiarthodi Rajendran and Mark Singleton for valuable hints for, comments on, and corrections to earlier drafts of the present chapter. 30 Philipp A. Maas Vaiṣṇava work, and on the other hand, the reuse of the PYŚ charged the objects of Māgha’s poetical descriptions as well as his poem with the philo- sophical and religious prestige of the śāstra. 1 The Pātañjalayogaśāstra The PYŚ, which is the oldest surviving systematic exposition of philosophi- cal Yoga, was probably composed at some time between 325 and 425 CE by an author-redactor named Patañjali.1 Comparatively late primary sources as well as quite a few works of modern secondary literature suggest that the PYŚ in fact consists of two works, namely the Yogasūtra by Patañjali and a later commentary called the “Yogabhāṣya,” by a (mythical) author-sage named Vyāsa or Veda-Vyāsa. In the context of the present chapter, there is no need to re-discuss the authorship problem of the PYŚ in any detail.2 As I shall demonstrate below, the stanzas of Māgha’s poem reusing the PYŚ draw equally upon sūtra and bhāṣya passages of Patañjali’s work. This shows not only that the poet regarded the PYŚ as a single whole, but also that he ex- pected his audience to share this view. Moreover, even for the commentator Vallabhadeva, who probably lived approximately two hundred years after Māgha, the PYŚ was a textual unit.3 In general, the philosophical and religious views of Pātañjala Yoga are similar to those of classical Sāṅkhya, as is known from the summary of the lost Ṣaṣṭitantra in the seventy (or slightly more) stanzas composed by Īśvara- kṛṣṇa (5th century CE) that are usually called Sāṅkhyakārikā.4 The philo- sophical systems of Yoga and Sāṅkhya are based on the ontological dualism of primal matter (prakṛti or pradhāna) and its products on one hand, and pure consciousness existing as an infinite number of subjects (puruṣa) on the other. There are, however, some noticeable doctrinal differences between Sāṅkhya and Yoga. Classical Sāṅkhya, for example, acknowledges the exis- tence of a tripartite mental capacity, whereas according to classical Yoga the 1 Maas 2006: xix. 2 On the authorship question of the PYŚ, see Maas 2006: xii–xix and Maas 2013: 57–68. 3 See below, sections 2 and 4.3. 4 According to Albrecht Wezler (2001: 360, n. 45), the title of the work as reflected in its final stanza is not Sāṅkhyakārikā but Sāṅkhyasaptati. The title Sāṅkhyakārikā found its way into the handbooks of Indian literature and philosophy possibly due to Cole- brooke’s seminal essay “On the philosophy of the Hindus,” in which the author states: “The best text of the Sánc’hya is a short treatise in verse, which is denominated Cáricá, as memorial verses of other sciences likewise are” (Colebrooke 1827: 23).  From Theory to Poetry 31 mental capacity is a single unit. Moreover, Yoga emphasizes the existence of a highest God (īśvara), who is described as an eternally liberated subject (puruṣa). The difference between God and other liberated subjects consists in that the latter are conceived as having been bound to matter in the cycle of rebirths prior to their liberation. In contradistinction to this, God was never bound in the past nor is there any possibility for him to be bound in the fu- ture. The transcendental status of God leads to the problem of how a transcen- dental subject, who is axiomatically considered totally free of any activity, can intervene in the world. The solution that Patañjali presented consists in postulating that God’s effectiveness is quite limited. At the beginning of each of the cyclically reoccurring creations of the world, God assumes a perfect mental capacity in order to provide instruction to a seer and to start a lineage of teachers and pupils. This process, according to Yoga, is not an activity in the full sense of the word. It is an event that takes place in accordance with God’s compassionate nature.5 Based on these philosophical and religious foundations, the PYŚ teaches meditations aiming at an unrestricted self-perception of the subject, in which consciousness becomes conscious exclusively of itself, unaffected by even the slightest content of consciousness.6 This special kind of cognition is be- lieved to be soteriologically decisive, because it removes the misorientation of the subject towards matter. This liberating insight is therefore the release of bondage in the cycle of rebirths. 2 Māgha’s Śiśupālavadha Māgha’s ŚPV is a work of a different literary genre than the PYŚ. It is not an authoritative exposition or system of knowledge (śāstra), but an epic poem belonging to the genre of kāvya literature, or, more specifically, to the cate- gory of mahākāvya.7 As such it is one of the most distinguished Sanskrit poetic compositions in which aesthetical purposes outweigh didactic ones.8 5 See Maas 2009: 265f. and 276f. 6 On yogic meditations, see Oberhammer 1977 and Maas 2009. 7 For a general introduction to kāvya literature, see Lienhard 1984 and Warder 1974– 1992. 8 Reusing the work of his predecessor Mammaṭa (11th c.), the 12th century poetologist and polymath Hemacandra specified in his Kavyānuśāsana (1.3) that the first and most important purpose of poetry is pleasure resulting from relishing poetry. Additional aims are fame for the poet and instructions that are delivered – as gently as only lovers 32 Philipp A. Maas The plot of the ŚPV is a modified and lengthy retelling of an episode from the second book of the Mahābhārata (i.e., MBh 2.33–42) that narrates the events leading Kṛṣṇa to kill his relative, the king Śiśupāla.9 Accordingly, the ŚPV as a whole is a case of adaptive reuse of a passage of the MBh as its literary exemplar.10 In his poetic creation, Māgha apparently had several interrelated inten- tions. One of these was providing his audience with a refined aesthetical ex- perience. Moreover, he aimed at glorifying the god Viṣṇu in his incarnation as Kṛṣṇa. Māgha took every effort to show his own poetic skills, his mastery of a large number of meters, and his learnedness in several branches of know- ledge, including literary criticism, metrics, grammar, music, erotology, philo- sophy, etc.11 As was already noted by Hermann Jacobi, Māgha’s literary agenda was also to outdo his predecessor and rival author Bhāravi, who had composed a glorification of the god Śiva in his great poem Kirātārjunīya.12 Modern critics have viewed Māgha’s extraordinary display of poetic and metrical skills as being disproportionate to the development of the plot of the ŚPV, which proceeds with a minimum of dramatic action. However, as Law- rence McCrea has convincingly argued, this slow development of an undra- matic plot and the plethora of embellishments work hand in hand to portray Kṛṣṇa as a consciously omnipotent being who is actually beyond any need of action to fulfill his role in the course of the universe, i.e., establishing and maintaining the Good.13 It is difficult to determine the date of the ŚPV’s composition. A still wide- ly accepted guess is that of Franz Kielhorn from 1906, who drew on informa- tion from the first stanza of the description of the poet’s family lineage (vaṃ- śavarṇana). This brief outline contains the name of a king under whom Mā- gha’s grandfather served as a minister.14 Kielhorn identified this king with a do – to the connoisseur. See Mammaṭa’s Kāvyaprakāśa 1.2 and Both 2003: 48. 9 For a brief summary of the plot of the ŚPV, see Rau 1949: 8f. 10 For a comparison of the ŚPV with its presumptive source, see Salomon 2014. 11 On the different branches of knowledge that a poet was supposed to master, see Kāvya- prakāśa 1.3 (p. 6) and its adaptive reuse in Hemacandra’s Kāvyānuśāsana 1.8 (Both 2003: 52–59). 12 Jacobi 1889: 121–135. According to Rau (1949: 52), Bhāravi and Māgha could at least theoretically both have relied on an unknown common source as their respective point of reference. On Māgha’s program, see also Tubb 2014. 13 See McCrea 2014. 14 sarvādhikārī sukṛtādhikāraḥ śrīdharmlābhasya babhūva rājñaḥ / asaktadṛṣṭir virājaḥ sadaiva devo ’paraḥ suprabhadevanāmā // 1 // (Kak and Shastri 1935: 305) “The glo- rious king Dharmalābha had a chief minister called Suprabhadeva (God of Good Ra- diance), who was chiefly obliged to virtuous actions, always liberal and pure, like a  From Theory to Poetry 33 certain Varmalāta who, according to epigraphic evidence, reigned “about A.D. 625.” This would establish that Māgha “must be placed in about the second half of the 7th century A.D.”15 However, the name of the patron of Māgha’s grandfather occurs in different versions of the ŚPV in twelve va- riants as Gharmalāta, Carmalāta, Dharmadeva, Dharmanātha, Dharmanābha, Dharmalāta, Dharmalābha, Nirmalānta, Varmanāma, Varmalākhya (=Varma- la), Varmalāta and Varmanābha.16 Already Wilhelm Rau observed that most of these variants can be ex- plained as scribal errors caused by the similarity of certain writing blocks or akṣaras in north Indian scripts.17 However, without additional evidence it is impossible to decide which variant (if any) was the starting point for the textual developments leading to the other eleven readings. The fact that “Var- malāta” is the only name attested in an inscription does not establish this reading as the original wording of the ŚPV.18 A final conclusion concerning the original name of the king could only be reached on the basis of research in the text genealogy of Māgha’s work.19 The same is also true for the ques- tion of whether the five stanzas making up the description of Māgha’s family lineage are an original part of the ŚPV or whether they were added in the course of its transmission, as Rau was inclined to believe on the basis of their absence in Mallinātha’s version of the ŚPV (1949: 56f.). Rau even thought that he could identify the commentator Vallabhadeva as the author of the vaṃśavarṇana. In order to arrive at this conclusion, he emended the appar- ently corrupt wording of the final colophon of Vallabhadeva’s Saṃdehavi- ṣauṣadhi (“The Antidote against the Poison of Doubt,” henceforth: Antidote) in such a way that it clearly states Vallabhadeva’s authorship of the five stanzas. This emendation may be unnecessary. According to the printed edition of Kak and Shastri, to which Rau did not have access, the 10th-century commentator from Kashmir actually introduced the section under discussion by stating that it was authored by the poet Māgha, and not by himself: second king (or: like a god).” 15 Kielhorn 1906: 146. McCrea (2014: 123) placed Māgha in the 7th century without further discussion. Bronner and McCrea (2012: 427) suggested the late 7th or early 8th century as the date of composition for the ŚPV, equally without providing any refer- ence. Salomon (2014: 225), who agreed with this dating, referred to Kielhorn 1906. 16 See Rau 1949: 54f. 17 Rau 1949: 55. 18 Hultzsch (1927: 224), however, stated that this is “the inscriptionally attested form of the name” (“die inschriftlich beglaubigte Form des Namens”). 19 Already Rau remarked that this question “can only be solved on the basis of the manu- scripts” (Rau 1949: 55 “läßt sich endgültig nur durch die Handschriften entscheiden”). 34 Philipp A. Maas adhunā kavir lāghavena nijavaṃśavarṇanaṃ cikīrṣur āha (Kak and Shastri 1935: 305,1.). Now the poet, desiring to briefly describe his own lineage, recites the following stanzas. However, even if it can be established that it was the poet Māgha who composed the description of his lineage, this part of his work does not allow for any definite conclusion concerning the date of composition of the ŚPV. At the present state of research, the dating of Māgha to ca. 750 CE, which George Cardona has suggested on the basis of the consideration that Māgha must have lived after Jinendrabuddhi,20 the author of a grammatical commen- tary to which the ŚPV apparently refers, may be the best educated guess.21 3 The Śiśupālavadha and Sāṅkhya Yoga in academic research Despite its high literary quality, the ŚPV has received until quite recently comparatively little scholarly attention. One of the few monographs on Mā- gha’s work is the dissertation of Wilhelm Rau from 1949, which was pub- lished posthumously only in 2012. Rau investigated the textual history of the ŚPV by comparing the text as transmitted in a transcript of a manuscript in Śāradā script containing Vallabhadeva’s Antidote22 with two printed edi- tions.23 In this context, Rau dealt, inter alia, with the historical relationship of two different versions of a passage from the fifteenth chapter of the ŚPV. One version, which was the basis of Vallabhadeva’s Antidote, consists of a series of stanzas that can be understood in two different ways. If interpreted in one way, these stanzas revile Kṛṣṇa. If the verses are understood in a dif- ferent way, they praise Viṣṇu. In contrast to this, the second version plainly denigrates Kṛṣṇa. Rau concluded that the version with two meanings (which Bronner and McCrea 2012 calls the “bitextual version”), that is, the version that Vallabhadeva commented upon, is probably of secondary origin, whereas the version with a single meaning (the “non-bitextual version,” in the termi- nology of Bronner and McCrea 2012), which was the basis of Mallinātha’s 15th-century commentary, was probably composed by Māgha himself. 20 See also Kane 1914: 91–95. 21 Cardona 1976: 281. 22 The exemplar of the transcript was divided in two parts and is now kept at the Staats- bibliothek in Berlin. See the editorial comment by Konrad Klaus und Joachim Sprock- hoff in Rau 1949: 11, n. 4. 23 These editions were Vetāl 1929, and Durgāprasāda and Śivadatta 1927.  From Theory to Poetry 35 In a recent article on this passage by Ygal Bronner and Lawrence McCrea (Bronner and McCrea 2012), the two authors have convincingly argued that using methods of literary criticism and narratology should become a standard for future research on Sanskrit kāvya literature. Bronner and McCrea applied these methods to the above-mentioned passage in the fifteenth chapter of the ŚPV that is transmitted in two different versions. In their discussion of these two divergent versions, they confirmed Rau’s conclusion that the bitextual version is probably of secondary origin, using a whole range of new argu- ments. In addition, they suggested that the younger version was probably composed in the ninth century in Kashmir. The anonymous author of the secondary version presumably considered Māgha’s original unacceptable because of its negative attitude towards Kṛṣṇa, the incarnation of Viṣṇu. Richard Salomon (2014) reviewed seven arguments that Bronner and McCrea adduced in favor of the conclusion that the bitextual version is of secondary origin. According to him, these arguments “have a cumulative force that is persuasive, though their individual power varies.”24 Salomon supplemented Bronner and McCrea’s work by comparing the two versions of the monologue in chapter 15 of the ŚPV with the passage MBh 2.33–42 that Māgha reused for his poem. He found the non-bitextual version to be closer to the MBh passage and in this way added an eighth argument in favor of the originality of the non-bitextual version. According to Salomon, these ar- guments taken collectively strongly suggest that the non-bitextual version is authentic.25 One of Bronner and McCrea’s arguments for the secondary origin of the bitextual version that Salomon does not discuss, probably because the argu- ment is not particularly strong, is that “the philosophical and Sāṅkhya-de- rived themes in the bitextual speech of chapter 15 echo nothing to be found elsewhere in the poem.”26 Although the secondary version of the speech in chapter 15 indeed con- tains many more allusions to Sāṅkhya and Yoga philosophy than the bulk of the text, references to Sāṅkhya and Yoga actually occur in other parts of the ŚPV as well. 24 Salomon 2014: 227. 25 Salomon 2014: 236. 26 Bronner and McCrea 2012: 447. 36 Philipp A. Maas 4 Pātañjala Yoga in the Śiśupālavadha The existence of references to Sāṅkhya and Yoga philosophy in the ŚPV was noticed quite early by scholars of Indology. Already more than one hundred years ago, James Haughton Woods pointed out that the two stanzas ŚPV 4.55 and 14.62, which will be discussed in more detail below, refer to Pātañjala Yoga.27 Eugene Hultzsch, the translator of the ŚPV into German, presented a list of references in Māgha’s work to Sāṅkhya and Yoga concepts in an article that appeared in the Festschrift dedicated to Richard Garbe in 1927.28 This article, which comprises just four and a half pages, mainly lists eighteen references that Hultzsch noted based on the explanations contained in Vallabhadeva’s Antidote. A detailed analysis of the credibility of Vallabha- deva’s information as well as of the nature of Māgha’s references and their respective relationship to the philosophical works of Sāṅkhya and Yoga was apparently beyond the scope of Hultzsch’s article.29 Seven of the eighteen references that Hultzsch listed occur within the bi-textual version of chapter 15. There remain, however eleven instances in the bulk of the text that, at least according to Vallabhadeva, refer to Sāṅkhya and Yoga concepts. Of these, the two stanzas 4.55 and 14.62 stand out, be- cause they do not only refer to Sāṅkhya and Yoga ideas in general; they adaptively reuse clearly identifiable text passages and ideas of the PYŚ. These stanzas therefore attest the thorough acquaintance of their author and his audience with the PYŚ. 4.1 The stanza Śiśupālavadha 4.55 The stanza ŚPV 4.55 is part of a long description of the mountain (or high hill with five peaks) Raivataka, the modern Girnār in Gujarat,30 to which Mā- gha dedicated the fourth chapter of the ŚPV. This chapter can be divided, according to the analysis of Gary Tubb, into three parts.31 The first and the second part consist of nine stanzas each, which constitute the introduction to the chapter and its extension. In these two parts, the voice of the author de- scribes the beauty of the mountain. Thereafter, Dāruka, Kṛṣṇa’s charioteer, 27 Woods 1914: xix. 28 Hultzsch 1927. 29 Hultzsch deals with the following stanzas of the ŚPV: 1.31–33, 2.59, 4.55, 13.23, 13.28, 14.19, 14.62–64, 14.70 and 15.15, 15.18, 15.20–21, 15.27, 15.28, 15.29 (of the bi-textual version). 30 On mount Girnār, see Rigopoulos 1998: 98 and the literature referred to in ibidem, n. 38. 31 Tubb 2014: 174.  From Theory to Poetry 37 takes over and describes again the excellence of the range in an additional fifty stanzas. As Tubb has demonstrated, the entire fourth chapter of the ŚPV (plus the initial strophe of the following canto) consists of twenty-three triads of stan- zas. The initial stanza of each triad is consistently composed in the Upajāti meter in the first two parts of the chapter, and in the Vasantatilakā meter in the third part, whereas the meters in the second and third stanzas of the triads vary. In general, Māgha had a tendency in the third part of the chapter to use comparatively rare meters for the second and third stanzas.32 More important than these metrical peculiarities are the stylistic characteristics of each stanza in a triad. The initial stanzas of a triad “usually have little or no ornamen- tation on the level of sound, and it is here that the poet, freed from the dis- tractions of elaborate rhyme und unusual meter, brings out his heavy guns of imagery.”33 The stanzas in the second position generally contain alliterations (anu- prāsa) and less lively and imaginary descriptions of the mountain, whereas the final stanzas of each triad frequently contain yamakas, i.e., structured repetitions of identical words or syllables with different meanings. Within the third part of the description of the mountain, i.e., within Dā- ruka’s description, stanzas 4.55, the first in a triad, reads as follows: maitryādicittaparikarmavido vidhāya kleśaprahāṇam iha labdhasabījayogāḥ | khyātiṃ ca sattvapuruṣānyatayādhigamya vāñchanti tām api samādhibhṛto nirodhum || (ŚPV 4.55, part 1, p. 146; meter: Vasantatilakā). And here absorption practicing yogis, knowing that benevolence et ce- tera prepare the mind, effect the removal of afflictions (kleśa) and reach an object-related concentration. They realize the awareness of the difference of mind-matter (sattva) and subject (puruṣa), and then they even want to let this cease.34 4.1.1 The reuse of the Pātañjalayogaśāstra in Śiśupālavadha 4.55 The stanza 4.55 of the ŚPV adaptively reuses concepts of Yoga soteriology and describes in a nutshell the yogic path to liberation. At an early stage of 32 Tubb 2014: 184. 33 Tubb 2014: 177. 34 This translation is based on the result of the analysis presented in the next sections of this chapter. Here and everywhere else in this chapter, I have refrained from using square brackets in order to enhance the readability of the translations. 38 Philipp A. Maas this path and as a preparation for more advanced types of attainments, the yogi practices meditations leading to mental stability or prolonged periods of attention. Once this aim is achieved, the aspirant is qualified for other con- tent-related forms of meditation, culminating in the awareness of the differ- ence between matter (which makes up the mind or citta) and subject. In order to gain the liberating insight, i.e., self-perception of the subject, even this ultimate content of consciousness has to cease. The stanza ŚPV 4.55 does not only reuse the PYŚ conceptually by de- scribing the just-mentioned path to liberation, it also reuses the terminology of – as well as phrases from – Patañjali’s authoritative exposition of Yoga. To start with, pāda a of stanza 4.55 draws heavily on the end of PYŚ 1.32 and the beginning of PYŚ 1.33, which read as follows: … tasmād ekam anekārtham avasthitaṃ cittam, yasyedaṃ śāstreṇa parikarma nirdiśyate. (32) maitrīkaruṇāmuditopekṣāṇāṃ sukhaduḥ- khapuṇyāpuṇyaviṣayāṇāṃ bhāvanātaś cittaprasādanam (sūtra 1.33) (PYŚ 1.32,24–33,2). Therefore, it has been established that the mind is a single entity refer- ring to multiple objects. The authoritative exposition teaches its prepa- ration: From cultivating benevolence, compassion, joyousness and dis- regard for beings experiencing happiness, suffering, merit and deme- rit, the mind becomes pure. In this passage, Patañjali states that the “authoritative exposition” or the “sys- tem of knowledge” (śāstra) teaches the cultivation of benevolence and other positive attitudes. To which exposition does this statement refer? Meditations aiming at the cultivation of virtually the same attitudes are prominent in different pre-modern South Asian religions and systems of knowledge. In Buddhism, these meditations are known as “The Four Immeasurable” (apra- māṇa) or “The Divine States of Mind” (brahmavihāra).35 The oldest syste- matic exposition of Jainism in Sanskrit, the Tattvārthasūtra, also teaches in sūtra 7.6 the cultivation of virtually the same attitudes to different kinds of beings. Moreover, Ayurvedic physicians, according to the Carakasaṃhitā, are also expected to develop similar attitudes towards different categories of patients.36 Although, accordingly, benevolence and so on play an important role also in non-yogic milieus, the lack of any reference to a non-yogic con- text in the passage cited above makes it probable that Patañjali referred with the word śāstra to his own authoritative exposition of Yoga or to a different 35 See Maithrimurthi 1999. 36 See Wujastyk 2012: 31.  From Theory to Poetry 39 śāstra of his own school of thought (including the authoritative expositions of Sāṅkhya).37 Cultivating benevolence, etc. occurs in different religions and systems of knowledge. This may in principle render it doubtful whether Māgha actually reused the PYŚ or an altogether different source. However, as James H. Woods noticed long ago, the manner in which ŚPV 4.55 a combines the text of the bhāṣya part of PYŚ 1.32 containing the word “preparation” (pari- karma) of the mind – which does not occur in any other source known to me – with the text of sūtra 1.33 indicates strongly that Māgha indeed reused the passage of the PYŚ cited above and not a similar formulation in a different work.38 The first word of pāda b is an additional case of a verbatim reuse, this time of the technical term “affliction” (kleśa), which refers in the context of Pātañjala Yoga to the set of five basic mental misorientations that sūtra 2.3 lists as “misconception, sense-of-I, craving, aversion and self-preservation.”39 As long as the mind (citta) is affected by these afflictions, the subject mistakenly identifies itself with the contents of mental activities. This process maintains and consolidates the bondage of the subject in the cycle of rebirths. Thus, to reach liberation the afflictions must be removed. In this connec- tion, Patañjali frequently used the word “removal” (hāna) and other deriva- tives of the verbal root hā, as for example in PYŚ 2.15, where he compared his authoritative exposition of Yoga with the science of medicine in the fol- lowing way: tad asya mahato duḥkhasamudāyasya prabhavabījam avidyā. tasyāś ca samyagdarśanam abhāvahetuḥ. yathā cikitsāśāstraṃ caturvyūham – rogo rogahetur ārogyaṃ bhaiṣajyam iti, evam idam api śāstraṃ caturvyūham eva. tad yathā – saṃsāraḥ saṃsārahetur mokṣo mokṣo- pāya iti. tatra duḥkhabahulaḥ saṃsāro heyaḥ. pradhānapuruṣasaṃ- yogo heyahetuḥ. saṃyogasyātyantikī nivṛttir hānam. hānopāyaḥ sam- yagdarśanam iti (PYŚ 2.15; Āgāśe 1904: 77,9–78,5). Therefore (the affliction) “misconception” is the seed for the growing of this huge mass of suffering. And the right view is the cause for its extinction. In the same way that the medical system of knowledge has four divisions – i.e., disease, the cause of disease, health and medicine 37 On the use of the word śāstra in different contexts and with different meanings within the PYŚ, see Wezler 1987: 343–348, which does not refer, however, to the occurrence of the word śāstra in the present context. 38 Woods 1914: xxi. 39 avidyāsmitārāgadveṣābhiniveśāḥ kleśāḥ (sūtra 2.3; ed. Āgāśe 1904: 59). 40 Philipp A. Maas – so also this system of knowledge of Yoga has four divisions, namely, the cycle of rebirths, the cause of the cycle of rebirths, libera- tion and the method leading to liberation. In this regard, the cycle of rebirth that is rich in suffering is what must be removed. The connec- tion of primal matter and the subject is the cause of what must be re- moved. The final dissolution of the connection is removing. The me- thod of removing is the right view.40 The word “removal” (prahāna) in ŚPV 4.55 b is a quasi-synonym of the word “removing” (hāna) and a clear allusion to Patañjali’s conception of the cancellation of the bondage of the subject by means of the removal of afflic- tions.41 The following word of stanza ŚPV 4.55 b, “absorption with an object” (sabījayoga), also refers to the PYŚ, where the phrase “these attainments are the object-related absorption” (tā eva sabījaḥ samādhiḥ) occurs as sūtra 1.46. The choice of the word -sabījayogaḥ instead of its synonym -sabījasamādhiḥ can probably be explained by metrical constraints. This change of terminol- ogy is unproblematic, because Patañjali introduced these two words as syn- onyms at the beginning of his work (PYŚ 1.1), where he explained that “yoga is absorption” (yogaḥ samādhiḥ).42 In addition, ŚPV 4.55 c reuses the central yogic concept of the awareness of the difference between mind-matter (sattva) and the subject (puruṣa).43 Patañjali mentioned this special awareness, which, as it were, paves the way to final liberation, seven times in his exposition, i.e., in PYŚ 1.2, 2.2, 2.26, 3.35, 3,49 (twice), and 4.27. Māgha reused also the next word of ŚPV 4.55 c, the verb-form “realize” (adhigamya), from the PYŚ. This verb or derivatives thereof occur at nine points in Patañjali’s work.44 Of these, the occurrences in PYŚ 1.29 and 2.32 40 For more detailed discussions of this passage, see Maas 2008: 127–130 and Maas 2014: 70f. 41 The addition of the preverbium pra- to -hāna in ŚPV 4.55 is presumably motivated by metrical requirements. 42 PYŚ 1.1,3. 43 My translation of Skt. sattva as “mind-matter” is based on the teaching of Sāṅkhya- Yoga that the mind (citta) consists of the luminous substance sattva, one of the three constituents of primal matter (pradhāna). The expression cittasattva “mind-sattva,” oc- curs, for example in PYŚ 1.2,2. The Pātañjalayogaśāstravivaraṇa explains the com- pound by stating that it is a descriptive determinative compound (kārmadhāraya) in which the first part is in an attributive relationship to the second (cittam eva sattvaṃ cittasattvam; Pātañjayogaśāstravivaraṇa 1.2, p. 10.25). 44 These nine instances are: PYŚ 1.29 (twice), 2.32 (as a quotation of sūtra 1.29), 2.41, 3.6, 3.25, 3.36, 3.48, and 4.23.  From Theory to Poetry 41 are probably the most pertinent cases in the present context, because just as in stanza ŚPV 4.55 they refer to the self-perception of the subject and the reali- zation of the ontological difference between the subject and matter. In addition, also the reference to the stopping of the awareness of the on- tological difference in ŚPV 4.55 d reuses PYŚ 1.2 as follows: ity atas tasyāṃ viraktaṃ cittaṃ tām api khyātiṃ niruṇaddhi. … sa nir- bījaḥ samādhiḥ (PYŚ 1.2,10f.). Therefore the mental capacity, when it becomes detached from this awareness, lets even this cease. This is absorption unrelated to an object. The description of the process of cessation as described in the ŚPV differs, however, from the description of the same development in the PYŚ. Whereas in stanza ŚPV 4.55 the final stopping of mental activity is the result of an act of volition on the side of yogis (vāñchanti), according to the PYŚ the practice of a so-called cessation-experience leads to the stopping of mental activity.45 The designation of yogis as “practitioners of absorption” (samādhibhṛt) in 4.55 d, although not taken directly from the PYŚ, reuses the central yogic term “absorption” (samādhi), which occurs in the final position of the series of terms sketching the yogic path to liberation that is called “ancillaries of Yoga” (yogāṅga).46 If this interpretation is correct, Māgha alluded with this term again to the equation of yoga and samādhi in PYŚ 1.1 (cf. above, p. 40). On the whole, the terminology of the stanza ŚPV 4.55 is similar to the technical yogic terminology of the PYŚ to such a degree that there can vir- tually be no doubt that Māgha consciously reused the PYŚ; he did not reuse a different work of pre-classical yoga or a different and now lost classical yoga treatise. 4.1.2 Śiśupālavadha 4.55 in context If one reads the stanza ŚPV 4.55 as a part of the monologue of Kṛṣṇa’s cha- rioteer Dāruka, it can hardly be overlooked that the yoga-related motifs figur- ing so prominently in stanza 4.55 differ from the literary motifs that Dāruka addresses in the remaining part of his speech. None of the remaining forty- eight stanzas describing the mountain addresses primarily religious or philo- sophic motifs. 45 See Maas 2009: 273f. 46 yamaniyamāsanaprāṇāyāmapratyāhāradhāraṇadhyānasamādhayo ’ṣṭāv aṅgāni. [sū- tra 2.29]. “The eight ancillaries are commitments, obligations, postures, breath control, withdrawing the senses, fixation, meditation and absorption.” 42 Philipp A. Maas Dāruka’s monologue contains an appealing description of the natural beauty of Raivataka, as for example the famous stanza 4.20, in which Māgha poetically depicts the simultaneous appearance of the sun and moon during their respective rise and setting as the appearance of two bells that shed radiance on the body of an eminent elephant.47 Moreover, Dāruka’s monolo- gue repeatedly emphasizes that the mountain abounds in precious stones and metals, filling the landscape with splendor.48 From the perspective of Yoga, these appealing visual features do not recommend the range as a suitable place for yogic practice, because the mountain as described by Dāruka pro- vides views that are too spectacular. At least according to the prescription of Śvetāśvatara-Upaniṣad 2.10, a place for yogic meditation should be pleasing to the mind, but not overwhelming to the eye.49 A further motif in Dāruka’s description of the mountain that Māgha may have connected with Yoga is “mountain caves.” The caves of Raivataka are, however, not – as one might expect based on stanza 4.55 – places for yogic meditation in reclusion,50 but the location of amorous pleasures that young women share with their lovers.51 The only two stanzas that might be regarded as providing a link to Yoga, because they contain the motif of world-renouncing anchorites, are ŚPV 4.54 and 4.64. The first of these, which appears immediately before the Yoga stanza under investigation, reads as follows: samīraśiśiraḥ śiraḥsu vasatāṃ satāṃ javanikā nikāmasukhinām | bibharti janayann ayaṃ mudam apām- 47 Stanza 4.20, which according to Vallabhadeva earned Māgha the name Ganthamāgha (Māgha of the Bells), reads as follows: udayati vitatordhvaraśmirajjāv ahimarucau himadhāmni yāti cāstam / vahati girir ayaṃ vilambighaṇṭādvayaparivāritavāraṇen- dralīlām // 20 // (Kak and Shastri 1935, part 1, p. 132; meter: Puṣpitāgara) “When the sun is rising as the moon is setting, each with its ropes of rays stretched upward, this mountain has the pomp of a lordly elephant caparisoned with a pair of hanging bells.” (Tubb 2014: 145). 48 See, for example, stanzas 4.21, 26, 27, 28, 31, 37, 40, 44, 46, 49, 53, 56, 65, 68. 49 same śucau śarkarāvahnivālukāvivarjite ’śabdajalāśrayādibhiḥ / mano’nukūle na tu cakṣupīḍane guhānivātāśrayaṇe prayojayet // (Śvetāśvatara Upaniṣad 2.10, Olivelle 1998: 418). “At a pure, level place that is free of grit, fire or sand, with noiseless water sources and so on, which is pleasant but does not press upon the eye, in a cave or a re- fuge that is protected from wind, he should concentrate.” 50 Pātañjayogaśāstravivaraṇa 2.46 mentions mountain caves (giriguhā) as a suitable place for the practice of yogic meditation (p. 225.15). 51 See ŚPV 4.67, p. 152.  From Theory to Poetry 43 apāyadhavalā balāhakatatīḥ || (ŚPV 4.54, part 1, p. 146; meter: Jalod- dhatagati). This mountain, cool by its breezes, pleases the ever-eased sages stay- ing on its summits by bearing bands of clouds that, through shedding their rain, turned into white curtains. Stanza 4.54 is a typical representative of a final stanza of triads that make up the fourth chapter of Māgha’s work. It is a verbal miniature painting creating a lively image of a cool, cloudy and beautiful mountain at the end of the rainy season. At the level of sound, the stanza contains a nice alliteration of sibi- lants in pāda a, combined with the yamakas or “structured repetitions and chiastic structures” that Tubb pointed out as a stylistic peculiarity of the third stanzas in the triads of the fourth chapter of the ŚPV (set in bold in the San- skrit text above). In contrast to what might be expected, however, Dāruka’s reference to sages living on the summits of mount Raivataka lacks any specific connota- tion of asceticism. On the contrary, the charioteer emphasizes that the moun- tain is pleasant by stating that the sages take special advantage of the cool clouds. On one hand, the clouds have already lost their water so that they do not make the sages too cold.52 On the other hand, the clouds are ultimately curtains that shelter the ascetics, possibly from the excessive heat of the sun- light. It appears that due to the cool wind and the clouds, asceticism on mount Raivataka is less painful than elsewhere. Just like many other verses in Dāru- ka’s monologue, stanza 4.54 creates a poetic sentiment of pleasure and ease. Unlike stanza 4.55, it neither refers, nor even alludes to any specific yogic soteriological concept. The same is true for the second stanza in Dāruka’s monologue referring to ascetics, i.e., stanza 4.64, which reads as follows: prāleyaśītam acaleśvaram īśvaro ’pi sāndrebhacarmavasanābharaṇo ’dhiśete | sarvartunirvṛtikare nivasann upaiti na dvandvaduḥkham iha kiñcid akiñcano ’pi || 64 || (ŚPV 4.64, part 1, p. 150; meter: Vasantatilakā). 52 Andrey Klebanov informed me (personal communication, September 2015) that the stanza ŚPV 4.54 possibly contains an allusion to stanza 1.5 of Kālidāsa’s Kumāra- saṃbhava. There, Kālidāsa describes the Himalaya as a place where perfected ascetics (siddhas) move to the sunny summits of the mountain range in order to avoid the rain in lower regions. By stating that sages are always at ease on Raivataka, Māgha may have implied that this mountain range was better suited to ascetics than the Himalaya. 44 Philipp A. Maas Even the Lord dwells on the snow-cold Lord of Mountains wearing a vesture of warming elephant-hide,53 whereas no renouncer is ever pained by the pair of opposites54 when he lives on this bliss-bringer throughout the year. This stanza contains again a number of alliterations (anuprāsa), as is cha- racteristic for the second stanzas in the triads of the fourth chapter of Māgha’s poem. The first repetition concerns the word “Lord,” referring to the Himalaya and the god Śiva (…eśvaram īśvaro). The second repetition in- volves the two indefinite pronouns in a sequence …a kiñcid akiñcano. On the level of meaning, the stanza contains a slightly ironic mocking of Śiva, who in other literary works frequently appears as the prototype of ascetics free of needs. Māgha, however, portrays Śiva as requiring warm clothes when he lives on the top of the Himalaya, whereas even ordinary ascetics on mount Raivataka never experience any needs at all. The message of the stanza con- sists again in a praise of the mountain as a pleasant location well worth being visited by Kṛṣṇa; it does not allude to any yoga-specific soteriological con- cepts. If one considers that Māgha consistently construes the characters in his poem in such a way that their speech mirrors their general character, also discussed in Bronner’s and McCrea’s recent article,55 stanza 4.55 does not fit well into the poem. In neither the ŚPV nor the MBh is Dāruka, Kṛṣṇa’s charioteer, related in any way to Sāṅkhya or Yoga theories of soteriological practices.56 An additional unusual feature of stanza 4.55 is that it does not refer to any sensually appealing quality of the mountain. This contradicts what is to be expected, because the very purpose of Dāruka’s monologue is making a stay at the mountain palatable for Kṛṣṇa by highlighting the various positive fea- tures of the place.57 53 Wearing an elephant hide is an attribute of Śiva in his appearance as the killer of the elephant-demon (gajāsurasaṃhāramūrti); see Haussig 1984: 166. 54 “The pair of opposites” (dvandva-) refers to troublesome sensations like heat and cold, hunger and thirst, etc. The PYŚ refers to this concept in the context of posture practice in section 2.48, which states that “[b]ecause of that (mastery of posture), one is not hurt by the pairs of opposites (sūtra 2.48). Because one masters the postures, one is not overcome by the pairs of opposites such as heat and cold” (tato dvandvānabhighātaḥ (sūtra 2.48) śītoṣṇādibhir dvandvair āsanajayān nābhibhūyate, iti.). 55 Bronner and McCrea 2012: 451. 56 See Sörensen 1904: 234b. 57 This can be concluded from the fact that Māgha expressedly states that Dāruka’s mo- nologue made Kṛṣṇa want to visit the mountain: “Hearing thus the true and beautiful  From Theory to Poetry 45 On the basis of these considerations one might suspect that the stanza ŚPV 4.55 is an interpolation in Māgha’s poem. However, the fact that the whole chapter is structured into triads of stanzas makes this conclusion im- possible, unless one is willing to argue that the stanza under discussion must have been a substitution for a different and unknown stanza at an unknown point of the transmission. Such an argument could, however, only be made on the basis of manuscripts that actually present an alternative to stanza ŚPV 4.55. For the time being, it seems that ŚPV 4.55 is an odd but probably ge- nuine part of Māgha’s composition. This unusual stanza adds an aspect to the description of the mountain that is not covered by any other stanzas of the ŚPV, namely its being the site of yogic practice leading to liberation from the bonds of rebirth.58 In this way, the description of a beautiful and charming mountain is supplemented by an element of ascetic value, or, in other words, the charming mountain is also a venue with a mystical flavor. Irrespective of whether one believes the author of this stanza made a lucky choice from an aesthetic point of view when introducing this additional characteristic of the mountain to the poem, it appears that the reuse of texts and concepts from the PYŚ in stanza 4.55 served a number of interrelated literary purposes. By explicitly mentioning that Raivataka was the place where yogis actually achieve liberation from the cycle of rebirths, the author implicitly identified the mountain as the place of fulfilling the religious aspi- rations of yogis. In this way, he created the notion of what might tentatively be called the sacredness of mount Raivataka. Moreover, the poet created sup- port for the claim that the PYŚ is the authoritative work on the theory of prac- tice of Yoga leading to liberation, possibly because this was the general view at his time and in the social circles to which he belonged. Accordingly, the śāstra and the notion of the mountain as a sacred space lend literary autho- rity, prestige and religious power to each other. The ŚPV contributed on one hand to reinforcing the recognition of the PYŚ as the authoritative work on words of his driver that were, as it were, incomparably sweet, he then, at their end, thus longed to live for a long time on that mountain that was dressed in a dress of the rows of its trees. (itthaṃ giraḥ priyatamā iva so ’vyalīkāḥ śuśruva sūtanayasya tadā vya- līkāḥ / rantum nirantaram iyeṣa tato ’vasāne tāsāṃ girau ca vanarājipaṭaṃ vasāne // 1 //) ŚPV 5.1, p. 153. 58 Mallinātha, the 15th century commentator on the ŚPV, highlighted this aspect of the stanza by saying: “The intention of the stanza is stating that this mountain is not only a place for sensual enjoyment, but also a place for achieving liberation.” (na kevalaṃ bhogabhūmir īyam, kiṃtu mokṣakṣetram apīti bhāvaḥ. Durgāprasāda and Śivadatta 1927: 108.) 46 Philipp A. Maas Yoga. On the other hand, the author of stanza 4.55 participated in shaping the public imagination of mount Raivataka as a place for fulfilling yogic practice. Quite interestingly, in the course of history, the mountain actually became a place of religious worship. Today, the mountain group harbors Jaina temples, of which the oldest can be dated to the 12th century, along with temples dedi- cated to the worship of Gorakhnāth and Dattātreya, both prominent figures in medieval forms of yoga (see Fig. 1).59 Figure 1: View of the Dattatreya Temple of Girnar (detail). Source: Sachinvenga <https://en.wikipedia.org/wiki/Girnar> (CC BY-SA 4.0). 4.2 The stanza Śiśupālavadha 14.62 The second stanza of the ŚPV containing clearly identifiable instances of reuse from the PYŚ is stanza 14.62. This stanza appears at the beginning of a speech of praise that Bhīṣma holds for Kṛṣṇa in order to introduce him as the only suitable guest of honor for the rājasūya because of his divine nature. The stanza reads as follows: sarvavedinam anādim āsthitaṃ dehinām anujighṛkṣayā vapuḥ | kleśakarmaphalabhogavarjitaṃ puṃviśeṣam amum īśvaraṃ viduḥ || (ŚPV 14.62, part 2, p. 123; meter: Rathoddhatā). 59 See Rigopoulos 1998: 98.  From Theory to Poetry 47 The yogis know him to be God, a special subject who doesn’t expe- rience afflictions, karma and its results. He, who is omniscient and without predecessor, embodied himself due to his wish to favor the embodied beings (or souls). 4.2.1 The reuse of the Pātañjalayogaśāstra in Śiśupālavadha 14.62 This stanza reuses conceptions found in the īśvara section of Patañjali’s work, more specifically, in PYŚ 1.24–25. Pādas c and d of ŚPV 14.62 are virtually a metrical paraphrase of sūtra 1.24. kleśakarmavipākāśayair aparāmṛṣṭaḥ puruṣaviśeṣa īśvaraḥ (PYŚ 1.24,2). God is a special subject who is unaffected by afflictions, kar- ma, its ripening and mental dispositions. In his reuse, Māgha substituted the word puruṣa “subject” from the sūtra with its synonym puṃs. Moreover, he changed the formulation -vipākāśayair aparāmṛṣṭaḥ “unaffected by ripening and mental dispositions” to the less technical but similar formulation -phalabhogavarjitaṃ “without experience of the results.” In addition, if interpreted through the lens of the PYŚ, it appears that Māgha reused conceptions occurring in the bhāṣya part of PYŚ 1.25 in pādas a and b of ŚPV 14.62. This section of the PYŚ deals, among other things, with the omniscience of God, which is established on the basis of the argu- ment that every increasable faculty must at some point reach a peak. This applies also, according to Patañjali, to knowledge, which reaches its peak in the state of omniscience. The argument concludes as follows: yatra kāṣṭhāprāptir jñānasya sa sarvajñaḥ (PYŚ 1.25,4f.). An omnis- cient person is somebody in whom the utmost limit of knowledge is reached. Apparently, Māgha changed the expression sarvajña from the PYŚ to the quasi-synonym sarvavedin, which, however can also be read to mean “know- ing all Vedas.” The same section of the PYŚ also deals with the problem of how God, a transcendental being and an eternally liberated subject, may be effective in the world. Patañjali presented the following solution: tasyātmānugrahābhāve ’pi bhūtānugrahaḥ prayojanam. “jñānadhar- mopadeśena kalpapralayamahāpralayeṣu saṃsāriṇaḥ puruṣān ud- dhariṣyāmi,” iti. tathā coktam – “ādividvān nirmāṇacittam adhiṣṭhāya kāruṇyād bhagavān parama ṛṣir āsuraye jijñāsamānāya provāca,” iti 48 Philipp A. Maas (PYŚ 1.25,8–11). Although he is beyond help for himself, helping living beings is his motive: “At the dissolutions at the end of an eon and at the Great Dissolutions of the universe, I shall rescue the sub- jects from the cycle of rebirth by teaching them knowledge and dhar- ma.” And in the same way it has been authoritatively stated: “The first knower, the venerable ultimate seer, assuming a mind of magical transformation, out of compassion taught Āsuri when he desired to know.”60 This passage explains the efficacy of God in the world by assigning to him the role of a primordial teacher. Periodically, at the beginning of each re- creation of the universe, he assumes a mind (citta) in order to help suffering beings by instructing them in the teaching of Yoga, which enables these be- ings to achieve liberation from the circle of rebirths.61 Māgha reused this passage in pādas a and b of stanza 14.62 by adapting the terminology as well as specifically yogic theorems to the needs of his poetry. He alluded to the yogic teaching of God, the original knower of yoga (ādividvān) who periodically re-disseminates the teaching of Yoga at the beginning of each re-creation, by simply stating that God is without a prede- cessor (anādi). Moreover, the poet adapted the specific yogic idea of God as- suming “a mind of magical transformation” (nirmāṇacitta) by stating in a much less technical tone – and in accordance with the needs of the episode that he depicts – that God assumed a body (āsthitaṃ vāpus). Similar con- siderations may have also lead Māgha to stating that God assumed a body due to his intention of helping embodied beings or souls (dehinām anujighṛ- kṣayā), instead of sticking to the yogic concept of God helping subjects (puruṣas) that are entangled in the cycle of rebirths (saṃsārin) due to the altruistic motive (prayojana) of helping beings (bhūtānugraha). These adap- tations did not only increase the intelligibility of the stanza, they also led to the creation of a metrical composition containing one of the previously noted stylistic features of Māgha’s poetry, structured repetitions (in this case of ve- dinam anādim, vedinam … dehinām, vapuḥ … viduḥ). Nevertheless, the simi- larities of Māgha’s stanza to the passages in the PYŚ discussed above clearly indicate that Māgha consciously reused Patañjali’s work. 60 In his commentary on the PYŚ, Vācaspatimiśra ascribed this fragment to the Sāṅkhya teacher Pañcaśikha (see Āgāśe 31.16). However, as Chakravarti (1951: 115f.), Ober- hammer (1960: 81f.), and others have argued, this ascription is almost certainly ahistoric. 61 Cf. Maas 2009: 277.  From Theory to Poetry 49 4.2.2 Śiśupālavadha 14.62 in context As mentioned above, the stanza ŚPV 14.62 is part of a speech of praise that Bhīṣma holds for Kṛṣṇa in order to introduce him as the only suitable guest of honor for the rājasūya. It is this speech that causes king Śiśupāla’s outbreak of rage, which in turn leads to his reviling speech against Kṛṣṇa in the next chapter of the ŚPV. This is the passage, as discussed above, that is transmit- ted in two versions in different recensions of the ŚPV. In Bhīṣma’s speech, which consists for the most part of the twenty-nine stanzas providing an ex- tended version of MBh 2.33.28–29, the following four stanzas precede stanza 14.62: atra caiṣa sakale ’pi bhāti māṃ praty aśeṣaguṇabandhur arhati | bhūmidevanaradevasaṅgame pūrvadevaripur arhaṇāṃ hariḥ || 58 || martyamātram avadīdharad bhavān mainam ānamitadaityadānavam | aṃśa eṣa janatātivartino vedhasaḥ pratijanaṃ kṛtasthiteḥ || 59 || dhyeyam ekam apathi sthitaṃ dhiyaḥ stutyam uttamam atītavākpatham | āmananti yam upāsyam ādarād dūravartinam atīva yoginaḥ || 60 || padmabhūr iti sṛjañ jagad rajaḥ sattvam acyuta iti sthitaṃ nayan | saṃharan hara iti śritas tamas traidham eṣa bhajati tribhir guṇaiḥ || 61 || (ŚPV 14.58–61, part 2, p. 122f.; meter: Rathoddhatā). 58 I see clearly that here in the whole congregation of gods on earth (brāhmaṇas) and gods of men (kings) Hari, the enemy of the previous gods, the abode of all good qualities, is worthy of this honor. 59Do not consider Him, who subdued the Daityas and Dānavas, a mere mortal! He is a part of the Creator, and although he is beyond the world, he abides in every being. 60The yogis state that he’s unique: the mentally inaccessible object of their meditation. They say that he’s perfect: the inexpressible object of their praise, the object of their diligent venera- tion, remaining in remotest distance. 61Through the three material qualities, this God is threefold: When he creates the world as rajas, he 50 Philipp A. Maas is Brahmā. When he maintains as sattva, he is Viṣṇu. When he de- stroys as tamas, he is Śiva. A closer reading of the opening stanzas of Bhīṣma’s monologue indicates that stanza 14.62 is in perfect agreement with its context. Bhīṣma mentions that Kṛṣna is not an ordinary human being but a divinity. He alludes to Viṣṇu’s victory over different classes of demonic beings or Asuras, before he ad- dresses aspects of Vaiṣṇava theology such as the paradoxically immanent and transcendent nature of Viṣṇu. In stanza 61, Māgha even lets Bhīṣma depict Kṛṣṇa as encompassing the three important deities of Brahmā, Śiva and Viṣṇu in their respective functions of creator, maintainer and destructor of the world by drawing on the sāṅkhyistic concept of the three qualities or con- stituents of primal matter, sattva, rajas and tamas. Even more important in the present context is stanza 60, which introduces yogis as a group of devo- tees to Viṣṇu-Kṛṣṇa, for whom God is the object of their meditation. It there- fore does not come as a surprise that stanza 14.62 refers more technically to Yoga theology as is outlined in the PYŚ. This in turn corresponds quite nicely to the literary figure of Bhīṣma, who also in the MBh delivers yoga- and sāṅkhya-related teachings. On the whole, the stanza ŚPV 14.62, unlike the previously discussed stanza ŚPV 4.55, is well integrated into the poem. Although the two stanzas differ from each other with regard to the degree to which they fit into their respective contexts, the purposes and the methods of the two cases of adaptive reuse are similar. In stanza 4.55, the poet identi- fied mount Raivataka as the place where yogis actually reach their aim of spiritual liberation. In this way, he had, on the one hand, reinforced the notion of the sacredness of the mountain. On the other hand, he had supported the claim of the PYŚ as the authoritative exposition of the practice of Yoga by identifying a geographical location where the aim of Yoga was actually reached. In the case of stanza 14.62, Māgha reused the PYŚ in order to rein- force the notion of Kṛṣṇa being a divine incarnation by identifying him with the unnamed transcendental God of classical Yoga theology. On the other hand, he appropriated the PYŚ, which Patañjali had consciously created as a trans-sectarian work, for his own project of venerating Viṣṇu. Moreover, by making Kṛṣṇa the high god of Yoga theology, Māgha even turned the PYŚ virtually into a work of Vaiṣṇava theology.62 62 The Kashmiri poet Ratnākara composed his Haravijaya in praise of Śiva in ca. 830 CE (according to Sanderson 2007: 425). In stanza 6.21 he reused virtually the same concepts of the PYŚ as ŚPV 14.62. In this way, he appropriated the PYŚ as a work of Śaiva theology. My thanks to Andrey Klebanov for drawing my attention to this parallel case of adaptive reuse of the PYŚ.  From Theory to Poetry 51 4.3 The passage Śiśupālavadha 1.31–33 The final passage that will be discussed in this chapter is a weak case of reuse; it might equally be interpreted as a reference or a strong allusion to the PYŚ. It appears in the first canto of the ŚPV, where Māgha sets the stage for the remaining part of the poem. Here, the heavenly seer Nārada arrives as a burning flame from the sky at Kṛṣṇa’s home, where the latter welcomes the divine ascetic with due respect before enquiring about the purpose of his visit. The poem continues in the following way: iti bruvantam tam uvāca sa vratī na vācyam itthaṃ puruṣottama tvayī | tvam eva sākṣātkaraṇīya ity ataḥ kim asti kāryam guru yoginām api || 31 || udīrṇarāgapratirodhakaṃ janair abhīkṣṇam akṣuṇṇatayātidurgamam | upeyuṣo mokṣapathaṃ manasvinas tvam agrabhūmir nirapāyasaṃśrayā || 32 || udāsitāraṃ nigṛhītamānasair gṛhītam adhyātmadṛśā kathaṃcana | bahirvikāraṃ prakṛteḥ pṛthag viduḥ purātanaṃ tvāṃ puruṣāṃ purāvidaḥ || 33 || (ŚPV 1.31–33, part 1, p. 19–20; meter: Vaṃśastha) 31 The sage replied to him, who had spoken thus: “Oh Highest Being, you may not speak like this. As even yogis have to visualize only you, which task could be more important for me?”32 For the wise man who wants to reach the path to liberation that is blocked by excited craving and inaccessible for ordinary people, because it remains constantly un- practiced, you are the final destiny that shelters without ill [(like) a far-away land to which only one liberating road leads, a road that is extremely difficult to travel, because robbers whom the people cannot drive away lurk there with excited desires]. 33With controlled minds the wise men of old realized that you are the ancient, totally passive subject. By seeing their inner self, they grasped with effort that you are different from matter and beyond its modifications. In these three stanzas Nārada introduces Kṛṣṇa to the audience of the poem. Initially, the divine seer addresses Kṛṣṇa with the term puruṣottama “Highest Being,” which can be understood as a general reference to the fact that Kṛṣṇa is the incarnation of God Viṣṇu. However, the term can also be understood as 52 Philipp A. Maas a technical term of Sāṅkhya-Yoga designating a transcendental subject, the faculty of pure consciousness (puruṣa), of which God (īśvara) is an ideal form.63 It is this meaning that Nārada alludes to when he states that “even yogis have to visualize only you” (tvam eva sākṣātkaraṇīya … yoginām api), because it is the aim of yogis practicing theistic meditation to realize the fundamental identity of God, conceptualized as an eternally liberated subject (puruṣa), and their own individual subject.64 This interpretation suggests itself even more if one considers the wording of the stanza ŚPV 1.33, in which the term puruṣa is clearly used as a technical term of Sāṅkhya-Yoga to refer to the faculty of consciousness that is ontologically different from (and opposed to) matter (prakṛti) and its modifications. An additional reference to the PYŚ is the expression adhyātmadṛśā “by means of the sight of the inner self” in stanza 1.33b, which is a conceptional parallel to PYŚ 1.29. This passage describes the result of theistic yogic me- ditation in the following way: kiṃcāsya bhavati tataḥ pratyakcetanādhigam[aḥ] … (sūtra 1.29). … svapuruṣadarśanam apy asya bhavati: “yathaiveśvaraḥ śuddhaḥ pra- sannaḥ kevalo ’nupasargas tathāyam api buddheḥ pratisaṃvedī ma- dīyaḥ puruṣaḥ,” ity adhigacchatīti (PYŚ 1.29,1–5). Moreover, from this yogic meditation the yogi acquires the realization of his inner consciousness (sūtra 1.29). He even acquires a vision of his own subject. He realizes: “As God is pure, clear, alone and free from trouble, so also is my subject here that experiences its mental capacity.”65 The two key terms of this passage are pratyakcetanādhigama “realization of inner consciousness” from the sūtra and its paraphrase svapuruṣadarśana “the vision of one’s own subject,” because adhyātmadṛś “seeing one’s inner self” that Māgha used in ŚPV 1.33b could be a synonym of these two com- pounds, used in order to describe the means by which the yogis of old rea- lized the ontological status of God as being different from matter. The fact that for Māgha a yogi practicing theistic meditation attains knowledge of the ontological status of God by seeing his inner self indicates that Māgha knew 63 See above, sections 1 and 4.2.1. A stanza occurring in Viṣṇupurāṇa 6.6.2 that Patañjali quotes in PYŚ 1.28,5–6 uses the expression para ātman “Highest Self,” a quasi-syn- onym of puruṣottama, to refer to God. 64 For a detailed exposition of theistic yogic meditation, see Maas 2009, especially pp. 276–280. 65 Translation based on Maas 2009: 279.  From Theory to Poetry 53 a form of theistic meditation similar to, or even identical with, the one taught in the PYŚ. 4.4 The reception of Māgha’s reuse in Vallabhadeva’s Antidote Vallabhadeva, who wrote his Antidote “in the first half of the tenth century” in Kashmir,66 fully recognized the reuse of the PYŚ in stanza 4.55, and he provided it with the most comprehensive commentary of a single stanza of the whole chapter. His gloss highlights the reuse of PYŚ 1.33 by quoting the sūtra almost verbatim and paraphrasing the bhāṣya passage.67 Moreover, the Kashmiri commentator explained the concept of afflictions (kleśa) with a brief summary of PYŚ 2.3–2.10.68 In addition, he provided a pregant des- cription of the yogic path to liberation. One of the few instances in which Vallabhadeva deviated from the PYŚ, to which he never referred by name, is his quotation of sūtra 2.29. This quote contains the already mentioned list of eight ancillaries. Possibly due to a slip of memory, the commentator pre- sented the last three ancillaries as “meditation, fixing the mind and absorp- tion” (dhyāna-dhāraṇā-samādhi), whereas the original sequence in the PYŚ is “fixing the mind, meditation and absorption” (dhāraṇā-dhyāna-samādhi). On the whole, however, Vallabhadeva demonstrated his detailed knowledge of Yoga philosophy and a clear understanding of Māgha’s reuse. At the end of his commentary on this stanza, however, the 10th-century commentator stated that his exposition was based on the explanations of his teacher Prakāśavarṣa.69 He added that the “understanding of this stanza (?) cannot exist in detail without knowledge derived from personal experience (anubhava).”70 This may imply that Vallabhadeva felt unable to explain the 66 Goodall and Isaacson 2003: xvii. On Vallabhadeva, see also Goodall and Isaacson 2003: xv–xxi. 67 Vallabhadeva paraphrases sūtra 1.33 (maitrīkaruṇāmuditopekṣāṇāṃ sukhaduḥkhapuṇ- yāpuṇyaviṣayāṇāṃ bhāvanātaś cittaprasādanam) with maitrīkaruṇāmuditopekṣāṇām sukhadu[ḥ]khapuṇyāpuṇyaviṣayāṇām abhyāsāc cetaḥprasādanaṃ cittaparikarma (Kak and Shastri, part 1, p. 147.11f.). 68 See Kak and Shastri, part 1, p. 147.16–19. 69 Prakāśavarṣa, Vallabhadeva’s teacher, was the author of a commentary (Laghuṭīkā) on the Kīrātārjunīya of Bhāravi. Andrey Klebanov, who is working on a critical edition of Prakāśavarṣa’s commentary, was kind enough to inform me (email 25 September 2016) that Prakāśavarṣa referred to Pātañjala Yoga in his commentary on Kirātārjunīya 3.26. 70 śrutvā prakāśavarṣāt tu vyākhyātaṃ tāvad īdṛśam. viśeṣatas tu naivāsti bodho ’trānu- bhād ṛte, iti (Kak and Shastri 1935, part 1, p. 147.22). In an email (4 November 2013) Dominic Goodall was kind enough to draw my attention to Goodall and Isaacson 2003: liii, where the two authors highlight the fact that Vallabhadeva “occasionally concedes that the poem takes him into areas of knowledge that are beyond his experience.…” In 54 Philipp A. Maas stanza in every detail, even with the help of his teacher Prakāśavarṣa. It would, however, be hazardous to draw any conclusions from Vallabhadeva’s statement as to the degree to which the PYŚ was known in Kashmir during Vallabhadeva’s lifetime. At least some circles of Kashmiri scholars knew the PYŚ quite well. This can be concluded from the fact that Abhinavagupta, the famous polymath who probably lived in Kashmir slightly later than Valla- bhadeva, and Rāmakaṇṭha (950-1000 CE) quoted the PYŚ repeatedly in their respective works.71 Also the poet Ratnākara, who lived approximately one hundred years before Vallabhadeva, reused the PYŚ in his own poetry.72 In his commentary on stanza 14.62, Vallabhadeva indicated clearly that he knew a distinct group (of theologians?) committed to the Yoga of Patañjali called “Pātañjalas.”73 Moreover, the commentator revealed that he was aware of the fact that this group had a peculiar exposition of their teaching, i.e., a śāstra.74 The same awareness of the PYŚ is also reflected in Vallabhadeva’s commentary on ŚPV 1.33, where he referred his reader for more information to the PYŚ by saying “etat tu sarvaṃ yogaśāstrād eva sujñānam.”75 This suggests that Vallabhadeva regarded Patañjali’s composition, just as Māgha had done before him, as a unified whole rather than a sūtra work together with a later commentary. Moreover, he apparently assumed that his reader would be able to access this work in some way or another, that is, either from memory or in writing. support of this, the two authors quoted the passage from Vallabhadeva’s commentary on ŚPV 4.55 cited above. Moreover, they referred to Vallabhadevas’s commentary on ŚPV 12.8, where the commentator admits that as a Kashmiri he does not know much about chariots. 71 For Abhinavagupta, see Maas 2006: 111. The dating of Rāmakaṇṭha follows Watson, Goodall and Sarma 2013: 15. On Rāmakaṇṭha’s references to the PYŚ see Watson, Goodall and Sarma 2013: 447–450. 72 See above, n. 62. 73 Vallabhadeva glossed the verb viduḥ “they know” in pāda d with pātañjalā avidan “The followers of Patañjali knew.” (Kak and Shastri 1935, part 2, p. 124.8). 74 “And it has been stated: ‘God is a special subject that is unaffected by afflictions, karma, its ripening and mental dispositions (YS 1.25),’ and this can easily be unders- tood from the [explanations in] the authoritative exposition (śāstra). If however, I would investigate the matter here, my work would become overloaded” (uktaṃ ca – “kleśakarmavipākāśayair aparāmṛṣṭaḥ puruṣaviśeṣa īśvaraḥ” ity etac ca tacchāstrād subodhyam, iha tu vicāre granthagauravaṃ syāt”) (Kak and Shastri 1935, part 2, p. 124.14f.). 75 Kak and Shastri 1935, part 1, p. 20, l. 14.  From Theory to Poetry 55 5 Conclusions The previous sections have examined how Māgha reused concepts and text passages of the PYŚ at two points in his ŚPV. The nature of the reuse makes it virtually impossible that Māgha reused a work different than the PYŚ. At a third point, Māgha merely referred to characteristic teachings of the PYŚ, teachings which he could theoretically also have known from a different yoga work that is today lost. The detailed nature of Māgha’s reuse of – and refer- ences to – Pātañjali’s work indicates, in any case, that the poet was thorough- ly familiar with the PYŚ. Apparently, Māgha expected that at least some of his audience was acquainted with the PYŚ to a similar degree, because other- wise the adaptive reuse would not have been recognizable.76 This finding suggests that the PYŚ was widely known in educated circles extending beyond specialists in Gujarat – if this region was indeed the home of the poet – at the time of the composition of the ŚPV, which was probably around the middle of the 8th century CE.77 The respective effects that the poet created with these two cases of adap- tive reuse of the PYŚ were similar to each other. They are related to the fact that Māgha expected his audience to share his view of the PYŚ as a presti- gious work at least to some degree. By reusing the PYŚ, Māgha reinforced its reception as the authoritative work on Yoga par excellence among the edu- cated audiences of his poem. Moreover, the poet transferred the prestige of the śāstra to the object of his poetical description, i.e., to a sacred mountain in stanza 4.55 and to Kṛṣṇa in stanza 14.62. This in turn may have contri- buted to the reception of the ŚPV as a prestigious poetic composition. Māgha’s reuse was recognized even about two hundred years after the composition of the ŚPV, that is, in the first half of the 10th century in Kash- mir. This indicates the PYŚ was known as an authoritative work on Yoga even outside yogic or philosophical circles for several centuries after its com- position.78 In fact, the PYŚ played an important role throughout South Asian cul- tural, philosophical and religious history. Already during the first hundred 76 On the multiple purposes of adaptive reuse, see the introduction to the present volume. 77 See above, section 2. 78 A detailed analysis of the reception history of the stanzas ŚPV 4.55 and 14.62 in later commentaries, as for example in Mallinātha’s Sarvaṃkaṣā (15th century) and in the more than fifty-six additional commentaries on Māgha’s poem listed in the NCC, could cast more light on the reception history of the PYŚ in pre-modern and early modern South Asia. Due to limitations of time and space, this work must be left to another oc- casion. 56 Philipp A. Maas years after its composition, the work emerged as the authoritative exposition of philosophical yoga. This is indicated by numerous references to the PYŚ, and quotations from it, from the fifth century onwards in various genres of South Asian literature.79 The first chapter of the PYŚ alone is quoted in more than twenty premodern, mainly philosophical Sanskrit works (Maas 2006: 111). Patañjali’s work was also well known in Buddhist circles. A mediaeval Singhalese chronicle provides the legendary account that the eminent fifth- century Buddhist commentator and author Buddhaghosa was a follower of Patañjali before he converted to Buddhism and emigrated to Sri Lanka (War- ren and Kosambi 1950: ix–xii and Hinüber 1997: 102, § 207). Additional tes- timony for the favourable reception of the PYŚ comes from the northwest of South Asia. There, the eleventh-century Perso-Muslim scholar al-Bīrūnī drew heavily on Patañjali’s work when he described the religion and culture of the people in his India.80 Al-Bīrūnī also rendered the PYŚ into Arabic.81 The virtually continuous relevance of Patañjali’s work in premodern South Asian philosophical and religious history is also indicated by the fact that the PYŚ became the subject of three commentaries: (1.) The Pātañjala- yogaśāstravivaraṇa (Vivaraṇa) possibly from the eighth century,82 by a certain Śaṅkara, (2.) the Tattvavaiśāradī or Pātañjalayogaśāstravyākhyā by the famous polymath Vācaspatimiśra I, who flourished around 950–1000 (Acharya 2006: xxviii), and (3.) the late sixteenth-century Yogavārttika by Vijñānabhikṣu (Nicholson 2010: 6). Thus, the various interpretations, re- interpretations and critical responses to the PYŚ that were produced over the last approximately 1600 years make the PYŚ an extremely important source for research in the history of South Asian philosophy and religion. 79 The earliest quotation from the PYŚ known to me occurs in the earliest commentary (vṛtti) on the Vākyapadīya of Bhartṛhari (ca. 450–510), which quotes a bhāṣya passage from PYŚ 2.6 in commenting on Vākyapadīya 2.31 (p. 67). Whether the vṛtti is an autho-commentary of Bhartṛhari or whether it was composed by one of Bhartṛhari’s students, is still a matter of debate in indological scholarship. 80 See the lists of al-Bīrūnī’s sources provided by Sachau (1888: I: xxxix–xl) and Shastri (1975). 81 See Maas and Verdon (forthcoming), who argue that al-Bīrūnī’s Kitāb Pātanğal is a free rendering of the PYŚ into Arabic and not at all a more or less literal translation of the “Yoga Sutra” together with an unknown commentary. 82 There are basically two arguments in favour of an early date of the Vivaraṇa. First, the Vivaraṇa does not refer to any author later than Kumārila, who lived in the 7th c. (Halbfass 1983: 120), and second, it can be demonstrated that the textual version of the PYŚ commented upon by the author of the Vivaraṇa goes back to an early stage of the transmission (Maas 2006: lxxii).  From Theory to Poetry 57 References Primary Sources and Abbreviations Antidote Vallabhadeva, Saṃdehaviṣauṣadhi, see Kak and Shastri 1935 Haravijaya Ratnākara, Haravijaya, see Dvivedī and Paraba 2005 Kāvyaprakāśa Mammaṭa, Kāvyaprakāśa, see Dwivedi 1966 Kumārasaṃbhava Kālidāsa, Kumārasaṃbhava, see Murti 1980 MBh Mahābhārata. The Mahābhārata. For the First Time Crit. ed. by V. S. Sukthankar, S. K. Belvalkar et al. 20 vols. Poona: Bhandarkar Oriental Research Institute, 1933 (1927)–1966. NCC New Catalogus Catalogorum, see Dash 2014 Pātañjayogaśāstravivaraṇa = Pātañjala-Yogasūtra-Bhāṣya-Vivaraṇa, see Rama Sas- tri and Krishnamurthi Sastri 1952 PYŚ Pātañjalayogaśāstra. For references to chapter 1, see Maas 2006. For references to other chapters, see Āgāśe 1904 Sarvaṃkaśā Mallinātha, Sarvaṃkaśā on the ŚPV, see Durgāprasāda and Śivadatta 1927 ŚPV Māgha, Śiśupālavadha, see Kak and Shastri 1935 Tattvārthasūtra Umāsvati, Tattvārthasūtra, see Premchand 1904 Vākyapādīya [Bhartṛhari, Vākyapadīya.] Vākyapādīyam. (Part II: Vāk- yakāṇḍam) by Bhartṛhari: With the Commentary of Puṇ- yarāja and Ambākartri by Raghunātha Sarmā. Varanasi: Sampurnanand Sanskrit Vishvavidyalaya, 1980. Sarasvatī- bhavana-Granthamālā 91. Secondary Sources Acharya 2006 Acharya, Diwakar. Vācaspatimiśra’s Tattvasamīkṣā: The Earliest Commentary on Maṇḍanamiśra’s Brahmasiddhi. Stuttgart: Steiner 2006. Nepal Research Center Publica- tions 25. Āgāśe 1904 Āgāśe, Kāśināth Śāśtri, ed. Vācaspatimiśraviracitaṭīkā- saṃvalitaVyāsabhāṣyasametāni Pātañjalayogasūtrāṇi, ta- thā BhojadevaviracitaRājamārtaṇḍābhidhavṛttisametāni Pātañjalayogasūtrāṇi <Sūtrapāṭhasūtravarṇānukramasū- cībhyāṃ ca sanāthīkṛtāni,> … tac ca H. N. Āpaṭe ity ane- na … prakāśitam, Puṇyākhyapattana [= Pune] 1904. Ānandāśramasaṃskṛtagranthāvaliḥ 47. Both 2003 Both, Leo. Hemacandras Kāvyānuśāsana: Kapitel 1 und 2 . Eine Einführung in die Grundlagen des indischen 58 Philipp A. Maas Dramas und der indischen Poesie. Wiesbaden: Harras- sowitz, 2003. Drama und Theater in Südasien 2. Bronner and McCrea 2012 = Bronner, Yigal, and Lawrence McCrea. “To Be or Not to Be Śiśupāla: Which Version of the Key Speech in Mā- gha’s Great Poem Did He Really Write?” Journal of the American Oriental Society 132.2 (2012): 427–455. Bronner, Shulman and Tubb 2014 = Bronner, Yigal, David Shulman and Gary A. Tubb, eds. Innovations and Turning Points: Toward a His- tory of Kāvya Literature. Delhi: Oxford University Press, 2014. South Asia Research Ser. Cardona 1976 Cardona, George. Pāṇini: A Survey of Research. The Hague: Mouton, 1976. Trends in Linguistics. State-of-the- Art-Reports. Chakravarti 1951 Chakravarti, Pulinbihari. Origin and Development of the Sāṃkhya System of Thought. Calcutta: Metropolitan Print- ing and Publishing House, 1951. Colebrooke 1827 Colebrooke, Henry Thomas. “On the Philosophy of the Hindus. Part 1: Sánkhya.” Transactions of the Royal Asiatic Society 1 (1827): 19–43. Dash 2014 Dash, Siniruddha, editor-in-chief. New Catalogus Catalo- gorum: An Alphabetical Register of Sanskrit and Allied Works and Authors. Vol. 34. Śāla-Śīrṣopaniṣad. Chennai: University of Madras, 2014. Durgāprasāda – Śivadatta 1927 = Durgāprasāda and Śivadatta, eds. The Śiśupāla- vadha of Māgha: With the Commentary (Sarvaṅkashā) of Mallinātha. 9th ed. revised by Laxman Shāstrī Paṇasīkar. Bombay: Nirṇaya Sāgara Press, 1927. Dvivedī and Paraba 2005 = Dvivedī, Durgāprasāda and Kāśinātha Pāṇḍuraṅga Paraba, eds. The Haravijaya of Rājānaka Ratnākara: With the Commentary of Rājānaka Alaka and English Introduction. Delhi: Bharatiya Kala Prakashan, 2005. Dwivedi 1966 Dwivedi, R. C., ed. and transl. The Poetic Light: Kāvya- prakāśa of Mammaṭa. Vol. 1. Ullāsas 1–6: Text with Transl. and Sampradāyaprakāśinī of Śrīvidyācakravartin. Delhi: Motilal Banarsidass, 1966. Goodall and Isaacson 2003 = Goodall, Dominic and Harunaga Isaacson. The Raghu- pañcikā of Vallabhadeva: Being the Earliest Commentary on the Raghuvaṃśa of Kālidāsa. Vol. 1. Groningen: Eg- bert Forsten, 2003. Groningen Oriental Studies 17. Halbfass 1983 Halbfass, Wilhelm. 1983. Studies in Kumārila and Śaṅ- kara. Reinbek: Inge Wezler, 1983. Studien zur Indologie und Iranistik, Monographie 9. Haussig 1984 Haussig, Hans Wilhelm, ed. Götter und Mythen des indi- schen Subkontinents. Unter Mitarbeit von Heinz Bechert et  From Theory to Poetry 59 al. Stuttgart: Klett-Cotta, 1984. Wörterbuch der Mytho- logie 1.5. Hinüber 1997 Hinüber, Oskar von. A Handbook of Pāli Literature. First Indian ed. New Delhi: Munshiram Manohar, 1997. Hultzsch 1927 Hultzsch, Eugene. “Sāṃkhya und Yoga im Śiśupālava- dha.” Aus Indiens Kultur. Festgabe Richard von Garbe dem Forscher und Lehrer zu seinem 70. Geburtstag darge- bracht von seinen Freunden, Verehrern und Schülern. Ed. Julius von Negelein. Erlangen: Palm und Enke, 1927. 78– 83. Jacobi 1889 Jacobi, Hermann. “Bhāravi and Māgha.” Wiener Zeit- schrift für die Kunde des Morgenlandes 3 (1889): 121– 145. Rpt. in Bernhard Kölver, ed., Hermann Jacobi. Klei- ne Schriften. Teil 1. Wiesbaden: Franz Steiner, 1970. 447– 478. Glasenapp Stiftung 4,1. Kak and Shastri 1935 Kak, Ram Chandra and Harabhatta Shastri, eds. Māgha- bhaṭṭa’s Śiśupālavadha: With the Commentary (Sandeha- Viṣauṣadhi) of Vallabhadeva. [Rpt. of the ed. Srinagar: Kashmir Mercantile Press, 1935]. Delhi: Bharatiya Book Corporation, 1990. Kane 1914 Kane, Pandurang Vaman. “Bhāmaha, the Nyāsa and Mā- gha.” Journal of the Bombay Branch of the Royal Asiatic Soc. 23 (1914): 91–95. Kielhorn 1906 Kielhorn, Franz. “Epigraphic Notes,” Nachrichten der Kö- niglichen Gesellschaft der Wissenschaften zu Göttingen, phil.-hist. Kl. 1906: 143–146. Rpt. in Wilhelm Rau, ed., Franz Kielhorn, Kleine Schriften. Mit einer Auswahl der epigraphischen Aufsätze. Teil 1. Wiesbaden: Franz Stei- ner, 1969. 428–431. Glasenapp-Stiftung 3,1. Lienhard 1984 Lienhard, Siegfried. A History of Classical Poetry: Sans- krit - Pāli - Prakrit. Wiesbaden: Harrassowitz, 1984. A history of Indian literature 3.1. Maas 2006 Maas, Philipp André. Samādhipāda: Das erste Kapitel des Pātañjalayogaśāstra zum ersten Mal kritisch ediert = The First Chapter of the Pātañjalayogaśāstra for the First Time Critically Ed. Aachen: Shaker, 2006. Studia Indolo- gica Universitatis Halensis. Geisteskultur Indiens. Texte und Studien 9. Maas 2008 — “The Concepts of the Human Body and Disease in Classical Yoga and Āyurveda.” Wiener Zeitschrift für die Kunde Südasiens 51 (2008): 123–162. Maas 2009 — “The So-called Yoga of Suppression in the Pātañjala Yogaśāstra.” Yogic Perception, Meditation, and Altered States of Consciousness. Ed. Eli Franco in collaboration 60 Philipp A. Maas with Dagmar Eigner. Wien: Verlag der Österreichischen Akad. d. Wiss., 2009. 263–282. Sitzungsberichte der phil.- hist. Klasse 794 = Beiträge zur Kultur- und Geistesge- schichte Asiens 64. Maas 2013 — “A Concise Historiography of Classical Yoga Philo- sophy.” Historiography and Periodization of Indian Phi- losophy. Ed. Eli Franco. Vienna: Institut für Südasien-, Tibet- und Buddhismuskunde, 2013. 53–90. Maas 2014 — “Der Yogi und sein Heilsweg im Yoga des Patañjali.” Wege zum Heil(igen): Sakralität und Sakralisierung in hinduistischen Traditionen? Ed. Karin Steiner. Wiesbaden: Harrassowitz, 2014. 65–90. Maas and Verdon forthcoming = Maas, Philipp A. and Noémie Verdon. “On the Hermeneutics of Yoga in al-Bīrūnī’s Kitāb Pātanğal.” Eds. Karl Baier, Philipp A. Maas, Karin Preisendanz. Yoga in Transformation. Vienna: Vienna University Press. Maithrimurthi 1999 Maithrimurthi, Mudagamuwe. Wohlwollen, Mitleid, Freu- de und Gleichmut: Eine ideengeschichtliche Untersuchung der vier apramāṇas in der buddhistischen Ethik und Spi- ritualität von den Anfängen bis hin zum frühen Yogācāra. Stuttgart: Steiner, 1999. Alt- und neu-indische Studien 50. McCrea 2014 McCrea, Lawrence. “The Conquest of Cool. Theology and Aesthetics in Māgha’s Śiśupālavadha.” In: Bronner, Shul- man and Tubb 2014: 123–141. Murti 1980 Murti, M. Srimannarayana with assistance of Klaus L. Ja- nert, ed. Vallabhadeva’s Kommentar (Śāradā-Version) zum Kumārasambhava des Kālidāsa. Wiesbaden: Steiner, 1980. Verzeichnis der orientalischen Handschriften in Deutschland 20.1. Nicholson 2010 Nicholson, Andrew J. Unifying Hinduism: Philosophy and Identity in Indian Intellectual History. New York: Colum- bia University Press, 2010. Oberhammer 1960 Oberhammer, Gerhard. “The Authorship of the Ṣaṣṭitan- tram.” Wiener Zeitschrift für die Kunde Süd- und Ost- asiens 4 (1960): 71–91. Oberhammer 1977 — Strukturen yogischer Meditation: Untersuchungen zur Spiritualität des Yoga. Wien: Verlag der Österreichischen Akad. d. Wiss., 1977. Sitzungsberichte der phil.-hist. Klas- se 322 = Veröffentlichungen der Kommission für Spra- chen und Kulturen Südasiens 132. Olivelle 1998 Olivelle, Patrick. The Early Upaniṣads: Annotated Text and Translation. New York: Oxford UP, 1998. South Asia Research Ser.  From Theory to Poetry 61 Premchand 1904 Premchand, Mody Keshavlal, ed. Tattvârthâdhigamasūtra by Umâsvati … . Vol. 1,2. Calcutta: Baptist Mission Press, 1904. Bibliotheca Indica, New Ser. 1079. Rama Sastri and Krishnamurthi Sastri 1952 = Rama Sastri, Polakam Sri and S. R. Krishnamurthi Sastri, eds. Pātañjala-Yogasūtra-Bhāṣya- Vivaraṇa of Śaṅkara-Bhagavatpāda: Crit. Ed. with Introd. Madras: Government Oriental Manuscript Library, 1952. Madras Government Oriental Ser. 94. Rau 1949 Rau, Wilhelm. “Vallabhadevas Komentar zu Māghas Śiśu- pālavadha: Ein Beitrag zur Textgeschichte des Māghakāv- ya.” Inaugural Dissertation zur Erlangung des Doktorgra- des der Philosophischen Fakultät der Philipps-Universität Marburg. Marburg 1949. Rept. Konrad Klaus and Joachim Sprockhoff, eds. Wilhelm Rau: Kleine Schriften. Teil 1. Wiesbaden: Harrassowitz, 2012. 1–204. Veröffentlichun- gen der Helmuth von Glasenapp-Stiftung 46. Rigopoulos 1998 Rigopoulos, Antonio. Dattātreya, the Immortal Guru, Yo- gin, and Avatāra: A Study of the Tranformative and Inclu- sive Character of a Multi-Faceted Hindu Deity. Albany: State University of New York Press, 1998. SUNY Series in Religious Studies. Sachau 1888 Sachau, Carl Edward. The Chronology of Ancient Nations: An English Version of the Arabic Text of the Athâr-ul- bâkiya of Albîrûnî or “Vestiges of the past.” London: W. H. Allen & Co, 1888. Salomon 2014 Salomon, Richard. “Māgha, Mahābhārata, and Bhāgavata: The Source and Legacy of the Śiśupālavadha.” Journal of the American Oriental Society 134.2 (2014): 225–240. Sanderson 2007 Sanderson, Alexis. “The Śaiva Exegesis of Kashmir.” Mélanges tantriques à la mémoire d’Hélène Brunner / Tantric Studies in Memory of Hélène Brunner. Eds. Domi- nic Goodall and André Padoux. Pondicherry: Institut français d’Indologie / École française d’Extrême-Orient, 2007. 231–442, and (bibliography) 551–582. Collection Indologie 106. Shastri 1975 Shastri, Ajay Mitra. “Sanskrit Literature Known to al-Bī- rūnī.” Indian Journal of History of Science, 10 (1975): 111–138. Sörensen 1904 Sörensen, Sören. An Index of the Names in the Mahābhā- rata with Short Explanations and a Concordance of the Bombay and Calcutta Editions and P. C. Roy’s Trans- lation. London: Williams and Norgate, 1904(–1925). Tubb 2014 Tubb, Gary. “Kāvya with Bells On. Yamaka in the Śiśu- pālavadha. Or, ‘What’s a flashy verse like you doing in a 62 Philipp A. Maas great poem like this?’” In: Bronner, Shulmann and Tubb 2014. 142–194. Vetāl 1929 Vetāl, Anantarāma Śāstrī, ed. The Śiśupālavadha by Ma- hākavi Māgha: With Two Commentaries, the Sandehavi- ṣauṣadhi by Vallabhadeva and the Sarvaṅkaṣā by Mallinā- tha. Benares: Chowkhamba Sanskrit Ser. Office, 1929. Kāshi Sanskrit Ser. 69, Kāvya Section 9. Warder 1974–1992 Warder, Anthony Kennedy. Indian Kāvya Literature. Vols. 2–6. Delhi: Motilal Banarsidass, 1974–1992. Warren and Kosambi 1950 = Warren, Henry Clarke and Damodar Dharmanand Ko- sambi. Visuddhimagga of Buddhaghosâcariya. Cam- bridge, MA: Harvard University Press, 1950. Harvard Ori- ental Series 41. Watson, Goodall and Sarma 2013 = Watson, Alex, Dominic Goodall and S. L. P. Anjaneya Sarma. An Enquiry into the Nature of Liebe- ration: Bhaṭṭa Rāmakaṇṭha’s Paramokṣanirāsakārikāvṛtti, a Commentary on Sadyojyotiḥ’s Refutation of Twenty Conceptions of the Liberated State (Mokṣa), for the First Time Crit. Ed., Translated into English and Annotated. Pondicherry: Institut Français de Pondichéry : École Française d’Extrême-Orient, 2013. Collection Indologie 122. Wezler 1987 Wezler, Albrecht. “Zu der ‘Lehre von den 9 Ursachen’ im Yogabhāṣya.” Hinduismus und Buddhismus: Festschrift für Ulrich Schneider. Ed. Harry Falk. Freiburg: Hedwig Falk, 1987. 340–379. Wezler 2001 — “Zu der Frage des Strebens nach äußerster Kürze in den Śrautasūtras.” Zeitschrift der Deutschen Morgenländi- schen Gesellschaft 151 (2001): 351–366. Woods 1914 Woods, James Haughton. The Yoga-System of Patañjali, Or the Ancient Hindu Doctrine of Concentration of Mind: Embracing the Mnemonic Rules, Called Yoga-Sūtras, of Patañjali and the Comment, Called Yoga-Bhāshya, Attri- buted to Veda-Vyāsa, and the Explanation, Called Tattva- Vaiçāradī, of Vāchaspati-Miçra. Cambridge, Mass.: Har- vard University Press, 1914. Harvard Oriental Ser. 17. Wujastyk 2012 Wujastyk, Dagmar. Well-Mannered Medicine: Medical Ethics and Etiquette in Classical Ayurveda. New York: Oxford University Press, 2012.  Creativity within Limits: Different Usages of a Single Argument from Dharmakīrti’s Vādanyāya in Vidyānandin’s Works* Himal Trikha In memoriam Helmut Krasser (1956–2014) Little is known about distinct textual references to other traditions of Indian thought in the works of the Jaina Digambara Vidyānandin (10th century).1 We can assume, however, that he referred to Dharmakīrti’s works in a similar way as Akalaṅka, his most prominent Digambara predecessor. Nagin Shah characterized Akalaṅka’s approach in the following way: Akalaṅka sometimes bodily takes the sentences of Dharmakīrti (some- times introducing minor changes therein) and makes use of them in connection with constructing his own ones. (Shah 1967: 39) Such a practice can be observed in Vidyānandin’s usage of an argument from Dharmakīrti’s Vādanyāya (VN). This technique of composition, however, is not employed by a mediocre imitator but by a creative, innovative and elegant mind: Vidyānandin’s adaptions of Dharmakīrti’s argument are cre- * Dear Helmut Krasser was the first scholar to comment on an early draft of this chapter. I am also very grateful to Elisa Freschi and Philipp Maas for their insightful remarks on the next-to-final version and to Horst Lasic and Anne MacDonald for their helpful comments on parts of earlier versions. The work on the article has been financed by the Austrian Science Fund (FWF) in the context of the FWF project P 24160–G15 (“Frag- ments of Indian Philosophy”), and I am obliged to the project leader Ernst Prets for his generous support of my work. I am further indebted to Vincent Eltschinger, Eli Franco, Nora Melnikova, Yasutaka Muroya and Torsten Wieser-Much for referring me to rele- vant literature and to Katharine Apostle for proofreading my English. The idea to write this article was prompted during the conference “Transmission and Tradition” in Mat- sumoto, Japan, in 2012 by Prof. Katsura’s kind, but doubting, interest in my claim that Vidyānandin, while using Dharmakīrti’s arguments against the Vaiśeṣika, did not me- chanically copy them, but slightly adapted them to fit the ontological framework of the Jainas. 1 See the overviews of textual references in Koṭhiyā 1949: 1–44, Jain 1964: 22–28, Ko- ṭhiyā 1977, prastāvanā, pp. 2–15, Borgland 2010: 77–92 and Trikha 2012a: 117f. 64 Himal Trikha ative with regard to the manner the inherited textual material is varied, inno- vative with regard to its application against Buddhist tenets and elegant with regard to how it is utilized to establish central Jaina philosophical tenets. Vid- yānandin’s compositional practice thus displays a characteristic described by Ashok Aklujkar as follows: The śāstrakāras in practically all areas seem to have seen nothing wrong even in adapting verses of others … Doing so was not a matter of inability but purely of convenience (and occasionally of being able to score points in debate through sarcasm). (Aklujkar 2000: 121) Sarcasm is a prominent feature of the argument central to this chapter. In his adaptions, Vidyānandin pays Dharmakīrti back in his own coin, and in some instances he scores with minimal changes to the original passage. Due to this remarkable adherence to the opponent’s very words and because of the restrictions Vidyānandin had to observe when transferring the argument from the Buddhist to the Jaina ontological framework, I would like to describe Vidyānandin’s creativity as having been constrained by and enfolded within the narrow confines of faithfulness to both textual transmission and ideologi- cal conviction.2 Adaptions of Dharmakīrti’s argument from the VN are not only found in Vidyānandin’s works, but also in works by Vācaspatimiśra, Prabhācandra and Paṇḍita Aśoka. We thus have before us what Ernst Steinkellner called ‘polemical parallels’ with a history of their own that is not necessarily related to a certain known author … [with] the difficult task remain- [ing] of determining the extent to which a certain idea or argument … has been taken up, answered or used by the other. (Steinkellner 2013: xxx) These parallels are clearly the result of the literary techniques summarized by Elisa Freschi and Philipp Maas with the terms “simple re-use” and “adaptive reuse” (see section 3.1 of the introduction, p. 13 above). It is, however, not always evident what exactly the “reused item” is or who exactly its reusing agent was. In the case at hand, it is not, for example, immediately clear whether Vācaspati got the argument directly from the VN or from another, as yet unidentified work. We find in the history of this argument nested layers of adaptions and our interpretation of whether a particular adaption is re- 2 Vidyānandin’s creativity is also addressed in Borgland 2010: 77. References to Dharmakīrti’s works are also discussed in Soni 1999: 155–157, Borgland 2010: 45–48, 51–54, 66f., 75f., 97 and Trikha 2012a: 138, 207–211. See also Balcerowicz 2011: 19.  Creativity within Limits 65 markably original (→ “adaptive reuse”) or a flawed reproduction (→ “simple re-use”) depends on our hypothesis as to what exactly the source of that particular adaption was. The examination of the “motives for reuse” and of the “purpose of innovating through reusing” (section 1 of the introduction, p. 11 above) is therefore closely intertwined with the study of the question of how a text was transmitted from one work to another. Here, with the con- siderable number of parallel texts for Dharmakīrti’s argument from the VN, a considerable number of alternatives for the transmission succession are pos- sible. The main aim of this chapter is to render some alternatives more plau- sible than others. In the first part of the chapter, I present the argument from the VN and an overview of eleven adaptions of the same in other works. In the second part (p. 73) I discuss three basic types for the succession of transmission of paral- lel texts and the adaptions of Dharmakīrti’s argument in works by Vācaspati- miśra and Paṇḍita Aśoka. In the third part (p. 82) I examine the adaptions in three works of Vidyānandin, concluding (p. 101) with an overview of my hy- potheses regarding the succession of transmission of the argument’s adap- tions. 1 A passage from the Vādanyāya and an overview of corresponding textual material idam eva ca pratyakṣasya pratyakṣatvam, yad anātmarūpavivekena svarūpasya buddhau samarpaṇam. ayaṃ punar ghaṭo ’mūlyadānakra- yī, yaḥ svarūpaṃ ca nopadarśayati pratyakṣatāṃ ca svīkartum iccha- ti. (VN 8,6–9) And the perceptibility of a perceptible entity (consists) precisely in this: that it transfers its nature to cognition by the exclusion of what is not its nature. But this, [your] pot, which does not show its nature and yet wants to acquire perceptibility, is a non-paying customer.3 3 See Much 1991: 22. Much translates svarūpa with “Gestalt.” In my translation “na- ture” I understand the English term in the sense of “the basic or inherent features, character, or qualities of something” (www.oxforddictionaries.com, September 12, 2014). The parallel passages examined in this chapter use svākāra and (sva-)ātman as equivalents for svarūpa. In my translation of viveka I follow Much’s translation of VN 5,3 (Much 1991: 10) and Franco 2012: 60. 66 Himal Trikha In the first statement Dharmakīrti determines a condition for perceptibility. In the second statement he ridicules the tenet of an opponent by claiming that it does not fulfill this very condition of perception. 1.1 The background of the argument The background of Dharmakīrti’s argument is his disagreement with the school of (Nyāya-)Vaiśeṣika concerning the ontological status of an aggre- gate, in this case, a pot. In Dharmakīrti’s opinion, only its components ac- tually exist. These are numerous finest atoms (paramāṇu) that, in close proxi- mity to each other, produce a specific homogeneous effect (ekārthakriyā). The effect may be addressed by convention of speech but it does not exist in reality (paramārthasat), only nominally (prajñāptisat).4 For the (Nyāya-)Vai- śeṣika, on the other hand, an aggregate cannot be considered to be a mere convention of speech. It reflects an actually existing entity, a whole (avaya- vin) that is delimited ontologically from its parts (avayava). Both, the parts and the whole, add up to the reality of the aggregate. With regard to this controversy, the passage from the VN offers a crite- rion for distinguishing between actually existing and alleged entities: per- ceptibility, defined as the transfer of the nature of the entity to the cognition. This transfer happens, according to Dharmakīrti, in a specific manner: “by the exclusion of what is not its nature (anātmarūpavivekena).” The compo- nents of an aggregate fulfill this condition, since their properties appear distinct from one another. Scent (gandha), for instance, appears as distinct from taste (rasa). The respective property’s nature is transferred by an exclu- sion of what it is not, namely, by the exclusion of the other properties’ na- tures. The alleged whole, however, does not fulfill this condition: it does not appear as distinct from its parts. Its alleged nature is not transferred by ex- cluding what it is not. On the contrary, instead of the nature of the whole, only the natures of the parts are transferred to the cognition.5 Dharmakīrti illustrates his conviction with an example from the domain of economic transactions. He calls into play an amūlyadānakrayin, who is to become the wretched hero of this chapter. An example of a definition of the terms used to characterize this agent is found in the Vyavahāranirṇaya (cited in Kane 1946: 495, n. 878): … loke jihāsitaṃ suvarṇādi mūlyam ucyate. upāditsitaṃ kṣetragṛhādi paṇyam ity ucyate. … mūlyatyāgapūrvakapaṇyasvīkāraḥ krayaḥ. 4 See Much 1991: 13f., n. 67; 14f., n. 78 and Dunne 2004: 37–45. 5 See Much 1991: 10 and 21, and Dunne 2004: ibid.  Creativity within Limits 67 According to common practice … what one wants to give away is called the “price,” for example gold. What one wants to achieve is called an “article,” e.g., a field, a house, etc. … Purchase is the acqui- sition of the article preceded by the discharge of the price. A purchase (kraya) is valid, if the price (mūlya) is given for the article (paṇya), such as gold given for a house (see Table 1). Therefore, a mūlyadā- nakrayin, i.e., “someone who buys (°krayin) by giving the price (mūlyadā- na°)” is a well-received customer. Table 1: Factors of a purchase with an example paṇya mūlya kraya gṛha suvarṇa suvarṇatyāga → gṛhasvīkāra By lining up these terms to Dharmakīrti’s description of perception (see Ta- ble 2), we get as the price for perceptibility the appearance of the nature (svarūpapratibhāsa). The components of an aggregate do appear and acquire perceptibility. The money is on the table, so to speak. Table 2: Factors of a purchase with an example paṇya mūlya kraya pratyakṣatva svarūpapratibhāsa svarūpaṃ upadarśayati → pratyakṣatāṃ svīkāraḥ In the case of the alleged whole, however (see Table 3), only the natures of the parts are offered (anātmarūpapratibhāsa), i.e., something which is not the price (amūlya). Table 3: Unsuccessful purchase/perception paṇya mūlya amūlya kraya pratyakṣatva anātmarūpapratibhāsa svarūpaṃ nopadarśayati + pra- tyakṣatāṃ svīkartum icchati → ∅ Someone who buys by giving what is “not the price (amūlya)” or, more gen- erally, “who does not buy by giving the price (a-mūlyadānakrayin)” corrupts his role in the transaction. Moreover, he makes a valid purchase impossible. A customer who does not pay for the desired good is not a “buyer.” Depend- 68 Himal Trikha ing on the action he pursues subsequent to his desire, he might be a “beggar,” a “thief,” or a “fraud.” What a convicted amūlyadānakrayin can get, at best, is a laugh. His intentions are simply funny, as is remarked by Śāntarakṣita and Vidyānandin (see next paragraph). The figurative usage of the element krayin suggests that the phrase serves as an ironic expression for an agent who fails to provide an indispensable condition for the action he wants to accomplish. Dharmakīrti uses the expression as a metaphor for a pot, i.e., the paradigm for the alleged whole. A “pot,” the subject of comparison (upameya), is ana- logous to a “non-paying customer,” the object of comparison (upamāna), because they share a mutual property (sādhāraṇadharma): both intend to do an action they are not qualified for. Applying the expression amūlyadānakra- yin to the argument points to still another figurative usage. The pot is not the agent who desires to acquire (svīkartum icchati) something and does not show up (nopadarśayati) with the proper reimbursement. It is a proponent of the Vaiśeṣika who fancies the concept of a whole without offering hard cur- rency, i.e., perceptual evidence. At this level of speech the argument is sar- castic, since it aims to mock the opponent’s view. This implication is also shown in Śāntarakṣita’s concluding words in his commentary on the argu- ment: … iti upahasati. (VA 37,15: “He [i.e., Dharmakīrti] ridicules [the posi- tion].”) 1.2 Overview of corresponding passages In addition to the attestation in the VN (attestation {1}), eleven further attes- tations ({2–12}) of the argument will be examined in this chapter.6 They are transmitted in five works by three authors who are all dated7 close to the turn of the first millennium: Vācaspatimiśra: – Nyāyavārttikatātparyaṭīkā: NVTṬ 342,8f. {2} Vidyānandin: – Tattvārthaślokavārttikālaṅkāra: TAŚVA 118,25f. {3}; 118,27f. {4}; 433,9–11 {5} – Aṣṭasahasrī: AS 79,12f. {6}; 176,4f. {7}; 176,8–10 {8} – Satyaśāsanaparīkṣā: SŚP 21,27f. {9}; II 14 {10}; 45,13f. {11} 6 After completing this investigation, I came across a further reference to this argument in the Prameyakamalamārtaṇḍa (PKM 546,11–13). In the context of this passage, Pra- bhācandra addresses a topic similar to the one Vidyānandin discusses in the context of TAŚVA 433,9–11, attestation {5} (see below pp. 85f.). 7 See p. 73, n. 9 and pp. 79 and 82 below.  Creativity within Limits 69 Paṇḍita Aśoka: – Sāmānyadūṣaṇa: SD 14,20f. {12}8 The above enumeration of the passages reflects the likely chronological order, which I will address in the course of this chapter. All further attesta- tions correspond to the argument in the VN to a very high degree, i.e., a com- parison of the respective passages reveals instances of literal or verbal cor- respondences. For example, in the comparison of the attestations in editions of the VN and the Tātparyaṭīkā we can single out the following exact literal and distinctive verbal correspondences: {1} idam eva ca pratyakṣasya pratyakṣatvam, yad anātmarūpaviveke- na svarūpasya buddhau samarpaṇam. ayaṃ punar ghaṭo ’mūlyadāna- krayī, yaḥ svarūpaṃ ca nopadarśayati pratyakṣatāṃ ca svīkartum icchati. (VN 8,6–9) {2} yad āhuḥ: so ’yam amūlyadānakrayī svākāraṃ ca jñāne samarpa- yati pratyakṣatāṃ ca svīkartum icchaty avayavīti. tad evaṃ paraṃ praty avayavino ’siddher vipratipattiḥ. (NVTṬ 342,8–10) The elements ayam amūlyadānakrayī and pratyakṣatāṃ ca svīkartum icchati match here literally in euphonic normalization. svarūpasya buddhau samar- paṇam (VN) is contracted in svarūpaṃ ca … upadarśayati (VN); and svākā- raṃ ca jñāne samarpayati (NVTṬ) mirrors the former semantically and the latter syntactically (see Figure 1). VN svarūpasya budhau samarpaṇam svarūpaṃ ca … upadarśayati semantic correspondence syntactic correspondence ↘ ↙ NVTṬ svākāraṃ ca jñāne samarpayati Figure 1: Two forms of verbal correspondence. The absence of na (NVTṬ) and the use of avayavin (NVTṬ) as a hypernym for ghaṭa (VN) are substantial variations to the text but do not diminish the textual dependency of the two passages. In a synopsis of the corresponding 8 In the secondary literature, this argument is referred to in Stcherbatsky 1932: 540, n. 5 (NVTṬ?), Shastri 1964: 255f. (NVTṬ), Much 1991: 22 (VN, NVTṬ, SD), Soni 1999: 149 (TAŚV 118,25–28 {3–4}) and Trikha 2012a: 209–212 (VN, NVTṬ, TAŚV 118,25f. {3}, SŚP II 14 {10}, SD). The sigla in brackets indicate which attestations have been referred to here. 70 Himal Trikha elements, the extent of the correspondence by way of expression can be de- marcated like this: {1} idam eva ca pratyakṣasya pratyakṣatvam, yad anātmarūpavive- kena svarūpasya buddhau samarpaṇam. ayaṃ punar ghaṭo ’mūlyadā- nakrayī, yaḥ svarūpaṃ ca nopadarśayati pratyakṣatāṃ ca svīkartum icchati. {2} yad āhuḥ: so ’yam amūlyadānakrayī svākāraṃ ca jñāne samar- payati pratyakṣatāṃ ca svīkartum icchaty avayavīti. tad evaṃ paraṃ praty avayavino ’siddher vipratipattiḥ. This loose but clearly distinct form of verbal correspondence with the VN prevails in the other attestations of the argument as well: {3} nāntar bahir vāṃśebhyo bhinno ’ṃśī kaścit tattvato ’sti. yo hi pratyakṣabuddhāv ātmānaṃ na samarpayati pratyakṣatāṃ ca svīka- roti, so ’yam amūlyadānakrayīty ayuktikam eva … (TAŚVA 118,25f.) {4} tatheme paramāṇavo nātmanaḥ pratyakṣabuddhau svarūpaṃ sa- marpayanti pratyakṣatāṃ ca svīkartum utsahanta ity amūlyadānakra- yiṇaḥ. (TAŚVA 118,27f.) {5} ime punā rūpādayo dravyarahitā evāmūlyadānakrayiṇaḥ, svarū- paṃ ca nopadarśayanti pratyakṣatāṃ ca svīkartum icchantīti sphuṭam abhidhīyatām. (TAŚVA 433,10f.) {6} kim avayaviparikalpanayā tasyāmulyadānakrayitvāt. sa hi praty- akṣe svātmānaṃ na samarpayati pratyakṣatāṃ ca svīkartum icchatīty amūlyadānakrayī, vikalpabuddhāv eva … (AS 79,12f.) {7} ta ime paramāṇavaḥ pratyakṣabuddhāv ātmānaṃ ca na samarpa- yanti pratyakṣatāṃ ca svīkartum icchantīty amūlyadānakrayiṇaḥ svā- vayavabhinnaikāvayavivat. (AS 176,4f.) {8} na ca, ete ’vayavā ayam avayavī samavāyaś cāyam anayor iti tra- yākāraṃ pratyakṣam anubhūyate sakṛd api, yato ’sāv apy amūlyadā- nakrayī na syāt, pratyakṣabuddhāv ātmānarpaṇena pratyakṣatāsvī- karaṇāviśeṣāt. tata eva parasparabhinnāvayavāvayavinām api pra- tyakṣe pratibhāsanād amūlyadānakrayiṇāv uktau samavāyavat. (AS 176,8–10) {9} ta ime paramāṇavaḥ pratyakṣabuddhāv ātmānaṃ na samarpayan- ti pratyakṣatāṃ ca svīkartum icchantīty amūlyadānakrayiṇaḥ. (SŚP 21,27f.)  Creativity within Limits 71 {10} na ca, ete ’vayavādaya ime ’vayavyādayaḥ samavāyaś ca teṣām ayam iti pratyakṣabuddhau visrasā bhinnāḥ sakṛd api pratīyante pra- tyakṣatāṃ ca svīkartum icchantīti. te ’mī amūlyadānakrayiṇaḥ praty- akṣabuddhāv ātmānarpaṇena pratyakṣatāsvīkaraṇāt. (SŚP II 15) {11} tad idaṃ paroditasvarūpaṃ sāmānyaṃ pratyakṣabuddhāv ātmā- naṃ na samarthayati pratyakṣatāṃ ca svīkartum icchatīty amūlya- dānakrayitvāt satām upahāsāspadam eva syāt. (SŚP 45,13f.) {12} idaṃ punar mūlyādānakrayi sāmānyam, svarūpaṃ ca nādarśa- yati pratyakṣatāṃ ca svīkartum icchati. (SD 14,20f.) 1.3 Groups of correlating elements Having these correspondences before us, it is clear that these passages are variations of one and the same argument. Through a comprehensive inspec- tion of the modifications, we get four groups of correlating elements with di- verse degrees of variation. The most stable element is the compound ex- pressing the metaphor; it has only one variation, found in the Sāmānyadūṣaṇa (group A): – amūlyadānakrayin {1–11} – mūlyādānakrayin {12} The second group consists of these correlates (group B): – pratyakṣatāṃ ca svīkartum iccha- {1–2, 5–7, 9, 11, 12} – pratyakṣatāṃ ca svīkara- {3} – pratyakṣatāṃ ca svīkartum utsaha- {4} – pratyakṣatāsvīkaraṇa {8, 10} More variations occur in these correlates (group C): – svarūpaṃ ca nopadarśaya-/nādarśaya- {1, 5, 12} – svākāraṃ ca jñāne samarpaya- {2} – pratyakṣabuddhāv ātmānaṃ (ca) na samarpaya-/samarthaya- {3, 7, 9, 11} – nātmanaḥ pratyakṣabuddhau svarūpaṃ samarpaya- {4} – pratyakṣe svātmānaṃ na samarpaya- {6} – pratyakṣabuddhāv ātmānarpaṇena {8, 10} With the exception of the missing negation in the Tātparyaṭīkā (group C {2}), the individual text versions of these three groups of correlates barely differ with regard to their semantic content. Many correlates of the following fourth group, however, point to significant alterations of the conceptual con- 72 Himal Trikha tent. This group consists of the correlates for the object(s) to which the amūlyadānakrayin is compared. The alterations are reflected in the variations of the metaphor with regard to their numerical expression (group D): – amūlyadānakrayī/mūlyādānakrayī: → ghaṭaḥ ~ avayavī ~ aṃśī {1–3, 6} → asau [samavāyaḥ] {8a} → sāmānyam {11–12} – amūlyadānakrayiṇau: → avayava & avayavin {8b} – amūlyadānakrayiṇaḥ: → paramāṇavaḥ ~ rūpādayaḥ {4–5, 7, 9} → avayava & avayavin & samavāya {10} The terms in group D represent the ontological concepts against which the variations of the argument are directed. Despite these somewhat different purports, we can observe in groups A to C that some attestations are closer to each other with regard to expression. Some variations characterize only the attestations in Vidyānandin’s works (group B {3, 4, 8, 10}), where, further- more, two attestations are connected more closely by using a nominal style (groups B and C {8, 10}). Most attestations in Vidyānandin’s works are con- sistent with regard to the use of the term pratyakṣabuddhau (except for the correlates C {5, 6}) and with regard to the use of ātman for svarūpa/svākāra in the non-Jain sources (except for C {4, 5}). C {5} is a particularly interest- ing correlate, since the causative stem darśaya- is used there, something otherwise attested only in the two Buddhist sources (C {1, 12}). Similarities between individual attestations suggest that these share a par- ticular stream of transmission. Before expressing the argument, an author may have referred to someone else’s copy of the work, remembered a teacher’s oral instruction or a skillful remark of a colleague or opponent, or may simply have taken up an earlier formulation in one of his own works. The latter can be safely assumed for some of the nine attestations in Vidyānandin’s works. Before coming to these, however, I will examine the context of the passages in Vācaspatimiśra’s and Paṇḍita Aśoka’s works with respect to their possible relation to the passage in the VN.  Creativity within Limits 73 2 The succession of transmission for the adaptions in Vācaspati’s and Aśoka’s works With regard to the succession of transmission of two particular attestations, I assume three basic types, which I will discuss using the example of their pos- sible relationship to the VN (attestation {1} above), the Nyāyavārttikatātpar- yaṭīkā (attestation {2}) and the Sāmānyadūṣaṇa (attestation {12}). 2.1 Basic types of the succession of transmission In comparison to the other attestations of the argument, a negation is missing in the three editions of the Tātparyaṭīkā I have at hand. A translation of the attested corresponding text could read: {2}… so ’yam amūlyadānakrayī svākāraṃ ca jñāne samarpayati pra- tyakṣatāṃ ca svīkartum icchaty avayavīti … (NVTṬ 342,8f., NVTṬT/A 478,16–18, NVTṬD 388,9–11) … “This one, [your] whole, is a non-paying customer as it transfers its form to cognition and wants to acquire perceptibility.” … This understanding could reflect an accidental transmission error, a (serious) misunderstanding or a deliberate distortion of the conceptual content of the initial argument (see sections 2.2.1–3 below). In pondering these alternatives, it is of relevance whom we regard as the creator of the particular expression of the argument in the Tātparyaṭīkā. One possibility is that it was Vācaspati himself: we can assume that he had a copy of the VN, or part of it, in front of him and that he adapted Dharmakīrti’s argument. With regard to the suc- cession of transmission, the relation of the two corresponding passages from the VN and the Tātparyaṭīkā could then be classified as a direct inclusion of text from one work into the other (see Figure 2). Figure 2: Direct inclusion of text from one work into the other. However, Vācaspati is separated from Dharmakīrti by a couple of centuries.9 In this period a large number of thinkers, both Buddhist and non-Buddhist, 9 Krasser (2012: 587) proposed “as a working hypothesis, the time of Kumārila and Dharmakīrti to be the middle of the sixth century,” against the consensual dating of Dharmakīrti in the 7th century. Acharya (2006: xxviii) considers Vācaspati “to have flourished between A.D. 950 and 1000,” and David (2012: 17f., n. 12) also thinks this 74 Himal Trikha were actively engaged with Dharmakīrti’s philosophical positions. Therefore, it might not have been Vācaspati who adapted the argument from the VN, but another person whose words Vācaspati referred to. This third source could have been available for Vācaspati in some form of textual evidence, but could just as well have been oral information from a teacher, a colleague or an opponent, whose words Vācaspati heard and preserved in a text that was to become a part of the Tātparyaṭīkā. In assessing the dissimilarities of otherwise closely corresponding passages – e.g., the missing negation in the attestation of the argument in the Tātparyaṭīkā – it can be meaningful to assume an intermediate source, as it might be exactly such as source that was responsible for the modification. If such (an) intermediate source(s) can be identified or demonstrated through an analysis of the context and of the argumentative function of two individual corresponding passages, their relation with regard to the succession of transmission would then have to be described as an indirect inclusion of text (see figure 3). Figure 3: Indirect inclusion of text via (an) intermediate source(s). A further basic type of transmission succession of two individual passages is both having transmitted the corresponding text independently from one another. This is unlikely here, but in theory both Dharmakīrti and Vācaspati could have referred to a third source and included it in an adapted form in their respective works (see Figure 4). Figure 4: Independent transmission. dating not improbable. Muroya (2011: 358f.) proposes that a “part in the NVTṬ may have been a response to the criticism by Jñānaśrī.” As Jñānaśrī, in turn, refers to the NVTṬ (see also Lasic 2000: 49), this would suggest that Vācaspati and Jñānaśrīmitra (“980–1030,” Kellner 2007: 7) were contemporaries.  Creativity within Limits 75 To substantiate such a theory, we would need textual evidence indicating that a significant number of the key semes of the argument (signified, e.g., by the lexemes pratyakṣa, svīkaraṇa, svarūpa, buddhi, samarpaṇa, mūlya, dāna, kraya) were already combined in the time before the VN’s composition. I am not aware of any such evidence; to the contrary, all attestations of the argu- ment point – in some way or another – to the context designed by Dharma- kīrti in the VN. I therefore assume for pragmatic reasons that the Tātparya- ṭīkā – and all other known attestations – depend on the argument in the VN. 2.2 The adaption in the Nyāyavārttikatātparyaṭīkā Vācaspati’s variation of the argument is found in a section called avayavi- parīkṣāprakaraṇa in Thakur’s edition (see NVTṬ 342,1). There, the section covers twelve pages and includes sūtras 2.1.33–36. The argument is found in the very beginning of this section, in the commentary on the first sūtra, which reads: sādhyatvād avayavini sandehaḥ. (NS 2.1.33) A doubt with regard to the whole is caused by the fact that it needs to be established. Vācaspati unfolds his examination with an interpretation of the word sādh- yatvāt. In the course of this interpretation he illustrates the position of an op- ponent, who denies that the notion of an avayavin can be justified. I under- stand the text as follows (NVTṬ 342,3–11, NVTṬT/A 478,12–18, NVTṬD 388,2–11):10 {2} [I] paraṃ1 praty asiddhatvam evāvayavinaḥ sādhyatvam. [II.1] sa khalv evaṃ mene: [II.2] a,b saṃvinniṣṭhā hi viṣayavyavasthitiḥ. a sa eva ca saṃvidā vyavasthāpyate, yas tasyā viṣayaḥ. sa eva ca2 viṣayaḥ, yaḥ 3 svākāram3 asyām arpayati. na ca nirantarotpannarūpādiparamāṇv- atiriktam avayavyākāraṃ bibhratīṃ saṃvidam īkṣāmahe, kiṃ tu nir- antarotpannarūpādiparamāṇvākārāṇām. sthaulyaṃ ca na yady api paramāṇūnāṃ pratyekam asti, tathāpi pratibhāsadharmo bahutvādi- 10 Here and below, text segments introduced by Roman numerals in superscript in square brackets indicate my segmentation of the argumentation structure. Arabic numerals in superscript refer to the variations mentioned in the apparatus. Latin letters in super- script indicate text segments with (hypothesized) corresponding passages. The symbol “=” used in the apparatus indicates that the text of the mentioned edition of the re- spective corresponding passage corresponds exactly, “~” that it corresponds literally with variations, and “#” that it is dominated by verbal correspondence. “//” indicates less significantly corresponding passages. In the text, literal correspondence with variations and distinct verbal correspondence are highlighted accordingly. 76 Himal Trikha vat, na 4punar avayavinam4 ekam avasthāpayitum arhati. [II.3.i] yad āhuḥ: [II.3.ii] cso ’yam amūlyadānakrayī svākāraṃ na5 jñāne samarpa- yati pratyakṣatāṃ ca svīkartum icchaty avayavīti.c,b [III] tad evaṃ pa- raṃ praty avayavino ’siddher vipratipattiḥ. a ~ AJP I 15,11 and 76,6; PrP 365,9; ŚVK II 131,25f. and III 149,24f.; ĪPVV 53,8; // NAVV 88,7; TĀV III 216,3; PASV 82,15 b unidentified parallel α? c # VN 8,8f. 1 na paraṃ NVTṬT/A 2 NVTṬD without ca 3 ākāram NVTṬT/A, D 4 punavaravayinam NVTṬ 5 na emended for ca NVTṬ, NVTṬ T/A, D [I] The fact that the whole needs to be established is precisely the fact that it is not established for the opponent. [II.1] He, indeed, thought the following: [II.2] “Because/Surely (?) the establishment of an object de- pends on cognition. And exactly that, which is the object of cognition, is established by it. And exactly that, which delivers its form in a cog- nition, is an object. And we do not observe a cognition bearing the form of a whole that surpasses the finest atoms like shape/color, etc., which occur without an intervening space [i.e., which do not collapse on each other], but (we observe) rather (a cognition) of the forms of the finest atoms like shape/color, etc., which occur without an inter- vening space. And even though concreteness (does not apply) to the finest atoms individually, (it) is still an appearing property (for them) like plurality and other [properties which only pertain to a group of things]; (this), however, cannot establish a unique whole. [II.3.i] As he/they (?) said: [II.3.ii] ‘This one, [your] whole, is a non-paying cus- tomer as it does not transfer its form to cognition and wants to acquire perceptibility.’” [III] In this way, therefore, dissent (follows) because the whole is not established for the opponent. In the following discussion of this passage I propose that the text segment indicated with “b” preserves the position of an unidentified Buddhist author. 2.2.1 On the main level of argumentation, sequence [I] states that the avayavin is not proved for an opponent of the Naiyāyikas, [II] illustrates this opponent’s position and [III] resumes the disagreement. The argumentation in [II] con- sists of an introductory sequence [II.1], an exposition of the opponent’s opi- nion in [II.2] and the reference to [II.3.i] and the mentioning of [II.3.ii] a corroborative statement corresponding to the passage from the VN. Consi- dering this argumentative framework, I think it very unlikely that the missing  Creativity within Limits 77 negation in [II.3.ii] reflects a deliberate distortion of the initial argument. Vācaspati intends to show here a disagreement with an opponent who is worthy of engaging in a debate on a serious philosophical problem. A distor- tion of the argument would compromise both the opponent’s expertise and the exposition of the philosophical problem. The conceptual content presented in [II.2] is similar to the context of the argument in the VN: An object is regarded as perceptible (pratyakṣā arthāḥ, VN 8,4) / demonstrable (viṣayavyavasthitiḥ, NVTṬ) only if its nature (svarū- pa, VN 8,7) / form (svākāra, NVTṬ) is transferred (samarpaṇa, VN 8,7) / delivered (arpaya-, NVTṬ) in a cognition (buddhi, VN 8,7; saṃvid, NVTṬ). The notion of an avayavin (→ ghaṭa, VN 8,8) does not fulfill this criterion, only the finest atoms (gandharasādayaḥ, VN 8,5; rūpādiparamāṇu, NVTṬ) do, which do not have a separate location (apṛthagdeśa, VN 8,4) / occur without an intervening space (nirantarotpanna, NVTṬ). This conceptual framework delimits the range of a misunderstanding, which is possibly reflected by the missing negation in [II.3.ii]: sequence [II.2] displays that its author is at least very well informed about the general direction of the argu- ment in [II.3.ii]. A misunderstanding would therefore not reflect a serious mistake with regard to the point of the argument, but can be delimited to the exact meaning of the rare compound amūlyadānakrayin. 2.2.2 With regard to the composition of the particular expression of the text in sequence [II], we can assume, first, that Vācaspati – based on his general knowledge of Buddhist philosophy – improvised a free rendering in sequence [II.2] in which he stereotyped “the” Buddhist position. The phrase sa khalv evaṃ mene would not introduce a particular opponent (saḥ …) and his thoughts (… evaṃ mene), but would be a literary device signaling the begin- ning of an imagined opponent’s position. Within this hypothesis on the composition of the textual material, we would have to understand hi at the beginning of sequence [II.2] as a modal particle (→ “surely”), since the more frequent use as a causal conjunction (→ “because”) would be redundant here.11 The interpretation of hi in its more frequent use requires an alternative theory on the composition of sequence [II.2]: Vācaspati could have included here material from another source in 11 As a causal conjunction, hi could only be related to the predicate of sequence [I] and its scope would reach as far as the end of sequence [II.3.ii]: “The avayavin is not estab- lished … because the establishment of an object depends on cognition …” But this connection of [I] and [II.3.ii] is already made clear by sequence [II.1]. 78 Himal Trikha which hi fits the argumentation structure better. And, in fact, hi is an intra- textual indication of what can be verified by ample intertextual evidence: saṃvinniṣṭhā hi viṣayavyavasthitiḥ (NVTṬ 342,3f.) is reflected in a number of sources, namely, in the works of Haribhadra, Śālikanātha, Sucaritamiśra and Abhinavagupta and his commentators (see sources indicated under “a”12). Haribhadra and the authors belonging to Kashmirian Śaivaism demarcate the sentence as external textual material, and in the three passages of the two Mīmāṃsā works the sentence is used in the context of discussions of Bud- dhist tenets: one passage among them (ŚVK III 149,23–150,13) also ad- dresses the term avayavin in connection with the examination of universals (sāmānya). A preliminary examination of these passages leads me to believe that they depend on another passage and that their correlate in the NVTṬ also stems from this as yet unidentified source. I therefore propose that Vācaspati introduces with sa khalv evaṃ mene a particular person, whose diction he followed closely also in the following sentences of sequence [II.2]. 2.2.3 Within this hypothesis, we have to consider at least the following agents with regard to the creation of the particular expressions (Figure 5): Vācaspati I II.1 III Unidentified person II.2 Dharmakīrti II.3.ii Figure 5: Layers of agency reflected in NVTṬ 342,3–11. Confronted with these nested layers of agency, it is unclear who was the responsible agent for the modifications in the argument from the VN in [II.3.ii]: any of the substitutions avayavin, svākāra, jñāna, samarpaya (NVTṬ) for ghaṭa, svarūpa, buddhi, upadarśaya- (VN) could have been carried out by the unidentified person or by Vācaspati. The same applies to the missing negation; but since the argument had to pass through the mind of the polymath Vācaspati, I think it unlikely that he was unaware of the con- notations of the word amūlyadānakrayin. I therefore regard the omission as a scribal transmission error and emend na for ca. 12 See above, p. 76. In these sources the sentence occurs with or without hi and always in the plural, e.g., tathā hi: saṃvinniṣṭhā viṣayavyavasthitaya iti sthitir iyam avivādā sar- vavādinām (PrP 365,9).  Creativity within Limits 79 With these layers of agency it is also unclear who supplemented [II.2] with [II.3] in the first place. Did already Vācaspati’s source sustain his argu- ment by paraphrasing the passage from the VN? The reference yad āhuḥ would then translate “as he (i.e., the respected teacher Dharmakīrti) said,” and the relation of the corresponding passages in the VN and the NVTṬ with regard to the succession of transmission would then be that of an indirect inclusion via an intermediate source. Or did Vācaspati illustrate the argument in [II.2] on his own accord by adding a trenchant formulation of the point in [II.3]? yad āhuḥ would then translate “as this/these (opponent/s) said,” and we could consider [II.3.ii] being a modified but direct inclusion of the VN passage. 2.2.4 Despite this somewhat opaque compositional structure, the alternative of an indirect inclusion should be considered more seriously, since the reasons for it are more conclusive. The reasons against an indirect inclusion and for a direct inclusion of the VN into the NVTṬ are, at first, that Vācaspati did not necessarily refer to a particular source in [II.2], because sa … evaṃ mene could merely introduce an imaginary position, hi can be understood as a modal particle and saṃvinniṣṭhā viṣayavyavasthitiḥ could simply be a com- mon place phrase. Secondly, if Vācaspati was referring to a particular source in [II.2] he could well have provided the textual material corresponding to the VN himself, as it was customary practice for scholars of his period not to demarcate borrowings from other works, and there would have been no need to place an iti at the joint between [II.2] and [II.3]. However, if we take the absence of an iti between [II.2] and [II.3] at face value and understand sa … evaṃ mene and hi in their precise meanings, while interpreting the many attestations of saṃvinniṣṭhā (hi) viṣayavyavasthitayaḥ as supporting these precise meanings, we have stronger evidence to conclude that Vācaspati re- produced the distinct thoughts of an individual, as yet unknown, pūrvapakṣa in [II.2–3] (the text segment indicated with “b” above). I therefore propose understanding the correspondence between the VN and the NVTṬ as the result of an indirect inclusion via an intermediate source. 2.3 The adaption in the Sāmānyadūṣaṇa Paṇḍita Aśoka, the composer of this short work, is dated to the first part of the 11th century, later than Vācaspati (see Steinkellner and Much 1995: 98), but his adaption of the amūlyadānakrayin argument is independent from the one in the NVTṬ. The argument appears in the second part of the work. There, Aśoka expresses the opinion that universals do not exist, since not 80 Himal Trikha existing is concomitant with not obtaining the characteristics of a cognition (yad yad upalabdhilakṣaṇaprāptaṃ san nopalabhyate, tad tad asad iti; SD 13,5). A universal would never obtain the characteristics of a cognition because its nature is not cognized (svabhāvānupalabdhi; SD 13,9f.). Aśoka explicates the potential of this latter reason by refuting objections which question it as being fallacious (→ hetvābhāsa). The argument from the VN follows one of these objections: {12} yat tūcyate, pratyakṣapramāṇasiddhasvabhāvatayā sāmānyasyā- siddha evāyam hetur iti, tad ayuktaṃ tasya svarūpāpratibhāsanāt. a idam eva hi1 pratyakṣasya pratyakṣatvam, 2yat2 svarūpasya 3svabud- dhau3 samarpaṇam. b idaṃ4 punar 5mūlyādānakrayi sāmānyam,5 sva- rūpaṃ ca nādarśayati6 pratyakṣatāṃ ca svīkartum icchati. b,a tathā hi … (SD 14,17–21) a ~ VN 8,6–9 b # NVTṬ 342,8f. VN: 1 ca for hi 2 yad ānātmarūpavivekena 3 buddhau 4 ayam 5 ghaṭo ’mūlyadānakrayī, yaḥ 6 nopadarśayati On the other hand, what is said, namely, “this reason [i.e., that the na- ture of a universal is not cognized] is in fact unestablished as the na- ture of a universal is established by a means of valid cognition, i.e., perception,” is not correct, because its (i.e., the universal’s) nature does not appear. Because the perceptibility of a perceptible entity con- sists precisely in this, that it transfers its nature to a cognition of its own. But this, [your] universal, is a non-paying customer: it does not show its nature and yet wants to acquire perceptibility. Because it is like this … When compared to the attestation in the NVTṬ, the sequence corresponding to the VN is longer here and contains the principle on which the amūlya- dānakrayin argument is based. Additionally, the passage in the SD corres- ponds more closely to the VN with regard to expression: the correspondence between the passages in the VN and the NVTṬ comprises literal, semantic and syntactic similarities in fairly equal proportions, whereas the correspon- dence between the VN and the SD is dominated by literalness. The few modi- fications pertain either to the substitution of lexemes with insignificant semantic variation (mūlya-adāna : amūlya-dāna and ādarśaya-: upadarśa- ya-), or reflect the embedding in a different argumentation structure (ca : hi and buddhau : svabuddhau, motivated by the omission of anātmarūpavive- kena). The substitution of sāmānya for ghaṭa (with the adjustment of the  Creativity within Limits 81 pronoun idam for ayam) represents the central element of this clear-cut adaption. Unlike in the NVTṬ, there is no clear indication that Aśoka referred to a third source while composing the section in which the material corresponding to the VN is embedded. We can therefore consider the corresponding text in the SD as bearing a slightly varied but direct inclusion of the passage from the VN. In concluding my discussion of different types of the succession of trans- mission, in Figure 6, I sketch the supposed relations of the passages discussed Figure 6: Different succession of transmission of the argument from VN 8,8f.13 13 Relations like the ones addressed here are sometimes expressed within the classifica- tion scheme and nomenclature established by Ernst Steinkellner for categorizing text witnesses (see, e.g., Lasic 2000: 25f. and Kellner 2007: 38–40). But only one of the relations hypothesized above can be adequately expressed with this system: the pas- sage in the SD is a modified “quotation” (citatum ex alio usus secundarii modo edendi): SD 14,19–21 Ce’e VN 8,6–9. Relations like the one proposed here for VN and NVTṬ cannot be expressed precisely, and relations like the one between NVTṬ and SD are not covered. I have therefore extended Steinkellner’s groundbreaking scheme and nomenclature to express also indirect and independent transmissions ade- quately (see Trikha 2012a: 127–140): the passage in the NVTṬ is, presumably, a me- diated paraphrase (NVTṬ 342,8f. <Re> VN 8,8f.). The verbally corresponding passages in NVTṬ and SD are independent from each other (NVTṬ 342,8f. Rp SD 14,20f.). 82 Himal Trikha 3 Vidyānandin’s use of the argument Vidyānandin can be dated between Vācaspati (second half of the 10th cen- tury) and Prabhācandra (first half of the 11th century).14 Therefore, all known adaptions of the amūlyadānakrayin argument from the VN emerged during a relatively short period, i.e., at the turn of the millennium. Vidyānandin refers to the argument in three works: in the Tattvārthaślokavārttikālaṅkāra (TAŚVA), the Aṣṭasahasrī (AS) and the Satyaśāsanaparīkṣā (SŚP). With regard to the relative chronology of these works, it can be assumed that the TAŚVA predates the AS15 and that the AS probably predates the SŚP.16 The argument has the following three functions (the numbers in curly brackets refer to the enumeration in part 1.2 of this chapter): – illustration of the Buddhist position (TAŚVA {3}, AS {6}) – rejection of the Buddhist position (TAŚVA {4, 5}, AS {7}, SŚP {9}) – utilization against common adversaries: (Nyāya-)Vaiśeṣika (AS {8}, SŚP {10}) and Mīmāṃsā (SŚP {11}) 3.1 The adaptions in the Tattvārthaślokavārttikālaṅkāra The Tattvārthaślokavārttika (TAŚV) is Vidyānandin’s extensive commentary in verse on the Tattvārtha(sūtra); the Alaṅkara (TAŚVA) is his own explana- 14 Vidyānandin cites the maṅgalaśloka of Vācaspati’s last work, the Bhāmatī (see Ak- lujkar 2000: 106, Acharya 2006: xxxi), in his allegedly last work, the Satyaśāsana- parīkṣā (SŚP 2,13–16). Prabhācandra refers to Vidyānandin by name in the Prame- yakamalamārtaṇḍa (PKM 176,4). See n. 9 above for Vācaspati’s dates and Jain 1959: 25 for those of Prabhācandra (980–1065 CE). 15 Close to the end of the AS, a work called “Tattvārthālaṅkāra” is mentioned for the further elaboration of the discussion: … vākyasphoṭasya kriyāsphoṭavat tattvārthā- laṅkāre nirastatvāt (AS 285,9). That this “commentary on the Tattvārtha” points to Vidyānandin’s own commentary, the TAŚV(A), is made clear later in the text by a further reference in which the Tattvārthālaṅkāra is mentioned together with a lost work of Vidyānandin, the Vidyānandamahodaya: … iti tattvārthālaṅkāre vidyānan- damahodaye ca prapañcataḥ prarūpitam (AS 289,24f.). 16 Soni (1999: 162) considers the SŚP to be Vidyānandin’s last work. Borgland (2010: 13 and 77) refers to textual similarities between the SŚP with the AS, but the respective passages have not been examined with regard to their transmission succession. There is some, as yet inconclusive, evidence that the SŚP presupposes the Āptaparīkṣāṭīkā (Trikha 2012a: 248). This work, in turn, presupposes the AS, since it mentions the AS by an alternative name: … iti devāgamālaṅkṛtau tattvārthālaṅkāre vidyānandamaho- daye ca vistarato nirṇītaṃ pratipattavyam (ĀPṬ 233,9f.). “Devāgama(-stotra)” is an alternative name for Samantabhadra’s Āptamīmāṃsā, the work commented on in the AS. Hence “Devāgamālaṅkṛti” is an alternative name for the AS.  Creativity within Limits 83 tion of the verses in prose. The adaptions of the argument from the VN occur three times: twice in the first adhyāya and once in the fifth adhyāya. 3.1.1 The first two adaptions are found in close proximity to each other, in a con- text where core theses of Jaina philosophy with regard to epistemology and ontology are incorporated and defended against Buddhist philosophical tenets. The adaptions occur in the commentary on TA 1.6, where two forms of cognition (adhigama) are mentioned, namely, means of valid cognition (pramāṇa) and viewpoint (naya). For Vidyānandin (see TAŚVA 1.6.1–4), the difference between these forms of cognition is that they obtain an object (ar- tha) either fully (kārtsnyataḥ) or in part (deśataḥ). A means of valid cogni- tion points out (ādeśin) the object together “with parts” (sakala), a viewpoint “without parts” (vikala), i.e., without parts other than the one focused on. A means of valid cognition can therefore determinate its object (svārthaniśca- ya), whereas a viewpoint cannot, since it only grasps part of its object (svār- thaikadeśa). This epistemological discussion shifts to the ontological level in TAŚVA 1.6.5–6, where the actual basis for a viewpoint is put into focus. This would be a particular portion (aṃśa) of a thing (vastu) that is neither the thing itself nor not the thing (avastu), but somehow both, just as a portion of the ocean (samudra) is and is not identical with the ocean. This assumption of a fluxionary difference and identity of portions and what they consist in (aṃśin) represents a characteristic notion of Jaina on- tology.17 In TAŚVA 1.6.7–8 this notion is tested against a Buddhist position (see Soni 1999: 148–158). The discussion starts with adaptions of the amūl- yadānakrayin argument: {3, 4} nāntar bahir vāṃśebhyo bhinno ’ṃśī kaścit tattvato ’sti. a yo hi pratyakṣabuddhāv ātmānaṃ na samarpayati pratyakṣatāṃ ca svī- karoti, so ’yam amūlyadānakrayī a ity ayuktikam eva sthaviṣṭhasyai- kasya sphuṭaṃ sākṣātkaraṇāt tadvyatirekeṇāṃśānām evāpratibhā- sanāt. b tatheme paramāṇavo nātmanaḥ pratyakṣabuddhau svarūpaṃ samarpayanti pratyakṣatāṃ ca svīkartum utsahanta ity amūlyadāna- krayiṇaḥ. b (TAŚVA 118,25–28) a # VN 8,8f. b # TAŚVA 118,25f.; VN 8,8f. 17 Elsewhere, Vidyānandin ascribes this balance of difference and identity also to other cases where the relation of an entity to (alleged) subordinated entities is questioned, e.g., to the relation of cause (kāraṇa) and effect (kārya), of substance (dravya) and qualities (guṇa) or of substance and mode (paryāya). See, e.g., AS ad ĀM 61, ĀPṬ 110,9f., YAṬ 22,8f., SŚP II 12; Shah 1999: 5, 19 and Trikha 2012a: 290–294. 84 Himal Trikha [Buddhist:] “In reality, there is no aggregate different from its por- tions, none whatsoever, neither internally nor externally.18 Because this one, [your aggregate], which does not transfer itself to a percep- tual cognition and acquires perceptibility, is a non-paying customer.” [Jaina:] This is simply unreasonable because – as a most concrete, sin- gle (entity) becomes clearly evident – through its exclusion the por- tions themselves do not appear. In this way, these, [your] finest atoms, are non-paying customers, as they themselves do not transfer their nature to a perceptual cognition and wish to acquire perceptibility. With regard to conceptual content, the first adaption of the argument from the VN (indicated with “a” in the text above) is a fair representation of the general idea conveyed in the VN. With the second adaption (indicated with “b”), Vidyānandin turns the tables. His argument relies on the widespread notion that “infinitesimal particles are too small to be perceived by ordinary persons; instead, the matter perceived by ordinary persons consists of par- ticles that have somehow been aggregated into an entity of perceptible size” (Dunne 2004: 24). Dharmakīrti holds that “‘aggregation’ refers to a particular state of those particles, namely, that their proximity enables them to causally support each other such that they can cause an image in the perceiver’s mind” (ibid.: 102).19 With the second adaption of the argument from the VN in the above passage, Vidyānandin takes a stance against Dharmakīrti’s position by turning the argument against its creator: even under the condition that the obvious evidence (sākṣātkaraṇa) of a single most concrete (sthaviṣṭha) ag- gregate is suspended, an analytical reduction of its perception to the percep- tion of its multiple factors (aṃśa) cannot be successful. The exclusion (vyati- reka) of the aggregate has the undesirable consequence that a medium for the perception of the components would be missing and hence the nature of the finest atoms could not appear. These would be the alleged entities with no correlation in the cognition.20 18 In parallel passages (see pp. 90f. below) the adverbs bahir and antar are conjoined with upa+√lambh (AŚSh 38,12) and pratyakṣa (AS 176,2). I therefore assume with Vaṃśīdhara (1915: 176, n. 7) that these adverbs refer to the activity (or, respectively, the result of an activity) of the external five senses (bahirindriya) and the inner sense (antaḥkaraṇa) during a perceptual process (see, e.g., Preisendanz 1994: 626). 19 We have encountered this subject matter already in the context of adaption {2} in the NVTṬ, where it was briefly touched upon: see above pp. 76f., where it is stated that numerous fine atoms occurring without an intervening space (nirantarotpanna) have concreteness and plurality as appearing properties (pratibhāsadharma). See also the positions in the contexts of adaptions {6} and {7} below. 20 Vidyānandin could be influenced here by a concept also expressed in LT 16. Cf. Bal-  Creativity within Limits 85 3.1.2 Vidyānandin uses this line of argumentation also in a further adaption of the argument from the VN in TAŚVA on TA 5.28. There the topos of the visibil- ity (cākṣuṣatva) of entities is discussed, and Vidyānandin places his opinion between Buddhist and Vaiśeṣika positions. For Vidyānandin, visibility per- tains to aggregates that appear (pratibhāsana) in a visual cognition (cakṣur- buddhi).21 This cognition cannot be reduced, as a Buddhist contends, to the appearance of the quality shape/color (rūpa) because, says Vidyānandin, a quality (guṇa) and what is qualified by it (guṇin) are not merely identical (abhedamātra) in every respect (sarvathā) but somehow different (ka- thañcidbheda), since they are subject to different modes.22 Besides, visibility is not to be confined to a particular type of substance (dravya), i.e., only to substances possessing the quality shape/color, as a Vaiśeṣika assumes, be- cause according to Vidyānandin, it is well known (prasiddha) that an indi- vidual substance is grasped by several senses. Moreover, a substance void of qualities (rūpādirahita) and, vice versa, qualities void of a substance (dravya- rahita) are not objects of perception (pratyakṣa).23 Vidyānandin elaborates this latter point by returning to the Buddhist contention with the following adaption of the argument from the VN: {5} a idam eva hi1 pratyakṣasya pratyakṣatvam, yad 2anātmany avive- kena2 3buddhau svarūpasya3 samarpaṇaṃ. 4ime punā rūpādayo drav- yarahitā evāmūlyadānakrayiṇaḥ,4 svarūpaṃ ca nopadarśayanti5 pra- tyakṣatāṃ ca svīkartum icchanti5 a iti sphuṭam abhidhīyatām. (TAŚVA 433,9–11) a ~VN 8,6–9 VN: 1 ca for hi 2 anātmarūpavivekena 3 svarūpasya buddhau 4 ayaṃ punar ghaṭo ’mūlyadānakrayī yaḥ 5 icchati Because the perceptibility of a perceptible entity consists precisely in this, that it transfers its nature to cognition without an exclusion of cerowicz (2006: 182): “An einem bestimmten Ort gibt es keine verbundenen mikrosko- pischen Atome (paramāṇu), weil es keinen makroskopischen Gegenstand (sthūla) zu sehen gibt.” 21 TAŚVA 432,33f. 22 TAŚVA 433,1–4. bheda and abheda, modified with both sarvathā and kathañcit, are central terms in the SŚP; see Trikha 2012a: 89f., n. 107. 23 TAŚVA 433,4–8. See Frauwallner 1956: 175f., 260f. and 272f. for the tenets dis- cussed. Further material for a more detailed analysis of this passage is found in the context of PKM 546,11–13. 86 Himal Trikha what it is not. But these, [your] properties like shape/color, etc., which are just void of a substance, are non-paying customers: They do not show their nature and yet want to acquire perceptibility. – Let it be said clearly like this. Here the argument is turned against its creator as in adaption {4} above, but it is remarkable how Vidyānandin twists Dharmakīrti’s very words. Adaption {4} is a loose paraphrase leaving considerable room for unexpressed concep- tual discrepancies with the initial argument and for the suspicion that the central substitution of paramāṇu for ghaṭa/avayavin is merely a skillful play on words. Adaption {5}, however, is dominated by a literal correspondence to the initial argument and it also captures the conceptual basis for the amūl- yadānakrayin analogy, i.e., the formulation of a criterion for perceptibility. Vidyānandin applies a significant change to the initial formulation here, i.e., the substitution of anātmarūpavivekena with anātmany avivekena. This “play on akṣaras” shows that Vidyānandin is clearly aware of Dharmakīrti’s sup- positions and that he is not inclined to share it. The VN reads: “The nature is transferred by the exclusion (°vivekena) of what it is not (anātmarūpa°).” The TAŚVA reads: “The nature is transferred without an exclusion (avivekena) of what it is not (anātmani).” For Dharmakīrti “the ultimate real is utterly unique … completely excluded or different from every other entity (sarvato bhinna, sarvato vyāvṛtta, ekāntavyāvṛtta)”24 and perception reveals this unique real entity in its singularity, i.e., by excluding everything else. This position implies, as Vidyānandin points out by subverting it, that a property that appears in the cognition, like rūpa, would have to be regarded as abso- lutely (sarvathā) identical (abhedamātra) to its substrate (guṇin), i.e., void of a substance (dravyarahita). Such an entity, however, does not appear in per- ception. What appears is an entity that can be differentiated from its onto- logical basis (→ kathañcidbheda) – be it a quality from a varying substance or a whole from its individual parts – but it appears without an exclusion (avivekena) of what it is not (anātmani): a quality transfers its nature to cognition together with a substance; the parts appear together with the whole. 24 Dunne 2004: 80f. For the mentioned terms, Dunne refers to a number of passages from the Pramāṇavārttika(-svavṛtti). Since the two terms sarvato bhinna and ekānta are ex- ceptionally prominent in the Sanskrit literature of Jaina philosophers, it would be a de- sideratum for the reconstruction of the development of Jaina philosophy to relate these passages from the Pramāṇavārttika(-svavṛtti) with similar discussions in Jaina Sanskrit works. A starting point could be identifying passages in Vidyānandin’s works that dis- cuss the respective tenets of the passages from the Pramāṇavārttika(-svavṛtti).  Creativity within Limits 87 3.1.3 Vidyānandin will get back to this point in adaptions {7} and {8} of the VN’s argument in the AS. Before discussing these adaptions, it is worthwhile to note the different character of the adaptions in the TAŚVA with regard to their relation to the VN. The succession of the transmission of the VN and adaption {3} cannot be determined exactly: Vidyānandin could either have referred directly to the VN and modified the text himself, or referred to other (oral) transmissions of the argument, with its catchy metaphor. The least that can be said with regard to adaption {3} is that Vidyānandin addresses, and is part of, a milieu familiar with the argument from the VN in some way or another. Adaption {4} is Vidyānandin’s individual contribution to this dis- course. Adaption {5}, however, represents a direct inclusion of the textual material from the VN. On the basis of the literal dependences it is tempting to speculate that Vidyānandin had a manuscript of the VN, or part of it, in hand when he composed the commentary on TA 5.28.25 With regard to Vidyā- nandin’s evaluation of Dharmakīrti’s thought, appraisal seems mixed with irony: by revising the conceptual content, reflected in the small rectification of the expression,26 the argument with its metaphor definitely has a point but, unfortunately, one that is directed against its originator. 3.2 The adaptions in the Aṣṭasahasrī The three uses of the argument from the VN in the AS mirror those in the TAŚVA: The first occurrence serves as an illustration of the Buddhist tenet, the second turns the tables, and in the context of the third the Jaina position is placed between the Buddhist and Vaiśeṣika positions. In contrast to the TAŚVA, where the point of the argument against the Vaiśeṣika only lingers in the background of the third occurrence, in its counterpart in the AS the argument is used explicitly against a Vaiśeṣika position. In the context of all three occurrences in the AS, the respective ontological presumptions are ex- 25 Within the nomenclature referred to in n. 13 above, the expression can be rendered as follows: The relation of attestation {3} to the VN is indeterminable (TAŚVA 118,25f. Re?/<Re>? VN 8,8f.), attestation {4} is based on attestation {3} (TAŚVA 118,27f. Re TAŚVA 118,25f.) and a mediated and conceptually altered paraphrase of the VN (TAŚVA 118,27f. <Re> VN 8,8f.). Attestation {5} is a variation on the citation of the VN (TAŚVA 433,9–11 Cee VN 8,6–9) and independent from the corresponding text in attestation {3} (TAŚVA 433,9–11 Rp TAŚVA 118,25f.) and {4} (TAŚVA 433,9–11 Rp TAŚVA 118,27f.). 26 The notions of rectification and revision are reflected in the meaning “corrected” for sphuṭa (Apte 1965: 1730a). iti sphuṭam abhidhīyatām can therefore be translated alter- natively as “[the argument] needs to be corrected in this way.” 88 Himal Trikha plicated more elaborately, conveying the impression that here a more mature scholar is deepening innovative ideas from an earlier stage in his life. 3.2.1 The first occurrence is found in the commentary on ĀM 7, which is inter- spersed with Dharmakīrti’s thought.27 The passage containing the adaption from the VN reads: {6} atha [I] a pratyāsannāsaṃsṛṣṭā rūpādiparamāṇavaḥ pratyakṣāḥ, te- ṣāṃ svakāraṇasāmagrīvaśāt pratyakṣasaṃvijjananasamarthānām evotpatteḥ, skandhasyāpi tata eva pareṣāṃ pratyakṣatopapatter anya- thā sarvaskandhānāṃ pratyakṣatvaprasaṅgāt skandhatvāviśeṣāt. tad- aviśeṣe ’pi keṣāñcit pratyakṣatve pareṣām apratyakṣasvabhāvatve pi- śācaśarīrādīnāṃ tathā svakāraṇād utpatteḥ, paramāṇūnām api keṣāñ- cit pratyakṣatvam anyeṣām apratyakṣatvaṃ tata evāstu. [II] kim avaya- viparikalpanayā tasyāmūlyadānakrayitvāt. b sa hi pratyakṣe svātmā- naṃ na samarpayati pratyakṣatāṃ ca svīkartum icchatīty amūlyadā- nakrayī, b vikalpabuddhāv eva tasya pratibhāsanād vicāryamāṇasya sarvathānupapannatvāt a iti matam. (AS 79,8–14) a unidentified parallel β? b # VN 8,8f. There is the (following) [Buddhist] opinion: [I] “Finest atoms like shape/color, etc., are perceptible, (when) they are adjacent (to but) unjoined (with each other), because due to the complex of their re- spective causes only their capacities to produce a perceptual cognition occur,28 since there is no distinctive property for being an aggregate, because of the undesired consequence that all aggregates would be perceptible [if being an aggregate is understood] in a different manner, because others29 obtain the perceptibility of an aggregate, too, due to 27 In his personal copy of the edition, Frauwallner identifies AS 77,3f. and AS 78,13f. as quotations from the Pramāṇavārttika. Vidyānandin equates the subject of yad āha in AŚVa 81,11 with “Dharmakīrti.” 28 “… a moment of sense perception is the result of a set of co-operating causes (sāmagrī) belonging to the immediately preceding moment, viz., knowledge, the sense- organ, an object, attention, light, etc. … It is essential for such a set of causes to be in approximation (saṃnidhi/pratyāsatti) in space and time with each other in order to produce their expected result.” (Katsura 1984: 217f.) 29 Vaṃśīdhara explicates pareṣām with naiyāyikānāṃ syādvādināṃ ca (AS 79, n. 20). In this sense, the ablative phrase would have to be understood as a concession to the op- ponent. But in this position, the possibility of the perceptibility of skandhas is taken to some extent seriously. According to “Buddhists such as Vasubandhu,” one perceives  Creativity within Limits 89 exactly this [complex of respective causes]. Even though there is no distinctive property for it [i.e., for being an aggregate], there is for some perceptibility, which has an imperceptible nature for others, (and hence) the bodies of demons30 and such like occur in this manner due to the respective cause; let (therefore) also the finest atoms be perceptible for some and imperceptible for others due to exactly this [complex of respective causes]. [II] [But] what (does one gain) by conceptualizing a whole, as it is a non-paying customer? Because as it does not transfer its own self to perception and wants to acquire perceptibility, it is a non-paying customer, since the investigated is in no way obtained, because it appears only in a conceptual cognition.” Adaption {6} corresponds loosely to the VN in the same way as the adaptions {2} in the NVTṬ and {3} in the TAŚVA, where the adaptions are likewise used as illustrations of the opponent’s tenet. With regard to the succession of transmission, the respective contexts of these three adaptions are independent from each other, i.e., the textual shape and specific conceptual content of the expounded positions vary to such an extent that we can rule out the possibility that Vidyānandin drew on the passage in the NVTṬ, or that Vācaspati and Vidyānandin both drew on the same passage in a work of a third author. Whereas the position in the context of adaption {3} in the TAŚVA is too short to determine whether Vidyānandin expressed it in his own words or included a passage from another work, the features of the position in the context of adaption {6} in the AS are clear cut due to its elaborate argumentation structure. We have seen in adaption {5} above, that Vidyānandin included a passage from the VN in the TAŚVA, and I have shown elsewhere (2012a: 141–157) that the text of the SŚP is characterized by lengthy inclusions from other works. I therefore think that due to the distinctive character of the arguments in the context of adaption {6}, we can take the atha … iti matam at face value and assume that here Vidyānandin included a distinctive passage of an as yet unidentified post-Dharmakīrtian Buddhist author (in the text segment indicated with “a” above).31 conglomerates (see Dunne 2004: 102). 30 A piśāca is considered an entity that cannot be perceived by an “ordinary cognizer” but “a piśāca can be perceived by a fellow piśāca … and a Yogin can also perceive a piśāca” (Kellner 1999: 195, n. 5). 31 See Ono 2000: 89–94 for a list of parallel passages in the AS and Prajñākaragupta’s Pramāṇavārttikālaṅkāra (PVA). For the close correspondence of a śloka transmitted in the PVA, the AS and the SŚP, respectively, see Trikha 2012a: 153 and 235f. Formally, the relations proposed above can be expressed as follows: AS 79,12f. <Re> VN 8,8f.; AS 79,12f. Rp NVTṬ 342,8f.; AS 79,12f. Rp TAŚVA 118,25f. 90 Himal Trikha The key argument of the opponent, i.e., that the presumption of the actual existence of aggregates has the untenable consequence that all aggregates would be perceptible, is retorted by Vidyānandin with the concept of varying types of aggregates (aṇumahattvādiparimāṇabheda, AS 79,17): some aggregates are perceptible, others are not. In addition, Vidyānandin again stresses, as in the context of adaption {4}, his position that finest atoms as such are never obtained in perception (°paramāṇūnāṃ tathātvena kasyacit kadācin niścayāsattvāt pratyakṣatānupapatteḥ; AS 79,14f.), only the aggregate clearly appears (skandhasyaiva sphuṭam adhyakṣe ’vabhāsanāt; AS 79,15). 3.2.2 The two other adaptions in the AS appear in close proximity to one another, just as the first two adaptions in the TAŚVA {3 and 4}. Also like these, they are expressed in the context of key notions of Jaina philosophy. Vidyānandin applies the adaptions in his explanation of Akalaṅka’s commentary on ĀM 36. The first part of the strophe reads: pramāṇagocarau santau bhedābhedau na saṃvṛtī | (ĀM 36ab) Difference and identity are not concealments (i.e., not conceptual), as they are in the scope of the means of valid cognition. The statement is directed against “Buddhists” and “Advaitins,”32 who maintain what the respective others deny: one group maintains the difference of the entities (artha) only and regards identity as a concealment (saṃvṛti); the others maintain the identity of phenomena (bhāva) only and regard difference as caused by conception (kalpanā).33 But for Samantabhadra, the means of valid cognition conveys that things (artha) consist of both difference and identity: the notion of identity is provided by a thing’s enduring substance (dravya) and the notion of difference, by its changing modes (paryāya).34 The exclusion of one of them constitutes a one-sided view (ekāntavāda), as Akalaṅka makes clear in his comment on the half strophe: pramāṇam avisaṃvādi jñānam anadhigatārthādhigamalakṣaṇatvāt. tad evaṃ sati bhedam abhedaṃ vā nānyonyarahitaṃ viṣayīkaroti pramāṇam. na hi bahir antar vā svalakṣaṇaṃ sāmānyalakṣaṇaṃ vā 32 Vaṃśīdhara 1915: 175, nn. 16 and 18. 33 See AS 175,6f. Cf. Shah 1999: 38: “Samantabhadra … feels that … the transcendental- ist is blind to the aspect of difference, the empiricist Buddhist to that of identity.” 34 See, e.g., YAṬ 21,16f., translated in Trikha 2012a: 315.  Creativity within Limits 91 tathaivopalabhāmahe yathaikāntavādibhir āmnāyate, sūkṣmasthūlā- kārāṇāṃ sthūlasūkṣmasvabhāvavyatirekeṇa pratyakṣādāv aprati- bhāsanāt. (AŚSh 38,11–14) A means of valid cognition is a cognition that is reliable, etc.,35 since it is characterized as obtaining a (previously) unobtained thing. Hence, because it is like this, the means of valid cognition objectifies neither difference nor identity as being separated from each other. Because neither externally nor internally do we cognize (something which has only) the characteristic of itself (i.e., a particular), or (something which has) a general characteristic (i.e., a universal) in exactly the same manner as it is considered by the propounders of one-sided (views),36 since in perception, etc., subtle or concrete forms do not ap- pear by an exclusion of the concrete or subtle nature.37 In his comment on the two notions expressed in the last ablative phrase of Akalaṅka’s argumentation, Vidyānandin applies the adaptions of the argu- ment from the VN. Vidyānandin first supports the notion [A] that subtle forms do not appear through the exclusion of a concrete nature (sūkṣ- ma…ākārāṇāṃ sthūla…svabhāvavyatirekeṇa apratibhāsana) and then the 35 For avisaṃvādin as one of five elements in Akalaṅka’s definition of pramāṇa, see Clavel 2008: I, 39–41 and 63–69. My translation for avisaṃvādin follows Clavel’s “fiabilité.” For English translations of the term in the context of Dharmakīrti’s defini- tion, see, e.g., Katsura 1984: 219 (“non-contradictionary”), Dunne 2004: 254 (“trust- worthy”) and Taber 2005: 32 (“confirmed”). 36 Terminologically, the two alternatives point to central Dharmakīrtian concepts, i.e., svalakṣaṇa (“particular,” Katsura 1984: 217) and sāmānyalakṣaṇa (“universal or gen- eral characteristic,” ibid.): “A moment or particular is the object of sensation, while a continuum or universal is the object of such conceptual knowledge as inference, judg- ment and verbal knowledge” (ibid.). 37 A considerable part of this passage corresponds verbally to LTV 25,4–6 ad LT 7cd: bhedābhedaikāntayor anupalabdheḥ arthasya siddhir anekāntāt. nāntar bahir vā svalakṣaṇaṃ sāmānyalakṣaṇaṃ vā parasparānātmakaṃ prameyaṃ yathā manyate paraiḥ, dravyaparyāyātmano ’rthasya buddhau pratibhāsanāt. Clavel (2008, III: 16) translates this as follows: “Puisqu’on n’appréhende pas [l’objet de connaissance] dans la [thèse] unilatérale de la différence, et dans celle de l’identité, l’objet est établi par la thèse non unilatérale. L’objet de connaissance n’est pas soit interne soit externe, il n’est pas soit un être individuel soit un universel, il n’est pas [non plus un objet com- posé des deux pôles] avec deux natures qui ne se mêleraient pas, comme d’autres le pensent, parce que l’objet qui apparaît dans l’intellect consiste en substance et en mode.” My partly deviating interpretation of the passage in the AŚ is based on the un- derstanding of antar and bahir as adverbs and on the position of vā in the phrase bhe- dam abhedaṃ vā nānyonyarahitaṃ. 92 Himal Trikha notion [B] that concrete forms do not appear through the exclusion of subtle natures (...sthūlākārāṇāṃ ...sūkṣmasvabhāvavyatirekeṇa apratibhāsana):38 {7, 8} [I.1] na hi pratyakṣe svalakṣaṇaṃ sūkṣmaṃ paramāṇulakṣaṇaṃ pratibhāsate sthūlasya ghaṭādyātmanaḥ pratibhāsanāt. [I.2] paramā- ṇuṣv evātyāsannāsaṃsṛṣṭeṣu dṛṣṭau pratibhāsamāneṣu kutaścid vi- bhramanimittād ātmani paratra cāsantam eva sthūlākāram ādar- śayantī saṃvṛtis tān saṃvṛṇoti keśādibhrāntivad iti cet. [I.3] naivam, bahir antaś ca pratyakṣasyābhrāntatvakalpanāpoḍhatvābhāvapra- saṅgāt, saṃvyavahārataḥ paramārthato vā a pratyakṣaṃ kalpanāpo- ḍham abhrāntam a iti lakṣaṇasyāsaṃbhavadoṣānuṣaṅgāt, paramā- ṇūnāṃ jātucid adhyakṣabuddhāv apratibhāsanāt. [I.4] b ta1 ime para- māṇavaḥ 2pratyakṣabuddhāv ātmānaṃ ca na2 samarpayanti praty- akṣatāṃ ca svīkartum icchantīty amūlyadānakrayiṇaḥ b svāvayava- bhinnaikāvayavivat. [II.1] na hi so ’pi sūkṣmasvāvayavavyatirikto ma- hattvopetaḥ pratyakṣe pratibhāsate kuṇḍādivyatiriktadadhyādivat. [II.2] samavāyāt tebhyo ’narthāntaram iva pratibhāsate iti cet. [II.3] na, avayavipratyakṣasya sarvatra bhrāntatvaprasaṅgāt. tathā ca c avya- bhicāritvaṃ pratyakṣalakṣaṇam c asambhavi syāt. [II.4] na ca, ete ’va- yavā ayam avayavī samavāyaś cāyam anayor iti trayākāraṃ praty- akṣam anubhūyate sakṛd api, yato d ’sāv apy amūlyadānakrayī na syāt, pratyakṣabuddhāv ātmānarpaṇena pratyakṣatāsvīkaraṇāvi- śeṣāt. d [II.5] tata eva parasparabhinnāvayavāvayavinām api pratyakṣe ’pratibhāsanād amūlyadānakrayiṇāv uktau samavāyavat. (AS 175,22–176,10) a = NB 1.4 b # VN 8,8f., ~ TAŚVA 118,27f. (adaption {4}) c // NS 1.1.4 d # TAŚVA 176,4f. (adaption {7}); VN 8,8f. 1 2 adaption {4} without ta nātmanaḥ pratyakṣabuddhau svarūpaṃ adaption {4} [I.1] [This is the case,] because a subtle particular having the character- istic of a finest atom does not appear in perception, since a concrete (entity), having the nature of a pot, etc., appears. [I.2] [The Buddhist ob- jection:] “When finest atoms only, which are adjacent (to, but) un- 38 Akalaṅka explicates the first notion elsewhere, namely in LTV 50,13–15 ad LT 23: … cakṣuṣā rūpaṃ saṃsthānātmakaṃ sthūlātmakam ekaṃ sūkṣmānekasvabhāvaṃ paśyati na punaḥ asādhāraṇaikāntaṃ svalakṣaṇam. Clavel (2008: III 48) has translated this as follows: “… on voit, grâce à la faculté visuelle, une forme qui consiste en une configu- ration, qui est grossière et qui possède une nature ténue et multiple, mais on ne voit pas l’être individuel dans sa singularité irréductible.”  Creativity within Limits 93 joined (with each other), appear in perception, a concealment display- ing a concrete form – (although this form) does not exist for them [i.e., for these atoms] and for others at all – conceals them due to a particular cause for error like the illusion of hairs (conceals an object for a person with a sight defect), etc.”39 [I.3] [Reply:] It is not like this, because of the undesired consequence that neither an external nor an internal perception would be non-erroneous and free of conceptual construction, since it necessarily follows the fault that the definition “perception is free of conceptual construction [and] non-erroneous” would not be possible either ultimately or practically, as the finest atoms never appear in a perceptual cognition. [I.4] These ones, [your] finest atoms, are non-paying customers, as they do not transfer themselves to a perceptual cognition and yet want to acquire percepti- bility, like a single whole which is different from its own parts. [II.1] Because also this (single whole), (since) it has a great extension, does not appear in perception as separated from its subtle parts, like sour milk, etc., separated from a pot, etc.40 [II.2] [The (Nyāya-)Vaiśeṣi- ka objection:] “Due to inherence the (whole) appears like a thing that is not different from its (parts).”41 [II.3] [Reply:] No, because of the undesired consequence that the perception of a whole would be erroneous in all cases. And so (an element of) the definition of per- ception, (i.e.,) that it is non-deviating, would not be possible. [II.4] And a threefold form of perception experienced at once – namely, “these 39 In the simile, the concept of a concrete form is compared to the non-conceptual error of a person with a sight defect. See Dunne (2004: 88f.) and Katsura 1984: 225–226, e.g.: “Perceptual judgment can be called ‘concealing’ (saṃvṛti) because it conceals the to- tality of an actual unique object by highlighting one of its universal characteristics” (ibid.: 226). 40 The elements of the simile kuṇḍa and dadhi are used in the Padārthadharmasaṅgraha in an example for the concept of relation (sambandha). The notion “in this pot here, (there) is sour milk” depends on the relation of pot and sour milk, therefore the notion “in these threads here, (there) is a piece of cloth” must also presuppose a relation. (Cf. PDhS §374: yatheha kuṇḍe dadhīti pratyayaḥ sambandhe sati dṛṣṭas tatheha tantuṣu pataḥ … iti pratyayadarśanād asty esāṃ sambandhaḥ iti jñāyate.) Whereas the con- nection between a pot and sour milk is arbitrary – sour milk might, but need not be, in a pot – the connection between threads and a piece of cloth or, more generally, the con- nection between parts and a whole is necessary: without a connection of the parts a whole cannot be produced. Pot and sour milk can appear separately from one another, since they are related by contact (saṃyoga); parts and a whole cannot appear sepa- rately, since they are related by inherence (samavāya); see Trikha 2012a: 204–207. 41 For anarthāntaram and the remaining part of the argumentation, see Trikha 2012a: 201, 204 and 207–213. 94 Himal Trikha are parts, this the whole and this the inherence of both (the parts and the whole)” – through which also that, [your] (inherence), would not be a non-paying customer, is not (the case) because the acquisition of perceptibility without transferring itself to a perceptual cognition (re- mains) unspecified. [II.5] Because of precisely this, also [your] parts and whole, which are different from each other, are called – like inherence – non-paying customers because they do not appear in perception. In the light of the previously discussed adaptions, the prominent feature of this text section is that we met in [I.2] a position that was referred to in the context of adaption {2}, touched upon in the context of adaption {3} and ela- borated in the context of adaption {6}. It seems also clear to me that before composing adaption {7} (the passage marked with “b”), Vidyānandin did not have the VN in mind, but remembered, or looked up, his earlier adaption {4} in the TAŚVA. Similarly, the process of composition would be misunder- stood, if we regarded adaption {8} (the passage marked with “d”) as a direct paraphrase of the text from the VN. Here Vidyānandin uses a variation of the argument he just used against the Buddhist to counter the Vaiśeṣika.42 But the striking feature of this text section is how Vidyānandin contrasts ‒ elegantly, in my opinion ‒ central concepts of the logico-epistemological tradition of Buddhism and of the Vaiśeṣika and shows the discrepancies of these concepts with prominent definitions of perception within the same traditions (indicated above with “a” and “c”). The amūlyadānakrayin argument here is in both cases a punchline driving home a point which has already been taken in the respective preceding argument. It is worthwhile to note that the style displayed here in the AS indicates years of learning, an experienced scholar and probably also a versed dispu- tant. In fact, Vidyānandin wrote at least what are today 696 edited pages43 before composing the text section containing the adaptions {7} and {8}. The assumption that we have to regard the composer of this text section as a ma- ture and somewhat independent thinker is further supported by the choice of the subject and the strategy of the argumentation. Samantabhadra and Aka- laṅka directed their arguments against what is today called the logico-episte- 42 Formally, the relations of the corresponding passages with regard to their successsion of transmission can be rendered as follows: attestation {7} is based on attestation {4} (AS 176,4f. Ce’e TAŚVA 118,27f.) and a mediated paraphrase of the text from the VN (AS 176,4f. <Re> VN 8,8f.). Attestation {8} is based on attestation {7} (AS 176,9 Re AS 176,4f.) and a multiply mediated paraphrase of the text from the VN (AS 176,9 < Re> VN 8,8f.). 43 The TAŚVA and the larger part of the AS, see n. 15 above.  Creativity within Limits 95 mological tradition of Buddhism on one hand and Advaita on the other. Vidyānandin, however, calls the Vaiśeṣika into play because he apparently thought that the ontological presumptions of this tradition would contrast more sharply with the Buddhist tenet (see notions indicated with [A] and [B] above p. 91).44 Elsewhere (2012a: 67–90 and 2012b) I have described Vid- yānandin’s interest in collecting, contrasting and refuting such contradictory alternatives as elements in his method of establishing epistemic pluralism through falsification: what remains after the refutation of such alternatives is the realization that somehow (kathaṃcit) both must be integrated in a complete account of the discussed phenomena. In the discussion of the text section containing the adaptions {7} and {8}, this discursively gained and therefore to a considerable extent rationally justified result is that perceptible entities appear inseparably with both subtle (→ bheda) and concrete (→ abheda) forms. 3.3 The adaptions in the Satyaśāsanaparīkṣā The common characteristic of the three adaptions in the SŚP is that they are no longer used as illustrations of a Buddhist teaching, but only as arguments against non-Jaina tenets, i.e., against Buddhist, Vaiśeṣika and Mīmāṃsā views. The relation of the SŚP to Vidyānandin’s other works is unclear with regard to their succession of transmission (see n. 16 above), and hence the relationship between the adaptions of the argument from the VN in the SŚP cannot be determined conclusively with regard to the adaptions discussed so far. But I think it probable that in the context of two adaptions in the SŚP, Vidyānandin utilized an argumentation structure which he had discovered in the TAŚVA and deepened in the AS. 3.3.1 The context of the first adaption is by now well known to the reader of this chapter, since much of the text corresponds literally to the context of adaption {7}:45 44 Samantabhadra uses Buddhist and Advaitic tenets as a background for the explanation of a Jaina tenet, Akalaṅka uses a distinct Buddhist terminology to clarify Samantabha- dra’s thoughts, and Vidyānandin uses a Vaiśeṣika tenet to specify Akalaṅka’s argu- ment. One gets the impression that these authors – apart from their indisputable adhe- rence to the Jaina faith and their evident objective of eliminating opposing belief sys- tems – are more interested in ideas and their adequate expression than in the affiliation to tradition and in the origination of these ideas. 45 The enumeration of the arguments follows the enumeration in the context of adaption {7} above. 96 Himal Trikha {9} [I.2] a nanu1 paramāṇuṣv evātyāsannāsaṃsṛṣṭeṣu dṛṣṭau pratibhā- samāneṣu kutaścid vibhramanimittād ātmani paratra cāsantam eva sthūlādyākāraṃ2 darśayantī 3 saṃvṛtis tān saṃvṛṇoti keśādibhrānti- vad iti cet. [I.3.i] naivam, bahir antaś ca pratyakṣasya 4bhrāntatvāpat- teḥ, tasya4 abhrāntatvakalpanāpoḍhatvābhāvaprasaṅgāt, 5 b pratya- kṣaṃ kalpanāpoḍham abhrāntam b iti lakṣaṇasyāsambhavadoṣānu- ṣaṅgāt. [I.3.ii] 6nanu naiṣa doṣaḥ paramāṇupratyakṣasya tallakṣaṇasam- bhavād iti cet. na,6 paramāṇūnāṃ jātucid adhyakṣabuddhāv aprati- bhāsanāt. [I.3.iii] 7na hi kaścil laukikaḥ parīkṣako vā deśakālaviprakṛṣ- ṭārthavat paramāṇūn sākṣāt pratyeti anyathā pratītyapalāpaprasaṅ- gāt.7 [I.4] c ta ime paramāṇavaḥ pratyakṣabuddhāv 8ātmānaṃ8 na sam- arpayanti pratyakṣatāṃ ca svīkartum icchantīty amūlyadānakrayi- ṇaḥ 9. c,a (SŚP 21,21–28) a ~ AS 175,23–176,5 (see the context of adaption {7}) b = NB 1.4 c # VN 8,8f. 1 AS without nanu 2 sthūlākā° AS 3 ādarśaya° AS 4 not in AS 5 saṃ- vyavahārataḥ paramārthato vā pratya° AS 6 not in AS 7 not in AS 8 ātmānam ca AS 9 svāvayavabhinnaikāvayavivat AS This extensive correspondence of two passages in different works by the same author establishes with a high degree of certainty that Vidyānandin took this argumentation from the earlier work and included it in the later. But which is which? Did Vidyānandin, broadly speaking, expand his argumenta- tion in the SŚP or did he contract it in the AS? More precisely: did he – while transferring the argumentation from the AS to the SŚP and on further reflection of the argument [I.3] – state the main consequence of the opponent’s tenet explicitly with bhrāntatvāpatteḥ (varia- tion 6) and did he drop the phrase saṃvyavahārataḥ paramārthato vā (v. 7), which at that time seemed somehow commonplace to him? Or did he – conversely, when he included the argumentation from the SŚP in the AS – think the main consequence (v. 6) to be self-evident so that he rather added a smooth phrase (v. 7), well known to the reader or listener? Does the passage in the AS, therefore, indicate an environment of learning or dispute in which thoughts had to be expressed straightforwardly and the three arguments of [I.3] in the SŚP were better summarized in one go to keep the attention of the audience? Or, conversely, did Vidyānandin, while composing the SŚP, find the time and leisure for the exposition of argument [I.3] from the AS by in- serting an objection (v. 8) and by creating a smooth junction (v. 9) for his adaption of the argument from the VN?  Creativity within Limits 97 3.3.2 A similar concomitance of either contraction or expansion can be observed in the context of the second adaption of the argument from the VN in the SŚP. This adaption corresponds to the context of adaption {8}, i.e., to the text which in the AS forms the bigger part of the continuation of the argumentation we have just examined. After fifteen pages of edited text, the SŚP reads (the enumeration of the arguments follows the context of adaption {8}): {10} [II.2] a nanu1 samavāyāt tebhyo ’vayavyādir2 anarthāntaram iva pratibhāsata iti cet. [II.3.i] na, avayavyādipratyakṣasya3 sarvatra bhrān- tatvaprasaṅgāt [II.3.ii] b 4timirāśubhramaṇanauyānasaṅkṣobhādyāhitavi- bhramasya b dhāvadhvādidarśanavad (?) asadākāraviśiṣṭārthagraha- ṇāt.4 [II.3.iii] tathā ca c avyabhicāritvaṃ pratyakṣalakṣaṇam c asambhavi syāt. [II.4] na ca, ete 5’vayavādaya ime ’vayavyādayaḥ5 samavāyaś 6ca teṣām ayam6 iti pratyakṣabuddhau visrasā bhinnāḥ sakṛd api pratī- yante d pratyakṣatāṃ ca svīkartum icchantīti. te ’mī amūlyadānakrayi- ṇaḥ pratyakṣabuddhāv ātmānarpaṇena pratyakṣatāsvīkaraṇāt7. d, a (SŚP II 14f.) a # AS 176,6–9 (see context adaption {8}) b ~ NB 1.6 c // NS 1.1.4 d # VN 8,8f. 1 AS without nanu 2 AS without ’vayavyādir 3 AS without ādi 4 not in AS 5 ’vayavā ayam avayavī AS 6 cāyam anayor AS 7 °svīkaraṇāviśeṣāt AS The most interesting modification here is variation 4.46 If Vidyānandin transferred these arguments from the AS to the SŚP, he expanded argument [II.3] with textual material from the Nyāyabindu (NB 1.6) [II.3.ii]. If he trans- ferred the arguments from the SŚP to the AS, he skipped this material. Did Vidyānandin – after remembering the argumentation in the AS where he had utilized NB 1.4 against the Buddhist tenet – remember NB 1.6 as a concept that he could utilize against the Vaiśeṣika in the SŚP? Or did he skip the argument containing NB 1.6 when including the arguments from the SŚP into the AS because he regarded it an unnecessary burden? These questions need to remain open and be asked again in future exami- nations of further parallel passages in the AS and the SŚP.47 The two alterna- 46 I have discussed the passage containing this adaption extensively elsewhere (2012a: 201–213). 47 The next steps in this regard would be an analysis of the parallel passages collected by 98 Himal Trikha tives for the relationship between the text passages examined in this chapter can be contrasted as follows: – The SŚP presupposes the AS, because the contexts of adaptions {9} and {10} represent, respectively, an extension of arguments that were origi- nally developed in the context of adaptions {7} and {8}. – The AS presupposes the SŚP, because the context of adaptions {7} and {8} represents a contracted conflation of arguments that were originally developed in the contexts of adaptions {9} and {10}.48 Although I see no decisive evidence in favor of one of these alternatives, I tend to assume the first to be more probable. Firstly, I am under the impres- sion that in general Vidyānandin is rather inclined to extend rather than con- tract arguments, since he was one of the most industrious Digambara authors and continuously improved on his edifice of thought. Secondly, the composi- tional technique of expansion is prevalent in Vidyānandin’s discussion of samavāya in the SŚP, where arguments are rather embellished, not succinct and to the point (see Trikha 2012a: 155). My examination of the latter argu- mentation has also shown that at least this part of the SŚP is a collage of already existing argumentation schemata and that the central method in the SŚP presupposes an elaborated theory on how to deal with non-Jaina tenets that was already explicated in the TAŚVA (ibid.: 109). Above, in section 3.2.2, we found that this well thought-out theory is also applied in the context of adaptions {7} and {8} in the AS. We have furthermore observed how Vidyānandin consolidated his usage of the argument from the VN in a lively intellectual exchange with the logico-epistemological tradition of Buddhism: The Buddhist usage of the argument reflected in adaptions {3} and {6] is parried, after playful beginnings in adaption {4}, with an increasing degree of Borgland (see n. 16 above) and an inspection of manuscripts in order to examine pos- sible cross-contaminations in the transmission of the texts of the AS and SŚP. I am grateful to Elisa Freschi for the latter observation, which points, ultimately, to the ne- cessity of collecting the quite large number of extant manuscripts of Vidyānandin’s works in order to study his oeuvre properly. 48 The previous scholarly assumption speaking against this relationship, namely, that the SŚP can be considered Vidyānandin’s last work, is merely supported by an argumen- tum ex silentio: “Since the work is incomplete one can consider whether it was V[idyā- nandin]’s last work” (Soni 1999: 162). Since the sole edition of the work is based on only two manuscripts (see Trikha 2012a: 107), its incompleteness could also be ex- plained by an insufficient transmission. Insufficiency of transmission pertains at least to one of Vidyānandin’s main works, the Vidyānandamahodaya, which is no longer extant (see Koṭhiyā 1949: 42f.).  Creativity within Limits 99 confidence in the contexts of adaptions {5} and {7}. As this more lively exchange with the Buddhist usage of the argument from the VN is missing in the SŚP, it seems implausible that the somehow mechanical processing of adaptions {9} and {10} in the SŚP should represent an earlier stage of de- velopment. I think it more likely that they were used when the matter was settled and when the amūlyadānakrayin argument was just another conve- nient point to be directed not only against the Buddhist, but also against the Vaiśeṣika.49 3.3.3 The notion of a continuous expansion of arguments and their scope holds good, in any case, with regard to the last adaption of the VN’s argument in the SŚP. Here it is utilized against the concept of universals (sāmānya) in the interpretation of the Mīmāṃsakas (mīmāṃsakākhyair bhāṭṭair prābhākaraiś ca; SŚP 45,8). A detailed analysis of the context of this adaption would re- quire delving into the dispute between Jainas and Mīmāṃsakas50; here I will only cite what were probably the last words Vidyānandin had to say on the amūlyadānakrayin argument: {11} na hi bhinnadeśāsu vyaktiṣu sāmānyam ekaṃ pratyakṣataḥ … pratīyate … a tad idaṃ paroditasvarūpaṃ sāmānyaṃ b pratyakṣabud- dhāv ātmānaṃ na samarpayati1 pratyakṣatāṃ ca svīkartum icchatīty2 b amūlyadānakrayitvāt a satām upahāsāspadam eva syāt. (SŚP 45,11– 14) a # VN 8,8f. b ~ SŚP 21,27f. (adaption {9}) 1 2 samarthayati SŚP 45,13; samarpayanti SŚP 21,27 icchanti SŚP 21,28 49 This hypothesis on the relationship of the contexts of adaptions {7}/{8} and {9}/{10} respectively confirms my earlier assumption that SŚP II 14f. (the context of adaption {10}) was remodeled from a text then unknown to me (see 2012a: 207, 213). This text “β1” (ibid.) would be AS 176,6–9. My assumption that the text is part of a more exten- sive passage “β” of Buddhist provenience (ibid.: 152–155 and 207) can be further as- sessed in the context of AS 175,22–176,10 (the context of adaptions {7} and {8}), which reverberates with various Buddhist concepts (see nn. 35–39 above). With regard to the SŚP, the formal expressions of my hypotheses read: SŚP 21,21–28 Ce’e AS 175,23–176,5; SŚP 21,27f. <Re> VN 8,8f.; SŚP II 14f. Re AS 176,6–9; SŚP II 14d <Re> VN 8,8f. 50 For a tentative translation of Vidyānandin’s examination of the mīmāṃsakamata in the SŚP, see Borgland 2010: 298–313. See also Borgland’s analysis, ibid.: 58–65, in which also Shah’s (1967) remarks on this dispute are considered. 100 Himal Trikha Because through perception a single universal is not cognized in the individuals, which have different locations … This one, a universal with a nature proposed by the opponent (Mīmāṃsakas), would be merely a source of amusement for decent persons51 since it would be a non-paying customer as it does not transfer itself to a perceptual cog- nition and wants to acquire perceptibility. We can assume that the reading samarthayati is a scribal transmission error, motivated by the confusion of the writing-block “थ” with “प,” and emend samarpayati against the background of the parallels in adaptions {2–4}, {6– 7} and {9}. Except for the two variations, the textual material marked with “b” is identical with the respective expression in adaption {9}, and we can ascertain that Vidyānandin took exactly this text from adaption {9} when he composed adaption {11}.52 51 The expression satām upahāsāspadam has a parallel in Haribhadra’s Anekāntajaya- patākā: upahāsasthānam āryāṇām (AJP II 160,1). This “condition for the honorable persons’ laughter” is reminiscent of the term “condition for defeat” (nigrahasthāna), the topic of the Vādanyāya and thus the context in which the amūlyadānakrayin argu- ment is first met. 52 SŚP 45,13f. Ce’e SŚP 21,27f.; SŚP 45,13f. <Re> VN 8,8f. In reviewing the applications of the classification system for corresponding text passages carried out in this chapter (see the previous nn. 13, 25, 31, 42 and 49), the system proves satisfactory with regard to the depiction of three intertextual parameters (see Trikha forthcoming): “degree of correspondence of expression” (→ literal, literal with variations, or verbal), “mode of reference in the immediate context” (→ various forms of demarcation, or no demar- cation) and “succession of transmission” (direct inclusion, indirect inclusion, and inde- pendent transmission). The system is, however, not fit for depicting a prominent feature of adaptions {4, 5, 7 and 9} discussed in this chapter, namely, the massive al- teration of content that comes along with small modifications of otherwise closely cor- responding textual material. This shortcoming of the system is particularly striking, if its possibilities for classification are considered in the comparison of adaptions {5} and {12}. Both the passages in the TAŚVA (above p. 85) and the SD (above p. 80) transmit the text from the VN literally with few variations, and it is, ironically, Vidyānandin and not Paṇḍita Aśoka, who demarcates the textual material from the VN: TAŚVA 433,9– 11 Cee VN 8,6–9 versus SD 14,19–21 Ce’e VN 8,6–9. The classification system is not transparent with regard to a critical distinction between the two passages, i.e., that one is contrary to and the other, congruent with the conceptual content of their mutual source. Conceptual variations of corresponding passages in Buddhist and Jaina works are reflected on by Kyo Kano (forthcoming).  Creativity within Limits 101 4 Conclusion My hypotheses on the succession of transmission of twelve attestations of the amūlyadānakrayin argument are depicted in Figure 7 below and can be sum- marized as follows: The argument from Dharmakīrti’s VN, attestation {1}, is used in Vācaspatimiśra’s NVTṬ {2}, in Vidyānandin’s TAŚVA {3} and AS {6} and in Paṇḍita Aśoka’s SD {12} to illustrate a tenet of the Buddhist lo- gico-epistemological tradition. The contexts of the attestations in the NVTṬ {2} and in the AS {6} preserve in all probability the distinct textual shape of two applications (α and β) of the argument by one or more unidentified Bud- dhist author(s). It is indeterminable (“?”), whether the source for Vidyānan- din’s first encounter with the argument in TAŚVA {3} was a further Bud- dhist application or the VN itself. Figure 7: Overview of the probable succession of transmission of twelve attestations of the amūlyadānakrayin argument. Vidyānandin modified the argument and directed it against the Buddhist tenet in TAŚVA {4}, TAŚVA {5}, AS {7} and SŚP {9}. Vidyānandin refined his first reaction to the argument in TAŚVA {4} with TAŚVA {5}, where he referred directly to the VN. But it is his first individual adaption of TAŚVA {4} that he reused first in AS {7} and then in SŚP {9}. 102 Himal Trikha Vidyānandin also used variations of the argument against the Vaiśeṣika in AS {8} and SŚP {10} and against the Mīmāṃsā in SŚP {11}. Vidyānandin employed the argument against the Vaiśeṣika first in AS {8}, where he used a tenet of the Vaiśeṣika as the counterpart of the Buddhist position he had par- ried with AS {7}. Vidyānandin later divided the inherently consistent argu- mentation structure of the AS and reused the respective arguments in his systematic refutation of Buddhist (SŚP {9}) and Vaiśeṣika (SŚP {10}) tenets in the SŚP. There, the use of the argument against the Mīmāṃsā, SŚP {11}, appears like an addendum to the long story of the argument’s adaptions. References Abbreviations and Primary Sources AJP Anekāntajayapatākā, Haribhadra. Anekāntajayapatākā by Haribhadra Sūri with his Own Commentary and Muni- candra Sūri’s Supercommentary. 2 vols. Crit. ed. H. R. Kapadia. Baroda: Oriental Institute, 1940–1947. Gaek- wad’s Oriental Ser. 88, 105. AŚVa Aṣṭaśatī, Akalaṅka, included in AS AŚSh —, see Shah 1999 AS Aṣṭasahasrī, Vidyānandin. Aṣṭasahasrī Vidyānandasvāmi- nā nirmitā Vaṃśīdhareṇa saṃśodhya … sampāditā. Bom- bay: Gāṃdhī Nāthāraṅgajī, 1915. ĀPṬ Āptaparīkṣāṭīkā, Vidyānandin. Vidyānandasvāmiviracitā … Āptaparīkṣā svopajñāptaparīkṣālaṅkṛtiṭīkāyutā. Ed. Darbārīlāl Jain Koṭhiyā. Sahārnapur, 1949. ĀM Āptamīmāṃsā, Samantabhadra, included in AS ĪPVV Īśvarapratyabhijñāvivṛtivimarśinī, Abhinavagupta. Īśva- rapratyabhijñāvivṛtivimarśinī by Abhinavagupta. Ed. Ma- dhusūdan Kaul Shāstrī. Bombay: Research Dep. Jammu & Kashmir State, 1938. Kashmir Ser. of Texts and Studies 60. TA Tattvārtha(-sūtra), included in TAŚV TAŚV Tattvārthaślokavārttika, Vidyānandin. Vidyānandisvāmivi- racitaṃ Tattvārthaślokavārtikaṃ Manoharlālnyāyaśāstri- ṇā sampāditaṃ saṃśodhitaṃ ca. 1st ed. Bombay, 1918. Ahmedabad: Saraswati Pustak Bhandar, R2002. Saraswati Oriental Research Sanskrit Ser. 16. TAŚVA Tattvārthaślokavārttika, Vidyānandin, included in TAŚV TĀV Tantrālokaviveka, Jayaratha. The Tantrāloka of Abhinava- gupta with Commentary by Rajanaka Jayaratha. Vol. III.  Creativity within Limits 103 Ed. Madhusūdan Kaul Shāstrī. Bombay, 1921. Kashmir Ser. of Texts and Studies 30. NAVV Nyāyāvatāravārttikavṛtti, Śāntisūri. Nyāyāvatāravārtika- Vṛtti of Śānti Sūri. Ed. Dalsukhbhāī Mālvaṇiyā. Ahmeda- bad: Saraswati Pustak Bhandar, R2002. Saraswati Oriental Research Sanskrit Ser. 14. NB Nyāyabindu, Dharmakīrti. Dharmakīrtikṛta-Nyāyabindor Dharmottara-kṛta-ṭīkayā … anuṭīkārūpaḥ … Durvekamiś- rakṛto Dharmottarapradīpaḥ. Ed. Dalsukhbhāī Mālvaṇiyā. 1st ed. 1955. Patna: Kashi Prasad Jayaswal Research Insti- tute, 21971. Tibetan Sanskrit Works Ser. 2. NVTṬ Nyāyavārttikatātparyaṭīkā, Vācaspatimiśra. Nyāyabhāṣya- vārttikatātparyaṭīkā of Vācaspatimiśra. Ed. Anantalal Thakur. New Delhi: Bhāratīyadārśanikānusandhāna Pari- ṣatprakāśitā, distributed by Munshiram Manoharlal, 1996. Nyāyacaturgranthikā 3. NVTṬT/A — Nyāyadarśanam with Vātsyāyana’s Bhāṣya, Uddyota- kara’s Vārttika, Vācaspati Miśra’s Tātparyaṭīkā and Viś- vanātha’s Vṛtti. 2 vols. Ed. Taranatha and Amarendra- mohan. 1st ed. Calcutta, 1936–1944. New Delhi: Munshi- ram Manoharlal, R1985. Calcutta Sanskrit Ser. 18, 29. NVTṬD — Nyāyavārtikatātparyaṭīkā by Vācaspatimiśra. Ed. Rā- jeśvara Śāstrī Drāviḍ. Varanasi: Chaukhambha Sanskrit Sansthan, 21989. Kashi Sanskrit Ser. 24. NS Nyāyasūtra, included in NVTṬ PASV Paramārthasāravivṛti, Yogarāja. The Paramārthasāra by Abhinavagupta with the Commentary of Yogarāja. Ed. J. Ch. Chatterji. Srinagar: Research Dept., 1916. Kashmir Ser. of Texts and Studies 7. PKM Prameyakamalamārtaṇḍa, Prabhācandra. A Commentary on Māṇikyanandin’s Parīkṣāmukhasūtra. Ed. by Mahen- dra Kumar Shastri. 1st ed. 1941. Delhi: Sri Satguru Publications, 31990. Sri Garib Dass Oriental Ser. 94. PDhS Padārthadharmasaṅgraha, Praśastapāda. Johannes Bronk- horst and Yves Ramseier. Word Index to the Praśastapā- dabhāṣya. A Complete Word Index to the Printed Editions of the Praśastapādabhāṣya. Delhi: Motilal Banarsidass, 1994. PP Pramāṇaparīkṣā, Vidyānandin. Vidyānanda praṇīta Pra- māṇaparīkṣā. Ed. Darbārilal Koṭhiyā. Vārāṇasī: Vīra Sevā Mandira Ṭrasṭa Prakāśana, 1977. Yugavīra Samantabhadra Granthamālā 14. PrP Prakaraṇapañcikā, Śālikanātha. Prakaraṇapañcikā of Śā- likanātha with Nyāya-Siddhi of Jayapuri Nārāyaṇa Bhaṭ- 104 Himal Trikha ṭa. Ed. Subrahmanya Sastri. Banaras: Banaras Hindu Univ., 1961. Banaras Hindu University Darśana Ser. 4. YAṬ Yuktyanuśāsanaṭīkā, Vidyānandin. Samantabhadrapraṇī- taṃ Yuktyanuśāsanam. Vidyānandaviracitayā ṭīkayā sam- anvitam Indralālaiḥ Śrīlālaiś ca sampāditaṃ saṃśodhitaṃ ca. Bombay 1919. Māṇikacandra Digambara Jaina Gran- thamālā 15. LT Laghīyastraya, Akalaṅka, see Clavel 2008 LTV Laghīyastrayavivṛti, see Clavel 2008 VA Vipañcitārthā, Śāntarakṣita. Vādanyāyaprakaraṇa of Dharmakīrti with the Commentary Vipañcitārthā of Śānta- rakṣita and Sambandhaparīkṣā with the Commentary of Prabhācandra. Ed. Dvārikādās Śāstrī. Vārāṇasī: Bauddha Bharati, 1972. Bauddha Bharati Ser. 8. VN Vādanyāya, Dharmakīrti. Michael Torsten Much. Dhar- makīrtis Vādanyāya. Teil I: Sanskrit Text. Teil II: Überset- zung und Anmerkungen. Wien: Verlag der Österreichi- schen Akademie der Wissenschaften, 1991. Veröffentli- chungen der Kommission für Sprachen und Kulturen Süd- asiens 25. ŚVK Ślokavārttikakāśikā, Sucaritamiśra. The Mīmāṃsāśloka- vārttika with the Commentary Kāśikā of Sucaritamiśra. Vol. 2–3. Ed. Sāmbaśiva Śāstri (Vol. 2) and Ramasvami Sastri (Vol. 3). Trivandrum 1929, 1943. Trivandrum Sans- krit Ser. 99, 150. SD Sāmānyadūṣaṇa, Paṇḍita Aśoka. Aśokanibandhau Avaya- vinirākaraṇaṃ Sāmānyadūṣaṇaṃ ca. Ed. Anantalal Tha- kur. Patna: Kashi Prasad Jayaswal Research Institute, 1974. Tibetan Sanskrit Works Ser. 15. SŚP Satyaśāsanaparīkṣā, Vidyānandin. Vidyānandikṛta-Satya- śāsanaparīkṣā. Ed. Gokulchandra Jain. Kalkattā etc.: Bhā- ratīya Jñānapīṭha Prakāśana, 1964. Jñānapīṭha Mūrtidevī Jaina Granthamālā: Saṃskṛta Grantha 30. (References with Roman numerals refer to the edition in Trikha 2012a). Secondary Sources Acharya 2006 Acharya, Diwakar. Vācaspatimiśra’s Tattvasamīkṣā: The Earliest Commentary on Maṇḍanamiśra’s Brahmasiddhi. Stuttgart: Steiner 2006. Nepal Research Center Publica- tions 25. Aklujkar 2000 Aklujkar, Ashok. “The Prologue and Epilogue Verses of Vācaspatimiśra I.” Rivista degli Studi Orientali 73.1–4 (1999 [2000]): 105–130.  Creativity within Limits 105 Apte 1965 Apte, V. M. The Practical Sanskrit-English Dictionary. 4th ed. 1965. Delhi: Motilal Banarsidass, R2000. Balcerowicz 2006 Balcerowicz, Piotr. “Akalaṅka und die buddhistische Tra- dition. Von der Nichtwahrnehmung von Unsichtbarem (adṛśyānupalabdhi) zur Allwissenheit.” Wiener Zeitschrift für die Kunde Südasiens 49 (2005 [2006]): 151–226. Balcerowicz 2011 — “Dharmakīrti’s Criticism of the Jaina Doctrine of Mul- tiplexity of Reality (anekāntavāda).” In: Krasser et al. 2011: 1–31. Borgland 2010 Borgland, Jens Wilhelm. “A Translation and Investigation of Vidyānandin’s Satyaśāsanaparīkṣā.” MA thesis. Depart- ment of Culture Studies and Oriental Languages, Univer- sity of Oslo 2010. Clavel 2008 Clavel, Anne. “La théorie de la connaissance dans le La- ghīyastraya d’Akalaṅka. Vol. I: Étude thématique. Vol. II: Matériaux pour une édition critique. Vol. III: Traduction intégrale des stances, de l’auto-commentaire et d’extraits choisis du Syādvādabhūṣaṇa d’Abhayacandra.” Diss. Université Lyon 3, 2008. David 2012 David, Hugo. “La parole comme moyen de connaissance. Recherches sur l’épistémologie de la connaissance verbale et la théorie de l’exégèse dans l’Advaita Vedānta.” Diss. École pratique des hautes études, Paris 2012. Dunne 2004 Dunne, John D. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publ., 2004 Franco 2012 Franco, Eli. “How to Distinguish between Non-existing Entities? Dharmakīrti and Prajñākaragupta on Universals as Objects of Knowledge.” Rocznik Orientalistyczny 65 (2012): 51–62. Frauwallner 1956 Frauwallner, Erich. Geschichte der indischen Philosophie. II. Band. Salzburg: Müller, 1956. Wort und Antwort 6.2. Jain 1959 Jain, Mahendrakumar, ed. Siddhiviniścayaṭīkā of Ananta- vīryācārya. The Commentary on Siddhiviniścaya and Its Vṛtti of Bhaṭṭa Akalaṅkadeva. 2 vols. Kāśī: Bhāratīya Jñā- nāpītha, 1959. Jñānapīṭha Mūrtidevī Jaina Granthamālā: Saṃskṛta Grantha 22–23. Jain 1964 see SŚP Kane 1946 Kane, Pandurang Vaman. History of Dharmaśāstra. Vol. III. Poona: Bhandarkar Oriental Research Institute, 1946. Government Oriental Ser. B.6. Kano forthcoming Kano, Kyo. “Syādvādaratnākara and Jñānaśrīmitra. A Style of Citation.” In: Marui and Prets forthcoming. Katsura 1984 Katsura, Shoryu. “Dharmakīrti’s Theory of Truth.” Journal of Indian Philosophy 12 (1984): 215–234. 106 Himal Trikha Kellner 1999 Kellner, Birgit. “Levels of (Im)perceptibility. Dharmotta- ra’s Views on the drśya in drśyānupalabdhi.” Dharma- kīrti’s Thought and Its Impact on Indian and Tibetan Philosophy. Proceedings of the Third International Dhar- makīrti Conference, Hiroshima, November 4–6, 1997. Ed. S. Katsura. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 1999, 193–208. Beiträge zur Kultur- und Geistesgeschichte Asiens 32. Kellner 2007 — Jñānaśrīmitra’s Anupalabdhirahasya and Sarvaśabdā- bhāvacarcā: A Critical Edition with a Survey of his Anu- palabdhi-Theory. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, 2007. Wiener Studien zur Tibeto- logie und Buddhismuskunde 67. Koṭhiyā 1949 see APṬ Koṭhiyā 1977 see PP Krasser 2012 Krasser, Helmut. “Bhāviveka, Dharmakīrti and Kumārila.” Devadattīyam. Johannes Bronkhorst Felicitation Volume. Ed. F. Voegeli, V. Eltschinger, D. Feller, M. P. Candotti, B. Diaconescu, and M. Kulkarni. Bern: Lang, 2012. 535– 594. Worlds of South and Inner Asia 5. Krasser et al. 2011 Krasser, Helmut, Horst Lasic, Eli Franco, and Birgit Kell- ner, eds. Religion and Logic in Buddhist Philosophical Analysis: Proceedings of the Fourth International Dhar- makīrti Conference, Vienna, August 23–27, 2005. Wien: Verlag der Österreichischen Akademie der Wissenschaf- ten, 2011. Beiträge zur Kultur- und Geistesgeschichte Asiens 69. Lasic 2000 Lasic, Horst. Jñānaśrīmitras Vyāpticarcā: Sanskrittext, Übersetzung, Analyse. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, 2000. Wiener Studien zur Ti- betologie und Buddhismuskunde 48. Marui and Prets forthcoming = Marui, Hiroshi, and Ernst Prets, eds. Transmission and Reflection: Proceedings of the International Conference on the Meaning and the Role of Fragments in Indian Philosophy, Matsumoto, August 20–24, 2012. Wien: Ver- lag der Österreichischen Akademie der Wissenschaften. Much 1991 see VN Muroya 2011 Muroya, Yasutaka. “Bhāsarvajña’s Interpretation of bhāva eva nāśaḥ and a Related Chronological Problem.” In: Krasser et al. 2011: 341–361. Ono 2000 Ono, Motoi. Prajñākaraguptas Erklärung der Definition gültiger Erkenntnis: Pramāṇavārttikālaṃkāra zu Pramā- ṇavārttika II 1–7. Teil 1. Wien: Verlag der Österreichi-  Creativity within Limits 107 schen Akademie der Wissenschaften, 2000. Beiträge zur Kultur- und Geistesgeschichte Asiens 34. Preisendanz 1994 Preisendanz, Karin. Studien zu Nyāyasūtra III.1 mit dem Nyāyatattvāloka Vācaspati Miśras II. Stuttgart: Steiner, 1994. Alt- und Neu-Indische Studien 46.1–2. Shah 1967 Shah, Nagin J. Akalaṅka’s Criticism of Dharmakīrti’s Phi- losophy. Ahmedabad: L. D. Institute of Indology, 1967. Lalbhai Dalpatbhai Ser. 11. Shah 1999 —, trans. etc. Samantabhadra’s Āptamīmāṃsā: Critique of an Authority. Along with English Translation, Intro- duction, Notes and Akalaṅka’s Sanskrit Commentary Aṣṭa- śatī. Ahmedabad: Jagruti Dilip Sheth, 1999. Sanskrit- Sanskriti Granthamālā 7. Shastri 1964 Shastri, Dharmendra Nath. Critique of Indian Realism: A Study of the Conflict between the Nyāya-Vaiśeṣika & the Buddhist Dignāga School. Agra: Agra Univ., 1964. Soni 1999 Soni, Jayandra. “Aspects of Jaina Epistemology with Special Reference to Vidyānandin.” Approaches to Jaina Studies. Philosophy, Logic, Rituals and Symbols. Ed. Olle Qvarnström, and N. K. Wagle. Toronto: University of To- ronto, Centre for South Asian Studies, 1999. 138–168. South Asian Studies, Papers 11. Stcherbatsky 1932 Stcherbatsky, Th. Buddhist Logic. Vol. I. Leningrad: Izda- telstvo Akademii Nauk SSR, 1932. Bibliotheca Buddhica 26. Steinkellner 2013 Steinkellner, Ernst. Dharmakīrtis frühe Logik: Annotierte Übersetzung der logischen Teile von Pramāṇavārttika 1 mit der Vṛtti. I. Introduction, Übersetzung, Analyse. II. Introduction, Anmerkungen, Anhänge etc. Tokyo: The International Institute for Buddhist Studies, 2013. Studia Philologica Buddhica Monograph Ser. 29a,b. Steinkellner and Much 1995 = Steinkellner, Ernst, and Michael Torsten Much. Texte der erkenntnistheoretischen Schule des Buddhismus II. Göttingen: Vandenhoeck & Ruprecht, 1995. Abhand- lungen der Akademie der Wissenschaften, Philologisch- Historische Klasse, 3.214. Taber 2005 Taber, John. A Hindu Critique of Buddhist Epistemology. New York: Routledge Curzon, 2005. Trikha 2012a Trikha, Himal Perspektivismus und Kritik. Wien: De No- bili, 2012. Publications of the De Nobili Research Library 36. Trikha 2012b — “Competing World Views. Perspectivism and Polemics in the Satyaśāsanaparīkṣā and Other Jaina Works.” Journal of Indian Philosophy 40.1 (2012): 25–45. 108 Himal Trikha Trikha forthcoming “Facets of a Fragment. Evaluation and Classification of Intertextual Elements in a Philosophical Jaina Sanskrit Work.” In: Marui and Prets forthcoming. Vaṃśīdhara 1915 see AS  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra Ivan Andrijanić 1 Introduction This chapter will first present some indications that Śaṅkara, for his com- mentary on the Brahmasūtra (BS), indeed reused older material from one or more lost sources, as was already argued by Ingalls (1952: 9–10; 1954: 292– 295) and Hacker (1953: 26), and then discuss some methods for identifying this older material. According to Ingalls and Hacker, the commentaries of both Śaṅkara and Bhāskara were based on earlier material composed within the framework of a bhedābheda theory. This can be concluded from various pieces of evidence, in addition to the fact that both commentaries on the BS share a great deal of textual material.1 Further, Hacker (1953: 26) remarked that Śaṅkara’s com- mentary contains many non-illusionistic similes that do not fit Śaṅkara’s illusionistic doctrine. Because of this, according to Hacker, parts of Śaṅkara’s text must be based on an older, well respected, non-illusionistic commentary on the BS. Hacker (ibid.) also doubted that Bhāskara would copy from the work of his hated enemy Śaṅkara. As a result, passages common to the works of Śaṅkara and Bhāskara would hint at the existence of an older commentary that both authors used independently as their respective source. Ingalls fol- lowed the same argument and called this earlier commentator the “proto- commentator” (1954: 294).2 The idea of Ingalls and Hacker that both com- mentaries are based on an older source was rejected by Rüping (1977), who claimed with convincing arguments that Bhāskara’s commentary draws only 1 Accordingly, Ingalls (1954: 295) characterized Śaṅkara’s commentary on the Bṛhadā- raṇyaka-Upaniṣad as “a far more original piece of writing than his Brahma-sūtra- bhāṣya.” 2 Van Buitenen (1971: 18) found Ingalls’s hypothesis unconvincing. For him, it is un- likely that Śaṅkara and Bhāskara used written sources; rather, their commentaries re- flect orally transmitted school traditions. 110 Ivan Andrijanić on Śaṅkara’s text and that Bhāskara did not have independent access to an older bhedābhedavāda commentary on the BS. Śaṅkara’s Brahmasūtra- bhāṣya (BSBh) contains sufficient bhedābheda arguments to explain their occurrence in Bhāskara’s Bhāṣya. Rüping (1977: 61) noted in fact some clear bhedābhedavāda arguments in Śaṅkara’s commentary, such as the one in BSBh 2.3.43, which might have been taken from an unknown older bhedā- bhedavāda commentary. Additional evidence for the fact that two partly contradictory attitudes can be detected in Śaṅkara’s BSBh was provided by Oberhammer in his study of Śaṅkara’s refutation of the Pāñcarātra doctrine in BSBh 2.2.42–45 (Oberham- mer 1977). There, Oberhammer noticed two conflicting standpoints in Śaṅka- ra’s critique. One acknowledges partial credibility of the Pāñcarātra, the other dismisses the whole system. Oberhammer also noticed in Śaṅkara’s text a critique of two Pāñcarātra standpoints that belong to different stages in the historical development of the Pāñcarātra doctrine. By drawing on external evidence derived from the works of Yāmuna, Rāmānuja and Sudarśanasūri, who mentioned an old commentary (bhāṣya) on the BS that also ascribed partial credibility to the Pāñcarātra, Oberhammer concluded that Śaṅkara relied on an older source by an anonymous author when he wrote his expla- nation of BS 2.2.42–43. In other words, Śaṅkara used the same source men- tioned by Yāmuna and others as an objection in his commentary on BS 2.2.44. And according to Oberhammer, Śaṅkara presented his own view in BSBh 2.2.44–45, where he dismissed the whole Pāñcarātra. As mentioned above, one aim of the present chapter is to investigate how the older material that Śaṅkara incorporated in his text can be identified. The material that will be examined can roughly be divided into three categories: 1) Quotations (or paraphrases)3 marked by Śaṅkara or by later sub-com- mentators as belonging to other authors. Śaṅkara marked these quotations with indefinite pronouns such as “others” (apare, anye) or “some” (kecit). The author of some of these doctrines is sometimes called the “author of the commentary” (Vṛttikāra or Vṛttikṛt) by later sub-commentators. In two passages Śaṅkara mentioned Upavarṣa as the author of certain views (BSBh 1.3.28, 3.3.53). While Śaṅkara sometimes criticized these views, sometimes he just mentioned them. 3 As these sources are not available, it remains unclear whether Śaṅkara provided quota- tions or paraphrases. I shall thus use the term “quotation” to cover both possibilities.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 111 2) Cases in which Śaṅkara provided more than one interpretation of sūtras, usually without criticizing them. Some of these divergent interpretations may originate from older, today lost, commentaries on the BS. 3) Material contradicting Śaṅkara’s Advaita which is neither marked by Śaṅkara nor by the sub-commentators as stemming from older sources. It may be argued that such passages originate from one or several well res- pected traditional sources. The material in the first two categories will be examined first, because the methodology developed there may help to define a possible methodology for recognizing the undesignated material belonging to the third category. I would suggest four ways to identify material that is possibly reused: The first is to search for passages marked by Śaṅkara as belonging to “others” or by sub-commentators as belonging to the (or: a) Vṛttikāra or Vṛttikṛt. The second way is tracing contradictions in Śaṅkara’s texts. There are many Brahmasūtras that interpret certain Upaniṣadic passages (especially in BS 1 and BS 3.3). When Śaṅkara’s interpretations of the sūtras dealing with a certain passage of an Upaniṣad are compared with his commentary on the same Upaniṣad, it is sometimes possible to find contradictions. If Śaṅkara’s interpretation of an Upaniṣadic passage occurring in his BSBh contradicts his interpretation of the same passage in his commentary on the respective Upa- niṣad, we may have an indication that Śaṅkara took his interpretation in the commentary on the BS from an older source. One may, however, object that such inconsistencies result merely from Śaṅkara’s intellectual development. This can be countered with the argument that some Upaniṣadic interpretations from the BSBh are alien to Śaṅkara’s Advaita doctrine; wherever we en- counter such an inconsistency, the interpretation in the commentary on the Upaniṣad is closer to Śaṅkara’s doctrine.4 An example for such an inconsis- 4 This approach is based on various assumptions regarding the question of authenticity of works that are ascribed to Śaṅkara, i.e., to the author of the BSBh. Padmapāda men- tioned Śaṅkara at the beginning of his Pañcapādikā as the author of the BSBh and as his teacher. Sureśvara, who in his Naiṣkarmyasiddhi 4.74 and 4.76 claimed that he served Śaṅkara’s lotus feet (as his direct disciple), composed a commentary on Śaṅkara’s Bṛhadāraṇyakopaniṣad-Bhāṣya, in which he mentioned that Śaṅkara was his teacher (Sureśvara ad BAUBh 6.5.25). Sureśvara also composed a sub-commentary on Śaṅkara’s commentary on the Taittirīya-Upaniṣad. Therefore, the BSBh, BAUBh and TaittUBh are quite clearly the works of the same Śaṅkara. On the other side, Hacker (1947) analyzed the colophons of the manuscripts of Śaṅkara’s works and concluded that the BSBh, the commentaries on the early Upaniṣads (with the exception of the Śvetāśvatara-Upaniṣad) and the commentary on the Bhagavad-Gītā are authentic works by Śaṅkara. However, on the methodological limitations of Hacker’s approach, see Maas 2013: 73f. 112 Ivan Andrijanić tency and for the fact that – in the case of divergence – Śaṅkara’s comments on the Upaniṣad are closer to his Advaita doctrine than the BSBh will be discussed below (see section 4.1). The third indication may be provided from Bhāskara’s commentary on the BS, which can be useful, although Bhāskara most probably had no indepen- dent access to older sources. Bhāskara regularly reused Śaṅkara’s text when it fit into his doctrine of bhedābhedavāda, but always departed from Śaṅkara’s work when the latter introduced some of his distinguishing Advaita teach- ings. Thus, whenever Bhāskara is following Śaṅkara, this may indicate that the text depends on more traditional, older material close to Bhāskara’s bhe- dābhedavāda. Although it is difficult to be sure about Bhāskara’s adherence to a teaching whose textual sources are no longer extant, it can be tentatively assumed that Bhāskara, who presented himself as a guardian of tradition, knew where Śaṅkara was still in contact with the tradition and where he di- verged from it. The fourth indication can be derived from peculiar terminological choices. Hacker (1950: 276–286) examined how Śaṅkara used terms like Īś- vara, Parameśvara, Brahman, Parabrahman and Paramātman, and concluded that in Śaṅkara’s works, these are synonyms. Moreover, Śaṅkara very often spoke about the qualities (guṇas or dharmas) of Brahman. Already Śaṅkara’s disciple Padmapāda did not use such qualifications when he spoke about Brahman, and after Vimuktātman (9th–10th century CE)5 such qualifications became totally unacceptable in Advaita (according to Hacker 1950: 286). It is possible that these terminological ambiguities result from the fact that Śaṅkara reused different sources for his BSBh, some (or all) of which may have consisted of older, non-Advaitic material. These older sources did not differentiate between a lower and the highest Brahman and they attributed positive qualifications to the Brahman. If this assumption is correct, it is possible that whenever qualifications for the highest Brahman appear in Śaṅ- kara’s works, he derived his exposition from an older source. In the first part of this chapter (sections 2 and 3), material marked as a quotation by Śaṅkara or by sub-commentators will be presented. In the second part (section 4), two examples from the BSBh will be discussed: 5 Vimuktātman must have lived after Sureśvara, because he quoted Sureśvara’s Naiṣ- karmyasiddhi, but before Yāmuna, who in his Ātmasiddhi quoted Vimuktātman’s Iṣṭa- siddhi. As Yāmuna lived in the second part of the 10th century (Mesquita 1973), Vi- muktātman must have lived between the late 8th or early 9th and the 10th century. Su- reśvara was a younger contemporary of Śaṅkara, who most probably can be dated to the middle of the 8th century (see Harimoto 2006). See also n. 22.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 113 a) Śaṅkara’s commentary on BS 1.1.12–19, where Śaṅkara’s sub-commen- tators identified a large part of the text as belonging to a Vṛttikāra, and b) Śaṅkara’s commentary on BS 1.3.1, where Śaṅkara himself indicated that the teaching he described derives from “another” (apara). 2 Material marked by Śaṅkara or by sub-commentators as being reused from other authors First, it must be said that in the text of the BSBh one can find numerous quo- tations from works belonging to śruti and smṛti texts, as well as lengthy dis- cussions of the positions of rival schools of thought. This chapter will neither discuss these passages nor their sources; the focus will be only on the ma- terial which may originate from one or more older, now lost, Vedānta texts. 2.1 Indefinite pronouns as markers of reuse Śaṅkara usually marked passages in which he quoted or paraphrased by using an indefinite pronoun. Kecit (“some”) predominantly (see Table 1) appears in reference to opponents external to Advaita Vedānta, whereas anye, apare (“others”) and eke (“some”) occur in reference to opponents within the Vedānta camp. A notable exception is found at the end of BSBh 1.1.1, where one encounters apare, eke and kecit referring to different heretical doctrines. Table 1: Pronouns as reuse-markers and their identification in BSBh and its commentaries6 Introductory phrase Identification Location kecit [followers of Prabhākara] BSBh 1.1.4 atra kecid udāharanti Vṛttikṛt (Govindānanda and BSBh 1.1.23 WOŚ “here some declare” Ānandagiri) pp. 50,4f. kecit [Buddhist Sarvāstivādins BSBh 2.2.18 and Vijñānāstitvamātravā- dins] 6 Square brackets are used to indicate identifications derived from the context, because they are not explicitly indicated in the commentaries. 114 Ivan Andrijanić Introductory phrase Identification Location kecit [followers of Sāṅkhya and BSBh 2.2.37 Yoga] kecit tāvad āhuḥ ekadeśin, i.e., a person with BSBh 3.1.8, “some indeed say” partial knowledge WOŚ p. 332,9 (Ānandagiri) kecit ekadeśin (Govindānanda) BSBh 3.2.21, WOŚ p. 360,3 kecit ekadeśin (Govindānanda) BSBh 3.3.38, WOŚ pp. 412,14f. kecit ekadeśin (Govindānanda, BSBh 3.3.57, Ānandagiri, Vācaspati) WOŚ p. 428,12 kecit BSBh 1.3.19, WOŚ p. 114,22 kecit BSBh 3.4.20, WOŚ p. 442,21 kecit BSBh 4.3.14, WOŚ p. 497,20 atrāpare pratyavati- Vṛttikāra (Govindānanda) BSBh 1.1.4 ṣṭhante “here others WOŚ p. 12,4 object” apara āha “someone Vṛttikāra (Govindānanda) BSBh 1.1.25, says” WOŚ p. 55,6 apara āha BSBh 1.1.27, WOŚ p. 56,13 apara āha BSBh 1.2.12, WOŚ p. 73,16  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 115 Introductory phrase Identification Location anye punar mayante Vṛttikṛt (Govindānanda) BSBh 1.2.23, “others, however, Vṛttikāra (Ānandagiri) ad WOŚ p. 86,1 think” BSBh 1.2.23 (BSŚWC p. 192) apara āha “someone BSBh 1.3.1, says” WOŚ p. 95,9 apara āha “someone BSBh 1.3.13, says” WOŚ p. 105,10 apare tu vādinaḥ “but BSBh 1.3.19, [according to] other dis- WOŚ p. 115,7 putants” anye tu varṇayanti “but Vṛttikṛt (Govindānanda) BSBh 1.4.3, others explain” Vṛttikṛt or Vṛttikāra WOŚ p. 149,14 (Ānandagiri) With the conjunction “or” (athavā), Śaṅkara introduced a different interpre- tation of BSBh 1.1.31 (WOŚ p. 61,24). Govindānanda and Vācaspati Miśra attributed this interpretation to the Vṛttikṛt; Ānandagiri ascribed it to the Vṛttikāra.7 2.2 Identifications of reuse by the sub-commentators A well-known case of attribution of a passage to a Vṛttikāra or Vṛttikṛt is found in the sub-commentaries ad BSBh 1.1.12–19. At the end of his com- mentary on BSBh 1.1.19 Śaṅkara introduced an alternative interpretation of sūtras 12–19 with the words idaṃ tv iha vaktavyam (“but this has to be said here”). Govindānanda and Ānandagiri regarded Śaṅkara’s text in BSBh 1.1.12–19 as having been authored by a Vṛttikṛt or Vṛttikāra, whereas they took the alternative interpretation, which starts in BSBh 1.1.19 with the words idaṃ tv iha vaktavyam, as genuinely by Śaṅkara. This example will be studied in some detail later in this chapter (section 4.1). 7 Govindānanda, Vācaspati and Ānandagiri ad BSBh 1.1.31 (BSŚWC p. 157). Deussen (1883: 30) thought that this passage might be an interpolation. See n. 6, 14, 15, 17 and 18. 116 Ivan Andrijanić 2.2.1 Reuse of the views of the Vṛttikāra The first problem one must deal with when facing the issue of the views attri- buted to the Vṛttikāra in the BSBh and its commentaries is the identity of this elusive authority. The appellation vṛttikāra appeared already in a famous passage of Śabara’s commentary on Mīmāṃsāsūtra 1.1.5, where we find a lengthy reused passage called Vṛttikāragrantha.8 Within this passage we also find a reused passage that the text attributes to Upavarṣa, and a secondary interpolation recognized by Frauwallner (1968: 109–110). Śabara also men- tioned the views of the Vṛttikāra in his commentary on MīmS 2.1.32–33, 2.2.26, 2.3.16, 3.1.6, 5.1.1 and 7.2.6. Ānandagiri and Govindānanda, in their commentaries on Śaṅkara’s BSBh 3.3.53, claimed that the Vṛttikāra is the same person as Upavarṣa. Nakamura (2004: 33–34), however, argued that Upavarṣa is not identical with the Vṛtti- kāra of Śabara’s Vṛttikāragrantha, for two reasons: (1.) The Vṛttikāragrantha quotes Upavarṣa (gakāraukāravisarjanīyā iti bhagavān upavarṣaḥ). (2.) Śaṅ- kara cites Upavarṣa approvingly, whereas he refutes the Vṛttikāra. The se- cond argument is problematic insofar as it is based on the claim that the Vṛtti- kāra of the Vṛttikāragrantha is the same person as the Vṛttikāra mentioned in the BSBh. Rāmānuja in his Śrībhāṣya ad BS 1.1.1 announced his intention to explain the BS in accordance with the previous teachers who had condensed the commentary (vṛtti) composed by Bodhāyana. Later he mentioned passages from the vṛtti or the Vṛttikāra six times (ŚBh 1.1.1 twice; 1.1.10; 1.2.1; 1.3.7; 1.3.32). Although these passages do not correspond exactly to the passages attributed by Śaṅkara’s sub-commentators to the Vṛttikāra, Nakamura (2004: 76–77) has argued that Rāmānuja’s Vṛttikāra is identical with the Vṛttikāra of Śaṅkara’s sub-commentators and that his name was Bodhāyana.9 Aklujkar (2010), however, has argued convincingly that Rāmānuja’s Vṛt- tikāra cannot be Bodhāyana, but must be a later scholar, namely Upavarṣa, who abridged Bodhāyana’s voluminous text. Aklujkar further claimed (2010: 18) that Śabara’s Vṛttikāra is also identical with Upavarṣa. Ānandagiri and Govindānanda mentioned a Vṛttikāra or Vṛttikṛt in their commentaries on ten passages of the BSBh.10 In one of them (1.1.4, Gov. in 8 See Frauwallner (1968: 24,16 ff.). 9 Thibaut (1890[I]: xxi) remarked cautiously that there is no reason to doubt that an ancient vṛtti (commentary) connected with the name Bodhāyana existed. 10 BSBh 1.1.1 (Gov. in BSŚWC, p. 19; this passage is also discussed in the Bhāmatī (BS- ŚWC, p. 38)); 1.1.4 (Gov. in BSŚWC, p. 65,15 and p. 66,4); 1.1.4 (Gov. in BSŚWC, p. 85,14); 1.1.12 (Gov. in BSŚWC, p. 119,19); 1.1.19 (Ānand. and Gov. in BSŚWC, p. 125), 1.1.23 (Ānand. and Gov. in BSŚWC, p. 140), 1.1.25 (Gov. in BSŚWC, p.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 117 BSŚWC, p. 85,14), the discussion of the views of the Vṛttikāra is directly followed by one about the opinions of the prābhākaras “followers of Prabhā- kara’s school of Pūrva Mīmāṃsā,” which might be an indirect hint at the affiliation of the Vṛttikāra.11 An additional piece of information concerning the philosophical orientation of the Vṛttikāra can be gathered from Bhāskara’s commentary on BSBh1.4.26, p. 144, which mentions that the Vṛttikāra accepted the theory of transformation of Brahman (pariṇāma). Let me now discuss the instances in which the sub-commentators claimed that Śaṅkara presented the views of the Vṛttikāra, which fall into two catego- ries. (1.) In some cases Śaṅkara seemingly refuted the position attributed by the commentaries to the Vṛttikāra. For instance, Vācaspati Miśra, in his Bhā- matī ad BSBh 1.1.1 (BSŚWC p. 38,26), claimed that Śaṅkara refuted a posi- tion of the Vṛttikṛt, according to which the word-stem brahma- within the tat- puruṣa compound brahma-jijñāsā (“a wish to know Brahman”) is to be un- derstood as being in the dative case. However, Śaṅkara only explained that he took the word-stem brahma- as being in the genitive case, and did not in- dicate that this explanation was meant as a refutation of a different interpreta- tion of the compound. (2.) In other cases, Śaṅkara does not seem to refute the Vṛttikāra. For instance, in his Bhāmatī ad BSBh 1.1.31 Vācaspati, as well as Ānandagiri and Govindānanda, attributed this entire portion of Śaṅkara’s text (WOŚ pp. 61,24–62,18) to the Vṛttikāra. The Vṛttikāra here discussed the threefold “meditation” (vidyā) on Brahman of KauU 3.2–4. This meditation takes place with the help of breath (prāṇa) and intelligence (prajñā), with Brahman considered the third part of this vidyā. Śaṅkara did not refute the explanation of the Vṛttikāra. It should be noted that Govindānanda attributed an objection in BSBh 1.1.12 to the Vṛttikāra (Gov. in BSŚWC, p. 119,19), namely that the sub- stance defined as ānandamaya “abundant of bliss” is the individual soul (jīva).12 It is, however, unlikely that this attribution is correct, because the first chapter of the BS and its commentaries focus on the harmonization of various Upaniṣadic passages by interpreting them all as referring to the Brahman. It would thus be odd to think that an ancient commentator of the BS might want to identify ānandamaya with a different entity.13 Moreover, 149,16), 1.1.31 (Vācasp., Ānand. and Gov. in BSŚWC, p. 158), 1.2.23 (Ānand. and Gov. in BSŚWC, p. 182), 1.4.3 (Ānand. and Gov. in BSŚWC, p. 298). 11 More precisely, according to Govindānanda in BSŚWC, p. 85, Śaṅkara finished the discussion started by the Vṛttikāra on p. 65, thereupon began a discussion with the prā- bhākaras. 12 On this translation of the word ānanadamaya, see below, section 4.1. 13 Throughout the first chapter of the Brahma-Sūtra, particular passages from different 118 Ivan Andrijanić this attribution is in strong contradiction to Ānandagiri’s and Govindānanda’s later claim that all of Śaṅkara’s text from 1.1.12–19 belongs to the Vṛttikāra. In fact, in BSBh 1.1.12–19, attributed by Ānandagiri and Govindānanda him- self to the Vṛttikāra, it is claimed that the ānandamaya is the highest Brah- man. This clear contradiction puts the reliability of Govindānanda as a source of information on the Vṛttikāra and his text into question. Since Govindā- nanda lived in the 17th century and thus much later than Śaṅkara, his claims must be taken with caution. 3 Different interpretations of the same sūtras There are some other passages in Śaṅkara’s BSBh that contain alternative interpretations of one and the same sūtra. These passages are not introduced with an indefinite pronoun and the sub-commentators did not identify them as views of other authors. Although Śaṅkara did not attribute these interpreta- tions to earlier sources, they may nevertheless go back to earlier authors whose views Śaṅkara recorded without criticism. An example occurs in BSBh 2.2.39 and 2.2.40, where Śaṅkara introduced an alternative explanation for both sūtras with the words anyathā vā sūtradvayaṃ vyākhyate14 (“alterna- tively, these two sūtras are differently explained [as follows]”).15 In addition, Śaṅkara introduced an alternative explanation of the sūtras 2.4.5 and 2.4.6 with the expression iyam aparā sūtradvayayojanā16 (“this is another way to construct these two sūtras”). Further, he introduced an alternative interpreta- tion of the second part of the sūtra 3.1.7 with the expression aparā vyākhyā17 (“another explanation”).18 Then, in BSBh 3.2.2419 we find an alternative in- terpretation of the second part of the sūtra. Finally, there is an alternative ex- Upaniṣads are discussed. Very frequently, the discussion is about whether a particular passage of the Upaniṣads is speaking about Brahman or something else. The commen- tators usually held the opinion that the Upaniṣads are speaking about Brahman, and object to the idea that they are speaking about some other entity such as the individual soul or pradhāna. 14 BSBh 2.2.40 (WOŚ p. 258,6). The first interpretation ranges from p. 257,24 to p. 258,6; the second from p. 258,6 to p. 258,15. 15 The sub-commentators did not identify the author of any of the explanations. 16 BSBh 2.4.6 (WOŚ p. 310,16). 17 BSBh 3.1.7 (WOŚ p. 339,8). 18 Ānandagiri (ad BSBh 3.1.7, BSŚWC, p. 602) states that the first interpretation was primary. 19 BSBh 3.2.22 (WOŚ p. 366,12).  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 119 planation of the word pādavat in BS 3.2.33, introduced with the conjunction athavā (“or”).20 For the purpose of identifying possible instances of reuse, BSBh 3.3 is also relevant because together with the first chapter, this section represents another chiefly exegetical part of Śaṅkara’s text.21 Most of BSBh 3.3 dis- cusses whether different śruti passages should be combined and used in meditation. In this connection, various tools are implemented for establishing whether different Upaniṣadic passages correspond to each other or not. The last four passages, in which Śaṅkara introduced alternative explanations of sūtras, also occur in this section. 4 Examples of reuse 4.1 The case of ānandamaya in Brahmasūtrabhāṣya 1.1.12–1.1.19 The first example of an older, reused text passage that I shall present here is the one most evident in Śaṅkara’s commentary. Moreover, its discussion is of relevance for establishing a relative chronology between different authorities in the area of Advaita Vedānta and beyond. In BS 1.1.12–1.1.19 we find two conflicting interpretations of the word ānandamaya in Taittirīya-Upaniṣad 2.5.22 The discussion focuses on whether ānandamaya refers to Brahman or 20 BSBh 3.2.33 (WOŚ p. 371,10). 21 Nakamura (1983: 436) has claimed that the first three pādas of the first chapter of the Brahma-Sūtra, called Samanvaya, and Brahma-Sūtra 3.3 represent the oldest core of the text and that it was composed at some time before the Christian era. 22 For Deussen (1883: 150–151) the second interpretation was possibly a later interpola- tion. If this were true, Deussen assumed, then the attribution of the Taittirīya-Upani- ṣad-Bhāṣya to Śaṅkara may not be correct, because the attribution is based on the identity of the teaching found in the Taittirīya-Upaniṣad-Bhāṣya with this second inter- pretation. Kanakura (1926: 383–385) held that this last part of BSBh 1.1.19 is not an interpolation due to the fact that Vācaspati Miśra commented on this text passage in his Bhāmatī. This argument is based on the claim that the time gap between Śaṅkara and Vācaspati was not long. Kanakura accepted the traditional date of Śaṅkara (788–820 CE), and he placed Vācaspati’s Bhāmatī to the mid-9th century based on his work Nyā- yasūcīnibandhana, which is dated to 898. Kanakura interpreted this as referring to the Vikrama age, which means the year 841. Acharya (2006: xviii–xxviii) has provided a review of the modern debate on Vācaspati’s dates, together with new evidence ac- cording to which the year 898 should be understood as Śaka, i.e., 976 CE, as Hacker (1951: 169) claimed. Acharya (2006: xxviii) has concluded that Vācaspati flourished between 950 and 1000 CE. The gap between Śaṅkara and Vācaspati may be much longer than Kanakura assumed. In my opinion, the second interpretation is certainly not an interpolation, because there is no doubt that the TaittUBh is a genuine work of 120 Ivan Andrijanić to the individual soul. In the first part, which comprises the commentaries on BS 1.1.12–1.1.19, it is claimed that ānandamaya designates Brahman as the supreme Self. This position is defended against objections, according to which ānandamaya does not refer to the supreme Self but to a secondary Self (amukhyātman) or bodily Self (śārīrātman). Śaṅkara defended this claim up to the second part of his commentary on BS 1.1.19 (WOŚ pp. 40,6ff.), when he offered a completely different interpretation of ānandamaya and claimed that the term does not refer to Brahman at all. 4.1.1 The introduction of the adhikaraṇa As was already pointed out above in section 2.2, Vācaspati, Ānandagiri and Govindānanda regarded the text of BSBh 1.1.12–19 up to idaṃ tv iha vakta- vyam (“but here this has to be said,” WOŚ p. 40,16) as belonging to an older commentary. Before starting his commentary on sūtra 1.1.12, Śaṅkara pro- vided a detailed introduction to the entire adhikaraṇa (WOŚ from p. 34,12 to the end of p. 35), in which he explained his well-known illusionistic doctrine of two forms of Brahman, a lower one differentiated by limiting adjuncts and a highest one devoid of any limiting adjuncts.23 Śaṅkara claimed at the end of his introduction to the adhikaraṇa that in some places the Upaniṣads teach that Brahman is distinguished by limiting adjuncts and is to be worshiped (upāsya), and that in others they teach that the highest Brahman must be known (jñeya) (WOŚ pp. 35,22ff.). I believe that the introduction is an origi- nal contribution of Śaṅkara’s, in which he foreshadowed his own, genuine interpretation, which he presented in full following BS 1.1.19. This is be- cause Bhāskara, in his introduction, immediately diverged from Śaṅkara and criticized him. This supposition might be corroborated by the fact that Vācas- pati, in his commentary on the introduction to the adhikaraṇa, did not men- tion an author other than Śaṅkara as the source, but immediately after the first sūtra (BS 1.1.12) claimed that Śaṅkara’s commentary on sūtra 1.1.12 con- veys the opinion of an ekadeśin (ekadeśimata). Śaṅkara. The main argument for this is that Sureśvara, who himself claimed that he was Śaṅkara’s direct disciple (BAUBhV 6.5.24, NaiS 4.76–77), wrote a commentary on the TaittUBh (see n. 5, above). Because of that, the time gap between Śaṅkara and Vācaspati is not particularly important. 23 Śaṅkara started with the words “Brahman is apprehended under two forms; in the first place as qualified by limiting conditions owing to the multiformity of the evolutions of name and form (i.e., the multiformity of the created world); in the second place as being opposite of this, i.e., free from all limiting conditions whatever” (dvirūpaṃ hi brahmāvagamyate, nāmarūpavikārabhedopādhiviśiṣṭaṃ, tadviparītaṃ ca sarvopādhi- vivarjitam BSBh 1.1.12, WOŚ, pp. 34,16f. as translated in Thibaut 1890: 61) and fur- ther explained this theory by strengthening it with quotations from śruti and smṛti.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 121 4.1.2 Brahmasūtrabhāṣya 1.1.12 The commentary on BS 1.1.12 strictly follows the five-fold structure com- mon in Mīmāṃsā texts, starting with (1.) an introduction (traditionally called viṣayavākya “statement of the topic”) to a passage from TaittU 2.1–5, which deals with a row of selves consisting of the essence of food (annarasamaya), of breath (prāṇamaya), of mind (manomaya), of intelligence24 (vijñānamaya) and of bliss (ānandamaya). (2.) Next, the commentary discusses the doubt (tatra saṃśayaḥ) whether ānandamaya is to be understood as the highest Brahman (param eva brahma) or something else that is similar to the other four selves. (3.) Immediately after the formulation of this doubt, an objection is raised according to which ānandamaya is a secondary self (amukhyātman). There are two arguments for this: (a) ānandamaya occurs in the same se- quence as annamaya (pravāhapatita); (b) some characteristics are attributed in the TaittU to ānandamaya which do not suit the Supreme Brahman, such as being embodied (śarīratva) and having pleasure as its head. (4.) These arguments are answered with a siddhānta, or settled conclusion, which in- cludes the sūtra and the claim that ānandamaya must be the highest Self (para evātmānandamayo bhavitum arhati). This is further explained in (5.), the part of the commentary called nirṇaya by Śaṅkara (BSBh 1.1.23, 1.2.1, 1.2.26 etc.), which explains this settled conclusion in detail. 4.1.3 Brahmasūtrabhāṣya 1.1.13–17 The text of the commentary on BS 1.1.13–17 tries to prove that ānandamaya is the highest Brahman against various objections, of which the most interest- ing is the claim that the suffix -maya denotes a modification (vikāra). This is answered already in BS 1.1.13 with the claim that the suffix -maya means “abundance” (prācurya); according to this argument, ānandamaya means “abundant bliss” or “in which bliss is abundantly established.”25 The argu- ment that -maya denotes abundance may originate from the Kāśikāvṛtti ad Pāṇini 5.4.21 (tat prakṛtavacane mayaṭ),26 where the word prakṛta is under- stood as prācuryeṇa prastutam (“abundantly established (?),” Bronkhorst 2004: 5). From the point of view of the relative chronology of the BS and its commentaries, it is important to note that the Kāśīkāvṛtti ad Pāṇini 5.4.21 uses the example of annamaya27 to illustrate that the suffix -maya means 24 Cf. Olivelle (1998: 303), who translated vijñānamaya as “consisting of perception.” 25 Bronkhorst 2004: 5. 26 “The taddhita affix mayaṭ occurs after a syntactically related nominal stem which ends in prathamā ‘nominative’ and signifies prakṛta ‘that which happens to be in abun- dance’” (Sharma 1999: 676). 27 The word annamaya appears also in ChU 6.5.4 and in TaittU 2.1 with the variation 122 Ivan Andrijanić abundance. Furthermore, the Kāśikāvṛtti provides the example annamayo yajñaḥ28 “a sacrifice in which food is abundant” from an unknown source which appears also in BSBh 1.1.13. From this parallel it follows that not only the commentator of BS 1.1.13 but also the author of this sūtra (where the word prācurya is mentioned) were acquainted with the interpretation preserved in the Kāśikāvṛtti. Bronkhorst (2004: 5–6) claims that at least one brahmanical commentary on Pāṇini on which the Kāśikā relied existed before BS 1.1.14 was composed. The Vṛttikāra most probably relied on such a lost source. In fact, if he reused the Kāśikāvṛtti, he would have to be dated after the 7th century CE, which would contradict the other evidence discussed above in section 2.2.1, which sug- gests a much earlier dating. 4.1.4 Brahmasūtrabhāṣya 1.1.17 The commentary on BS 1.1.17 can be divided into two parts. Its first part follows the sūtra, which most probably establishes a distinction (bheda) be- tween ānandamaya and the transmigrating individual soul (saṃsārin). This part of the text was interpreted in the same way by Śaṅkara and Bhāskara, whose commentary is simply an abridged version of Śaṅkara’s work. Rüping (1977: 34–35) presented this example in order to show how Bhāskara’s text is secondary to Śaṅkara’s. However, in the second part of Śaṅkara’s text we encounter Śaṅkara’s typical illusionistic doctrine that is also found in the introduction to BS 1.1.12. In this second part, Śaṅkara claimed that the dis- tinction between the ānandamaya and the saṃsārin is real only according to a worldly view (laukika dṛṣṭa) and that it depends on limiting adjuncts (upadhi) in the same way as a pot limits the air it contains. Here, Bhāskara abruptly stopped following Śaṅkara and criticized his idea that the difference between the individual soul and God is illusory and that it depends on limiting ad- juncts. He introduced his criticism with the words atra kecit svamatikalpitadarśanaparitrāṇāya sūtrārthaṃ vinaśayanto vyācakṣate (Kato 2011: 42). Here some, in order to protect their own fabricated view, explain by destroying the meaning of the sūtra. annarasamaya. 28 There, annamayo yajñaḥ is glossed with annaṃ prakṛtam (“food that is abundantly es- tablished”), and the Kāśikā-Vṛtti glosses prakṛta with prācuryeṇa prastutam (“abun- dantly established”).  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 123 Bhāskara’s disapproval of Śaṅkara’s view may indicate that Śaṅkara’s teaching deviates from the traditional interpretation of BS 1.1.17. 4.1.5 Brahmasūtrabhāṣya 1.1.19 In his commentary on BS 1.1.19, Śaṅkara concluded his teaching of ānanda- maya as being identical with the highest Self (tasmād ānandamayaḥ para- mātmeti sthitam).29 Immediately thereafter, Śaṅkara introduced a contrary interpretation with the words idaṃ tv iha vaktavyam “but here this has to be said.” He started his critique with the same arguments presented by the ob- jector in BSBh 1.1.12–13. Śaṅkara’s arguments that ānandamaya is not the highest Self are: a) the suffix -maya in the sequence annarasamaya, prāṇa- maya, manomaya and vijñānamaya cannot at first denote a modification and then suddenly mean “abundance” in the compound ānandamaya; b) ānanda- maya occurs in the same sequence as the other four terms and thus belongs to the same ontological category; c) ānandamaya is mentioned in TaittU 2.5 as having pleasure as its head,30 but since pleasure is not a predicate of Brah- man, ānandamaya and Brahman cannot be identical; d) ānandamaya is not designated as Brahman in the TaittU and Brahman is actually mentioned in TaittU 2.5 as brahma pucchaṃ pratiṣṭhā (“the bottom on which it rests is the brahman.” [Trans. Olivelle 1998: 305]). According to Śaṅkara, Brahman in the highest sense is only this bottom or tail, on which ānandamaya rests. These arguments correspond closely to the three arguments the objector raised at the beginning of BSBh 1.1.12 (cf. section 1.1.2): (1.) ānandamaya occurs in the same sequence of terms as annamaya, etc. (this corresponds to Śaṅkara’s argument b); (2.) some characteristics of ānandamaya, such as being embodied and having pleasure as its head cannot be applied fittingly to the Supreme Brahman (this corresponds to Śaṅkara’s argument c); (3.) the suffix -maya denotes a modification (vikāra) (this corresponds to Śaṅkara’s argument a). In his TaittUBh 2.5, Śaṅkara provided the same four arguments for the claim that ānandamaya is not the highest but the lower Self (kāryātman; “self which has to be accomplished” or “active self”) with minor modifications.31 Śaṅkara added here one more argument that we do not find in the BSBh, namely that ānandamaya is something to be attained. 29 The first interpretation starts in WOŚ on p. 39,21 and finishes on p. 40,6. 30 In his translation of TaittU 2.5, Olivelle (1998: 305) has translated priya as “pleasure.” 31 Regarding argument a), Śaṅkara’s expression is adhikārapatita, while the objector in BSBh uses the word pravāhapatita. 124 Ivan Andrijanić From all this it becomes clear that Śaṅkara used the arguments of an ob- jector to dispute some old, traditional interpretations which are preserved in parts of his commentary on BSBh 1.1.12–19 (WOŚ pp. 34,12–39,3; pp. 39,15–40,6). Therefore I think it safe to assume that Śaṅkara’s second inter- pretation from BSBh 1.1.19 and his interpretation in TaittUBh 2.5 represent his genuine understanding of ānandamaya, while he borrowed his previous claims in BSBh 1.1.12–19 from a different source. There are two more arguments that support this assertion. The first is that Vācaspati and Ānandagiri held the same opinion in their respective commen- taries. According to Vācaspati, the commentary on BS 1.1.12 begins the ex- position of the opinion of a person called ekadeśin, which runs up to the words idaṃ tv iha vaktavyam. After this, Śaṅkara begins with his own inter- pretation (svamata). Ānandagiri claims that the part of the commentary start- ing with the word taittirīyaka32 (the first word of Śaṅkara’s commentary on BS 1.1.12) and ending with tasmāt (tasmād ānandamayaḥ paramātmeti sthi- tam “therefore it stands that ānandamaya is the highest Self”) in BSBh 1.1.19 was composed by a person called Vṛttikṛt (vṛttikṛtāṃ matam upasaṃharati – tasmād iti “he summarizes the view of the Vṛttikṛt in the sentence starting with tasmāt”).33 The second argument for the assertion that we are dealing with a reused text passage can be derived from Bhāskara’s commentary. He began his commentary on BS 1.1.19 using almost the same words as Śaṅkara. Then, however, he departed from Śaṅkara’s text as soon as Śaṅkara presented his own interpretation:34 Table 2: The commentaries of Śaṅkara and Bhāskara ad BS 1.1.19 Śaṅkara ad BS 1.1.19 (WOŚ pp. Bhāskara ad BS 1.1.19 (BSBh[Bh] 39f.) p. 43) itaś ca na pradhāne jīve vānanda- itaś cānandamayaḥ | paro na saṃsā- mayaśabdaḥ | yasmād asminn āna- rī na pradhānam | asminn ānanda- ndamaye prakṛta ātmani pratibud- maye prakṛte ’sya jīvasya pratibud- dhasyāsya jīvasya tadyogaṃ śāsti | dhasya tadyogaṃ tena yogaṃ muk- tadātmanā yogas tadyogaḥ, tad- tiṃ śāsti śāstram | “yadā hy evaiṣa bhāvāpattiḥ | muktir ity arthaḥ | etasminn adṛśye ’nātmye ’nirukte tadyogaṃ śāsti śāstraṃ- “yadā hy ’nilayane ’bhayaṃ pratiṣṭhāṃ vin- 32 BSŚWC, p. 119,31. 33 BSŚWC, p. 124,30. 34 Identical words are underlined, quotations from BS 1.1.19 are set in bold.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 125 evaiṣa etasminn adṛśye ’nātmye date | atha so ’bhayaṃ gato bhava- ’nirukte ’nilayane ’bhayaṃ prati- ti” iti (TaittU 2.7) | tad etasmin muk- ṣṭhāṃ vindate | atha so ’bhayaṃ tiśāsanaṃ paramātmaparigrahe ’va- gato bhavati | yadā hy evaiṣa etas- kalpate nānyatheti sthitam | minn udaramantaraṃ kurute | atha tasya bhayaṃ bhavati” (TaittU 2.7) iti | etad uktaṃ bhavati – yadaitasminn ānandamaye ’lpam apy antaram atādātmyarūpaṃ paśyati tadā saṃ- sārabhayān na nivartate | yadā tv etasminn ānandamaye nirantaraṃ tādātmyena pratitiṣṭhati tadā saṃ- sārabhayān nivartata iti | tac ca paramātmaparigrahe ghaṭate, na pradhānaparigrahe jīvaparigrahe vā | tasmād ānandamayaḥ para- mātmeti sthitam | idaṃ tv iha vaktavyam ... atra kecid imaṃ siddhāntaṃ dūṣa- yitvā ... The opening paragraph of the commentary is almost identical in the two versions, although Bhāskara omitted the latter part of Śaṅkara’s text. The situation changes when Śaṅkara starts his real interpretation of ānandamaya with the words idaṃ tv iha vaktavyam “but this has to be said here”: this interpretation is rejected by Bhāskara, by his saying atra kecid imaṃ siddhān- taṃ dūṣayitvā “here some35 (i.e., Śaṅkara) corrupt this settled conclusion ....” The rest of Bhāskara’s text departs completely from Śaṅkara’s, advocates a bhedābheda theory, and criticizes Śaṅkara’s differentiation between a lower and a higher Brahman. Bhāskara especially criticized Śaṅkara’s claim that the phrase puccham pratiṣṭhā in the TaittU refers to Brahman. This example fulfills the criteria for the identification of reused material set at the beginning of this chapter (section 1): a) one of the two conflicting interpretations is designated by all three sub-commentators as being taken from some other source, which Ānandagiri and Govindānanda identified as the work of the Vṛttikāra; b) this interpretation is in conflict both with Śaṅ- kara’s other interpretation and with his interpretation of the same passage in 35 Bhāskara used the honorific plural form when addressing Śaṅkara. 126 Ivan Andrijanić the TaittUBh; c) Bhāskara followed the first interpretation and criticized the second; d) the first interpretation deviates from the traits typically associated with Śaṅkara’s doctrine. Furthermore, it is possible to propose a structural analysis of the Ānanda- mayādhikaraṇa by distinguishing five segments of the text. (1) The intro- duction (WOŚ pp. 34,11–35,26) appears to be Śaṅkara’s genuine contribution to this section, because it contains philosophical positions that are typical of him. Moreover, Bhāskara criticized these positions, and Vācaspati and Ānan- dagiri set this section apart from the BSBh text occurring after the sūtra BS 1.1.12. (2) The commentaries from BSBh 1.1.12 (WOŚ p. 35,1) up to the end of the first part of BSBh 1.1.17 (WOŚ p. 39,3), by contrast, appear to be reused from an unknown source. (3) After this passage, Śaṅkara again pre- sented his own teaching, which Bhāskara criticized. Śaṅkara’s genuine ex- position ends at p. 39,13, where (4) he reused again an older source up to WOŚ p. 40,6, where (5) we finally encounter Śaṅkara’s real interpretation and a critique of everything stated in segment (2). Assuming that the interpretation in segment 2 comes from an older com- mentary (and that there is thus no need to harmonize its philosophical content with Śaṅkara’s teachings), the doctrine represented in it is the following: It does not distinguish between a lower and the highest Brahman, but fo- cuses on the difference between the individual soul (jīva, saṃsārin) and Brahman; it contains no trace of an argument for the illusory nature of the world, and relies on a grammatical interpretation found in the Kāśikāvṛtti ac- cording to which the suffix -maya stands for abundance (see above, section 4.1.3). Bhāskara interpreted this older part of the commentary in the frame- work of his bhedābhedavāda, and it is possible that the philosophical stand- point of this earlier commentary indeed represents some form of bhedābhe- davāda. It should be noted that the text passage designated by Śaṅkara’s sub- commentators as belonging to an older commentary displays some formal peculiarities: The commentary on the opening sūtra in the adhikaraṇa bears in fact a five-part structure. This same structure is followed throughout the first chapter of the BSBh and in BSBh 3.3. It is exactly in these parts that we find almost all references to other interpretations and all references to a Vṛttikāra. 4.2 The “bridge” (setu) from BS(Bh) 1.3.1 and MU(Bh) 2.2.5 In BSBh 1.3.1 we find a reference to the view of “others,” which the sub- commentators did not attribute to the Vṛttikāra. The question in BSBh 1.3.1–  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 127 7 is whether the expression “the abode (āyatana)36 of heaven, earth and oth- ers” (dyubhvādi) from Muṇḍaka-Upaniṣad 2.2.5 refers to the highest Brah- man or not. At the end of BSBh 1.3.1, Śaṅkara referred to someone else’s interpretation of the word “bridge” (setu), which also appears in the MU. The text reads (WOŚ pp. 95,9ff.): apara āha – “tam evaikaṃ jānatha ātmānam” (MU 2.2.5) iti yad etat saṃkīrtitam ātmajñānam, yac caitad “anyā vāco vimuñcatha” (MU 2.2.5) iti vāgvimocanam, tad atrāmṛtatvasādhanatvāt, “amṛtasyaiṣa setuḥ” (MU 2.2.5) iti setuśrutyā saṃkīrtyate na tu dyubhvādyāyata- nam. Someone else says that what is referred to with the words “him alone know as the one Soul”37 (MU 2.2.5) is knowledge of the Self, and [what is referred to with the words] “other words dismiss” (MU 2.2.5) is the [command] to put away words, not the abode of heaven, earth and so on. In fact, [these two] are referred to through the mention of the word “bridge” in “he is the bridge to immortality” (MU 2.2.5), be- cause here (in MU 2.2.5) [these] are the means for immortality. From this brief account we understand that knowledge of the Self and the abandoning of talk are said to be a “bridge” because they lead to immortality, whereas the word “bridge” does not refer to āyatana (abode of heaven, etc.). It seems that this additional interpretation is not at odds with what was stated in earlier parts of the text, namely that āyatana is Brahman. Śaṅkara apparently introduced this interpretation to answer the objection that the term “bridge” cannot define the abode of heaven, etc. as Brahman. Śaṅkara’s first answer to this objection is that the word setu (“bridge”) comes from the verbal root √SI, “to bind” and that here a bond is meant, which means that the word “bridge” should be taken metaphorically as something that binds or supports. Thus, Śaṅkara reused this interpretation of the word setu according to which the word setu should not be connected with the first part of MU 2.2.5, which speaks about the abode of heaven, but with the second part, which speaks about reaching immortality. If we look at Śaṅ- kara’s commentary on MU 2.2.5, we find an interpretation which is actually the same as the opinion attributed to “someone else” in BSBh 1.3.1: setu is glossed with the word ātmajñāna (“knowledge of the Self”), as in the text- 36 The word “abode” (āyatana) from BS 1.3.1 does not appear in the Upaniṣad itself. The word āyatana in the sūtra can be understood as a gloss on the indefinite pronoun yas- min (“on which”) in MU 2.2.5. 37 Translations of the quotations from MU are according to Hume (1965: 372). 128 Ivan Andrijanić passage analyzed above, and the second pāda of MU 2.2.5 is understood as a way to reach immortality. This means that Śaṅkara embedded this inter- pretation in his commentary on the MU. It cannot be the other way around, because it is hard to believe that Śaṅkara would take his own interpretation from his MUBh and then embed it in BSBh 1.3.1 as an opinion of someone else. This might be an indication that the MUBh is later than the BSBh. The first part of the text of BSBh 1.3.1 resembles the text designated as belonging to an older source in the Ānandamayādhikaraṇa, because it is structured in the same way as the first interpretation of ānandamaya: a) text example, b) doubt, c) objection, d) settled conclusion and e) explanation of the settled conclusion. Further evidence in support of the hypothesis that it was taken from an older source is the fact that it does not share any typical tenets of Śaṅkara’s philosophy and that Bhāskara followed this part of the commentary almost verbatim. By contrast, after this beginning part, Bhāskara said that others consider a bridge to represent the knowledge of the Self (in- deed Śaṅkara did gloss setu as ātmajñāna in MUBh 2.2.5) and then criticized this interpretation as unsuitable (ayukta). It should be noted that Bhāskara abbreviated Śaṅkara’s second interpretation and that he omitted an important part of this interpretation, namely the part which speaks about the abandon- ment of speech. Because of this I suppose that Bhāskara did not know the source to the same extent as Śaṅkara did, and that he thus just shortened Śaṅ- kara’s reference and added his critique. As in the case discussed in section 4.1, Bhāskara agreed with Śaṅkara in the first interpretation but fought the second one. In both cases, Śaṅkara in- corporated the second interpretation in his Upaniṣad commentary. The differ- ence is that the second interpretation in the Ānandamayādhikaraṇa is Śaṅkara’s own, whereas here it is, like the first interpretation, also taken from somewhere else. If we suppose that the first part of this interpretation, followed by Śaṅkara and Bhāskara, was taken from an older source, to whom does the second interpretation belong? It is possible that Śaṅkara had more than one source available; he did not criticize this alternative interpretation in the BSBh and followed it closely in the MUBh. This may suggest a source different from the one he used in composing the first part of BSBh 1.3.1. It is also possible that this text-passage was taken from a source other than the traditional source Śaṅkara used in the interpretation of ānandamaya, since he ultimately adopted it. The third possibility is that the view of “others” in BSBh 1.3.1 was already part of the old source Śaṅkara reused. Further, we may suggest that Śaṅkara’s commentary on the MU is later than the BSBh, since Śaṅkara  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 129 adopted and incorporated this second interpretation in his Upaniṣad com- mentary. 5 Conclusions and outlook for further research The goal of this investigation was to throw some light on the history of early Vedānta philosophy and its lost bhedābhedavāda period before Bhāskara. It also aimed at identifying possible reasons of seeming ambiguities and impre- cisions in Śaṅkara’s terminology. If it were possible to prove that these contradictions and terminological imprecisions are the result of a complex history leading to the BSBh being compiled also on the basis of earlier sources, we would know more about the history of the Advaita doctrine. But why did Śaṅkara embed this older material that is alien to his doctrine in his works? I would speculatively say that it was a “must” for the young Śaṅkara to embed traditional, well-respected source(s) into his text in order to root his Advaita doctrine firmly in the Vedāntic tradition, while at the same time introducing his monistic and illusionistic ideas. Elaborating upon this conjecture may throw further light on the role of authoritative texts apart from śruti and smṛti in early Vedānta. Moreover, was this material only orally transmitted or did it belong to written sources? The strict five-fold structure of the adhikaraṇas, which is recognizable in the supposed older material, may suggest that we are dealing with written texts. The presence of attestations of the same passage in differ- ent Sanskrit sources may also strengthen this hypothesis. Possible candidates for the original composer of the passages reused by Śaṅkara are the authors mentioned by later Vedāntins. Rāmānuja spoke of Bodhāyana’s commentary and the commentaries of older teachers in ŚBh 1.1.1. In Vedārthasaṅgraha 130 he mentioned again the ancient commenta- tors Bodhāyana, Ṭaṅka, Dramiḍa,38 Guhadeva, Kapardin and Bhāruci. The Prapañcahṛdaya (Aklujkar 2010: 9f.) mentions the commentaries of Bo- dhāyana, Upavarṣa, Devasvāmin, Bhavadāsa and Śabara, before the ones by Bhagavatpāda (Śaṅkara), Brahmadatta and Bhāskara.39 Yāmuna in the Sid- dhitraya (ST, p. 4, quoted in Nakamura 2004: 3f.) also mentioned earlier 38 Rāmānuja also quoted the commentary of Dramiḍa in ŚBh 2.1.14 and 2.2.3 as well as in Vedārthasaṅgraha 173, 186, 195. 39 The Prapañcahṛdaya states that Devasvāmin and Bhavadāsa authored a commentary on the MīmS together with the Saṃkarṣa-Kāṇḍa, although Śabara did not comment on the latter. Śaṅkara, Brahmadatta and Bhāskara composed commentaries only on the BS. 130 Ivan Andrijanić teachers who wrote on Vedānta.40 From all this it is possible to infer that older (possibly written) sources existed, although the question remains of whether and to what extent they were available to Śaṅkara. References Abbreviations Ānand. Ānandagiri BAUBh Bṛhadāraṇyaka-Upaniṣadbhāṣya (Śaṅkara) BAUBhV Bṛhadāraṇyaka-Upaniṣadbhāṣyavārttika (Sureśvara) BS Brahmasūtra BSBh Brahmasūtrabhāṣya (Śaṅkara) BSBh[Bh] Śārīrakamīmāṃsābhāṣya (Bhāskara) in Kato 2011 BSŚWC see Shastri 1980 ChU Chāndogya-Upaniṣad Gov. Govindānanda KauU Kauṣītaki-Upaniṣad MīmS Mīmāṃsā-Sūtra MU Muṇḍaka-Upaniṣad MUBh Muṇḍaka-Upaniṣadbhāṣya (Śaṅkara) NaiS Naiṣkarmyasiddhi (Sureśvara) ŚBh Śrībhāṣya (Rāmānuja) ST Siddhitraya (Yāmuna), in Aṇṇaṅgarācārya 1944 TaittU Taittirīya-Upaniṣad TaittUBh Taittirīya-Upaniṣadbhāṣya (Śaṅkara) Vācasp. Vācaspati Miśra WOŚ Brahmasūtra with Śāṅkarabhāṣya (Works of Śaṅkarācārya) in Śaṅka- rācārya 1965 Primary sources Aṇṇaṅgarācārya 1944 Aṇṇaṅgarācārya, P. B., ed. Siddhitrayam. BhagavadYāmu- namunibhir anugṛhītam. Madras: Liberty Press, 1944. Kato 2011 Kato, Takahiro. “The First Two Chapters of Bhāskara’s Śārīrakamīmāṃsābhāṣya ‒– Critically Edited with an In- troduction, Notes and an Appendix.” Diss. Martin-Luther- Universität, Halle-Wittenberg. 26 October 2016. <http://d- nb. info/1031915001/34> Raghavachar 2002 Raghavachar, S. S. Vedārthasaṅgraha of Śrī Rāmānujā- cārya. Kolkata: Advaita Ashrama, 2002. 40 These are Śrīvatsāṅka Miśra, who composed a commentary on the BS, Ṭaṅka, Bhartṛ- prapañca, Bhartṛmitra, Bhartṛhari, Śaṅkara, Brahmadatta, Śrīvatsāṅka and Bhāskara.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 131 Śaṅkarācārya 1965 Śaṅkarācārya. Brahmasūtra with Śāṅkarabhāṣya, Works of Śaṅkarācārya in Original Sanskrit. Vol. 3. Delhi: Motilal Banarsidass, 1965 (reprint 2007). Shastri 1980 Shastri, J. L., ed. Brahmasūtra-Śāṅkarabhāṣyam: With the Commentaries Bhāṣyaratnaprabhā of Govindananda, Bhāmatī of Vācaspati Miśra, Nyāyanirṇaya of Ānandagiri. Delhi: Motilal Banarsidass (Reprint [2010] Delhi: Motilal Banarsidass). Vedārthasaṅgraha see Raghavachar 2002. Secondary sources Acharya 2006 Acharya, Diwakar, ed. Vācaspatimiśra’s Tattvasamīkṣā: The Earliest Commentary on Maṇḍanamiśra’s Brahma- siddhi. Stuttgart: Franz Steiner Verlag, 2006. Publications of the Nepal Research Centre 25. Aklujkar 2010 Aklujkar, Ashok. “Vṛtti and Vṛttikāra in Rāmānuja’s Śrī- bhāṣya.” Anantaṁ Śāstram. Indological and Linguistic Studies in Honour of Bertil Tikkanen. Ed. Klaus Karttu- nen. Helsinki: Finnish Oriental Society, 2010. 3–20. Stu- dia Orientalia 108. Bronkhorst 2004 Bronkhorst, Johannes. “More on the Sources of Kāśikā.” Problems in Vedic and Sanskrit Literature (Ganesh Uma- kant Thite Felicitation Volume). Ed. Maitreyee Deshpan- de. Delhi: New Bharatiya Book Corporation, 2004. 47–54. 27 October 2016. <http://my.unil.ch/serval/document/ BIB_BC98BFE01B6F.pdf> 1–8. van Buitenen 1971 van Buitenen, J. A. B. Yāmuna’s Āgamaprāmāṇya or Treatise on the Validity of Pāñcarātra. Madras: Ramanuja Research Soc., 1971. Deussen 1883 Paul Deussen. Das System des Vedānta. Leipzig: F. A. Brockhaus, 1883. Frauwallner 1968 Frauwallner, Erich: Materialen zur ältesten Erkenntnis- lehre der Karmamīmāṃsā. Graz etc.: Böhlau, 1968. Öster- reichische Akademie der Wissenschaften. Sitzungsberich- te. Phil.-hist. Klasse 259.2. Veröffentlichungen der Kom- mission für Sprachen und Kulturen Süd- und Ostasiens 6. Hacker 1947 Hacker, Paul. “Śaṅkarācārya and Śaṅkarabhagavatpāda. Preliminary Remarks Concerning the Authorship Prob- lem.” New Indian Antiquary 9 (1947): 175–186. Hacker 1950 — “Eigentümlichkeiten der Lehre und Terminologie Śaṅ- karas: Avidyā, Nāmarūpa, Māyā, Īśvara.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 100 (1950): 246–286. (Rpt. in Schmithausen 1978: 69–109.) 132 Ivan Andrijanić Hacker 1951 — “Jayantabhaṭṭa und Vācaspatimiśra, ihre Zeit und ihre Bedeutung für die Chronologie des Vedānta.” Beiträge zur indischen Philologie und Altertumskunde. Walter Schu- bring zum 70. Geburtstag dargebracht von der deutschen Indologie. Hamburg: Cram, de Gruyter, 1951. 160–169. Alt- und Neu-Indische Studien 7. (Rpt. in Schmithausen 1978: 110–119.) Hacker 1953 — Vivarta: Studien zur Geschichte der illusionistischen Kosmologie und Erkenntnistheorie der Inder. Wiesbaden: Verlag der Akad. d. Wiss., 1953. Akad. d. Wissen. u. d. Lit. in Mainz. Abh. der geistes- und sozialwiss. Kl., Jahrgang 1953, Nr. 5. Harimoto 2006 Harimoto, Kengo. “The Date of Śaṅkara: Between the Cāḷukyas and the Rāṣṭrakūṭas.” Journal of Indological Studies 18 (2006): 86–111. Hume 1965 Hume, Robert Ernest, trans. The Thirteen Principal Upa- nishads. 6th impression of the 2nd rev. ed. from 1931 (1st ed. London 1921). London: Oxford University Press, 1965. Ingalls 1952 Ingalls, Daniel H. H. “The Study of Śaṅkarācārya.” Annals of the Bhandarkar Oriental Research Institute 33 (1952): 1–14. Ingalls 1954 — “Śaṅkara’s Arguments Against the Buddhists.” Philo- sophy East and West 3 (1954): 291–306. Kanakura 1926 Kanakura, Yensho. “Über die Interpolation des Śāṅkara- bhāṣya zum Brahmasūtra.” Beiträge zur Literaturwissen- schaft und Geistesgeschichte Indiens, Festgabe Hermann Jacobi. Ed. Willibald Kirfel. Bonn: Kommissionsverlag Fritz Klopp, 1926. 381–383. Maas 2013 Maas, Philipp André. “A Concise Historiography of Clas- sical Yoga Philosophy.” Periodization and Historiography of Indian Philosophy. Ed. Eli Franco. Vienna: Sammlung de Nobili, 2013. Publications of the De Nobili Research Library 37. 53–90. Mesquita 1973 Mesquita, Roque. “Yāmunamuni: Leben, Datierung und Werke.” Wiener Zeitschrift für die Kunde Südasiens 17 (1973): 177–193. Nakamura 1983 Nakamura, Hajime. A History of Early Vedānta Philoso- phy. Vol. 1. English Trans. (Japanese edition Tokyo 1950). Delhi: Motilal Banarsidass, 1983. Nakamura 2004 — A History of Early Vedānta Philosophy (Vol. II). English trans. (Japanese edition Tokyo 1950). Delhi: Motilal Banarsidass, 2004.  Traces of Reuse in Śaṅkara’s Commentary on the Brahmasūtra 133 Oberhammer 1977 Oberhammer, Gerhard. “An Unknown Source in Śaṅ- kara’s Refutation of the Pāñcarātra.” Annals of the Bhan- darkar Oriental Research Institute 58–59 (Diamond Jubilee Vol.) (1977–1978): 221–233. Olivelle 1998 Olivelle, Patrick. The Early Upaniṣads: Annotated Text and Translation. Oxford: Oxford University Press, 1998. Rüping 1977 Rüping, Klaus. Studien zur Frühgeschichte der Vedānta- Philosophie: Philologische Untersuchungen zu den Brah- masūtra-Kommentaren des Śankara und des Bhāskara. Wiesbaden: Steiner, 1977. Alt- und Neu-Indische Studien 17. Schmithausen 1978 Schmithausen, Lambert, ed. Paul Hacker: Kleine Schrif- ten. Stuttgart: Steiner, 1978. Glasenapp-Stiftung 15. Sharma 1999 Sharma, Rama Nath. The Aṣṭādhyāyī of Pāṇini. New Delhi: Munshiram Manoharlal, 1999. Thibaut 1890 George Thibaut, trans. Vedânta-sûtras with the Commen- tary by Saṅkarâkârya. Part I and II. Oxford: Clarendon Press, 1890. Sacred Books of the East 34 and 38. On Parallel Passages in the Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara* Yasutaka Muroya Vācaspati Miśra I (ca. 10th c.) is a well-known Indian philosopher1 and his Nyāyavārttikatātparyaṭīkā (NVTṬ) is one of the best known sub-commenta- ries on the Nyāyasūtra (NS). One aspect of Vācaspati’s work that has not received much scholarly attention up to now is its relationship to Bhaṭṭa Vāg- īśvara’s direct commentary on the NS called Nyāyasūtratātparyadīpikā (NTD). The work is, as Karin Preisendanz formulates, “the only completely preserved direct commentary on the Nyāyasūtra between the Nyāyabhāṣya and the fifteenth-century Nyāyatattvāloka.”2 Nonetheless it has in general not received the scholarly attention it deserves.3 This may result from the fact * Work on this chapter was supported by the the German Research Foundation (DFG) in the context of a research project (FR 2531/4-1 “Logic, Dialectics and Epistemology of the Nyāya Tradition”) at the University of Leipzig, and by the Austrian Science Fund (FWF) in the context of a research project (P27863 “Fragments of Indian Philosophy”) at the Austrian Academy of Sciences. A part of this chapter is a revised version of Mu- roya 2013 (in Japanese). I am much indebted to the late Muni Shree Jambuvijayajī as well as to the following institutions: Jaisalmer Lodravapur Parsvanath Jain Svetambara Trust (Jaisalmer), Government Oriental Manuscripts Library (Chennai), Oriental Re- search Institute & Manuscripts Library (Thiruvananthapuram), and Government Sans- krit College (Tripunithura, Ernakulam) for permitting access to their manuscript mate- rials. Moreover, I would like to express my gratitude to Professor Karin Preisendanz for reading an earlier draft and to her FWF Project (P24388 “Metaphysics and Episte- mology of the Nyāya Tradition III”) for permitting me to utilize copies of relevant ma- nuscripts. My sincere thanks are also due to Dr. Elisa Freschi and Dr. Philipp A. Maas for their detailed comments and valuable criticism. 1 For an overview of the assessment of Vācaspati’s commentarial activity by scholars such as Erich Frauwallner (1898–1974), Otto Strauss (1881–1940) and Helmuth von Glasenapp (1891–1963), see Preisendanz 2008: 604–606. For a detailed philological survey of the chronological and philosophical relationship of Vācaspati’s commenta- ries to other works, see Acharya 2006: xviii–xxx, xlv–lxvii. 2 Preisendanz 2005: 62. 3 However, for a philological examination of the sūtras from the first daily lesson (āhni- ka) of the third chapter (adhyāya) of the NS as transmitted in the NTD, see Preisendanz 1994 (in a number of notes; see, e.g., 254, n. 52 and 558, n. 184 on the textual problem 136 Yasutaka Muroya that the NTD contains many passages that are similar to, or parallel with, pas- sages found in Vācaspati’s work. This textual agreement has been interpreted as a sign for the strong indebtedness of the NTD to the NVTṬ. However, a closer look at the relevant passages shows that the assumption of Vācaspati’s antecedence must be questioned. The textual evidence on which this supposi- tion is based requires historical and philological analyses from a wider per- spective. The present chapter takes a step in this direction and examines the historical and intellectual relationship between Vācaspati and Vāgīśvara by drawing upon parallel passages in their two commentaries. 1 Bhaṭṭa Vāgīśvara’s Nyāyasūtratātparyadīpikā Vāgīśvara’s commentary was edited by Kishore Nath Jha and published in Allahabad in 1979 under the title of Nyāyatātparyadīpikā. In his editorial notes, the learned pandit and scholar from Mithila surveyed Vāgīśvara’s al- leged textual sources. He presented a number of parallel and thematically related passages found in other Nyāya works such as Vātsyāyana’s Nyāya- bhāṣya (NBh), Uddyotakara’s Nyāyavārttika (NV) and Vācaspati’s NVTṬ. Anantalal Thakur, who had procured a copy of the manuscript from Chen- nai that served as the basis of the edition, provided a detailed introduction (prastāvanā) to the edition, in which he correctly called the work Nyāya- sūtratātparyadīpikā in correspondence to the information attested in the colo- phons. Thakur made some assertions that received wide acceptance in subse- quent scholarly literature. In accordance with Jha, Thakur said that Vāgīśvara aimed at providing a summary of the main points (sārasaṃgraha) of the three quasi-canonical commentaries of the NS (nyāyamūlagranthatritaya).4 He also proposed dating Vāgīśvara after Vācaspati, because he thought that the NTD reused passages from the NVTṬ and earlier commentaries. He then placed Vāgīśvara before Udayana, because Vāgīśvara appears to be unfami- of the edition of the NTD). Krishna (1997: 110–143) discussed the unique transmission of the NS that the NTD presents. 4 NTD, prastāvanā (by Thakur), p. (ṭa), under (7): iyaṃ tāvad vṛttir atisaṃkṣiptā na vi- cārabahulā; nyāyabhāṣyasya, nyāyavārttikasya nyāyavārttikatātparyaṭīkāyāś ca mite- na vacasā sārasaṃgraha evāsyāḥ kṛtyam (“In the first place, this commentary is ex- tremely succinct and not rich in examination; the objective of this [commentary] is nothing but the summary of the main points of the Nyāyabhāṣya, Nyāyavārttika and Nyāyavārttikatātparyaṭīkā by means of brief statements.”). For the same assessment emphasizing the usefulness of the NTD, see the brief introduction of Jha to his edition: prāstāvikaṃ kiñcit, p. (kha).  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 137 liar with Udayana’s commentary.5 Moreover, Thakur highlighted a common feature that the NTD shares with Bhaṭṭa Jayanta’s (ca. 9th c.) Nyāyamañjarī (NM). Namely, both works refer to Śiva’s praise of Akṣapāda, the legendary author of the NS.6 5 NTD, prastāvanā (by Thakur), p. (ṭa), under (7): bahutra śabdaikyam api tair mūla- granthair atrāvalokyate […] tātparyaṭīkāntasya nyāyamūlagranthatritayasya vyavahā- ro ’tra bhūyasā dṛśyate. nātra nyāyanibandhasya na vā tatkartur ācāryasyodayanasya parāmarśo ’py atrāyāti. etavatā[!] udayanād api prāgbhavatvam asya granthakāras- yety ābhāti. (“Literal agreement with these basic works is also observed in many places here [in the NTD] […]. The verbal usage of the triplet of the basic Nyāya works [from the Bhāṣya] up to the Tātparyaṭīkā is abundantly seen here [in the NTD]. But the Nyā- yanibandha (= Udayana’s Pariśuddhi) and its author, the teacher Udayana, are not re- ferred to here [in the NTD]. Accordingly, it appears that this author of the work (= Vāgīśvara) predates even Udayana.”) Some scholars regard Vāgīśvara’s dates as unset- tled. For instance, G. C. Tripathi states in the foreword of the edition of the NTD that regarding Vāgīśvara, “very little, if anything, is known except that he appears to be a Southerner.” Cf. also EIPh I/879: no. DU806, where the NTD is registered under the “Authors of Unknown Dates” (Part Two) without explanation. I could not verify the al- leged existence of another edition of the NTD mentioned as “Allahabad 1976” under “DU806.1.1.” 6 See NTD, prastāvanā, p. (ḍa): antimaśloke akṣapādaṃ prati śrīśaṅkarakṛtaḥ sādhu- vādaḥ samuddiṣṭaḥ. kavitārkikeṇa bhaṭṭajayantenāpi – jayanti purajiddattasādhuvāda- pavitritāḥ | nidānaṃ nyāyaratnānām akṣapādamuner giraḥ || ity anena ślokena sa eva viṣayaḥ samudghoṣitaḥ (“In the final stanza, it is mentioned that Śiva exclaimed ‘ex- cellent’ to Akṣapāda. The same content is proclaimed by the poet-logician Bhaṭṭa Jayanta, too, in the stanza ‘Victorious are the words of the sage Akṣapāda that have been purified by the exclamation “excellent” made by the Conqueror of Fortresses (Śi- va), [and which are] the essence of the jewels of Nyāya.’”). For the verse in the NM, see NM I 2,5–6 = Kataoka 2007: 187,2–3 (reading nidhānaṃ ‘treasury’ instead of ni- dānaṃ ‘cause, original form, essence’; see Kataoka 2008: 81, n. 10 on his selection); for Kataoka’s Japanese translation, see Kataoka 2008: 69. On the last verse (in Mālinī metre) in the NTD, see NTD 158,14–17 (indicating a lacuna marked with brackets “[…]”): iti jagati janānām īśasāyujyabhājām anupajanita [...] bhāṣaṇeṣu | paśupatir api yasmai sasmitaḥ sādhu sādhv ity avadad avatu so ’smān akṣapādo munīndraḥ ||; this lacuna is indicated in the modern transcript (dated 1920–1921) that served as the basis of the edition, i.e., NTD (M), p. 197,12, with a series of three dots highlighting physical damage to the exemplar. See also Thakur’s conjectural reconstruction filling in the lacuna with five akṣaras (p. [ḍa]: anupajanita[dākṣyaṃ ditsate]); for an English translation and explanation of Thakur’s version, see Preisendanz 2005: 62–63, and n. 26: “May Akṣapāda, the Indra of sages, protect us, [he] to whom the Lord of Animals for his part smilingly said ‘Excellent, excellent!’ (when the former formed the wish to endow) those persons in the world who partake of intimate union with God (with dex- terity) in speech which had not yet arisen [in them], …” Round brackets indicate the translation of the Sanskrit passage that Thakur reconstructed. Another modern tran- script in a modern book format of the NTD (NTD [E]) kept at the Government Sanskrit College (Tripunithura, Ernakulam) transmits a complete version of the b-pāda as fol- 138 Yasutaka Muroya 2 Parallel passages in the Nyāyasūtratātparyadīpikā, the Nyāyabhāṣya and the Nyāyavārttika Vāgīśvara did not mark any passages that he quoted from the NBh and the NV, as for example by using the particle iti. Rather, he embedded these pas- sages in – and freely adapted them to – the main text of his commentary, as was relatively common in dealing with works of one’s own school (see Fres- chi 2015). Nonetheless, Vāgīśvara clearly stated how he drew upon the NBh and the NV in a passage directly after the benedictory verse. There, he said: I shall compose this “illumination of the meaning and intention of the Nyāyasūtra” after having examined the Bhāṣya word by word and also followed the Vārttika (NTD 1,7–8 [Anuṣṭubh]: anvīkṣyānupadaṃ bhāṣyam apy anukramya vārttikam | nyāyasūtrārthatātparyadīpi- keyaṃ vidhāsyate ||).7 This explicit statement clearly shows that the author studied the NBh and NV carefully and that these works served as the basis for his independent commentary. For the following discussion in this chapter, it is worth noting that Vāgīśvara neither mentioned Vācaspati nor any of his works. 3 Parallel passages in the Nyāyasūtratātparyadīpikā and the Nyāyavārttikatātpāryaṭīkā Parallel passages in the NTD and the NVTṬ are not marked by the respective authors as quotations. In itself this absence of marking would not rule out the possibility that one author borrowed from the work of the other. There are in principle three explanations for textual agreements and similarities between the NTD and the NVTṬ. Either Vāgīśvara reused the NVTṬ, or Vācaspati referred to the NTD, or both works drew upon a common source. Jha and Thakur chose the first hypothesis. The following part of this chapter is devo- ted to critically examining their assessment. lows: anupajanitakaṭākṣaṃ mārggam ābhāṣaṇeṣu (unmetrical); cf. NTD (E), unknown page number, lines 7–8. However, the upper part of the text at question in the common exemplar of both transcripts, NTD (T) f. 92r,7, is physically lost, and the remaining lines can confirm the presence of mārggam ā-, preceded by two lost akṣaras, but not that of kaṭākṣaṃ ‘a glance or side look’ (Monier-Williams 1899: 243c, s.v. kaṭa). 7 Cf. Preisendanz 2005: 62.  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 139 3.1 Vāgīśvara and Vācaspati on Nyāyasūtra 1.1.1 The table below shows an example of a parallel between the NVTṬ and NTD taken from the commentary on the first sūtra of the NS, which presents the sixteen categories or cardinal topics (padārtha) as well as the goal of Nyāya.8 Table 1: Vāgīśvara and Vācaspati on Nyāyasūtra 1.1.19 NTD 1,13–17 NVTṬ 2,23–3,8 tattvaṃ (1)jñāyate ’neneti (2)tattva- abhimataḥ sambandho ’bhisamban- jñānaṃ śāstram. tasmāt (3)pramā- dhaḥ, (4)śāstraniḥśreyasayor hetuhe- ṇādipadārthatattvāvagamāvānta- tumadbhāvaḥ. “tasya” [= NV 1,9] ravyāpārān niḥśreyasasyādhiga- idaṃ (4´)sūtravākyam abhisamban- maḥ prāptir bhavatīti (4)śāstraniḥ- dhavākyam. [NV = 1,9–10]. “pra- śreyasayor hetuhetumadbhāvapra- māṇādipadārthatattvajñānāt” [=NV tipādanārthaṃ (4´)sūtram. evaṃ ca 1,9] iti. atra hi (1)jñāyate ’neneti jñā- sati (5)pramāṇādipadārthaśāstra- nam iti vyutpattyā (2)*śāstram ucya- yor jñāpyajñāpakabhāvaḥ pramā- te, pañcamyā ca tasya hetutvam. ṇādipadārthatattvajñānayoḥ kār- (6)na hi viṣaghnamantravat svarūpa- yakāraṇabhāvalakṣaṇaś ca sam- mātreṇa tad avivakṣitārthaṃ niḥśre- bandhaḥ sūcito bhavati, (6)viṣa- yasahetur iti padārthatattvāvagama- ghnamantravat svarūpeṇāvivakṣi- karaṇatayā śāstram apadiśati, *na tārthasya śāstrasya niḥśreyasahe- svarūpeṇa. tena śāstrasya niḥśreyase tutvānabhyupagamāt.10 kartavye (3)pramāṇāditattvāvagamo 8 NS 1.1.1: pramāṇaprameyasaṃśayaprayojanadṛṣṭāntasiddhāntāvayavatarkanirṇaya- vādajalpavitaṇḍāhetvābhāsacchalajātinigrahasthānānāṃ tattvajñānān niḥśreyasādhi- gamaḥ. 9 Underlined text headed by an Arabic numeral is parallel in the NTD and NVTṬ, whe- reas underlined text headed by a letter of the Roman alphabet is related to other textual sources that will be mentioned below separately. 10 Translation: “The knowledge of the true nature [mentioned in the sūtra] is the doctrinal edifice insofar as the true nature is known by it. The acquisition, namely, the attain- ment of the highest good, arises from this [doctrinal edifice] (tasmāt), which has as [its] intermediate operation the understanding of the true nature of the categories begin- ning with the means of valid cognition. Hence, the sūtra has the purpose of making un- derstood (pratipādana) the relationship of cause and its possessor (hetuhetumadbhāva) between the doctrinal edifice and the highest good. And such being the case, the rela- tionship of that which is made known and that which makes known (jñāpyajñāpaka- bhāva) between the categories beginning with the means of valid cognition and the doctrinal edifice, as well as the relation (sambandha) characterized by the relationship of cause and effect (kāryakāraṇabhāva) between the categories beginning with the means of valid cognition and the knowledge of their true nature, is indicated [by this sūtra], because it is not accepted that the doctrinal edifice, without the intended mean- 140 Yasutaka Muroya ’vāntaravyāpāra ity uktaṃ bhavati. tathā ca (A)pramāṇādipadārthatat- tvaṃ pratipādyaṃ pratipādakaṃ ca śāstram iti (5)śāstrapramāṇādipadār- thatattvayor jñāpyajñāpakabhāvaś ca pramāṇādipadārthatattvajñāna- niḥśreyasayoḥ kāryakāraṇabhāva- lakṣaṇaś ca sambandhaḥ sūcito bha- vati.11 In the two passages cited above, the degree of textual agreement cannot be explained as being mere coincidence. Jha, the editor of the NTD, remarks, accordingly, that the NVTṬ is the source (upajīvya) of the NTD.12 ing [being understood by those who study it (cf. NVTP 8,19-9,2 and ŚṬ 4,1-3: śiṣyasambaddhavyāpāravattā)], becomes the cause of the highest good through its own nature/form, like an antidotal mantra [which becomes the cause of the cure for poison through its own nature/form].” 11 v.l.: śāstram] J2; tattvajñānaṃ śāstram ed. – na] J2; na tu ed. (the oldest dated manu- script of the NVTṬ, J1, of which the quite faithful copy, J2, was utilized by Thakur, is not available for this passage). For an English translation, see Perry 1995: 255–256, who appears to read *tadvivakṣitārthaṃ for tad avivakṣitārthaṃ against the edition: “By ‘right relation’ is meant ‘the intended relation’: the śāstra and the highest good stand in the relationship of cause and effect. ‘Of this’ <śāstra> this statement of the sūtra is the ‘statement of the right-relation.’ For in this statement, ‘from the knowledge of the true-nature of the categories of the means of knowledge etc.,’ knowledge of the true-nature means the ‘śāstra’, according to the etymological analysis, ‘knowledge’ is ‘that through which <something> is known’. And because of <the use of> the fifth case affix, this <śāstra> is the cause. For the intended meaning of this <statement> is not that it is the cause of the highest good by virtue of its mere intrinsic nature, as is the case with spells that counteract <the effects of> poison; hence he teaches that the śāstra <is the cause of the highest good> in that it is the instrumental cause for under- standing the true-nature <of the categories of the means of knowledge etc.>, but not by its intrinsic nature. Therefore, what is meant is: as the highest good is to be produced by the śāstra, the understanding of the true-nature <of the categories> of the means of knowledge etc. is an intermediate operation. Moreover, the true-nature of the catego- ries of the means of knowledge etc. is what is to be taught, while the śāstra is that which teaches them; hence, he indicates both that the śāstra and the true-nature of the categories of the means of knowledge etc. stand in the relation of ‘what makes known’ and ‘what is to be made known’, and that the relationship between the knowledge of the true-nature of the categories of the means of knowledge etc. and the highest good is that of cause and effect.” 12 NTD, pariśiṣṭam (1) “ālocanātmikā ṭippaṇī” on NS 1.1.1.  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 141 The two text passages deal with the first sūtra of the NS in different ways and from different perspectives. Vāgīśvara analyzes the connection of the categories (padārtha) within the larger perspective of the role of śāstra (“doctrinal edifice”). By contrast, Vācaspati focuses on commenting upon the word abhisambandhavākya (“statement that demonstrates an intended rela- tionship”), a term introduced by his predecessor Uddyotakara. At the same time, he clarifies the internal coherence of the central concepts of the first sūtra.13 The two underlined text-passages no. 5 deal with two kinds of relation- ships, namely, (1.) the relationship between that which is made known and that which makes known (jñāpyajñāpakabhāva), and (2.) the relationship between effect and cause (kāryakāraṇabhāva). The NTD assigns the former relationship to the categories and the doctrinal edifice (śāstra). In contrast, the NVTṬ states that the same relationship exists between the true nature of the categories (padārthatattva) and the doctrinal edifice (śāstra). Vāgīśvara assigns the second type of relationship to the categories and the knowledge of the true nature (tattvajñāna), whereas Vācaspati holds that this relationship exists between the knowledge of the true nature and the highest good (niḥ- śreyasa). The text passage (A), which is only found in the NVTṬ, contains an addi- tional designation of the relationship, i.e., pratipādyapratipādakabhāva. This relationship links the true nature of the categories with the doctrinal edifice. The same relationship is described by Vācaspati as the epistemic causal rela- tionship (jñāpyajñāpakabhāva). The reason why Vācaspati added a further designation for the same relationship is unclear. In short, Vāgīśvara regards the categories (padārtha) as the object to be made known (jñāpya) by the doctrinal edifice (śāstra). Vācaspati is more elaborate and illustrates the twofold classification according to which the true nature (tattva) of the categories, not the mere categories (as implied by Vāg- īśvara), is the object to be made known (jñāpya) or to be made understood (pratipādya). Does this difference reflect a philosophical disagreement? Was Vāgīśvara aware of Vācaspati’s formulations? Did he reuse them in an inten- tionally modified manner? 13 For Uddyotakara’s usage of abhisambandhavākya, to which Vācaspati refers, see NV 1,9–10: pramāṇādipadārthatattvajñānān niḥśreyasādhigama ity etac chāstrādisūtraṃ tasyābhisambandhavākyam. 142 Yasutaka Muroya 3.1.1 Udayana’s theory of categories To answer these questions,14 it is relevant to take Udayana’s refined classi- fication of the aforementioned elements into account. Udayana agrees with Vāgīśvara in that the doctrinal edifice (śāstra) makes the categories under- stood (Udayana) or known (Vāgīśvara). However, Udayana defines the rela- tionship between the categories and the doctrinal edifice as pratipādyapra- tipādakabhāva (the relationship between that which is understood and that which causes understanding; see underline A), the same term that Vācaspati uses.15 In the case of the relationship kāryakāraṇabhāva (the relationship of effect and cause) Udayana agrees with Vācaspati in that he applies this rela- tion to the knowledge of the true nature of categories and the highest good.16 Accordingly, Udayana in one case shares a position with Vāgīśvara, and in the second case agrees with Vācaspati. It is therefore possible that Udayana was acquainted with the interpretation as presented by Vāgīśvara in the NTD, although it is possible that he developed the same idea independently.17 14 Udayana developed an exegetical theory about the role of śāstra and its relation to the other key terms that are listed in the first sūtra, namely (a) śāstra (as vyāpārin or vyāpāravant-), (b) padārthatattvajñāna or -tattvāvagama (as śāstra’s vyāpāra), (c) pramāṇādipadārtha (as its viṣaya) and (d) niḥśreyasa (as its phala). In NVTP 9,7–13, these elements are presented as being subject to the following five relationships: (1) kāryakāraṇabhāva (between [a]–[b]), (2) pratipādyapratipādakabhāva (between [a]– [c]), (3) viṣayaviṣayībhāva (between [c]–[b]), (4) kāryakāraṇabhāva (between [b]–[d]) and (5) hetuhetumadbhāva (between [c]–[d]). The first and fourth relationship seem to be identical, but as glossed in the so-called Śrīkaṇṭhaṭippaṇaka ascribed to Śrīkaṇṭha, the pairing elements are different; on the second relationship in NVTP 9,7–9 (for the text, see n. 16), cf. ŚṬ 4,15: vyāpāravyāpāriṇor iti. tattvajñānaṃ vyāpāraḥ, śāstraṃ vyāpāri. For the terms used for the fivefold classification, see NVTP 9,14 (pañcasu vaktavyeṣu) and ŚṬ 4,23 (udayanoktasambandhapañcakamadhyāt). 15 On the equivalence of pratipādyapratipādakabhāva and jñāpyajñāpakabhāva, see ŚṬ 2,24: dvitīyo ’yaṃ jñāpyajñāpakabhāvaḥ pratipādyapratipādakabhāvāparanāmnā. 16 For Udayana’s gloss on NVTṬ 3,4–5 (tena śāstrasya niḥśreyase kartavye), see NVTP 9,7–9: evaṃ ca sati śāstrasya niḥśreyase kartavye padārthatattvajñānasya niruktiba- lena vyāpāratve darśite vyāpāravyāpāriṇoḥ kāryakāraṇabhāvaḥ. Translation by Perry 1995: 430–431: “And if this is the case, ‘as the highest good is to be produced by the śāstra,’ since the knowledge of the true-nature of the categories has been shown to be the operation by dint of its etymology, the relationship of cause and effect holds be- tween what embodies the operation (vyāpārin; YM) and the operation.” 17 An indirect piece of evidence for Udayana’s acquaintance with Vāgīśvara could be derived from a study of the readings of the NS mentioned by Udayana. On the textual tradition of the NS adopted by Vāgīśvara and other commentators as known to Udayana, see Preisendanz 1994: 181, n. 13.  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 143 3.2 Vāgīśvara and Vācaspati on Nyāyasūtra *5.2.15(16) A comparison of the NTD on NS 5.2.15(16)18 with Vācaspati’s commentary on the same sūtra provides additional elements for examining the issue of reuse.19 Table 2: NTD and NVTṬ ad NS *5.2.15 (ed. 5.2.16) NTD 156,20–26 NVTṬ 692,9–19 (1)prativādinā na tāvat sarvaṃ dū- syād etat – (1)*prativādinā na tāvat ṣaṇīyam, anyatamāvayavadūṣaṇe- sarvaṃ dūṣaṇīyam, anyatamāvaya- naiva sādhanasya dūṣitatvena do- vadūṣaṇenaiva sādhanasya dūṣita- ṣāntarābhidhānavaiyarthyāt. na tvena doṣāntarābhidhānasya vai- khalu mṛto mṛtyunā punaḥ(1) śakyo yarthyāt. na khalu mṛto mṛtyunā ra- mārayitum, evaṃ dūṣitam api na kṣitaḥ (1)*śakyo mārayitum, evaṃ śakyaṃ dūṣayitum. (2)tasmād anya- dūṣitam api *na śakyaṃ dūṣayitum. tamasya (2)dūṣaṇasya pariṣadā vi- (2)tasmāt saty api *dūṣaṇabāhulye jñātasya vādinā trir abhihitasya (2)’nyatamaṃ (2)dūṣaṇīyam. tathā ca tacchabdena śabdāntareṇa vā yad (B´)sarvānubhāṣaṇe sarvasyādūṣya- apratyuccāraṇaṃ *tad ananubhā- tvād (C)yad dūṣaṇīyaṃ tat punar ṣaṇaṃ nigrahasthānam, anyathā anūdyam iti (3)(B)sarvānuvāde dvir- (A)*dūṣaṇasyāśrayasyānirdeśena anuvādo ’dūṣyānuvādaś ceti *ni- 18 The sūtra is numbered as NS 5.2.16 according to Thakur’s and Jha’s edition of the NVTṬ and NTD, respectively. However, Vāgīśvara regarded the two sūtras (NS 5.2.14–15) devoted to the definition of punarukta (“repetitive statement”) as a single sūtra. Therefore, I have tentatively renumbered the sūtra at issue as NS 5.2.15 with an asterisk. This position of Vāgīśvara’s is not apparent in the edition of the NTD, because at the beginning of the commentary on NS 5.2.14–15 the editor filled an alleged lacuna with the text “[tad anena sūtradvayena] punaruktam,” using square brackets. The editor took the additional text from the NVTṬ. However, the Trivandrum manuscript (NTD [T]), as well as its modern transcript utilized for the edition (NTD [M]), does not have any lacuna at all. They read instead tad evārthapunaruktam; cf. NTD 156,12 = NTD (T), f. 90v,8 and NTD (M), p. 94,17 (tadevā rtha; the blank space as it appears). Vācaspati, in contrast to Vāgīśvara, explicitly refers to two sūtras; cf. NVTṬ 691,1: tad anena sūtradvayena punaruktam ekam eva nigrahasthānam. This may indicate that he wanted to rectify the view that there was only a single sūtra. In the NTD, the alleged two sūtras are adduced in immediate sequence; cf. NTD 156,10–11, where the given numbers, 5.1.14 and 5.1.15, in the edition must be corrected to 5.2.14 and 5.2.15. Vāgīśvara’s commentary on this unified sūtra is concluded with the ex- pression iti sūtrārthaḥ (NTD 156,17), an expression that shows that he dealt with a single sūtra. For Vāgīśvara’s marking system, cf., e.g., NTD 64,10 (iti sūtradvita- yārthaḥ for NS 2.2.54–55) and 55,26–27 (iti sūtratritayārthaḥ for NS 2.2.18–20). 19 This passage was not taken into account by Jha. 144 Yasutaka Muroya “*kasyāvayavasyedaṃ dūṣaṇam” iti grahasthānadvayam. (D)tasmād saṃśayaprasaṅgāt. na ca (3)(B)sar- anubhāṣaṇam eva nigrahasthānam vānuvādo ’bhimataḥ, (3)(B)tathā saty iti viparītam āpatitam iti. ata āha – adūṣyānuvādadviranuvādādidoṣa- “apratijñānāc ca” [NV 527,10] prasaṅgād iti sūtrārthaḥ.20 iti.21 The two passages cited above deal with the concept of ananubhāṣaṇa (“in- capability of repeating a debater’s proof”) as one of the “points of defeat” 20 v.l.: tad] M, T; na tad ed. – dūṣaṇasyāśraya-] M, T; dūṣaṇāśraya ed. – kasyāvayava-] em.; tasyāvayava ed., M, T. Translation: “First, the contestant does not have to refute all [constituents of the disputant’s proof], because it is pointless to mention additional faults, inasmuch as the [disputant’s] proof was already refuted solely by means of the refutation of just one of the constituents [of the logical formula]. As you should know, someone who died cannot be killed again by Death, and likewise, anything that has al- ready been refuted cannot be refuted any further. Therefore, the incapability of repeat- ing – by using the disputant’s words or different expressions – one of the refutations that has been comprehended by the assembly and asserted thrice by the disputant, is the ananubhāṣaṇa, a point of defeat. [It is so,] because it would otherwise be unclear which (kasya) constituent [of the disputant’s proof] is refuted, because the basis of the refutation is not indicated. Furthermore, the repetition of all [of the disputant’s proof] is not intended [in the definition of ananubhāṣaṇa], because there would otherwise oc- cur, as undesirable consequences, such faults as the repetition of what is not to be dis- proved, a redundant repetition and so on. This is the meaning of the sūtra.” My above emendation kasya for tasya is based on the assumption that Vāgīśvara alludes here to Vātsyāyana’s formulation apratyuccārayan kimāśrayaṃ parapakṣapratiṣedhaṃ brūyāt (“On which basis would the person who does not repeat [the disputant’s proof] state the rejection of the other’s [i.e., the disputant’s] thesis?”); cf. NBh 217,1–2 ≈ VN 52,9– 10. I am obliged to Professor Kei Kataoka for discussing this problematic text portion. 21 v.l.: prativādinā na] J1; na prativādinā ed. – śakyo mārayitum] J1; mārayituṃ śakyaḥ ed. – na śakyam] J1; om. ed. – dūṣaṇa-] J1; dūṣya ed. – nigrahasthāna-] J1; dūṣakani- grahasthāna ed. Translation: “There would be the following [objection]: First, the con- testant does not have to refute all [constituents of the disputant’s proof], because it is pointless to mention additional faults, inasmuch as the [disputant’s] proof was already refuted solely by means of the refutation of just one of the constituents [of the logical formula]. As you should know, someone who died cannot be killed [again] while ruled by Death, and likewise, anything that has already been disproved cannot be disproved any further. Therefore, even though many refutations may be possible, only one con- stituent needs to be refuted. And likewise, if one repeats all [of the disputant’s proof at the beginning], while not all has to be refuted, what actually should be refuted needs to be repeated again [later at the time of its refutation]. Thus, when repeating everything, there are two points of defeat, namely, the redundant repetition and the repetition of something not to be refuted. Therefore, the paradoxical consequence arises that the very repetition is already a point of defeat. Therefore [Uddyotakara] says, ‘And be- cause it is not [our] assertion.’”  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 145 (nigrahasthāna).22 Vāgīśvara’s text covers the entire part of his commentary containing the definition of ananubhāṣaṇa. By contrast, the passage in Vā- caspati’s work is found in the midst of the commentary on Uddyotakara’s NV. In fact, the last words marked in the above quotation, apratijñānāc ca, are a so-called pratīka that introduces Uddyotakara’s statement. The passages underlined from (1) to (3) are found in both the NTD and the NVTṬ. The parallel passage of the NVTṬ and the NTD is headed in Vācaspati’s work by the expression syād etad “there could be the following [view].” In general, this formulation introduces a different opinion or an objection. Thus, the introductory words syād etad apparently indicate that Vācaspati knew Vāgīśvara’s interpretation of ananubhāṣaṇa, to which, however, he did not subscribe. An alternative scenario, according to which Vāgīśvara, intention- ally and without further justification, adopted the very position that Vācaspati criticized, is highly unlikely. 3.2.1 Dharmakīrti’s discussion of ananubhāṣaṇa The content of the shared text passages in the NTD and NVTṬ as presented in Table 2 above can be outlined in the following way. The underlined pas- sage (1) deals with the topic of the partial disproof of a disputant’s argument. In order to be successful, a contestant only needs to disprove a single compo- nent (anyatamāvayava) of the syllogism, and not the other components. This is summarized in the passage with underlining (2) as well as in the surround- ing text. The passage with underlining (3) illustrates the two kinds of faults (doṣa) that occur when a contestant must repeat the whole proof of the dis- putant. It is possible to hypothetically identify some sources for the passages with underlining (A) and (3/B) in Vāgīśvara’s work. Most probably, Vāgīśvara al- ludes here to Uddyotakara’s apologetic theory of ananubhāṣaṇa as a nigra- hasthāna and to Dharmakīrti’s (ca. 7th c.) criticism of this view. This guess is supported by the fact that Dharmakīrti’s Vādanyāya (VN) literally quotes a number of passages from the NBh and the NV in order to criticize the definitions of nigrahasthānas presented in the fifth book of the NS. In the following passage, Uddyotakara argues in defense of ananubhā- ṣaṇa as a nigrahasthāna against anonymous opponents (kecit), who were possibly Buddhists and whom Vācaspati referred to with the title bhadanta.23 22 For a general explanation of the concept of ananubhāṣaṇa, see, e.g., Solomon 1976: 236–238. 23 For Uddyotakara’s anonymous opponents, cf. NV 527,4: na, uttareṇāvasthānān nedaṃ nigrahasthānam iti kecit ≈ VN 52,11 (uttareṇāvasānān nedaṃ nigrahasthānam iti cet). For Vācaspati’s identification of the opponent as bhadanta (“Buddhist mendicant”), cf. 146 Yasutaka Muroya Table 3: The NV and the VN on the topic of ananubhāṣaṇa NV 527,10–11 VN 53,1–3 apratijñānāc ca: nedaṃ pratijñāya- apratijñānāc ca: nedaṃ pratijñāyate te – pūrvam uccārayitavyaṃ paś- – pūrvaṃ sarvam uccārayitavyaṃ cād uttaram abhidheyam, api tu ya- paścād uttaram abhidhātavyam iti, thākathaṃcid uttaraṃ vaktavyam. api tu yathākathaṃcid uttaraṃ vak- uttaraṃ (A´)cāśrayābhāve na yuk- tavyam. uttaraṃ (A´)cāśrayābhāve tam iti.24 ’yuktam iti. Vāgīśvara adaptively reused the last part of Uddyotakara’s explanation, i.e., āśrayābhāve, with the expression dūṣaṇasyāśrayasyānirdeśena that is marked with underlining (A) in Table 2. In this part, Uddyotakara stated that a rejoinder is not correct if its target is not indicated (uttaraṃ cāśrayābhāve na yuktam). In order to provide a valid rejoinder, the disputant’s proof has to be repeated. Therefore, the inability to repeat it is a point of defeat (nigraha- sthāna).25 Supporting Uddyotakara, Vāgīśvara argued that the failure to re- peat the disputant’s proof leads to the undesirable consequence that the dis- putant and the assembly are left in doubt as to which part of the syllogism is actually the target of the contestant’s refutation. Dharmakīrti, in his turn, refuted Uddyotakara’s defense of the Nyāya defi- nition, by using Uddyotakara’s formulation with the additional determination sarvam in pūrvaṃ sarvam uccārayitavyam (“the proof of the disputant should be articulated completely in advance”), as underlined in the above quotation. As is shown below with underlining (B´´), Dharmakīrti further specified his interpretation of the Nyāya definition with the word sakṛt (“in one go”). VN 54,9–11: (B´´)sakṛtsarvānubhāṣaṇe ’pi (C´)doṣavacanakāle punar viṣayaḥ pradarśanīya eva, apradarśite doṣasya vaktum aśakyatvāt.26 NVTṬ 692,4: tam etaṃ bhadantākṣepaṃ samādhatte. According to Much, this Bud- dhist opponent is Dignāga; cf. Much 1991a: 211–220, esp. 211 and 217 (Fragment #17). 24 Translation: “And because it is not [our] assertion: [Namely,] it is not asserted that he (i.e., the contestant) should articulate in advance [the disputant’s proof], [and] subse- quently present [his] rejoinder. Rather, the rejoinder should be presented in a certain manner. And if there is no basis [of this rejoinder], the rejoinder is not correct.” For an additional English translation, see Jha 1984: 4/1763. For Vācaspati’s gloss, see NVTṬ 692,15–17. 25 Cf. NV 527,11–12: iti yuktam apratyuccāraṇaṃ nigrahasthānam. 26 “Even if one has repeated all [of the disputant’s proof] at once, the object [of rejection]  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 147 Dharmakīrti held that such a repetition is superfluous, because the spe- cific points targeted by the refutation will necessarily be presented on each relevant occasion. Furthermore, the point of defeat called ananubhāṣaṇa can be subsumed under the point of defeat called apratibhā (“non-appearance [of an answer]” or “inapprehension of one’s own reply”) as defined in NS 5.2.18.27 3.2.2 Vāgīśvara’s and Vācaspati’s references to Dharmakīrti In the case of Vāgīśvara’s allusion to Dharmakīrti’s argumentation, the phrase sakṛtsarvānubhāṣaṇe in the VN (see underlining [B´´]) immediately above), corresponds to the part of the NTD with underlining (B). Vācaspati, in contrast, seems to have reused – directly or indirectly – Dharmakīrti’s for- mulations that are marked as (B´´) and (C´) in the passage cited above in the part of his NVTṬ with underlining (B´) and (C). Moreover, he paraphrased the text (B´) with sarvānuvāde in (3/B), which may be an allusion to Vāg- īśvara. Dharmakīrti’s concluding criticism runs as follows: VN 54,13–15: dūṣaṇavādinā dūṣaṇe vaktavye yan na tatropayujyate tasyābhidhānam (B´´)adoṣodbhāvanaṃ dviruktiś ceti (D´)sakṛtsarvānu- bhāṣaṇaṃ parājayādhikaraṇaṃ vācyam.28 should inevitably be illustrated again at the time when one indicates its fault, because one cannot present a fault in reference to what is not illustrated.” For English and Ger- man translations, see Gokhale (1993: 124) and Much (1991b: 95), respectively. 27 Cf. VN 55,1–2: nānanubhāṣaṇaṃ pṛthag nigrahasthānaṃ vācyam, apratibhayā gata- tvāt; cf. also VN 56,11–12. NS 5.2.18 is uttarasyāpratipattir apratibhā (= VN 58,15). 28 “When a rejection should be presented by a confuter, the presentation of something that does not contribute to it constitutes a case of the non-indication of faults and of a reiteration. Hence, the presentation of the whole [proof] at once is to be regarded as the ground for defeat.” For English and German translations, see Gokhale (1993: 125) and Much (1991b: 95), respectively. Vāgīśvara and Vācaspati paraphrase adoṣodbhāvana with adūṣyānuvāda (“repetition of what is not to be disproved”). This suggests that they possibly took the term as referring to the indication of non-faults, namely, in the way that Dharmakīrti explained in VN 23,7–10. This interpretation, however, goes against the above translators’ understanding. (Gokhale: “Not pointing out the fault,” and Much: “das Nichtaufzeigen eines Fehlers”) and against the Tibetan translation (P 393b4, D 350a5: skyon brjod pa med par “not stating a/the fault”). Dharmakīrti dis- cusses this interpretation in VN 21,14–17 (cf. Much 1991b: 95, n. 397). For his two in- terpretations of adoṣodbhāvana, see Solomon (1976: 251–253), Much (1986: 135–136 and 1991b: Einleitung, XI–XII), Gokhale (1993: Introduction, xxiii–xxiv) and Sasaki (2013). 148 Yasutaka Muroya The two kinds of faults (underlining [B´´]) that Dharmakīrti mentioned, namely adoṣodbhāvana (“non-indication of fault”) and dvirukti (“repeti- tion”), are envisaged by Vāgīśvara as adūṣyānuvāda and dviranuvāda (under- lining [B] in Table 2). Vācaspati’s text dviranuvādo ’dūṣyānuvādaś ca may be an allusion to Vāgīśvara’s formulation. Dharmakīrti’s final remark (un- derlining [D´]) is reflected only in Vācaspati’s text (underlining [D]). 4 On the relative chronology of Vāgīśvara and Vācaspati On the backdrop of the above observations, it may be concluded that Vācas- pati was addressing in fact two opponents, namely, Dharmakīrti as well as Vāgīśvara, and cleverly refuted both. If this is correct, we have to reverse the chronological sequence of Vāgīśvara and Vācaspati that Thakur and Jha pro- posed. It appears much more likely that Vācaspati reused Vāgīśvara’s com- mentary than the reverse. However, it remains unresolved whether Vācaspati had direct access to Vāgīśvara’s commentary, or whether he knew it from other sources. This is mainly because the possible intermediate sources, the works of the Mañjarīkāra29 and of Trilocana30 (ca. 10th c.), are lost. The latter was Vācaspati’s teacher. The former is known as the author of some Nyāya- 29 For the close relationship between Vāgīśvara and the Mañjarīkāra as recorded by Aniruddha, cf. Muroya (forthcoming). According to Aniruddha’s Vivaraṇapañjikā, a commentary on the NBh, NV and NVTṬ, Vācaspati was familiar with the interpreta- tions of the Mañjarīkāra and sometimes rejected them. I find it plausible to identify the Mañjarīkāra with Trilocana, as Thakur (1969: v–vii) does, because Vācaspati was ob- viously closely related to them both. However, Aniruddha referred to the two authors with different names in different contexts. There is neither compelling supportive nor contradictory evidence for their identity. 30 For Udayana’s remark on the relationship of Trilocana and Vācaspati, cf. NVTP 3,8– 10: kiṃ nāmātra trilocanaguroḥ sakāśād upadeśarasāyanam āsāditam amūṣāṃ punar- navībhāvāya dīyata iti yujyate. For an English translation, cf. Perry 1995: 417: “It is appropriate <to say> here: ‘However, the elixir of teaching which he had obtained from his guru Trilocana is being given in order to give new life to those <com- mentaries>’”; Perry’s rendering of kiṃ nāma with “however” appears to be based upon Śrīkaṇṭha’s gloss (ŚṬ 1,29–30: kiṃ nāmeti kiṃ tv ity arthaḥ. evam uttaratrāpy asya pa- dasyāyam eva paryāyo deyaḥ. iti yujyata iti. dīyata iti yat tad yuktaṃ bhavatīty ar- thaḥ.). For a contextualization of Udayana’s remark as well as Trilocana’s works con- taining relevant bibliographical information, see Preisendanz 2005: 64–65 and notes 29–30, where the above cited passage is paraphrased as follows: “Why then, some fictitious partner in discourse asks, should they not be rejuvenated by means of directly administrating the life-giving elixir consisting in the readily obtained teaching of Vācaspati’s guru Trilocana? Udayana considers this suggestion to be appropriate[.]” Cf. also Randle 1930: 106, n. 1.  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 149 mañjarī, of which only quotations found in works by Aniruddha (ca. 10–11th c.) have been preserved. My recent study of the fragments from the work of the Mañjarīkāra as reported by Aniruddha shows that Vāgīśvara shared cer- tain ideas with the Mañjarīkāra, whereas the same views were criticized by Vācaspati. Although it cannot be excluded that Vācaspati and Vāgīśvara had inde- pendent access to a common source, the hypothesis that Vāgīśvara predates Vācaspati is supported, among other things, by a few archaic features of the NTD.31 As I have discussed elsewhere,32 the NTD regards the phrase trai- kālyagrahaṇāt as an authentic independent sūtra and interprets it as being directly related to the threefold division of inference defined in NS 1.1.5 (tri- vidham anumānam). The fact that Dignāga criticized the same phrase may in- dicate that he also took this text to be an independent sūtra.33 Moreover, the version of the NS that Vāgīśvara commented upon contains some ancient readings that agree with certain readings found in the works of Bhaṭṭa Jayanta and Bhāsarvajña (ca. 10th c.), who had access to an older version of the NS than Vācaspati.34 The new hypothesis on Vāgīśvara’s relative chronology, for which I have demonstrated its greater plausibility, will to a certain extent have conse- quences in the study of classical Nyāya. Historically, the analysis of Vācas- pati’s mode and extension of reuse as well his indebtedness to Vāgīśvara’s ideas certainly deserves attention. Philologically, the transmitted text of the NTD may be weighed more heavily than the NVTṬ in terms of its text- critical value in the case of parallels to the NS, the NBh and the NV. In the case of the parallel passages between the NTD and the NVTṬ, the text of the NTD will testify to the original status of the text found in the Jaisalmer mss. 31 Preisendanz (1994: 545 [included in n. 181 on pp. 526–551]) discusses an instance in which Vāgīśvara’s interpretation of a sūtra most probably corresponds to the “intention of the sūtrakāra.” 32 Cf. Muroya 2006: 37–40. The editions of the NS do not contain the phrase traikālya- grahaṇāt as a seperate sūtra. Vāgīśvara, however, clearly considered it as a part of the NS (NTD 4,17). He referred to this text as a sūtra also at a different place (NTD 35,7 on NS 2.1.42: traikālyagrahaṇād ity anena sūtreṇa). The editor of the NTD suggested that sūtreṇa in fact means bhāṣyeṇa and referred the reader to the commentary on NS 1.1.5. 33 For Dignāga’s reference to the phrase in question in his Pramāṇasamuccayavṛtti, see Jinendrabuddhi’s PSṬ 2 72,13. According to the editors, Dignāga probably cited one of the lost commentaries on the NS (“Ce ? [commentary on NSū]”); cf. also Muroya: 2006: 39 and nn. 65–66. 34 Cf. Muroya 2006: 29–30, 35. 150 Yasutaka Muroya of the NVTṬ, but not to the text as known in the printed edition of the NVTṬ, where they differ (see the variant readings for Table 2). References Primary Sources NTD Bhaṭṭavāgīśvarapraṇītā Nyāyatātparyadīpikā. Ed. Kishor Nath Jha. Allahabad: Gaṅgānāth Jhā Kendrīya Saṃskṛta Vidyāpīṭham, 1979. NTD (E) Manuscript of the NTD preserved at the Government Sanskrit College, Tripunithura, Ernakulam, Transcript Ms. 315 (“Nyāyasūtratātparyadīpikā”), paper, bound, 296 (?) leaves, Malayalam, undated. NTD (T) Manuscript of the NTD preserved at the Oriental Research Institute & Manuscripts Library, University of Kerala, Ms. 14670 (“Nyāyasūtratātparyadīpikā”), palm leaf, 92 leaves, Malayalam, undated. NTD (M) Manuscript of the NTD preserved at the Government Ori- ental Manuscripts Library, Chennai, Ms. R. 3405 (“Nyāya- sūtratātparyadīpikā”), paper, 198 pages, Devanagari, copied in 1920–1921. NBh Gautamīyanyāyadarśana with Bhāṣya of Vātsyāyana. Ed. Anantalal Thakur. New Delhi: Indian Council of Philoso- phical Research, 1997. Nyāyacaturgranthikā 1. NM Nyāyamañjarī of Jayantabhaṭṭa, with Ṭippaṇī – Nyā- yasaurabha by the Editor. Ed. K. S. Varadacharya. Msore: Oriental Research Institute, University of Mysore, 1969– 1983. Oriental Research Institute Ser. 116 and 139. NV Nyāyabhāṣyavārttika of Bhāradvāja Uddyotakara. Ed. Anantalal Thakur. New Delhi: Indian Council of Philoso- phical Research, 1997. Nyāyacaturgranthikā 2. NVTṬ Nyāyavārttikatātparyaṭīkā of Vācaspatimiśra. Ed. Ananta- lal Thakur. New Delhi: Indian Council of Philosophical Research, 1996. Nyāyacaturgranthikā 3. NVTṬ (J1) Manuscript of the NVTṬ preserved at the Jain Jñāna- bhaṇḍār (Baḍā Bhaṇḍār), Collection of Jinabhadrasūri, Jaisalmer, “Nyāyatātparyaṭīkā ṭippaṇī saha,” Ms. No. ji. kā. 1274/1–2, paper, 245 leaves (ff. 5–280; f. 37 is missing), Devanagari, dated 1222 CE. NVTṬ (J2) Manuscript of the NVTṬ preserved at the Jain Jñāna- bhaṇḍār (Baḍā Bhaṇḍār), Collection of Jinabhadrasūri, Jai- salmer, “Nyāyatātparyavṛtti ṭippaṇīsaha,” Ms. No. ji. kā.  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 151 69, paper, 201 leaves (ff. 201–409), Devanagari, dated 1444 CE. NVTP Nyāyavārttikatātparyapariśuddhi of Udayanācārya. Ed. Anantalal Thakur. New Delhi: Indian Council of Philoso- phical Research, 1996. Nyāyacaturgranthikā 4. NS Nyāyasūtra. In: NBh. PSṬ 2 Jinendrabuddhi’s Viśālāmalavatī Pramāṇasamuccayaṭīkā. Chapter 2. Part I: Critical Edition. Ed. Horst Lasic, Hel- mut Krasser and Ernst Steinkellner. Beijing–Vienna: China Tibetology Publishing House, Austrian Academy of Sciences Press, 2012. Sanskrit Texts from the Tibetan Autonomous Region 15/1. VN Dharmakīrti’s Vādanyāyaḥ. Ed. Michael Torsten Much. Teil I: Sanskrit-Text. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 1991. Veröffentlichungen der Kommission für Sprachen und Kulturen Südasiens 25. VN (Tibetan) A Tibetan Translation of the VN. Peking edition: Cat. No. 5715, mDo ce 364b8–400a7. sDe dge edition: Cat. No. 4218, mDo che 326b4–355b5. ŚṬ Nyāyadarśane Śrīkaṇṭhaṭippaṇakam [A Commentary on the Major Nyāya-texts] by Śrīkaṇṭhācārya. Ed. Anantalal Thakur. Calcutta: The Asiatic Society, 1986. B. I. Ser. 313. Secondary Sources Acharya 2006 Acharya, Diwakar. Vācaspatimiśra’s Tattvasamīkṣā: The Eearliest Commentary on Maṇḍanamiśra’s Brahmasiddhi. Criti. Ed. with an Introd. and Crit. Notes. Stuttgart: Steiner, 2006. Nepal Research Centre Publications 25. EIPh Encyclopedia of Indian Philosophies. Ed. Karl H. Potter. Vol. 1: Bibliography, Section 1. Third Revised Edition. Delhi: Motilal Banarsidass, 1995 (1st ed. Delhi 1970; reprint Delhi 1974). Freschi 2015 Freschi, Elisa. “The Reuse of Texts in Indian Philosophy: Introduction.” Journal of Indian Philosophy 43 (2015): 85–108. Gokhale 1993 Gokhale, Pradeep P. Vādanyāya of Dharmakīrti: The Logic of Debate. Delhi: Sri Satguru Publications, 1993. Bibliotheca Indo-Buddhica Ser. 126. Jha 1984 Jha, Ganganatha, trans. The Nyāya-Sūtras of Gauṭama: With the Bhāṣya of Vātsyāyana and the Vārṭika of Uḍḍyoṭakara. 4 vols. Delhi: Motilal Banarsidass, 1984. Kataoka 2007 Kataoka, Kei. “Critical Edition of the Śāstrārambha Section of Bhaṭṭa Jayanta’s Nyāyamañjarī.” The Memoirs 152 Yasutaka Muroya of the Institute of Oriental Culture, University of Tokyo 150 (2007): 204–170 = (123)–(157). Kataoka 2008 — “Jayanta ni yoru ronrigaku no ichizuke – Nyāyamañjarī ‘josetsu’ wayaku” [The role of Nyāya according to Jayanta: A Japanese translation of the Śāstrārambha section of the Nyāyamañjarī]. Tetsugaku Nenpō (Annual of Philosophy) 67 (2008): 55–90. Krishna 1997 Krishna, Daya. Indian Philosophy – A Counter Per- spective. Paperback edition. Delhi: Oxford Univ. Press, 1997. Monier-Williams 1899 Monier-Williams, Monier. A Sanskrit-English Dictionary. Etymologically and Philologically Arranged with Special Reference to Cognate Indo-European Languages. New Ed. Greatly Enlarged and Improved with the Collaboration of E. Leumann … C. Cappeler … [et al.] Oxford: Claren- don Press, 1899. Much 1986 Much, Michael Torsten. “Dharmakīrti’s Definition of ‘Points of Defeat’ (Nigrahasthāna).” Buddhist Logic and Epistemology. Studies in the Buddhist Analysis of In- ference and Language. Eds. B. K. Matilal and R. D. Evans. Dordrecht [u.a.]: Reidel, 1986. 133–142. Much 1991a — “Fragments from Dignāga? Traces of a Pre-Dharma- kīrti Buddhist Polemic against the Nyāya-nigrahasthāna.” Studies in the Buddhist Epistemological Tradition, Pro- ceedings of the Second International Dharmakīrti Con- ference Vienna, June 11–16, 1989. Ed. Ernst Steinkellner. Wien: Verlag der Österreichischen Akademie der Wissen- schaften, 1991. 211–220. Beiträge zur Kultur- und Geistesgeschichte Asiens 8. Much 1991b —, trans. Dharmakīrtis Vādanyāyaḥ. Teil II. Übersetzung und Anmerkungen. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 1991. Veröffentlichungen der Kommission für Sprachen und Kulturen Südasiens 25. Muroya 2006 Muroya, Yasutaka. “Some Observations on the Manu- script Transmission of the Nyāyabhāṣya.” Journal of Indo- logical Studies 18 (2006): 23–62. Muroya 2013 — “On the Nyāyasūtratātparyadīpikā.” Journal of Indian and Buddhist Studies 62.1 (2013): 288–282 = (241)–(247). (Japanese) Muroya forthcoming — “Aniruddha’s Reference to ‘Mañjarīkāra’ Fragments and Their Relation to Vācaspati Miśra and Bhaṭṭa Vāg- īśvara.” Proceedings of the Japan-Austria International Symposium “Transmission and Tradition: The Meaning  Nyāya Commentaries of Vācaspati Miśra and Bhaṭṭa Vāgīśvara 153 and the Role of ‘Fragments’ in Indian Philosophy,” 20–24 August 2012, Matsumoto. Perry 1995 Perry, Bruce Millard. “An Introduction to the Nyāyacatur- granthikā: With English Translations.” Diss., University of Pennsylvania. 1995. UMI No. 9532256. Preisendanz 1994 Preisendanz, Karin. Studien zu Nyāyasūtra III.1 mit dem Nyāyatattvāloka Vācaspati Miśras II. Stuttgart: Steiner, 1994. Alt- und Neu-indische Studien 46. Preisendanz 2005 — “The Production of Philosophical Literature in South Asia during the Pre-colonial Period (15th to 18th Cen- turies): The Case of the Nyāyasūtra Commentarial Tradition.” Journal of Indian Philosophy 33 (2005): 55– 94. Preisendanz 2008 — “Text, Commentary, Annotation: Some Reflections on the Philosophical Genre.” Journal of Indian Philosophy 36 (2008): 599–618. Randle 1930 Randle, H. N. Indian Logic in the Early Schools: A Study of the Nyāyadarśana in Its Relation to the Early Logic of Other Schools. 1st Indian ed. and rpt. New Delhi: Oriental Books Repr. Corp., 1976 (first edition, Oxford [et al.]: Ox- ford University Press, 1930). Sasaki 2013 Sasaki, Ryo. “Dharmakīrti’s Interpretation of nigraha- sthāna (2): On adoṣodbhāvana.” Kuwon: Research Papers 4 (2013): 55–75. (Japanese) Solomon 1976–1978 Solomon, Esther A. Indian Dialectics: Methods of Philo- sophical Discussion. 2 vols. Ahmedabad: B. J. Institute of Learning and Research, Gujarat Vidya Sabha, 1976–1978. Sheth Bholabhai Jeshingbhai Institute of Learning and Re- search, Research Ser. 70 and 74. Thakur 1969 Thakur, Anantalal. “Introduction.” Nyāyabhāṣyavārtti- kaṭīkāvivaraṇapañjikā [II–V] of Aniruddhācārya. Ed. Anantalal Thakur. Darbhanga: The Director, Mithila Insti- tute of Post-graduate Studies and Research in Sanskrit Learning, 1969. Mithila Institute Ser., Ancient Text No. 19.  Adaptive Reuse of the Descriptive Technique of Pāṇini in Non-Pāṇinian Grammatical Traditions with Special Reference to the Derivation of the Declension of the 1st and 2nd Person Pronouns* Malhar Kulkarni In this chapter I argue that not just terms or passages from certain earlier texts can be adapted and reused by later texts, but also the techniques and methods of describing data. I will show this by demonstrating how later Sanskrit non- Pāṇinian grammatical texts adapted and reused the technique of substitution from Pāṇinian grammar to describe linguistic facts. I will demonstrate this with an example, namely, the declension of the 1st and 2nd person pronouns of Sanskrit and its derivation according to Pāṇinian and non-Pāṇinian gram- mars. I begin by quoting a verse from Kālidāsa’s Raghuvaṃśa: sa hatvā vālinaṃ vīras tatpade cirakāṅkṣite | dhātoḥ sthāna ivādeśaṃ sugrīvaṃ saṃnyaveśayat || 12.58 That warrior having killed Vāli, established Sugrīva on his throne, which he wished for a long time, like an ādeśa (substitute) in the place of a verbal base (dhātu)1. In this verse, one clearly sees that Kālidāsa has used the technique of sub- stitution, widely employed in the description of linguistic forms by the Vyā- karaṇa school, as an upamāna, a standard of comparison. It is believed tra- ditionally that a standard of comparison facilitates the comprehension of the object to be compared and hence must be well known. With regard to the above stated stanza, this means that for Kālidāsa the use of the technique of ∗ I wish to thank Dr. Anuja Ajotikar and Dr. Tanuja Ajotikar for a stimulating discussion on this topic and help for formatting this chapter. I wish to express my gratitude to Ei- vind Kahrs, who went through the draft of this chapter and made valuable suggestions. Needless to say that thanks are also due to the editors for their patience and diligence. 1 See Nandargikar 1897: 371. 156 Malhar Kulkarni substitution in grammatical description was a standard practice and well known even to a common reader to such a degree that he could use it as a standard of comparison. This technique is stated by Aṣṭādhyāyī 1.1.56, 57, 58, where it is called sthānivadbhāva; the essential concepts related to this technique are discussed in detail in Kahrs (1998: 175−267) and, more re- cently, in Freschi and Pontillo (2013: 65−129). According to Kahrs (1998: 175), in simple words, the Indian grammarians adopted a model whereby stages in the linguistic derivational process are accounted for by say- ing ‘Y occurs in the place of X’ as opposed to ‘X becomes Y’. Kahrs (1998: 176) further says: The substituend, the element which is to undergo the substitution op- eration, is called the sthānin, literally ‘place-holder; placing-posses- sor’, and the substitute, the element which replaces the sthānin, is called ādeśa, literally, ‘instruction; specification’. I will now show that this same technique was adapted also by later non- Pāṇinian Sanskrit grammarians in their description of Sanskrit. In doing so, the inherent problems that characterized the Pāṇinian description and brought certain limitations to Pāṇinian methodology were addressed only partially, and the major limitation never received any serious attention. The 1st and 2nd person pronouns in Sanskrit have peculiar forms. For both pronouns, the inflected forms of each number are phonetically different to such a degree that it is difficult to determine a single common pronominal base. I hereby present them below in Tables 1 and 2, with the numbering scheme with which I will refer to them hereafter. A particular form will be referred to for example as 1/3, which means 1st line, 3rd column, i.e., nomina- tive plural. Table 1: Forms of yuṣmad Singular (1) Dual (2) Plural (3) 1 tvam yuvām yūyam 2 tvām yuvām yuṣmān 3 tvayā yuvābhyām yuṣmābhiḥ  Adaptive Reuse of the Descriptive Technique of Pāṇini 157 4 tubhyam yuvābhyām yuṣmabhyam 5 tvat yuvābhyām yuṣmat 6 tava yuvayoḥ yuṣmākam 7 tvayi yuvayoḥ yuṣmāsu Table 2: Forms of asmad Singular (1) Dual (2) Plural (3) 1 aham āvām vayam 2 mām āvām asmān 3 mayā āvābhyām asmābhiḥ 4 mahyam āvābhyām asmabhyam 5 mat āvābhyām asmat 6 mama āvayoḥ asmākam 7 mayi āvayoḥ asmāsu In each of these tables there are three possible candidates that can be called a “base” insofar as it provides a starting point for the derivation process leading to these forms. They are: tva, yuva and yuṣma in the case of the pronoun of the second person as displayed in Table 1 and aha, āva and asma in case of the pronoun of the 1st person as displayed in Table 2. These forms are candi- dates for being the bases of the respective rows, namely, tva and aha for singular, yuva and āva for dual, and yuṣma and asma for plural. Pāṇini how- ever, adopted only one base for all three numbers, namely, yuṣmad for Table 1 and asmad for Table 2, and substituted these candidates in place of them in the respective columns when the derivation process begins to derive forms in that respective column. I present below the entire derivation process for yuṣmad and present the methodology employed by Pāṇini. The same is applicable also for the word asmad. In this Table 3, information about the various rules required for de- riving these forms is presented. 158 Malhar Kulkarni Table 3: Rules required for derivation of yuṣmad in Pāṇini’s grammar No. Pāṇini Rule Application 1. 7.1.27 yuṣmadasmadbhyāṃ ṅaso ’ś yuṣmad/asmad [ṅas → aś] 2. 7.1.28 ṅeprathamayoram yuṣmad/asmad [su-au-jas- am-au-ṅe→ am] 3. 7.1.29 śaso na yuṣmad/asmad [śas → ns] 4. 7.1.30 bhyaso bhyam yuṣmad/asmad [bhyas (4/3) → bhyam] 5. 7.1.31 pañcamyā at yuṣmad/asmad [bhyas (5/3) → at] 6. 7.1.32 ekavacanasya ca yuṣmad/asmad [ṅas (5/1) → at] 7. 7.1.33 sāma ākam yuṣmad/asmad [sām → ākam] 8. 7.2.86 yuṣmadasmador anādeśe [yuṣmad → yuṣmā] anādeśa vibhakti [asmad → asmā] anādeśa vibhakti 9. 7.2.87 dvitīyāyāṃ ca [yuṣmad → yuṣmā] am/au/śas [asmad → asmā] au/śas/am 10. 7.2.88 prathamāyāśca dvivacane [yuṣmad → yuṣmā] au bhāṣāyām [asmad → asmā] au 11. 7.2.89 yo 'ci [yuṣmad → yuṣmay] ajādi vibhakti [asmad → asmay] ajādi vibhakti 12. 7.2.90 śeṣe lopaḥ [yuṣmad → yuṣmø] where there is no ātva/yatva [asmad → asmø] where there is no ātva/yatva 13. 7.2.91 maparyantasya [yuṣmad → yuṣm] [asmad → asm] 14. 7.2.92 yuvāvau dvivacane [yuṣmad → yuvad] dual suffixes [asmad → āvad] dual suf- fixes  Adaptive Reuse of the Descriptive Technique of Pāṇini 159 15 7.2.93 yūyavayau jasi [yuṣmad → yūyad] jas [asmad → vayad] jas 16. 7.2.94 tvāhau sau [yuṣmad → tvad] su [asmad → ahad] su 17. 7.2.95 tubhyamahyau ṅayi [yuṣmad → tubhyad] ṅe [asmad → mahyad] ṅe 18. 7.2.96 tavamamau ṅasi [yuṣmad → tavad] ṅas [asmad → mamad] ṅas 19. 7.2.97 tvamāv ekavacane [yuṣmad → tvad] ekavacane [asmad → mad] ekavacane In Table 4 below, I present the entire derivation process: Table 4: The derivation of yuṣmad according to Pāṇini’s grammar Form Procedure Rules 1/1 yuṣmad su 1.2.45, 4.1.1, 4.1.2 yuṣmad am 7.1.2, 7.2.91 tva ad am 7.2.94 tvad am 6.1.197 tva am 7.2.90 tvam 6.1.107 1/2 yuṣmad au 1.2.45, 4.1.1, 4.1.2 yuṣmad am 7.1.28 yuva ad am 7.2.91, 7.2.92 yuvad am 6.1.97 yuva ā am 7.2.88 yuvā am 6.1.101 yuvām 6.1.107 1/3 yuṣmad jas 1.2.45, 4.1.1, 4.1.2 yuṣmad am 7.1.28 yūya ad am 7.2.91 yūyavayau jasi 7.2.93 yūyad am 6.1.97 yūya am 7.2.90 yūyam 6.1.107 160 Malhar Kulkarni 2/1 yuṣmad am 1.2.45, 4.1.1, 4.1.2 yuṣmad am 7.1.28 tva ad am 7.2.92 tvad am 6.1.97 tva ā am 7.2.87 tvā am 6.1.101 tvām 6.1.107 2/2 yuṣmad au 1.2.45, 4.1.1, 4.1.2 yuṣmad am 7.1.28 yuva ad am 7.2.91, 7.2.92 yuvad am 6.1.97 yuva ā am 7.2.88 yuvā am 6.1.101 yuvām 6.1.107 2/3 yuṣmad śas 1.2.45, 4.1.1, 4.1.2 yuṣmad ns 7.1.29 yuṣma ā ns 7.2.87 yuṣm ā ns 6.1.101 yuṣmān 0 8.2.23 yuṣmān 3/1 yuṣmad ṭā 1.2.45, 4.1.1, 4.1.2 yuṣmad ā 1.3.7, 1.2.9 tva ad ā 7.2.92 tvad ā 6.1.97 tvay ā 7.2.89 tvayā 3/2 yuṣmad bhyām 1.2.45, 4.1.1, 4.1.2 yuva ad bhyām 7.2.92 yuvad bhyām 6.1.97 yuva ā bhyām 7.2.86 yuvābhyām 6.1.101 3/3 yuṣmad bhis 1.2.45, 4.1.1, 4.1.2 yuṣma ā bhis 7.2.86 yuṣmā bhis 6.1.101 yuṣmābhiḥ  Adaptive Reuse of the Descriptive Technique of Pāṇini 161 4/1 yuṣmad ṅe 1.2.45, 4.1.1, 4.1.2 tubhya ad am 7.2.95 tubhyad am 6.1.97 tubhya a 7.2.90 tubhyam 6.1.107 4/2 yuṣmad bhyām 1.2.45, 4.1.1, 4.1.2 yuva ad bhyām 7.2.92 yuvad bhyām 6.1.97 yuva ā bhyām 7.2.86 yuvā bhyām 6.1.101 yuvābhyām 4/3 yuṣmad bhyas 1.2.45, 4 .1.1, 4.1.2 yuṣmad bhyam 7.1.30 yuṣma 0 bhyam 7.2.90 yuṣmabhyam 5/1 yuṣmad ṅasi 1.2.45, 4 .1.1, 4.1.2 yuṣmad at 7.1.32 tva ad at 7.2.97 tvad at 6.1.97 tva0 at 7.2.90 tvat 6.1.97 5/2 yuṣmad bhyām 1.2.45, 4 .1.1, 4.1.2 yuva ad bhyām 7.2.92 yuvad bhyām 6.1.97 yuva ā bhyām 7.2.86 yuvā bhyām 6.1.101 5/3 yuṣmad bhyas 1.2.45, 4 .1.1, 4.1.2 yuṣmad at 7.1.31 yuṣma0 at 6.1.97 yuṣmat 6/1 yuṣmad ṅas 1.2.45, 4 .1.1, 4.1.2 tava ad as 7.2.96 tavad as 6.1.97 tavad aś 7.1.27 tava0 a 7.2.90 tava 6.1.97 162 Malhar Kulkarni 6/2 yuṣmad os 1.2.45, 4 .1.1, 4.1.2 yuva ad os 7.2.92 yuvad os 6.1.97 yuvay os 7.2.89 yuvayo ḥ 6/3 yuṣmad ām 1.2.45, 4 .1.1, 4.1.2 yuṣmad ākam 7.1.33 yuṣma ākam 7.2.90 yuṣmākam 6.1.101 7/1 yuṣmad ṅi 1.2.45, 4 .1.1, 4.1.2 tva ad i 7.2.86 tvad i 6.1.97 tvay i 7.2.89 tvayi 7/2 yuṣmad os 1.2.45, 4 .1.1, 4.1.2 yuva ad os 7.2.92 yuvad os 6.1.97 yuvay os 7.2.89 yuvayo ḥ 7/3 yuṣmad sup 1.2.45, 4 .1.1, 4.1.2 yuṣma ā su 7.2.86 yuṣmā su 6.1.101 yuṣmāsu Here Pāṇini’s grammar uses substitution in place of the base as well as of the suffix. There are in all eleven rules stating the substitution of the base and seven rules stating the substitution of the suffix. The substitution of the base can be classified under two headings: substitution of the final consonant -d- and substitution of the part up to m (i.e., yuṣm in yuṣmad). Pāṇini, who is famous for his brevity, needed nineteen rules to derive these twenty-one forms. Even by formulating such a comparatively large number of rules, Pā- ṇini did not manage to make the derivation process entirely transparent. There is at least one form whose derivation process still remains incomplete. That form is 4/3. In the derivation of this form, after the application of 7.2.90, the scope of the application of 7.3.103 is created, but there is no rule prohi- biting the application of 7.3.103. In order to avoid this problem, the Pāṇinian tradition resorted to two means: (1.) The tradition interprets the rule bhyaso bhyam 7.1.30 as bhyaso abhyam and then substitutes abhyam (and not bhyam) for bhyas, thereby not letting the scope of 7.3.103 come into being.  Adaptive Reuse of the Descriptive Technique of Pāṇini 163 (2.) The tradition adopts a meta-rule, namely, aṅgavṛtte punarvṛttāvavidhiḥ |2 which means: “once [an operation on an aṅga] has been applied, one is not supposed to apply another operation on the same [stem].” In the following, I present the way in which later non-Pāṇinian grammars adapted this technique of substitution and reused it. I shall focus only on one case, namely, 4/3. Kātantra (K)3 – yuṣmad bhyas 2.2 – yuṣma0 bhyas 2.148 – yuṣma abhyam 2.157 – yuṣmabhyam 2.17 Observations: K assumes yuṣmad to be the base. It adapts the strategy of substituting the base as well as the suffix. It substitutes the final d by 0, the suffix bhyam by abhyam and finally substitutes a by 0. Here, K avoids the problem faced by the Pāṇinian system by clearly stating that abhyam is the substitute. Cāndra (C)4 – yuṣmad bhyas 2.1.1 – yuṣmad abhyam 2.1.29 – yuṣm 0 abhyam 5.4.57 – yuṣmabhyam Observations: As did Pāṇini, C assumes yuṣmad to be the base. It adapts the strategy of substituting the base as well as the suffix. It substitutes the final ad by 0 and the suffix bhyam by abhyam. Here, C avoids the problem faced by the Pāṇinian system by clearly stating that abhyam is the substitute. 2 For more details, see Dikshit 1987: 353. 3 The Kātantra grammar is the oldest grammar among the post-Pāṇinian systems of grammar. It is belived that the rules of this system were first uttered by Kumāra Kārtikeya. Later Śarvavarman compiled these rules and presented them in the form of a system. Śarvavarman is dated between 150 and 100 CE. For more details, see Kul- karni, Ajotikar and Ajotikar 2010a. 4 The author of this grammar is Candragomin. The name of the author is found variously quoted as Candra, Candrācārya, etc. He is considered to have flourished in the 5th c. CE. For more details, see Kulkarni, Ajotikar and Ajotikar 2012. 164 Malhar Kulkarni Jainendra (J)5 – yuṣmad bhyas 3.1.2 – yuṣmad abhyam 5.1.26 – yuṣma 0 abhyam 5.1.149 – yuṣmabhyam 4.3.84 Observations: As did Pāṇini, J assumes yuṣmad to be the base. It adapts the strategy of substituting the base as well as the suffix. It substitutes the final d by 0 and the suffix bhyam by abhyam. Here, J avoids the problem faced by the Pāṇinian system by clearly stating that abhyam is the substitute. Śākaṭāyana (S)6 yuṣmad bhyas 1.3.135 yuṣmad abhyam 1.2.178 yuṣma0 abhyam 1.2.181 yuṣm0 abhyam 1.2.106 yuṣmabhyam Observations: Also S takes yuṣmad to be the base. It adapts the strategy of substituting the base as well as the suffix. It substitutes the final d by 0 and the suffix bhyam by abhyam and then again the base final a by 0. Here, S avoids the problem faced by the Pāṇinian system by clearly stating that abhyam is the substitute. After having presented the data from these four non-Pāṇinian systems, I would like to add that also other systems, such as the Sarasvatīkaṇṭhābha- raṇa (SKB)7, the Haima (H)8, the Mugdhabodha (M)9, the Sārasvata (Srv)10 and the Saṃkṣiptasāra (Ss)11 follow a similar procedure. The entire picture in this regard can be presented graphically in the following way: 5 It is believed that the Jainendra grammar was written by Devanandi, alias Pūjyapāda, who is believed to have flourished around the 5th−6th c. CE. For more details, see Kul- karni, Ajotikar and Ajotikar 2010c. 6 It is believed that the Śākaṭāyana grammar was written by Pālyakīrti. He flourished in the first half of the 9th c. CE. For more details, see Kulkarni, Ajotikar and Ajotikar 2010b. 7 It is believed that the SKB is written by King Bhoja, who flourished in the 10th−11th c. CE. For more details, see Mīmāṃsaka 1984: 553. 8 It is believed that the Haima grammar was written by Hemacandra, who lived 1088−1172 CE in northern Gujarat. For more details, see Belvalkar 1915: 61. 9 This grammar was composed by Vopadeva, who flourished in the 13th c. CE. For more details, see Belvalkar 1915: 87. 10 The origin of this school is not earlier than the 13th c. CE. For more details, see Belval- kar 1915: 77. 11 The origin of this school is not earlier than the 13th c. CE. For more details, see Belval-  Adaptive Reuse of the Descriptive Technique of Pāṇini 165 yuṣmabhyam Substitution of Substitution Gaps in base of suffix derivation Final d substituted ad substituted ad substituted bhyam abhyam SKB, H, Ss by zero by zero by a P, K, J, S, SKB C, M Srv P, Srv, Ss K, C, J, S, SKB, H, M Figure 1: Derivational strategies in different systems of Sanskrit grammar. In this overview, all systems are shown to have adapted the procedure of substitution of the base as well as the suffix. Some have substituted the final d of the base with 0 and others have substituted the final ad with 0. Almost all have tried to solve the problem faced by PāP ṇini by substituting abhyam in the place of the suffix bhyas. There are only two grammars that have substi- tuted bhyam in the place of bhyas, namely, Srv and Ss. Ss has a gap in the derivation and Srv substitutes the final ad of the base with a, thereby giving rise to the same problem as in P Pāṇini’s grammar. Apart from these two, all other grammar systems addressed the issue and came up with a solution. When adapting the technique of substitution, substitution, some paid attention to the minutest details of the derivational procedure, while others did not. On the basis of the outline above, I conclude that the post-Pāṇinian and non- Pāṇinian grammarians successfully adapted the technique of substituting one verbal element by another in describing linguistic facts. Most of them re- shaped the manner of substitution in order to remove the problem faced by Pāṇini. The post-Pāṇinian inian systems also adapted and reused the same base, namely, yuṣmad and asmad. It is indeed a surprise that no grammatical sys- tem uses tvad or yuvad as the base to derive the same forms. Nobody seems to have arrived at the idea that the derivation could also start with the other two elements as bases. This absence of innovation can be said to have been the cost of this adaptation and reuse in the case of the Sanskrit grammatical tradition. kar 1915: 78. 166 Malhar Kulkarni References Belvalkar 1915 Belvalkar, Shripad Krishna. Systems of Sanskrit Grammar (An Account of the Different Existing Systems of Sanskrit Grammar, Being the Vishwanath Narayan Mandlik Gold Medal Prize-Essay for 1909). Reprint. Delhi: The Bharatiya Book Corporation, 1997. Dikshit 1987 Dikshit, Girijesh Kumar, ed. Paribhāṣenduśekhara of Śrī Nāgeśa Bhaṭṭa: With the Commentary Sarvamaṅgalā. Va- ranasi: Sampoornanand Sanskrit Vishvavidyalaya, 1987. Sarasvatībhavana Adhyayanamālā 36. Freschi and Pontillo 2013 = Freschi, Elisa, and Tiziana Pontillo. Rule-extension Stra- tegies in Ancient India: Ritual, Exegetical and Linguistic Considerations on the tantra- and prasaṅga-Principles. Frankfurt a. M.: Lang, 2013. Kahrs 1998 Kahrs, Eivind. Indian Semantic Analysis, the ‘nirvacana’ Tradition. Cambridge, UK: Cambridge University Press, 1998. University of Cambridge Oriental Publications 55. Kulkarni, Ajotikar and Ajotikar 2010a = Kulkarni, Malhar, Anuja Ajotikar, and Tanu- ja Ajotikar. “Derivation of Declension of yuṣmad and asmad in Kātantra Grammar.” Alaukika Jñānayogī Dr. Gaṇeśa Tryambaka Deśapāṇḍe Smṛtigrantha. Ed. Manju- shree Deshpande and P. Deshpande. Vidyavihara, Rana- pratapnagar, Nagpur: G. T, 2010. Deshpande Smṛtigrantha Samitī. 143−157. Kulkarni, Ajotikar and Ajotikar 2010b = — “Derivation of Declension of yuṣmad and asmad in Śākaṭāyana Grammar.” Journal of Sanskrit Academy, Osmania University (Sanskrit Academy, Hy- derabad) 20 (2010): 97−109. Kulkarni, Ajotikar and Ajotikar 2010c = — “Derivation of Declension of yuṣmad and asmad in Jainendra Grammar.” Journal of the Oriental Institute, M. S. University of Baroda (Oriental Institute, Baroda) 60/1−2 (2010): 41−45. Kulkarni, Ajotikar and Ajotikar 2012 = — “Derivation of Declension of yuṣmad and asmad in Cāndra Grammar.” Devadattīyam. Johannes Bronkhorst Felicitation Volume. Ed. F. Voegeli, V. Elt- schinger, D. Feller, M. P. Candotti, B. Diaconescu and M. Kulkarni. Bern: Lang, 2012. 123−141. Worlds of South and Inner Asia 5. Mīmāṃsaka 1984 Mīmāṃsaka, Yudhiṣṭhira. Saṃskṛta Vyākaraṇa-Śāstra kā Itihāsa. 4th ed. Bahalgarh (Sonipat): Ram Lal Kapoor Trust, 1984. Nandargikar 1897 Nandargikar, Gopal Raghunath, ed. The Raghuvaṁśa of Kālidāsa, with the Commentary of Mallinātha, Engl. Transl. and Notes. 3rd ed. Bombay: Radhabai Atmaram Sagoon, 1897.  Section 2: Adaptive Reuse of Tropes  The Steadiness of a Non-steady Place: Re-adaptations of the Imagery of the Chariot Elena Mucciarelli This chapter points at the different meanings that the motif of the chariot as- sumed in different historical strata of the Vedic corpus. Moreover, it aims at supplying a picture of the semantic fields that words for “chariot” and its parts have been associated with. This provides the background for an analysis of the ways in which the motif was reused in later times, as for example, in South Asian medieval contexts. Throughout Vedic times, the “chariot was not merely a practical instrument for conveying persons, but an object vested with religious significance and symbolic values” (Sparreboom 1985: 1). More precisely, this vehicle functioned as a sacred space “in motion.” In the Ṛgvedic period, the chariot, representing an allegory of movement, figures in various poetic semantic fields. In the late Vedic period, against the back- ground of a wider social and political re-casting, both the use and the image of this vehicle undergo a process of specialization: the chariot plays a role within the great royal rituals, already as an attribute of power, as it will re- main in medieval times. Premise This contribution originated within a wider research project focusing on traces of alien elements within the ritual and social picture of sovereignty determined by the Brāhmaṇic worldview as presented in the Vedic corpus.1 These elements could be interpreted as remnants of a condition prevalent before the Middle Vedic reform, which entailed a restructuring of the social and ritual set-up (see section 2.1) and which is known as the śrauta reform. Within this context, together with my colleague Cristina Bignami, I started to work on the role of the motif of the chariot in connection with sovereignty. 1 “Traces of a Heterodox Concept of Kingship in Ancient, Medieval and Modern India” led by Tiziana Pontillo at the University of Cagliari. 170 Elena Mucciarelli We focused on how this motif was re-used in different South Asian cultural contexts.2 Looking at the motif of the chariot in different historical eras, we won- dered whether the idea of re-use might be a useful parameter for examining the recurrence of this motif in South Asian history and textual production. Indeed, this vehicle played a decisive role in royal rituals in both the Vedic and medieval periods. Some patterns, such as the relationship between the circular movement (displayed by the form of the wheel as well as by the movement of the chariot during ritual races) and the evocation of fertility,3 are strikingly similar in both contexts. Yet the conceptual frameworks in which the motif of the chariot is inscribed in each period are different. The present contribution represents a preliminary analysis of the motif of the chariot and its imagery in Vedic times, whereby it aims at collecting the pieces of evidence needed to establish a comparison with its use in the me- dieval period. At the same time, I will also chart the development of the cha- riot itself within the Vedic period and show the various changes it underwent. As will be shown below, the passage between the early and the middle Vedic periods saw a restriction of the different values attached to the chariot. This shrinking of the scope of meanings correlates with a ritualization of the cha- riot under the influence of a change in the lifestyle of the Vedic tribes. Later, in the early medieval and pre-modern periods, the imagery and role of this vehicle was again strongly modified and adapted in order to fit different needs more fully. Before any further analysis, a clear distinction must be made between the different types of chariots: the wagon, the war chariot, the royal chariot, the vipatha (a kind of chariot used by certain types of troops) and the rathavāhana (a movable stand to carry chariots). In this study I will focus on the word ratha “chariot”: a “light, fast, two-wheeled vehicle with spoked wheels, usually drawn by horses and used for warfare, hunting, racing and ceremonial purpose” (Littauer and Crouwel 1979). An interesting subcate- gory of the ratha is the śamaratha, used for an old chariot, a chariot in peacetime, or a disassembled or broken chariot. This is contrasted with the aśamaratha, a safe and running chariot. It is hard to imagine what a ratha may have looked like, all the more since many descriptions are related to the rathas of the gods and not to those of 2 This research was undertaken as part of the DFG-financed project “Kings of the Wild: Re-use of Vedic and local elements in the legitimation process of medieval Karna- taka.” The work of the author is being supported by the Institutional Strategy of the University of Tübingen (Deutsche Forschungsgemeinschaft, ZUK 63). 3 See below, section 2.2 and the chapter in this volume by Bignami, section 4.  The Steadiness of a Non-steady Place 171 humans. The immortal chariot of the Aśvins is “honey bringing”; it could have three wheel rims as well as three supports,4 perhaps to hold something on it (ṚV 1.34.2). The chariot of Savitṛ in ṚV 1.35.4 is covered with pearls and golden yoke-pins. And it is no surprise that the means of transportation for Agni is bhānumat, luminous (as, e.g., in ṚV 5.1.11). Notwithstanding the focus on divine vehicles, from the verses praising and describing chariots one can gain a glimpse of how they were conceived and imagined. Moreover, a number of hymns provide concrete details and features of worldly chariots.5 For instance, in ṚV 1.30.16 Indra gives men a golden ratha, in ṚV 6.47.26 it is stated that the wood of the war chariot is knotted with cows, meaning that the chariot was constructed with joints and leather straps. A wonderful de- scription can be found in ṚV 10.85, where the chariot, decorated with flowers and wood, plays a central role in the marriage of Sūryā. 1 The Ṛgvedic ratha: The chariot as a living prismatic metaphor I would like to begin by charting the main notions linked to the chariot in the Ṛgvedic poetic corpus. In this way, I aim at showing the multiple meanings evoked by the image of chariots in this anthology of hymns. In fact, its se- mantic values extend from the concrete battle-field to the poetic domain of rituals, thus giving rise to a tangle of metaphors and reciprocal connections.6 1.1 ratha and swiftness The first feature associated with a ratha is its swiftness, which it shares with soma (9.38.1 arṣati). Being swift is a modality of the notion of “going,” which had a key importance in the semi-nomadic society of early Vedic times. The following mantras provide examples of this feature.7 ūrdhvó naḥ pāhy áṃhaso ní ketúnā | víśvaṃ sám atríṇaṃ daha | kṛdhī́ na ūrdhvā́ ñ caráthāya jīváse | vidā́ no dúvaḥ || (ṚV 1.36.14) 4 In 1.34.2 the chariot has tráyaḥ paváyaḥ and tráya skambhā́ saḥ. 5 In particular in 3.53.17–19, quite a few details are mentioned about the different com- ponents of the chariot, including the kinds of wood to be used. 6 Many scholars, e.g., Edgerton (1919) and Windisch (1893), have already pointed out this semantic richness. 7 Unless otherwise stated, all translations are my own. 172 Elena Mucciarelli Upright protect us from the enclosure8 with (your) sign9, burn the evil devourer (Atri) completely; make us upright for a vagrant life (a life on the move)10, may you find for us preference among the gods. The strophe above prays for a carátha life, while in the following strophe 1.72.6 the contraposition between carátha and sthā́ tṛ emerges clearly: both are to be protected and together they represent the entire livestock: tríḥ saptá yád gúhyāni tvé ít padā́ vidan níhitā yajñíyāsaḥ | tébhī rakṣante amṛ́taṃ sajóṣāḥ paśū́ ñ ca sthātṝ́ñ caráthaṃ ca pāhi || (ṚV 1.72.6) As those worthy of worship found the three times secret paths con- tained in you, with these they protect the immortal one acting in har- mony; protect the stationary cattle and the [wild animals that] move.11 Again, in 1.68.1 we find the binomial set sthātṛ-caratha, referring to the activity of Agni as the common fire:12 śrīṇánn úpa sthād dívam bhuraṇyú sthātúś carátham aktū́ n vy ū̀ rṇot | (ṚV 1.68.1) Burning, he approaches the sky quivering, he uncovers the one still and the one moving during the nights. The same reference to the activity of Agni is found also in 1.70.7, where it is paralleled by the expression sthatuś ca ratha, used also in strophe 3 of the same hymn. 8 On the meaning of the word áṃhas conveying the idea of a concrete “narrowness” as a mental distress, see Gonda (1957: 254). Taking into account the pāda c, I would sug- gest keeping the spatial idea of something small and closed vs. the urú kṣáyāya (1.36.8) summoned up with carátha. 9 ketú, derived from the verbal root cit-, conveys the idea of a sign of recognition through its illuminating qualities; see Renou 1955 and EWAia I, 399. 10 Proferes (2003: 172) translates this “for carrying out a long life,” but I would say that here the idea of moving implied in carátha cannot be left out, because this was a cen- tral element of the life of the Vedic tribes; cf. also the translations of Geldner, and the one by Witzel and Gotō (1.36.14d) “Richte uns auf zum Wandel, zum Leben!” 11 Here sthātṛ might also refer to plants, as is the case with similar expressions in the Pāli Canon (see Schmithausen 1991). However, the reference to plants would be an anach- ronism in the Ṛgveda, which reflects the life of nomadic tribes: plants have more signi- ficance in the later period of semi-nomadic life. See also the translation in Jamison and Brereton 2014. 12 See Proferes (2007: 395) for the idea of a common fire and many fires as a social and political metaphor.  The Steadiness of a Non-steady Place 173 várdhān yám pūrvī́ ḥ kṣapó vírūpā | sthātúś carátham ṛtápravītam || (ṚV 1.70.7) Him they made grow during the many nights of different form, the one still and the one moving, the one conceived through order. 1.1.1 ratha as a means for crossing fields In the context of the semi-nomadic existence of the carátha life, the chariot provides more than just an instrument to move: it has the ability to go through, and in this sense, it helps to cross difficulties. This aspect is more than one of the many features of the motif of the chariot; the capacity to tra- verse (pṛ- / tṛ-) is indeed the underlying rationale connecting the motif of the chariot to poetry, as we will see below, and to ritual. Such an ability is also that of the boat,13 as found in 10.101.2; 1.46.8 where the tīrthe sindhunām is clearly related to the ritual. In 10.116.9 and finally in 7.70.2, it is not a ship that is being spoken about, but a drink, gharma, is mentioned, which, in turn, is said to be yujanaḥ like a chariot and able to traverse the waters (samudram piparti).14 1.2 The godly character of the ratha Elsewhere the chariot is not regarded as an instrument for transportation; the poets rather emphasize its sacred and supernatural essence. The godly charac- ter of the chariot is affirmed both in the sense of it having a divine status in itself, as well as by the use of the word ratha in apposition to, or as an appel- lative for, the gods, especially for Agni. In the following strophe, the fire-god is identified with a series of other gods, either as a part of their bodies or through their attributes; the last epithet of the list is déva ratha. índrasya vájro marútām ánīkam mitrásya gárbho váruṇasya nā́ bhiḥ | sémā́ ṃ no havyádātiṃ juṣāṇó déva ratha práti havyā́ gṛbhāya || (ṚV 6.47.28) You are the vajrá of Indra, the row of the Maruts, the son of Mitra, the navel of Varuṇa. Rejoicing in this oblation given by us15 take in / in- gest the oblation, o divine chariot. 13 The analogy of ritual as a ship is also found in late Vedic literature; see ŚB 4.2.5.10, AB 4.13; TS 5.3.10.1 (cf. Hauer 1927: 247). For different types of boats in the Vedic period see Klaus 1990. 14 On the meaning of the noun samudrá in the Vedic texts, apart from the controversy connected with the river Sarasvatī, see Klaus 1989. 15 hayvadāti, “das Geben der Opfertränke” (Grassmann 1875 [1996: 1657]), emphasizes the action of giving. As an adjective it is especially used for Agni, who conveys the 174 Elena Mucciarelli As it could be expected, there is frequently a connection between chariots and the sun gods such as the Aśvins, Sūrya and Savitṛ, which traces back to the Indo-European imagery of the sun as a wheel carried through the sky by a horse.16 1.3 ratha and conquest Another semantic world evoked by the chariot is the one of conquest, battle and victory. The ratha is strongly connected to competitions,17 both as con- quering expeditions or raids (gaviṣṭī) and as races (āji). In this connection it is noteworthy that the word that will later indicate a ritual race, āji, as a tech- nical term, in the Ṛgveda Saṃhitā conveys the general meaning of “race” or “battle.” 1.4 ratha in the ritual context The victory, the booty to be gained, represents the overlapping aim of both warfare and sacrifice, as found in 1.123.5. jáyema táṃ dákṣiṇayā ráthena (ṚV 1.123.5d) May we win over this one with a dakṣiṇā as a chariot – with the dakṣiṇā, the chariot. Furthermore, soma is often equated with the chariot, not only because of the swiftness of both, but also their shared ability to bring wealth: prá svānā́ so ráthā iva / árvanto ná śravasyávaḥ | sómāso rāyé akramuḥ (ṚV 9.10.1) Resounding like a chariot the soma juices seeking glory like coursers proceeded towards wealth. The sacrificial fee, which stands metonymically for the sacrifice, is also equated with the chariot, insofar as both are said to be instruments for attain- ing a secure success. Moreover, Franklin Edgerton (1919: 180–181) suggested reading the mo- tif of the chariot in a few hymns, such as ṚV 10.53 and ṚV 10.70, from a metaphoric perspective. He speculated that no real vehicle is intended here, but rather a symbol either for parts or for the entirety of the rite. The link oblations to the other gods. In this use, havyadāti alludes to the role of Agni as a mes- senger. 16 See already von Schroeder 1916: 65–69. 17 Its warlike character is also expressed by its association with weapons like vájra, a comparison that later thrives in post-Ṛgvedic texts.  The Steadiness of a Non-steady Place 175 between the ratha and sacrifice is established very clearly later on, since in the post-Ṛgvedic Saṃhitās and Brāhmaṇas, the yájña is said to be a ratha. In post-ṚV times, a ratha can also be said to be made of meters and is therefore called a chandoratha; in particular, this correlation also has a prac- tical application insofar as an actual chariot can be used as a unit of measure for constructing the sacred space, as, e.g., in the nirūḍhapaśubandha, the separate offering of animals or the offering of an eviscerated animal, in which the chariot serves as the measurement for establishing the sacred area. To sum up, the Ṛgvedic and Brāhmaṇa associations of sacrifice with the vehicle are similar to one another, but in fact the basis of such relations are quite different: in the latter the space of the ritual procedure is the underlying rationale, while in the former the two elements are associated through poetic imagery. 1.5 ratha and poetry The connection with ritual is indeed part of a larger set of correlations that link the motif of the chariot to speech (vāc), especially the sacred speech of the hymns that is considered to shape reality. The relation to poetry of the ratha-motif is indeed expressed through a variety of connections that relate to a number of semantic levels. Among them is frequently found the compari- son between the chariot builder and the poet, often expressed with the verb takṣ-, as in ṚV 5.2.11: etáṃ te stómaṃ tuvijāta vípro | ráthaṃ ná dhī́ raḥ svápā atakṣam | yádī́ d agne práti tváṃ deva háryāḥ | súvarvatīr apá enā jayema || (ṚV 5.2.11) This praise of you, born of power, I, inspired, fashioned: a chariot like a skillful artisan; if this will delight you, o Agni, we may then win wa- ters and sun. Likewise, the competitive aspect of poetry, later highlighted in the brahmod- ya, is associated with horse races, as in ṚV 8.3.15, where the honeyed songs are said to compete for the prize like chariots. In the same context a further association, already mentioned above, is established between the motif of the chariot and the idea of passing over difficulties. Additionally, poetic skills are often associated with the motif of the chariot: vayám u tvā pathas pate ráthaṃ ná vā́ jasātaye | dhiyé pūṣann ayuj- mahi || (ṚV 6.53.1) 176 Elena Mucciarelli O Puṣan, we have yoked you for inspired poetry, like a chariot for vi- gor18, o lord of the path. Similarly, in 1.113.17 we find the expression “the reins of speech” (syūmanā vācaḥ), while in 7.24.5 a prayer is compared to a horse at the chariot-pole: dhurī̀ vā́ tyo ná vājáyann adhāyi. This image is elaborated in 7.64.4, where the mind itself shapes the chariot; thus, the vehicle is not only compared to a hymn, but it is also created through identical means. In the same verse, one who might be sprinkled by the benevolence of Mitra and Varuṇa is said to be equally capable of creating a chariot seat (this is again expressed with the verb takṣ-) and of producing inspired visions: yó vāṃ gártam mánasā tákṣad etám ūrdhvā́ ṃ dhītíṃ kṛṇávad dhā- ráyac ca | ukṣéthām mitrāvaruṇā ghṛténa tā́ rājānā sukṣitī́ s tarpaye- thām || (ṚV 7.64.4) Who will fashion for you two this chariot seat with his mind, who will make a high inspired vision and support it, may you, o Mitra and Va- ruṇa sprinkle him with ghee, may you bless the secure dwelling, o kings. In 2.31.7 the hymn in its entirety is related to a well-equipped chariot, which is also in other instances strongly related to inspired chants. In the first example one finds again the verb takṣ-: etā́ vo vaśmy údyatā yajatrā átakṣann āyávo návyase sám | śrava- syávo vā́ jaṃ cakānā́ ḥ sáptir ná ráthyo áha dhītím aśyāḥ || (ṚV 2.31.7) I wish these being raised up for you, the men fashioned into a new one; may they get an inspired vision like a chariot’s steed seeking fame, desiring the prize. The following stanza is similar: prá śukraítu devī́ manīṣā́ asmát sútaṣṭo rátho ná vājī́ (ṚV 7.34.1) Let the bright divine inspired thoughts come forth from us, like a cha- riot seeking the prize. 1.6 ratha and generative power Through the noun vāja, used in both of the stanzas quoted above, we arrive at another motif connected to the ratha: its generative power, to which the word vāja is also connected. This term, often referred to as the boon that must be won at a ceremonial chariot race after the winter in order to let the earth re- 18 For vā́ ja as vigour or vegetative power, see Gonda 1954.  The Steadiness of a Non-steady Place 177 joice,19 is meant, according to Gonda (1954: 194), as vital power, namely vegetative energy. We shall see below the same concept in the context of ritual races (section 2.2.3). Another element that links the chariot and vāja is the sun, in its turn a creator, as in ṚV 9.90.1 ab: prá hinvānó janitā́ ródasyo rátho ná vā́ jaṃ saniṣyánn ayāsīt (ṚV 9.90.1b) The creator of the two worlds being pushed forward has advanced like a chariot striving to gain the vāja. By contrast, in ṚV 3.38.1 ab the motif of fertility is associated to the chariot, and both chariot and fertility are included in a simile for poetic work, an act like that of the taṣṭṛ, thus interweaving these two semantic fields into a single representation. abhí táṣṭeva dīdhayā manīṣā́ m átyo ná vājī́ sudhúro jíhānaḥ | (ṚV 3.38.1ab) Like a carpenter I reflected on my poetic inspiration20, as a horse seek- ing for the prize that bounds forward, well drafted to a yoke. The chariot is associated with fertility both in the Ṛgveda Saṃhitā and in the Atharvaveda tradition. Mostly, the idea of circular movement embodied by the spin of the wheels represents a symbol of new life, as in many cultural contexts, and this imagery is developed in the royal rituals as well (see be- low). Furthermore, chariots are associated with marriage, as in the famous hymn 10.85 where Sūryā is brought to her husband on a chariot.21 Likewise, in the much debated sūkta 10.102, a chariot race is mentioned as part of a plot that, although obscure in many details and open to different interpretations,22 19 See also Kuiper 1960: 240. 20 The word manīṣā́ , “geistige Erregtheit” (EWAia II 308–309), does not only represent the conjuring up of an idea, but also the work of the mind in connection with inspira- tion. According to Thieme (1967), who explains the noun as a compound from a root- noun man and the verbal noun īṣā́ “die Aufwallung,” it conveys the idea of “dichte- rische Ekstase,” thus pointing to the idea of inspiration that resembles the Greek en- thusiasmós. Gonda translates the term with “inspired thought”; see Gonda 1963. 21 This same hymn is re-used in the liturgy of the royal rites (see section 2.2.1). 22 See the different readings of Brereton (2002), who proposes connecting it with a niyoga ritual, and Jamison (2011), who reckons it with the dialogic discussions about the ritual innovations that took place at the end of the early Vedic period. According to the latter interpretation, the matter under consideration was the participation of the patnī in the solemn rituals. 178 Elena Mucciarelli is nonetheless linked to the theme of fertility and the issue of begetting offspring for an impotent husband. 1.7 Summing up: The many semantic values of the ratha in the Ṛgveda Saṃhitā To sum up, the main semantic fields associated with the motif of the chariot in the Ṛgvedic anthology are as follows: swiftness, divine status, conquest, victory and sacrifice, poetry and fertility. It is important to focus on this semantic variety in order to contrast it with the features associated with the chariot in the following periods and to ex- plore the choices made within conscious reuses of this element. While in the Ṛgveda Saṃhitā the use of the chariot motif is quite diverse, in the late Vedic period, the chariot and its corresponding motif undergo a process of speciali- zation in the framework of a wider social and political re-casting. Within the great royal sacrifices, the chariot comes to play a role as an attribute of power, as is also clearly the case in the medieval period. Accordingly, the vehicle contains a metaphoric as well as a concrete status. As for their con- nection, as maintained also by Sparreboom (1985: 6), already in the śrauta period, the chariot was a thing of the past and its driving, a lost art. This change created room for the chariot to assume a more confined ritual role. 1.8 The medieval adaptive reuse of the ratha compared to its Vedic use Before considering the middle Vedic period, I shall dwell briefly on the Ṛg- vedic use of the chariot in the light of the post-Vedic devotional development of this use, as analyzed in the chapter by Bignami in this volume. Although there is no evidence for a processional use of the chariot in the Ṛgvedic cor- pus, nonetheless a form of procession that could well have been a real jour- ney accompanied by rituals is conceivable. Taking into account the medieval and modern processional chariot, it is worth noting that what in the Ṛgvedic period was a real voyage ‒ re-used in a simple and linear way during the less nomadic time of the following centuries ‒ changed substantially in medieval times, when we witness its adaptive reuse. If compared to Vedic times, the ritual role of the ratha is adaptively reused in connection with royal power as an instrument for negotiating politi- cal claims. As anticipated at the beginning of this chapter, this kind of reuse does not entail a strong modification of the semantic or cultural values con- nected with the idea of a ratha. In this sense, what happens within the Vedic period is the linear re-use of the chariot. In contrast, as can be seen in Bignami’s chapter, the motif of the chariot in the medieval period, while remaining an important component of royal  The Steadiness of a Non-steady Place 179 legitimation, was subject to a strong adaptation within different religious and political contexts. Its imagery was deeply modified due to new issues: the chariot as a model for the temple, as suggested by Bignami, indeed represents a completely new use of the motif. 2 The linear re-use of the ratha in the middle Vedic period: The symbolic chariot 2.1 The socio-political context of the re-use The major social and political shift that took place around the turn of the first millennium BC involves some aspects23 that must be recalled here, since they probably determined the modification of the imagery of the chariot. While the Ṛgvedic period sees the warring of everybody with everybody ‒ a time in which we may imagine a political structure composed of some fifty small tribes in constant conflict with each other and against other indi- genous peoples ‒ at the beginning of the so-called Yajurvedic period, a more stable political situation had emerged. In the new set-up, which corresponds to the composition of the liturgic Saṃhitās, the division of power has been reshaped. In particular, the two social figures of the kṣatriyas and rājans begin to emerge from the group of “equals” or sajāta.24 In some rites attested in the liturgical corpus, such as the “optional or wish offerings” (kamyeṣṭi), it is possible to pinpoint tensions between a chieftain lord (rājan) who is ruling, or trying to rule, over fellow rājans and kṣatriyas and the viś. The political instability increases in the following Brāhmaṇa period, due to the shift from a semi-nomadic to a resident culture. In fact, as a consequence of this shift, the state formation shows a development from below, while the leaders increase their efforts to establish a pyramid-like political structure (Kulke 1992). These contrasting tendencies are further intensified by the social fragmenta- tion of the region. All of these elements lead to a high degree of instability, which might resemble that of the “elusive”25 south South Asian medieval kingdoms, whose rulers engaged in an endless effort to maintain their power. The political dynamics at the emergence of the rājans and kṣatriyas is well represented in the rāṣṭrabhṛt ceremony: the rāṣṭrabhṛt offerings should 23 For a complete picture, see, among others, Witzel 1999 and Bryant and Patton 2004. 24 See the description of this mechanism in Kulke 1992. 25 As to the similarity between state formation in the Vedic and medieval periods, see Kulke 1995, and Kulke and Rothermund 2010. The term “elusive” has been used by Shulman (1985) to refer to kings whose kingdoms were always close to disappearance. 180 Elena Mucciarelli be offered by someone who desires the kingdom, as well as by someone who desires a village. In the text of Taittirīyasaṃhitā 3.4.8 (cf. also KS 37.11: 91.16) the grāmakāma indeed seems to match Kulke’s description of the new role that came with the first state formations and the need of chiefdoms to establish their power among the sajātas (see Kulke 1992: 190–191). rāṣṭrákāmāya hotavyā̀ rāṣṭráṃ vái rāṣṭrabhṛ́to rāṣṭréṇaivā́ smai rāṣ- ṭrám áva runddhe rāṣṭrám evá bhavati | ātmáne hotavyā̀ rāṣṭráṃ vái rāṣṭrabhṛ́to rāṣṭrám prajā́ rāṣṭrám paśávo rāṣṭráṃ yác chréṣṭho bhá- vati rāṣṭréṇaivá rāṣṭrám áva runddhe vásiṣṭhaḥ samānā́ nām bhavati grā́ makāmāya hotavyā̀ rāṣṭráṃ vái rāṣṭrabhṛ́to rāṣṭráṁ sajātā́ rāṣṭré- ṇaivā́ smai rāṣṭráṁ sajātā́ n áva runddhe grāmī́ || (TS 3.4.8.1) They should be offered by someone desiring a kingdom. The rāṣṭra- bhṛt are the kingdom. With the kingdom he obtains the kingdom for him. He becomes the kingdom. They should be offered to himself. The rāṣṭrabhṛt are the kingdom. The people are the kingdom, the cat- tle are the kingdom. In that, he becomes the best, he is the kingdom, he obtains the kingdom with the kingdom. He becomes the wealthiest among similar men. They should be offered by someone desiring a grāma. The rāṣṭrabhṛt are the kingdom. The related men are the king- dom. With the kingdom he wins for him the kingdom, the related men. He is the grāmin. The two parallel structures of this paragraph stress the two different cha- racteristics of the rāṣṭrákāma and the grāmakāma. Both long for a rāṣṭra, but while the former is relying on the prājas, a term also used for subjects, the grāmakāma has his support in the sajātas – thus implying a less pyramid-like society. The rituals that will be examined below belong to the middle Vedic pe- riod, when state formation had already taken place. The texts depicting them result from the effort to create a shared ritual set and sustain the social struc- ture that had developed during that stage of state formation. Nonetheless some traces of the conflicts that took place in the intermediate period, such as can be seen in the rāṣṭrabhṛt ceremony, are still apparent in the texts. From this perspective, I will analyze the role of the chariot and the corresponding motif in this new social panorama, especially the way this vehicle is used within rituals involved in the legitimation process of the new social structure. 2.2 The chariot in the middle Vedic sacrifices The creation of homogenous ritual procedures was part of the political agenda of the chieftains taking control over the different tribes. This was all  The Steadiness of a Non-steady Place 181 the more true in the case of royal sacrifices, in which the chariot was used. In this sense, the ratha imagery underwent a recasting process that preserved only a few features attested in Ṛgvedic hymns. As seen above, in that earlier period, chariots were the main instruments with which contesting parties engaged in battle (section 1.3) and they were also related to the ability of going through obstacles (section 1.1.1). The former aspect, as argued by Sparreboom,26 is the main reason for the connection yajña-ratha, and, we may add, for the pivotal role of the chariot in royal rituals. However, the ritual use of the ratha shows that its role had become problematic and that the actual practices of racing or raiding were slowly being replaced or subli- mated. As will become evident in the following examples, the chariot was used only as a ritual token, whereas the concrete function underlying the ritual use had been lost. 2.2.1 The chariot in non-royal sacrifices First, I shall consider rituals other than the great royal sacrifices in which the chariot plays a role. The ratha was conceived as a dakṣinā in the Śrauta Sūtra (LātŚS 2.7.20; DrāhŚS 5.3.22) and in the Gṛyha Sūtra literature. Moreover, within the rathārohana rituals, there is the prescription of a drive as an initia- tion for a new chariot (ĀśvGS 2.6.1–15). This ritual drive is also inserted into the marriage ceremony as it is described in the Vārāhagṛhyasūtra (VGS) and the Mānava Gṛhya Sūtra. Quite interestingly, a sort of post – more precisely a tree or a funerary monument (caitya in VGS 15.4) – is again mentioned in these texts, as in the earlier race contexts (see section 2.2.2). Additionally, the Ṛgvedic mantra to be recited during such wedding ceremonies is a modifica- tion of 10.85.20, where Sūryā, the daughter or wife of the sun, is mentioned along with a rich description of a chariot that is here again involved in a sort of wedding procession: sukiṃśukaṃ śalmaliṃ viśvarūpaṃ hiraṇyavarṇaṃ sudhuraṃ sucak- ram āroha sūrye amṛtasya panthāṃs tena yāhi gṛhānsvasti ityāropa- yet | 2 | (VGS 15.2) “Mount the good, golden chariot made of śalmali, adorned with kiṃśuka flowers, colorful, that moves smoothly, o Sūryā – o bride – (mount) the paths of immortality, along with it go to the house hap- pily!” saying this, may he make her mount. 26 Sparreboom 1985, in particular p. 82: “the chariot and the sacrifice were inseparable, being the instrument for the winning of booty and the conquering of new land.” 182 Elena Mucciarelli The use of this strophe containing the image of a procession of vehicles framed within a solar context seems to allude to the idea of circular move- ment as found in the Ṛgveda Saṃhitā. This aspect, not prominent in the Brāhmaṇical representation of the chariot, seems to be an aspect of the cha- riot that survived the middle Vedic reform. Another use of the chariot that is not connected with the royal sacrifice is to transport the fire in the Agnyādheya, the ritual installation of the sacrificial fire. Notably, in this case, a horse, the fire and the chariot are presented sepa- rately: first the horse is led to the āhavanīya (the sacrificial fire), then it is led around it with a turn and is sprinkled; afterwards it is let loose to the north, and eventually, it is brought back to its starting position. The adhvaryu (the priest entitled to perform the sacrifice) carries the fire, and then a chariot wheel, or a chariot, is led from the gārhapatya (the domestic fire) to the āhavanīya (the sacrificial fire).27 The three components of the possible func- tion of the ratha as a means of transportation have been disassembled, and we might say that they are “no longer in use.” The chariot is slowly changing its function, and consequently its value, being more a symbol of conquest and fortune than an actual instrument for attaining desired boons. 2.2.2 The chariot in the royal sacrifices Turning now to the royal sacrifices, in the Aśvamedha a chariot drive takes place just before (Kātyāyanaśrautasūtra [KātŚS], Mānavaśrautasūtra, Bau- dhāyanaśrautasūtra [BauŚS]) or after (Āpastambaśrautasūtra) the anointing and adorning of the sacrificial horse by the three wives on the second soma- pressing day, the day on which the horse is sacrificed. The chariot is driven to a pond that lies in the south according to BauŚS 15.24: 228.6 atha dakṣiṇaṃ hradam abhiprayāti, in the east (Kātyāyana), or in the north (Āpastamba). In some texts the horses enter the water ā kroḍebhyo ’śvān abhidhāvayanti (BauŚS 15.24: 228.7), while in the Āpastambaśrautasūtra they can also just smell the waters. Afterwards, the chariot makes a turn to the right, which seems to resemble the turn around the end post in a chariot race.28 Consider- ing these descriptions, it is clear that the role of the chariot in the Aśvamedha is already merely symbolic; a similar level of abstraction can be found also in the aforementioned rituals, where the chariot does not seem to have any practical purpose. Let us now turn our attention to the rituals in which, already in the Vedic period, the use of a chariot was connected with the consecration of sove- 27 Only Vaitānaśrautasūtra 2.5.18 has a real transport rathenāgnau praṇīyamāne. 28 BauŚS 15.24: 228.8 athaitaṃ rathaṃ pradakṣiṇam āvartya śālām ānayati.  The Steadiness of a Non-steady Place 183 reignty. The two main royal rituals in which the ratha plays a more func- tional role are the Vājapeya and the Rājasūya.29 In the descriptions of the Vājapeya (BauŚS 11.6–11.8) the chariot is used to perform a race, and the main ritual sūtras give a similar account of this with only minor variations. It is a real race, entailing a prize and a winner: the chariot of the rājan must go around a post made of udumbara wood. On the other hand, the Rājasūya presents some peculiarities. In the KātŚS, the Mānava Śrauta Sūtra and the Āpastamba Śrauta Sūtra, the role of the chariot is described differently than in the BauŚS. In the former three texts, which are dated to a later period, only the preparation of the chariot occurs vājapeyavat “as in the Vājapeya,” whereas the actual race does not take place: it is rather a symbolic horse drive that is referred to. According to the KātŚS, only one chariot driven by the yajamāna is present; it is led amidst cows that belong to a relative of the sacrificer (sva). A cattle raid of relatives and the shooting of an arrow against them is also referred to in LātŚS 9.1.14– 22. In contrast, according to ĀpŚS, HirŚS and MānŚS the sacrificer raids a kṣatriya or rājaniya armed with a bow. The sacrificer shoots at him. On the contrary, in the BauŚS, the oldest of these textual sources, a chariot race is indeed performed and, unlike in the other texts, it is joined with a symbolic cattle turn of a single chariot by the adhvaryu. This symbolic drive is done twice, at the beginning and the end of the rite. We can summarize the variants regarding the procedure of the Rājasūya as follows: a) Single chariot: 1) Āpastamba, Hiraṇyakeśin, Mānava: symbolic cattle and raids of kṣatriya. 2) Shooting at kṣatriya. 3) Kātyāyana: cattle raid of cattle belonging to a relative of the sacri- ficer. 4) Lāṭyāyana: the king shoots his arrows at his relatives. b) Several chariots: Baudhāyana30: 1) The adhvaryu performs a symbolical drive. 29 For a thorough description and analysis of the Rājasūya (especially that of the Yajur- veda school), see Heesterman 1957. 30 BauŚS 12.7; 12.12–14. 184 Elena Mucciarelli 2) The sacrificer puts on boar skin sandals and pays homage to the earth. 3) The race is settled and the rājaputra takes the role of the sacrificer. The adhvaryu instructs him to shoot at the rājanya. 3a) The race begins, and during the race drums are played and loud cries are made, as during the Mahāvrata (see below). 3b) The race ends and the adhvaryu makes another symbolic drive. The first thing to be noted here is the substitution of the yajāmana with the adhvaryu in the symbolic drives. Another interesting difference is that the shooting is performed in this ritual along with the chariot race by the prati- hita “the heir apparent.” Indeed, the rājaputra, for whom the Rājasūya estab- lishes a strong lineal connection,31 is instructed by the adhvaryu to shoot at a rājanya and miss him, in order to gain a thousand cows: atha pratihitāya dhanuḥ prayacchann āha rājanya eṣa ṣaṭtriṃśatsu śamyāpravyādheṣu nirjayena sahasreṇāvasitas tasmā iṣum asyatād apainaṃ rādhnutāj jitvainaṃ dakṣiṇāpathenātyākurutāt iti | rājanya eva saṃśiṣṭo bhavati rājaputras ta iṣum asiṣyati sa tvāparātsyati tasmā uttaravargyeṇa saṃmṛjyeṣu prayacchatāt | iti (BauŚS 12.12: 103). Then he gives the bow to the heir apparent and says, “a rājanya is standing within thirty-six throws (śamyā) with a thousand conquests (i.e., cows), throw to him the arrow and miss him, having won him bring them here through the dakṣinā path.” The rājanya is also in- structed: “the rājaputra will throw an arrow at you and miss you; hav- ing removed it with the upper garment, grant it to him.” Although it is not clear exactly at which point this raid is performed (before the race, immediately after the instruction sa tathā karoti, or during the chariot race), the passive role of the yajamāna is striking. If we look at the combination of these elements, what we get from the ac- count in the BauŚS is a quite complex and confused picture, particularly as a guide to actions to be performed; Sparreboom already wondered how these different actions could be performed together.32 If the chariot doing the symbolical drive belongs to the sacrificer, for instance, it can be the same chariot that is used to shoot at the rājanya, and how could it possibly take 31 On the role and significance of the heir apparent, see Jamison 1996: 110–114. 32 See Sparreboom 1985: 49–50.  The Steadiness of a Non-steady Place 185 part in the race? If we consider the relative chronology33 of these texts, Baudhāyana reflects the most archaic state we can have access to, and it may enable us to shed light on an older and not attested form of rājasūya. As Heesterman has pointed out, in the accounts we have for rituals, there are traces of a “genuine” antagonism.34 It is therefore worth noting the “confu- sion” or “the impression of juxtaposition” of the descriptions presented in particular by Baudhāyana. We can speculate that this feature is a trace of a moment when the “new” and polished form of rājasūya had not yet been completely developed. The redactors of these texts aimed at re-shaping the use of the ratha into a pre-determined dynamic whose process and result could be controlled. Thus, within the construction of the Brāhmaṇic ritual system, the role of the chariot changes towards a more specialized function, leaving out some central aspects of the Ṛgvedic ratha, as for example the connection with poetry or its dynamic character. The motif of the chariot is indeed reused in a new way, by which it becomes a symbolic instrument connected to a ficti- tious ride. This process can be observed not only in the royal rituals, but also in the other ones I examined earlier in this section. If we consider some features of the rites in which the chariot is used, we might find other signs pointing to this sort of reuse. In the texts dealing with the chariot as a dakṣinā, it is men- tioned that the chariot race can be avoided (LātŚS 2.8.16; DrāhŚS 5.4.16) by reciting the proper mantras. This establishes a sort of equation between the chariot and the liturgic formula. Going back to the Agnyādheya, the chariot is disassembled and separated from the horse and thus deprived of its concrete function. Furthermore, in the Vājapeya the formal race is “concurring” with the brahmán singing the sāman;35 in BauŚS 11.8: 77.4 it is said gāyati brahmā vājināṃ sāma taṃ ya eva kaś ca parikarmy āveṣṭayati,36 commented upon by the gloss bhramayati cakram; similarly, in the ĀpŚS,37 the brahmán is on a wheel and symboli- cally performs a race: 33 As to the relative chronology and especially the position of the Baudhāyana Śrauta Sūtra, see Gonda (1975: 482) and Jamison and Witzel (1992: 19–20). 34 This hypothesis does not need a strong defense. However one of the scholars who has strongly brought forward the idea of a competitive and antagonistic society in early Vedic times is Heesterman, see, e.g., Heesterman 1993. 35 JB 2.193 = 3.113 rathantara. 36 Caland takes āveṣṭayati to be a corruption of āceṣṭayati “to set in motion”; cf. the relevant TB passage. 37 ĀpŚS 18.4.3–11. 186 Elena Mucciarelli audumbaraṃ rathacakraṃ brahmārohati || tam āha vājināṃ sāma gā- yeti | tasya cakraṃ triḥ pradakṣiṇam āvartayati | vartamāne brahmā gāyati (ĀpŚS 18.4.8–11) The brahmán mounts the wheel of a chariot made of udumbara wood; he says “sing this sāman of the vāja’s winners”; he makes its wheel perform a pradakṣiṇa three times; the brahmán sings while the wheel turns.38 All these attestations present the motif of the chariot as a symbolic token that may even function as a unit in the substitution mechanism of the Brāhmaṇic ritual system. To conclude, I shall take into account a third royal ritual in which a cha- riot is actually involved in a raid, namely the Gavām Ayana39 ritual, the an- nual sattra, which according to Baudhāyana and Āpastamba also includes a chariot race and a cattle raid blending into each other during the Mahāvrata day.40 uttareṇāgnīdhraṃ kaṭasaṃghāte tejanasaṃghāte vārdraṃ carma vya- dhanārthaṃ vitatyocchrayanti | agreṇāhavanīyaṃ ratheṣu kavacinaḥ saṃnahyante (ĀpŚS 21.18.5–6) “To the north of the Āgnīdhra, a wet piece of leather is raised as a tar- get on a frame made either of wicker or of reeds; in front of the Āhavanīya the armed ones dress themselves.” After preparing for the shooting, the adhvaryu lets the lute “sing” and a copulation between a courtesan (puṃścalī) and a man from Māgadha takes place: saṃvarttete puṃścalī māgadhaś ca | ājiṃ dhāvanti || ĀpŚS 21.19.6–7. Directly thereafter the chariot race starts, and then the adhvaryu instructs each of the kṣatriyas: “Do not throw away” “Do not throw asunder”: teṣām ekaikaṃ saṃśāsti māparātsīr māti vyātsīr iti || ĀpŚS 21.19.14.41 In this ritual, another element plays an important role in the structure of the performance: the loud soundings of voices and instruments, as in the 38 With regard to the role of the priest taking over that of the yajamāna, it is noteworthy that in BauŚS 11.8: 78.3 ff. the rewards are first presented to the runners, but imme- diately afterwards they are collected and offered to the brahmán: tāni (kṛṣṇalam) sārdhaṃ samādāya hiraṇyapātraṃ madhoḥ pūrṇam ity ekadhā brahmaṇa upaharati” BauŚS 11.9. 39 See BauŚS 16.13–23; ĀpŚS 21.15–23; KātŚS 13. 40 Described in BauŚS 16.20: 266.5–268.7; ĀpŚS 21.18.1–6; 21.19.7; 21.19.13–17; MānŚS 7.2.7.17; KātŚS 13.3.13; TS 7.5.9.2; KS 34.5: 39.15; PB 5.5.21. 41 A similar account is given in KātŚS 13.3.10–13.  The Steadiness of a Non-steady Place 187 Rājasūya’s chariot race. The KātŚS offers the following description of these sounds on the Mahāvrata day: sadaḥsraktiṣu dundubhīn vādayanti || āgnīdhramapareṇa śvabhraṃ savāladhānena carmaṇāvanahya vāladhānenāhanti || godhāvīṇākāḥ kāṇḍavīṇāścāḥ patnyo vādayanti || upagāyanti || anyāṃś ca śabdān kurvanti || 19 || (KātŚS 13.3.15–19) Some people play on the drums at the corners of the Sadas. Having covered a pit to the west of the Āgnīdhra with the skin [of a bull] and its tail, they beat it with the tail. The wives play the vīṇā made of strings and reeds, they sing, they produce different sounds.42 2.2.3 The chariot and the evocation of fertility There is an idea brought forth in connection with the music, chanting and copulation that is related to the use of the ratha during all three sacrificial procedures, namely the Rājasūya, Vājapeya and Gavām Ayana: the evocation of fertility. This aspect is also present in the Śrauta Sūtra texts dealing with marriage rituals, where we have found again the use of a chariot and a sort of race. Moreover, as Heesterman has already argued, one function of the royal rituals is the regeneration of productive forces.43 The circular motion of the wheel and that of the pradakṣiṇa drive represent the circularity of time and seasons, i.e., the year, as can be seen for instance in the Taittirīyabrāhmaṇa. 1.7.9.1–2. devarathám evā́ smai yunakti | tráyó ’śvā bhavanti | ráthaś caturtháḥ | dváu savyeṣṭhasārathī́ | ṣáṭ sáṃpadyante || ṣáḍ vā́ ṛtávaḥ | (Taittirīyabrāhmaṇa 1.7.9.1–2) He yokes the chariot of gods, there are three horses, the chariot is the fourth, and the two charioteers, one on the right, one on the left, they amount to six; there are indeed six seasons. Additionally, the vāja – the prize to be conquered (in both the Vājapeya and the Rājasūya, the horse is called “vāja winner” [vājajit]) – is a form of renew- ing force, as we have already seen in the Ṛgvedic context (see section 1.6).44 42 Commented upon by Karka as mardalabherīpaṭahādijān anyān api, thus including different drums. 43 See Heesterman 1957: 133–139. 44 On the connection between fertility and the motif of the chariot in the Ṛgvedic context, see section 1.6. 188 Elena Mucciarelli 2.3 Shrinking of meanings in middle Vedic reuse If one recalls the variety of images discussed at the beginning, one will easily notice that only a few of the motifs attached to the Ṛgvedic chariot – a con- crete object working as an “objective correlate” – are still found in the middle Vedic reuse of the motif. At this stage, the symbol of the chariot only stands for power and divine status, and is connected to fertility. The Ṛgvedic motifs are still present, but they convey different values insofar as they are inserted in the ritual structure. Most notably, fertility is no longer regarded as merely the generative energy of nature: it is a matter of power and is strictly con- nected to the human – specifically royal – role. 3 Conclusion The full-fledged range of imagery of the ratha in the Ṛgvedic corpus de- creased in the Brāhmaṇa texts. The motif of the chariot became a symbolic token of power. That is to say, while in the early Vedic period the motif of the ratha was connected with many different semantic fields, during the later period it seems to be restrained to that of power and sovereignty. Considering this development carefully, it seems that the motif of the chariot was reused in a linear way. The motif did not lose the connection with its previous func- tion completely, and yet its role became substantially different. Is it possible to argue that priestly groups made a conscious selection of the Ṛgvedic fea- tures of the chariot to reuse this important and meaningful token to convey new meanings? Although I cannot offer a conclusive answer yet, the present chapter has shown that the concrete use of the chariot was lost. More pre- cisely, the earlier actual use of the chariot was sublimated into a symbolic value, thus creating the possibility for it later to become an object of devo- tion. The connection of the motif of the chariot to kingship appears again throughout the medieval period (see the chapter by Bignami in this volume), when Vedic elements were reused to legitimize sovereignty and royal power. During this politically unstable period, kings indeed developed an array of different means to establish their power; as part of this agenda, they asserted in inscriptions that they had, for instance, performed the great royal rituals such as the Aśvamedha, the Vedic horse sacrifice mentioned above (section 2.2.2). Another aspect found again in the medieval period is a strong connec- tion between the motif of the chariot and fertility, which is also an instrument  The Steadiness of a Non-steady Place 189 for establishing or proving royal power. For instance, the rathayātras45 are carried out still today in the spring to celebrate the restoration of Kāma, whose friend Vāsanta drives about the earth to proclaim Kāma’s renewed reign over the world. Although the two types of ceremonies are different, since the Vedic royal rituals are a yajña and the medieval ones, as for exam- ple the chariot festivals, are pūjās, nonetheless the strong connection of the motif of the chariot to the construction of royal power, based on a continuous exchange with the divine, appears to be a common feature throughout differ- ent periods and different cultural frames. References Abbreviations AB Aitareyabrāhmaṇa see Aufrecht 1879 ĀpŚS Āpastambaśrautasūtra see Garbe 1983 ĀśvGS Āśvalāyanagṛhyasūtra see Sharma 1976 BauŚS Baudhāyanaśrautasūtra see Caland 1904 and Kashikar 2003 DrāhŚS Drāhyāyaṇaśrautasūtra see Reuter 1904 HirŚS Hiraṇyakeśiśrautasūtra JB Jaiminiyabrāhmaṇa see Raghu Vira and Lokesh Chandra 1954; for JB 2.334-370 see Tsuchida 1979 KātŚS Kātyāyanaśrautasūtra see Weber 1859 KS Kāṭhasaṃhitā see von Schroeder 1900 LātŚS Lāṭyāyanaśrautasūtra see Ranade 1998 MānŚS Mānavaśrautasūtra see Gelder 1985 PB Pañcaviṃśabrāhmaṇa see Vedāntavagīśa 1870–1874 ṚV Ṛgvedasaṃhitā see Müller 1849–1874, Geldner 1951 and Witzel and Gotō 2007 TB Taittirīyabrāhmaṇa see Mitra 1862 TS Taittirīyasaṃhitā see Weber 1871–1872 Vaitānaśrautasūtra see Vishvabandhu 1967 VGS Vārāhagṛhyasūtra see Raghu Vira 1932 and Rolland 1971 Primary sources Aufrecht 1879 Aufrecht, Theodor, ed. Das Aitareya Brāhmaṇa. Mit Aus- zügen aus dem Commentare von Sāyaṇācārya und anderen Beilagen. Bonn: Adolph Marcus, 1879. 45 On the rathayātra, see Meyer 1937 II: 199. 190 Elena Mucciarelli Caland 1904 Caland, Willem, ed. The Baudhāyana Śrauta Sūtra Be- longing to the Taittirīya Saṃhitā. Calcutta: Asiatic Society, 1904. Garbe 1983 Garbe, Richard, and C. G. Kashikar. The Śrauta Sūtra of Āpastamba, Belonging to the Taittirīya Saṃhitā with the Commentary of Rudradatta. 3 vols. 2nd ed. New Delhi: Munśīrāma Manoharalāla, 1983. Gelder 1985 Gelder, J. M. van and C. G. Kashikar, eds. The Mānava Śrautasūtra Belonging to the Maitrāyaṇī Saṃhitā. Delhi: Sri Satguru Publ., 1985 [1961]. Sri Garib Dass Oriental Ser. 31. Geldner 1951 Geldner, Karl Friedrich. Der Rig-Veda aus dem Sanskrit ins Deutsche übersetzt und mit einem laufenden Kom- mentar versehen von K. F. Geldner. 3 vols. Cambridge Mass.: Harvard University Press, 1951. Kashikar 2003 Kashikar, C. G., ed. and trans. Baudhāyanaśrautasūtram. Crit. ed. and transl. by C. G. Kashikar. New Delhi: Indira Gandhi National Centre for the Arts, Motilal Banarsidass, 2003. Mitra 1862 Mitra, Rājendralāla, ed. The Taittirīya Brāhmaṇa with the commentary of Sāyaṇācārya. Calcutta: Asiatic Society of Bengal, 1862 [1859]. Müller 1849–1874 Müller, Friedrich Max, ed. Rig-veda-sanhita: The Sacred Hymns of the Brahmans, together with the Commentary of Sayanacharya. London: Allen, 1849–1874. Raghu Vira 1932 Raghu Vira, ed. Vārāhagṛhyasūtram with Short Extracts from the Paddhatis of Gaṅgādhara and Vasiṣṭha. Lahore: The University of the Panjab, 1932. Ranade 1998 Ranade, H. G., ed. and trans. Lāṭyāyana-Śrauta-Sūtra. Critically Edited and Translated. New Delhi: IGNCA, Motilal Banarsidass, 1998. Reuter 1904 Reuter, J. N., ed. Drāhyāyaṇa Śrautasūtra, with the Com- mentary of Dhanvin. Part 1. London 1904. Rolland 1971 Rolland, Pierre, ed. Un rituel domestique védique. Le Vā- rāhagṛhyasūtra. Aix en Provence: Ophrys, 1971. von Schroeder 1900 Schroeder, Leopold von, ed. Kāṭhakam. Die Saṃhitā der Kaṭha-śākha herausgegeben von Dr. Leopold von Schroe- der. Leipzig: Brockhaus, 1900. Sharma 1976 Sharma, Narendra Nath, ed. Āśvalāyana gṛhyasutram, with Sanskrit Commentary of Nārāyaṇa. 1st ed. Delhi: Eastern Book Linkers, 1976. Tsuchida 1979 Tsuchida, R., ed. Das Sattra-Kapitel des Jaiminīya Brāh- maṇa (2,334-370). Nach den Handschriften herausge-  The Steadiness of a Non-steady Place 191 geben, ins Deutsche übersetzt und erklärt. Marburg: Mauersberger, 1979. Vedāntavagīśa 1870–1874 = Vedāntavagīśa Ānandacandra, ed. Tāṇḍya Mahābrāh- maṇa, with the Commentary of Sāyaṇa ācārya. Calcutta: Asiatic Society of Bengal, 1870–1874. Vishvabandhu 1967 Vishvabandhu, ed. Vaitāna-śrauta-sūtram, with the Com- mentary of Somāditya. Hoshiarpur: Vishveshvaranand Institute, 1967. Woolner Indological Ser. 13. Weber 1859 Weber, Albrecht, ed. The Śrauta Sūtra of Kātyāyana with Extracts from the Commentaries of Karka and Yājñika- deva. Berlin: Dümmler, 1859. Weber 1871–1872 — , ed. Die Taittirîya-Saṃhitâ. Leipzig: Brockhaus, 1871– 1872. Witzel and Gotō 2007 Witzel, Michael and Toshifumi Gotō. Rig-Veda: das heilige Wissen. Erster und zweiter Liederkreis. Frankfurt am Main: Verlag der Weltreligionen, 2007. Secondary Sources Brereton 2002 Brereton, Joel P. “The Race of Mudgala and Mudgalānī.” Journal of the American Oriental Society 122 (2002): 224–234. Bryant and Patton 2004 = Bryant, Edwin F. and Laurie L. Patton, eds. The Indo- Aryan Controversy: Evidence and Inference in Indian His- tory. London: Routledge, 2004. Edgerton 1919 Edgerton, Franklin. “The Metaphor of the Car in the Rig- vedic Ritual.” American Journal of Philology 40 (1919): 175–193. EWAia Mayrhofer, Manfred. Etymologisches Wörterbuch des Alt- indoarischen. Heidelberg: C. Winter, 1992–1996. Gonda 1954 Gonda, Jan. Aspects of Early Visnuism. Utrecht: Oosthoek, 1954. Gonda 1957 — “‘Attraction’ and Co-ordination in the Veda.” Bulletin of the School of Oriental and African Studies 20 (1957): 279–289. Gonda 1963 — The Vision of the Vedic poets. The Hague: Mouton, 1963. Gonda 1975 — Vedic Literature (Saṃhitās and Brāhmaṇas). Wies- baden: Harrassowitz, 1975. A history of Indian literature 1.1. Grassmann 1875 Grassmann, Hermann. Wörterbuch zum Rig-veda. 6., über- arbeitete und ergänzte Auflage von Maria Kozianka. Wiesbaden: Harrassowitz, 1996. 192 Elena Mucciarelli Hauer 1927 Hauer, Jakob Wilhelm. Der Vrātya: Bd. 1. Die Vrātya als nichtbrahmanische Kultgenossenschaften arischer Her- kunft. Stuttgart: Kohlhammer, 1927. Heesterman 1957 Heesterman, J. C. The Ancient Indian Royal Consecration: The Rājasūya Described According to the Yajus Texts and Annotated. ’s-Gravenhage: Mouton, 1957. Disputationes Rheno-Trajectinae. Heesterman 1993 — The Broken World of Sacrifice: An Essay in Ancient Indian Ritual. Chicago: University of Chicago Press. Jamison 1996 Jamison, Stephanie W. Sacrificed Wife Sacrificer’s Wife: Women, Ritual, and Hospitality in Ancient India. New York: Oxford University Press, 1996. Jamison 2011 — “The Secret Lives of Texts.” Journal of the American Oriental Society 131 (2011): 1–7. Jamison and Brereton 2014 = Jamison, Stephanie W., and Joel Brereton. The Rigveda: The Earliest Religious Poetry of India. New York: Oxford University Press, 2014. Jamison and Witzel 1992 Jamison, Stephanie W. and Michael Witzel. “Vedic Hin- duism.” Avaliable at: <http://www.people.fas.harvard.edu/ ~witzel/mwpage.htm> accessed 18 March 2007. Klaus 1989 Klaus, Konrad. “Samudrá im Veda.” XXIII. Deutscher Orientalistentag vom 16. bis 20. September 1985 in Würz- burg: Ausgewählte Vorträge. Ed. Einar von Schuler. Stutt- gart: Steiner, 1989. 364–371. Klaus 1990 — Die Wasserfahrzeug im vedischen Indien. Mainz: Aka- demie der Wissenschaften und der Literatur; Steiner, 1989. Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse 1989, 13. Kuiper 1960 Kuiper, F. B. J. “The Ancient Aryan Verbal Contest.” Indo-Iranian Journal 4 (1960): 217–281. Kulke 1992 Kulke, Hermann. “The Rājasūya. A Paradigm of Early State Formation.” Ritual, State and History in South Asia. Essays in honour of J. C. Heesterman. Ed. A. W. van den Hoek, D. H. A. Kolff, and M. S. Oort. Leiden: Brill, 1992. 188–198. Kulke 1995 —, ed. The State in India, 1000–1700. Delhi: Oxford Uni- versity Press. Kulke and Rothermund 2010 = Kulke, Hermann and Dietmar Rothermund. Geschich- te Indiens von der Induskultur bis heute. München: Beck, 2010. Littauer and Crouwel 1979 = Littauer, M. A. and J. H. Crouwel. Wheeled Vehicles and Ridden Animals in the Ancient Near East. Leiden: Brill, 1979. Handbuch der Orientalistik: 7. Abt., Kunst und  The Steadiness of a Non-steady Place 193 Archäologie: 1. Bd. Der alte vordere Orient: 2. Abschnitt, die Denkmäler, B-Vorderasien Lfg. 1. Meyer 1937 Meyer, Johann Jakob. Trilogie altindischer Mächte und Feste der Vegetation. Ein Beitrag zur vergleichenden Reli- gions- und Kulturgeschichte, Fest- und Volkskunde. 3 vols. Zürich: Max Niehans, 1937. Proferes 2003 Proferes, Theodore Nicholas. “Poetics and Pragmatics in the Vedic Liturgy for the Installation of the Sacrificial Post.” Journal of the American Oriental Society 123.2 (2003): 317–350. Proferes 2007 — Vedic Ideals of Sovereignty and the Poetics of Power. New Haven, Conn.: American Oriental Society, 2007. Renou 1955 Renou, Louis. “Les pouvoirs de la parole dans le Ṛgveda.” Études védiques et pāṇinéennes 1. Paris: E. de Boccard, 1955. 1–27. Schmithausen 1991 Schmithausen, Lambert. The Problem of the Sentience of Plants in Earliest Buddhism. Tokyo: The International Institute for Buddhist Studies, 1991. von Schroeder 1916 Schroeder, Leopold von. Arische Religion. Vol. 2B. Leip- zig: Haessel, 1916. Shulman 1985 Shulman, David Dean. The King and the Clown in South Indian Myth and Poetry. Princeton, N. J.: Princeton Uni- versity Press, 1985. Sparreboom 1985 Sparreboom, Marcus. Chariots in the Veda. Leiden: Brill, 1985. Thieme 1967 Thieme, Paul. “Vedisch manīṣā́ .” Beiträge zur Indogerma- nistik und Keltologie: Julius Pokorny zum 80. Geburtstag gewidmet. Ed. Wolfgang Meid. Innsbruck: Sprachwissen- schaftliches Institut der Universität Innsbruck, 1967. 99– 106. Windisch 1893 Windisch, Ernst. “Eine vedische Wettfahrt? RV II.31.” Festgruß an Rudolph von Roth zum Doktor-Jubiläum. Ed. Ernst Kuhn. Stuttgart: Kohlhammer, 1893. 139–144. Witzel 1999 Witzel, Michael. “Substrate Languages in Old Indo-Aryan (Ṛgvedic, Middle and Late Vedic).” Electronic Journal of Vedic Studies 5.1 (1999) 1–67.  Chariot Festivals: The Reuse of the Chariot as Space in Movement* Cristina Bignami Since Vedic times, the chariot (ratha) has been an element of various reli- gious ceremonies in different cultures of South Asia. Its use in ritual contexts, accordingly, has a long tradition, and even today, the chariot is the main feature in large chariot festivals (rathayātrā or rathotsava). The present chapter focuses on the reuse of the symbol of the Vedic chariot as “space in movement” in ancient, medieval and contemporary contexts. Although the chariot seems to have preserved some general ritual functions over the centu- ries, its concrete use has undergone many modifications. These changes concern not only the outer form of the chariots (see section 8), but, more interestingly, the interpretation of the chariot as a religious symbol. In the course of history, the chariot preserved its role as a “space in movement,” but it acquired new meanings that turned it into an instrument for legitimizing sovereignty and generating social integration. 1 Introduction A South Asian chariot festival is generally a part of large temple celebrations, in the course of which processional idols are carried through the streets and deities are hosted in highly decorated chariots that resemble huge temples on wheels (Fig. 1). These celebrations are usually called rathayātrā or rathot- sava. The term rathayātrā (chariot journey) refers to the journey of the cha- riot that leads the gods from one resident temple to another. A prominent ∗ This chapter was partly composed in the context of the International Project “Traces of an Heterodox Concept of Kingship in Ancient, Medieval, and Modern India” led by Tiziana Pontillo at the University of Cagliari and partly within the framework of the DFG-financed project “Kings of the Wild: Re-use of Vedic and local elements in the legitimation process of Medieval Karnataka,” which I lead with my colleague Elena Mucciarelli. This project has been generously supported by the Institutional Strategy of the University of Tübingen (Deutsche Forschungsgemeinschaft, ZUK 63). 196 Cristina Bignami example for such a journey takes place at the festival held in Puri, Orissa (Kulke and Schnepel 2001: 66–81). The term rathotsava is related to the term brahmotsava, which designates an annual religious festival. This festival starts with smaller chariots and then continues with bigger and bigger ones, such as the vāhana-ratha or “the chariot of the [divine] vehicles,” (Śeṣa, Haṃsa, Garuḍa, Hanumān), until through a continuous progression in size the climax is reached with the largest chariot, a duplicate of the temple.1 These vāhanarathas are wheeled vehicles in the form of the animals that, according to mythological accounts, serve as the vehicles (vāhana) of gods. Before the procession takes to the road, statues of the divinities are installed in the middle of these peculiar chariots. The principal sacred object of these festivals is evidently the chariot (Fig. 2). While its shape can take different forms, the elements used to construct it are always identical. The upper part, which is temporarily reassembled every year some days prior to the festival (Fig. 3–6), represents the dome (śikhara) of the temple and is thus similarly shaped. It consists of wooden and bamboo sticks together with simple ropes that are covered with a shroud and crowned with a pinnacle or round top-piece called kumbha in Tamil Nadu (Kulkarni 1994: 44) or kalasa in Orissa (Kulke and Schnepel 2001: 72). The internal structure of the chariot is hidden behind carved panels that are fixed to the body as if they were a curtain wall. The sculptural decorations are based on the same module and therefore they correspond to the proportions of the elements in the chariot’s structure. The upper deck is covered horizontally by a wooden panel. It accommodates the sacrificial altar (vedi) of the idol.2 The four corners of each chariot are bedecked with wooden images that mostly have a decorative character and depict fabulous animals, floral ornaments, or mythological-symbolic scenes. The reliefs on the four sides, however, mainly depict gods and goddess as individual images or in connection with groups (Fig. 7). The iconography of these group images, which varies across regions and from one chariot to another, seems to be based on ancient local traditions that were later superimposed on each other in different ways (Kulkarni 1994: 44). The lower and permanent part of the chariot is made of wood. It represents the jaṅghā, the walls of the temple. Finally, the wheels as the pi- votal element of the chariot provide the chariot with its most important cha- racteristic: motion. 1 L’Hernaut 1984: 265–269, Ramesan 1981. 2 Sometimes the gods are placed in swings tied with cloth, as for example at the brah- motsava festival at Belur, Karnataka.  Chariot Festivals 197 The approach of this study is based on a chronology of the ritual use of chariots. I shall start with a description of the role of chariots in Vedic times. The Vedic chariot will serve as the standard of comparison for the ritual use of chariots in subsequent contexts, starting from a Buddhist record, moving to a medieval account and then to modern festivals in South Asia and in the diaspora. 2 The origins of chariot processions in the Vedic period3 The Vedic literature reflects (at least to some extent) the life of part of the Vedic society in a period that can roughly be dated to the middle or late second millennium BCE. The Vedic tribes, which practiced a semi-nomadic life-style, had a mixed pastoral and agricultural economy, in which cattle played a major role. In fact, these tribes used carts and chariots for various purposes, probably ranging from warfare to travel and from hunting to ritual.4 Already at this early time, the chariot had also become a symbol along with its concrete role, as described by Sparreboom: “From the texts it is clear that the chariot was not merely a practical instrument for conveying persons, but an object vested with religious significance and symbolic values” (Sparre- boom 1985: 1). Indeed, the Ṛgvedic corpus reveals that the chariot was used for multiple purposes and played an important role in the poetic imagery of the ṛṣis, being connected with various semantic aspects (see in this volume Mucciarelli, section 1 for further references). In the course of time, the range of use was, however, narrowed more and more, until in the later Vedic period, the chariot played a role as an attribute of power within the large royal rituals. As is well known, at the end of the first millennium BCE, the formation of the great clan of the Kuru-Pañcāla and the consequent ritual reformation took place. As Sparreboom (1985: 6) aptly highlights, the cultural complex of the chariot as an operative object was already in this period an artifact of the past. According to Elena Mucciarelli (see section 3), at this time the chariot had lost its functional use and had thus become a royal symbol for the litur- gical culture of the middle and late Vedic society. In this context, the reuse of 3 For an outline of the development of chariots during the Vedic period, see Mucciarelli’s contribution to the present volume. 4 On the Vedic culture, see, among others, Erdosy 1995 and especially Witzel 2003, which contains a broad bibliography on the topic. In this volume, one may also refer to the chapter by Mucciarelli. 198 Cristina Bignami the chariot can be seen as a device to keep the original tradition alive, so that the chariot shifted from a means of transportation into a sacred object used in ritual contexts. Already in the earliest Ṛgvedic attestations the chariot is equated with the gods, an equation that is the origin of its sacred meaning. As has been described by Mucciarelli: “The godly character of the chariot is affirmed both in the sense of it having a divine status in itself, as well as by the use of the word ratha in apposition to, or as an appellative for, the gods, especially for Agni.”5 This equation furnished a basis for the devotional development in the South Asian medieval period. 3 Faxian’s record of chariot festivals In contrast to what might be expected, the earliest surviving evidence for a chariot festival comes neither from South Asia nor from a Brahmanical back- ground, but from the kingdom of Khotan. There, the Chinese Buddhist pil- grim Faxian witnessed and recorded a Buddhist chariot festival around the year 400 CE.6 The very fact that Faxian described a second Buddhist festival that he witnessed in the ancient city of Pataliputra (the modern city of Patna in Bihar) suggests that chariot processions were widespread ritual practices throughout South Asia (and beyond) at the time of Faxian. In Khotan, Faxian saw the great Mahāyāna Buddhist festival that he de- scribed as follows: In this country [Khotan] there are fourteen great sangharamas, not counting the little ones. From the first day of the fourth month they sweep and water the thoroughfares within the city and decorate the streets. Above the city gate they stretch the great awning and use every kind of adorning. This is where the king and the queen and court ladies take their place. The Gomati priests, as they belong to the Great Vehicle, which is principally honoured by the king, first of all take their images in procession. About three or four li from the city they make a four-wheeled image-car about thirty feet high, in appearance like a moving palace, adorned with the seven precious substances. They fix upon it streamers of silk and canopy curtains. The figure is placed in the car with two Bodhisattvas, as companions, whilst the Devas attend of them; all kinds of polished ornaments made of gold 5 See Mucciarelli’s chapter in the present volume, p. 173. 6 See Beal 1884: XI–XV.  Chariot Festivals 199 and silver hang suspended on the air. When the image was a hundred paces from the gate, the king takes off his royal cap, and changing his cloths from new ones, proceeds barefoot with flowers and incense in his hand, from the city, followed by his attendants. On meeting the image he bows down his head and worships at his feet scattering the flowers and burning his incense. On entering the city, the queen and court ladies from above the gate-tower scatter about all kinds of flow- ers and throw them down in wild profusion. So splendid are the ar- rangements of worship. The cars are all different and each sangharam had the day for its image procession. They begin on the first day of the fourth month and go on to the fourteenth day, when the procession ends. The procession ending, the king and the queen then return to the palace (Beal 1884: XXVI–XXVII). The main item of the procession described above is a four-wheeled chariot that is approximately nine meters tall. This was pulled by the devotees (probably in the same way as done today) in a fourteen-day long procession from the Buddhist monasteries on the outskirts of the city into its center. The chariot was richly adorned with images of the Buddha, Bodhisattvas and De- vas made of gold and silver, and it is said to have resembled a movable “sa- cred shrine of the Buddha.” Every monastery of the town had its own unique chariot. The king together with his queen and court participated as spectators of the ceremony. They worshipped the images by placing offerings in front of the city gates, and the king finally joined the procession as it progressed through the town. The procession at Pataliputra was similar. Faxien recorded the following: Every year on the eighth day of the second month there is a procession of the images. On this occasion they construct a four-wheeled car, and erect upon it a tower of five stages composed of bamboos lashed to- gether, the whole being supported by a centre-post resembling a large spear with three points in height twenty-two feet and more. So it looks like a pagoda. They then cover it with fine white linen, which they af- terward paint with gaudy colours. Having made figures of the dēvas and decorated them with gold, silver and glass, they place them under canopies of embroidered silk. Then at the four corners (of the car) they construct niches (shrines) in which they place figures of Buddha in a sitting posture with a Bodhisattva standing in attendance. There are perhaps twenty cars thus prepared and differently decorated. During the day of the procession both priests and laymen assemble in great numbers. There are games and music, whilst they offer flowers and 200 Cristina Bignami incense. […] Such is the custom of all those who assemble on this oc- casion from the different countries round about (Beal 1884: LVI– LVII). Faxian saw a Buddhist procession of about twenty four-wheeled and five- storied chariots. Each of these vehicles was over six meters in height and had the shape of a temple. Here again, the chariots carried silver and golden im- ages of the Buddha and other gods. In his descriptions, Faxian remarked that the cars looked like pagodas. If the cars resembled Buddhist temples, this remark possibly indicates that in the 5th century CE, the Buddhist temples of India were multistoried wooden buildings similar to the coeval sacred build- ings of China. 4 A record of the chariot festival in the southern kingdom According to Kulke (1993), Schnepel (2002) and Stein B. (1984), territorial segmentation and political development of local chiefs are central characte- ristics of the political situation leading to the development of regional king- doms in India during the period from the 11th to the 15th century. In this pe- riod, the political situation is marked by the clear emergence and disappear- ance of many political centers in the course of more or less constant internal power struggles. This political instability made concepts such as “alliance” and “power legitimation” relevant for political discourses. One of the main issues during this period was the relationship of the Hindu rājās – themselves often descendants of tribal chiefs – to the tribes living in the surrounding, isolated forest areas of the kingdom, because the societies of the regional kingdoms depended on the support of these tribes for the security of their borders and internal communication. The partial integra- tion of local or tribal cultures on the part of the Hindu rājās was achieved by enrolling parts of the male population into the kingdom’s militia, referred to as “kṣatriyaization” by Kulke (1993: 5) and, to a certain extent, by religious means. Scholars agree that the earliest evidence for the existence of the chariot festival at Puri occurs in the late Ganga period of Somavaṃśa rule in the 10th–11th century. The oldest iconographical evidence of the festival and its temple chariots (ratha) comes from the later Ganga period of the 13th–14th century (Kulke 1993: 69). In his historical study of the Orissa chariot festival, Kulke emphasizes the special importance of the festival as a device of royal legitimation (Kulke  Chariot Festivals 201 1993: 66–81). Kulke analyzed the participation of the king in this ritual cele- bration from a political point of view. He suggested that the participation of kings in this grand celebration may have been confined to casual visits at first, but later, in the 15th century, when the usurper kings crowned them- selves in Orissa, the rulers were in need for a special legitimation of their rule. During the Vijayanagara Empire, chariot processions became the most important celebration of kingship.7 The information of the epigraphic sources suggests8 that in the chariot festivals the king played an important role in the procession himself and became associated with the gods. The relevance of the festival for the rulers of Vijayanagara becomes evident also from the fact that the kings offered land in order to sustain the chariot festival economi- cally. Moreover, in some case even the road system was extended prior to a festival taking place (Rice 1879: 225), most likely to provide the celebration with more urban space. Moreover, the public display of the deities during the festival apparently lead to an egalitarian sentiment in the population of the kingdom. According to Nandakumar: [t]he car festivals of gods and goddesses are popular in South India probably because one of the important methods of disseminating reli- gious and spiritual fervour in Hinduism has been the institution of the public appearance of the deities. Though one’s approach to the gods in the temples is subject to some restrictions, perfect equality is assured to the citizens when enclosing rulers are set aside and the deity emerges out of the temple in various mounts. The car festival, in par- ticular, is the community festival in which almost everyone takes part. Here cast-born differences are forgotten, class division holds no meaning. Everyone feels close to the god/goddess and all participants can touch the ropes and pull the chariot (Nandakumar 2003: 432). What emerges as particular to this celebration is the participation of rural as well as urban communities of the kingdom. It appears that the transformation and increased popularity of this festival may have been connected to political agendas of important kings. One of the reasons for such a popularization could have been the chance provided by the celebration for the community to move away from their villages or cities to the pilgrimage center and gather 7 For the rathotsava of Vijayanagara, see Brückner 2014: especially 109–112. 8 Rice 1879: 224–225; EC III: Sr 91; EC III: Md 71–2; EC V: Bl 4–5; EC V: Hn 2; EC VI: Cm 48, 153; EC IX: Ma 1; EC X: Kl 34; EC XI: Dg 30, 83; EC XVII: Mr 147. 202 Cristina Bignami once a year. In this context, the king could then prove, as an act of legitima- tion, his power in front of all the different peoples that composed his kingdom: The royal participation in the car festival helped to proclaim the new idea of Hindu kingship to the rural and tribal population. [...] At this point of development the significance of the Hindu temple in the legi- timation of political power began to change. It had been a royal status symbol of external legitimacy with little implication for the relation- ship between the rājā and the rural population. Now the participation of the rājā and his prajā (people) in the car festivals increasingly in- fluenced internal or vertical legitimation, too (Kulke 1993: 106). With regard to the role and the meaning of the chariot, Kulke mentions that it is seen as a duplication of the temple and that its movement serves a ritual function. By means of the motion of the chariot, the participants witness the extension of the sacred sphere of the temple into a major part of the town.9 In addition to the two pivotal concepts of “motion” and “sacred object” that are linked to the chariot, we may also add that of fertility. In fact, in the course of the celebration of modern chariot festivals, an important event is the wedding of the gods that takes place twice a day after the rite of aṅkurārpaṇa (“the rite of seeds and their germination”) by means of a circumambulation (pradakṣi- ṇa) of the temple. 5 The modern ritual of rathotsava at the Cennakeśava Temple of Belur, Karnataka The rathotsava at the Cennakeśava Temple of Belur is part of one of the largest temple festivals (brahmotsava) in South Asia.10 It takes place in the month of Caitra (March‒April) and lasts for fourteen days. During this time the festival is structured in the following seven distinct phases of different length. 9 “During the ratha yātrā […] the ‘Lord of the Universe’ [Jagannath] leaves his […] ‘lion throne’ in order to appear to his devotees, the ratha thus transforms the separate temple building of the ‘divine palace’ into an ideal type of Hindu temple, drawn by devotees from all social strata and pilgrims from all quarters of the Hindu world. The rathas are an example of ‘mobile architecture’ […] [they] extend the ritual and sacred sphere of the temple into major parts of the town” (Kulke 1993: 71). 10 See Kersenboom 1987: 133–136, L’Hernault 1984: 265–269.  Chariot Festivals 203 1) Preliminary rites that are meant to ward off of evil influences and to pro- pitiate the cardinal directions by means of cooked rice offerings that are presented at major and intermediary sites around the temple. 2) The rite of aṅkurārpaṇa, “the rite of seeds and their germination,” (Kram- rish 1976: 15): various grains are sowed from vessels in order to secure fertility, prosperity and abundance. 3) Dhvajarohana: the hoisting of the Garuḍa flag at the dhvajastambha, the Garuḍa pillar in front of the temple, which signifies the commencement of the brahmotsava proper. 4) The wedding of the gods. 5) Several days of procession in the mornings and evenings in which the gods are carried on the vāhanaratha. 6) Rathotsava: the temple chariot race. 7) Mahābhiṣeka: the final bath and the lowering of the flag, which symbo- lizes the conclusion of the brahmotsava. During the two days of the chariot race (point no. 6, above), the chariot is worshipped as directly connected with a deity, just like a temple. In the course of field studies during the rathotsava at Belur, my colleague Muccia- relli and I noted two different places of worship and religious rituals prior to the chariot race: (1.) the temple of Cennakeśava, and (2.) the big tank outside of the temple. In the days before the chariot race around the temple area, people from villages in the countryside come to the city and settle near a tank on one side of Belur. This space is an area separated from that of the temple. There, the visitors spend a few days adoring a large tree that is a symbol of fertility and purification, offering animal sacrifices as part of the ceremonies. These prac- tices would not be accepted in an orthodox ritual and thus their performance close to the chariot’s depository clearly makes evident the differences among the various communities and their customs, and how the presence of the chariot allows also non-orthodox practices. Although different social groups live in the sacred space of the temple, these “different worlds” mix and mingle. The social relationships are charac- terized by what, according to Turner (1973: 191–230), is an “anti-structure.” The creation of a sacred space leads to the disappearance of social distinc- tions and the direct, immediate and total confrontation of human identities. Further, during the chariot’s procession, the creation of a sacred space is shaped by the movement of the chariot and the participants of the ritual, and this movement emerges as the pivotal element of the procession. 204 Cristina Bignami In my view, this movement is the key element of the ritual, because the movement of the chariot creates a sacred space. The other key element of the rathotsava celebration is the chariot itself. The fact that hundreds of people from several different communities draw the chariot along by its long ropes shows the involvement of the whole community in the creation of the sacred space. 6 The modern ritual of rathayātrā at Puri, Orissa The most popular rathayātrā festival is celebrated at Puri, Orissa, on the second day of the month of Ashadh (June‒July). It involves the participation of thousands of local and international devotees of Śrī Jagannath, who lead the procession through the main streets of the town. Lord Jagannath is taken out of the temple on an enormous chariot, while two other chariots host the god’s brother and sister, Balabhadra and Subhadrā. The festival commemo- rates Kṛṣṇa’s journey from Gokul to Mathura in response to the invitation of Kansa.11 The chariot procession first proceeds on a broad avenue until it reaches Gundicha Mandir, the Lord’s summer house and garden, where the gods spend seven days before they are brought back to the temple. Upon completion of the ceremony, the chariots are dismantled and religious relics are thus produced for the devotees. For this reason, the chariots must be re- constructed every year. The days before the processions of the chariots, the images of Jagannath, Balabhadra and Subhadrā12 are taken out of the sanctuary and placed on a platform near the outer wall where they can be seen from the street below. As described already in 1908 by O’Malley, “here they are bathed with 108 pitch- ers containing water taken from a well near northern gateway which is used only once in the year. In consequence, the paint is so much damaged that they have to be removed to a side room in order to be renovated for the Car Fes- tival, when they next make their public appearance” (O’Malley 1908:104). The images of Jagannath, Balabhadra and Subhadrā are quite atypical, since they are wooden masks instead of the usual metal mūrtis carried in other chariot festivals. This peculiarity probably signals the local origin of the cult of Jagannath, which accordingly may have originally been a local deity adaptively reused in a more Sanskritized ritual framework. 11 See O’Malley 1908:106. 12 For the origin of Jagannath’s wooden idol at Puri, see the Puruṣottama Māhātmya of the Skanda Purāṇa that contains the Indradyumna legend.  Chariot Festivals 205 The origin of the visual representation of Jagannath, Balabhadra and Su- bhadrā has been the object of a long debate. The Orissa Project, which was directed by Heinrich von Stietencron,13 suggested a tribal origin of the icons; against this, Starza argued that the three wooden figures may have had their origin in “orthodox” Hindu traditions and that the “conceptual and icono- graphic basis for the Puri triad is found in the motif of Brahma, Visnu and Mahesvara emerging from Para Siva, and in the symbol of the ekamukha- linga (Sadasiva/Purusottama)” (Starza 1993: 133). This point of view takes into consideration the rite of the renewal of the Jagannath icons as derived from orthodox funerary rites in memory of a deceased brahmana (Starza 1993: 90). Starza argued that “the great Car Festival held for seven days on the bank of the Indradyumma Lake in a mandapa to which the Puri icons are brought would originally have commemorated the death ceremony performed by the Eastern Gangas in memory of Codaganga, the founder of the Jagan- natha temple” (Starza 1993: 79). According to Kulke (Kulke and Schnepel 2001: 66–81), the foremost cha- riot festival of India, the Puri festival, has two unique features. Firstly, in contrast to the practice in South India, the chariots in Puri are reconstructed each year, because they are demolished after the rathayātrā. Only their up- permost portion (kalaśa), the small painted wood carvings and the wooden horses that are attached to each chariot, are retained to be used again. Se- condly, the Jagannath cult of Puri is more directly associated with kingship than most of India’s great pilgrimage sites. Both peculiarities have had direct economic and political implications. The resurge of the Jagannath cult in the 19th century and the consequent development of the importance of the annual festival of the rathayātrā at Puri represent a pivotal example of a reuse process: this celebration became the symbol of Oriya nationalism during the independence struggle (for other political reuses of chariot festivals, see also the next section). 7 Applying the concept of reuse: The chariot in the diaspora The awareness of the fact that the chariot was reused over centuries allows scholars to interpret the use of the chariot in religious festivals in the South Asian diaspora in a new light, namely as an instance of adaptive reuse in which an old item, the chariot used in religious processions, acquires a new 13 See Eschmann, Kulke, and Tripathi 1978 and Kulke and Schnepel 2001. 206 Cristina Bignami meaning, i.e., that of a symbol of a marginal community aimed at gaining visibility. As Jacobsen has observed: In the diaspora, much work and energy are spent in preserving and in- stitutionalizing cultural heritage. This is true in particular of the reli- gious traditions of the diaspora groups and these traditions often attain new functions in the diaspora. An important function of religion in the diaspora is to secure emotional attachment to the culture of origin. Religions often function as preservers of traditions inherited from the past notably because their rituals are repeated and their norms consi- dered eternal or transcendent. In the diaspora, processions often be- come ritual events of celebration of the religious tradition of the country of origin that confirm identity and transfer it to the next gen- eration (Jacobsen 2008: 200). In the context of the contemporary diaspora, the chariot festival is the largest annual celebration of the Indian community outside its homeland. Chariot processions or festivals are an important form of devotion in the regions where Tamil people have settled, sometimes already in the colonial era. Today, these are also held at numerous sites in the West.14 During these fes- tivals, the deities are removed from their usual abode within the temples and paraded on adorned palanquins through public streets and past crowds of adoring devotees, “dissolving the boundaries that normally prevail between divinity and humanity” (Nabokov 2000: 8–9). This allows for extended access to the blessings of darśana, the seeing and being seen by the deities. A growing number of Tamil Hindu processions that include such chariot pro- cessions are today held in London. In his analysis of such festivals, David has pointed to the public performance of these “embodied customs” as evidence for an increased confidence and assurance in articulating a “specific Tamil identity” in the migration setting (David 2009: 218). Luchesi also described the increased conspicuousness of ritual practices of Tamil Hindus in urban Germany as a process of “leaving invisibility” and “claiming their own place in German religious plurality” (Luchesi 2008: 180). A study undertaken by Vineeta Sinha (Sinha 2011) has also focused on the chariot festival as an aspect of Hindu cultural heritage in the diaspora. The historical and empirical project of Sinha is grounded in her desire to theorize religion‒state relations in the multi-ethnic, multi-religious and secu- 14 See Jacobsen (2008) on these festivals in Norway, Fuller Collins (1997) and Willford (2007) in Malaysia, Sinha (2011) in Singapore, and Trouillet (2008) in Tamil Nadu.  Chariot Festivals 207 lar city of Singapore. Her main research questions have emerged from the confluence of the domains “religion, law, and bureaucracy” and “religion and colonial encounters.” The study focuses on Hindu temple management, the observance of Hindu festivals, and processions as enacted within these ad- ministrative and bureaucratic contexts. By reconstructing the 19th-century Hindu landscape, Sinha shows that the construction of Hindu temples in Sin- gapore can be dated back to the 1820s. It is very likely that the earliest cha- riot processions amongst Hindus in Singapore were initiated by groups of devotees associated with Hindu temples and their desire to mark religious festivals. In these early times, images of deities were placed on wooden cha- riots and devotees pulled these around within the temple grounds or out in the streets. However, such chariot processions did not occur only in relation to Hindu festivals. On the occasion of the commemoration of fifty years of Queen Victoria’s rule over Singapore, the Hindu community decided to cele- brate this event with a grand ceremony at Mariamman Temple, where chariot processions were held by three Hindu temples (Sinha 2011: 63). With this background, we may assume that the reuse of the chariot pro- cession as an annual parade created by the movement of people became one of the most important means for gaining public recognition. Since size is an important factor for minority groups seeking public recognition, such ga- therings of large groups on public streets became a sign of strength and pride. As stated by Jacobsen: Processions have become important occasions for large groups of sev- eral different […] communities to gather together for a religious pur- pose. These processions bring people together in order to put religion and religious identity on display. People moving with sacred objects [or images] on public streets make their religious traditions, their identities and their concerns visible to each other and to an audience. Processions also make visible the size of the groups and in this way confirm collective identities and ambitions for influence and power. […] In the diaspora, processions often become ritual events of cele- bration of the religious tradition of the country of origin that confirm identity and transfer it to the next generation (Jacobsen 2008: 191, 200). Conventional methods for transmitting religious ideas depend on a number of social factors. When such supportive structures decline in modern society and especially in the context of migration and resettlement in places far away from the original homeland, either new strategies need to be invented or old ideas and values must be allowed to decline. It is precisely in this cultural 208 Cristina Bignami context that strategies of reutilization are applied as a device to manage the survival of the original tradition. As Hegewald and Mitra have emphasized: History has shown, however, that re-use as a compromise in which not all is lost but enough is retained to establish a connection with the past, can lead to hybridity, to assimilation and finally to integration (Hegewald and Mitra 2012: 4). What emerges in the adaptive reuse of the rathayātrā is the chariot and its movement in the parade: the chariot stands for the sacred building in motion and the routes, which often lead through the city streets, act as a large social container, simultaneously making the different strata of the social community publicly visible. In the modern rathayātrā in the diaspora context, the adaptive reuse im- plies some important modifications of the celebration. Most evident surely is the comparatively short duration of the festival, usually only a few days, which is certainly determined by the urban lifestyle and the need to organize the use of city streets. The route of the parade also undergoes adaptation. Whereas in the Indian context, the journey of the chariot starts in one temple with the pradakṣiṇa around the sacred building, in Western countries this movement is often impossible, because Hindu temples usually do not have their original architectural forms and the sacred ritual space in the diaspora is often accommodated as part of modern structures and not within proper temples. The pivotal elements that remain unchanged in this adaptive reuse are the chariot and the parade itself, but also the consequent encounter of different social entities. During this exceptional time, this mix of different people creates a “liminal space” in which the prevalent social order is inter- rupted (Younger 2002: 4). The movement of the chariot, the great flow of people following it and moving to the rhythm of drums, as well as the large number of devotees in front of the vehicle pulling the ropes, incited by the leader’s voice setting the pace ‒ all of these are the representation of a larger emotional container within which these separate elements are mixed to create a coherent unit. The reuse of the chariot festival in the diaspora is an example of simple re-use (as defined in the introduction to this volume), but at the same time it is also one of the tools used by the community to maintain the cultural traits of its country of origin. Thus, some elements of adaptiveness can be detected in the re-semantization of chariot processions in the diaspora, which are re- interpreted as moments of cultural identity rather than as mere religious fes- tivals.  Chariot Festivals 209 8 Conclusions The chariot as a sacred structure that hosts and even embodies divinities has ancient roots.15 Its gradual transformation is connected, during the process of reuse, with its original sacred meaning. In fact, its special characteristic, namely movement, has enabled it to adapt to the emergent needs through history of displaying the divinities in different religious rituals. Motion plays a pivotal role in the adaptive reuse, because this feature allows the celebration to take the form of a parade in which social and political elements are in- volved. The movement of gods on the chariots transforms the route through the streets into a sacred ground and simultaneously creates a liminal container of which all kinds of people can partake. The strong connection of divinities with the chariot could also suggest that the chariot lies at the basis of the establishment of the first Hindu temples as permanent external structures, a topic I will address in a separate study. The first substantial transformation of the chariot in ritual contexts took place in the early centuries CE. In its shift from Vedic to Buddhist contextu- alization, the sacred object of the chariot was reinterpreted as a means for legitimizing the ruler and as a purely religious item, and thus it was adap- tively reused. Faxian’s description of the chariot in the Buddhist tradition is strikingly similar to the modern chariot festival in the Hindu tradition at Belur. At Be- lur, just as in Faxian’s account, the chariot has a towering structure con- structed out of wood and bamboo sticks which is covered with red and yellow silk. While in the Buddhist context its four corners provided space for figures of Buddhas and Bodhisattvas, in the modern Hindu context the same space is reserved for fertility symbols, while images of the gods are carved along the wooden panels that form the base of the chariot. Moreover, Faxian described the local ruler as being part of the chariot processions he witnessed. This aspect has a parallel in the medieval period, where the presence of the king and the queen served as an instrument of the legitimation of kingdom. An additional aspect of chariot festivals of the medieval period that reap- pears in modern rituals is the temporal dissolution of social boundaries in the course of the ritual. In the diaspora, chariot processions have assumed the 15 A similar use of the chariot, which also links religious and legitimization purposes, can also be found outside India: in Egypt (see Blyth 2006, Thomas 2003) and in the Grae- co-Roman civilisation. See Sharma 2008: 134 for the hypothesis of a link between these occurrences of the chariot. 210 Cristina Bignami additional meaning of creating an identity for the community that is indepen- dent of its religious beliefs. Figures Figure 1: rathotsava at Belur, Karnataka. Photo: Cristina Bignami Figure 2: pūja at the chariot, Belur, Figure 3: Construction of the upper part KA. of the chariot, Belur, KA. Photo: Cristina Bignami Photo: Cristina Bignami  Chariot Festivals 211 Figure 4: Construction of the upper Figure 5: Construction of the upper part part of the chariot, Belur, of the chariot, Belur, KA. KA. Photo: Cristina Bignami Photo: Cristina Bignami Figure 6: Construction of the upper part of the chariot, Belur, KA. Photo: Cristina Bignami 212 Cristina Bignami Figure 7: Reliefs on one side of the chariot, Belur, KA. Photo: Cristina Bignami References Epigraphical Sources EC III = Rice, B. L. 1894. Inscriptions in the Mysore District. Bangalore: Mysore Gov. Central Press. EC VI = Rice, B. L. 1901. Inscriptions in the Kadur District. Mysore: Mysore Gov. Central Press. EC V = Rice, B. L. 1902. Inscriptions in the Hassan District. Bangalore: Basel Mission Press. EC XI = Rice, B. L. 1903. Inscriptions in the Chitaldroog District. Mysore: Mysore Gov. Central Press. EC IX = Rice, B. L. 1905. Inscriptions in the Bangalore District. Bangalore: Mysore Gov. Central Press. EC X = Rice, B. L. 1905. Inscriptions in the Kolar District. Manglore: Basel Mission Press. EC XVII = Seshadri, M. 1965. Supplementary Inscriptions in the Kolar District. Mysore: Mysore Gov. Central Press. Rice, B. L. 1879. Mysore Inscriptions. (1st ed. Bangalore 1879). New Delhi: Navrang.  Chariot Festivals 213 Secondary Sources Beal 1884 Beal, Samuel, transl. Si-yu-ki: Buddhist Records of the Western World, Translated from the Chinese of Hiuen Tsiang (A.D. 629). Reprint (1st ed. London 1884). Delhi: Motilal Banarsidass, 1994. Blyth 2006 Blyth, Elizabeth. Karnak: Evolution of a Temple. London: Routledge, 2006. Brückner 2014 Brückner, Heidrun. “Sakrales Königtum? Bildliche, litera- rische und performative Hinweise auf die Rolle des Kö- nigs in der Festkultur Vijayanagaras.” Wege zum Heil(i- gen)? Sakralität und Sakralisierung in hinduistischen Traditionen. Ed. Karin Steiner. Wiesbaden: Harrassowitz, 2014. 91–118. David 2009 David, Ann R. “Performing for the Gods? Dance and Em- bodied Ritual in British Hindu Temples.” South Asian Po- pular Culture 7.3 (2009): 217–231. Erdosy 1995 Erdosy, George, ed. The Indo-Aryans of Ancient South Asia: Language, Material Culture and Ethnicity. Berlin: de Gruyter, 1995. Indian Philology and South Asian studies 1. Eschmann, Kulke and Tripathi 1978 = Eschmann, A., H. Kulke and G. C. Tripathi. The Cult of Jagannath and the Regional Traditions of Orissa. New Delhi: Manohar, 1978. Fuller Collins 1997 Fuller Collins, Elizabeth. Pierced by Murugan’s Lance: Ritual, Power, and Moral Redemption among Malaysian Hindus. DeKalb: Northern Illinois University Press, 1997. Guy 1991 Guy, John. “The Mahābodhi Temple: Pilgrim Souvenirs in Buddhist India.” The Burlington Magazine 133.1059 (1991): 356–367. Hegewald and Mitra 2012 = Hegewald, Julia A. B. and Subrata Kumar Mitra. Re-use: The Art and Politics of Integration and Anxiety. New Delhi: SAGE Publications, 2012. Jacobsen 2008 Jacobsen, Knut A. “Processions, Public Space and Sacred Space in the South Asian Diasporas in Norway.” South Asian Religions on Display: Religious Processions in South Asia and in the Diaspora. Ed. K. A. Jacobsen. Lon- don: Routledge, 2008. 191–204. Jacobsen 2013 — Pilgrimage in the Hindu Tradition: Salvific Space. London: Routledge, 2013. Kersenboom 1987 Kersenboom, Saskia C. Nityasumaṅgalī: Devadasi Tradi- tion in South India. Reprint (1st ed. 1987). Delhi: Motilal Banarsidass, 1998. Kramrisch 1976 Kramrisch, Stella. The Hindu Temple. Vols. I–II. (1st ed. 1976). Delhi: Motilal Banarsidass, 1996 214 Cristina Bignami Kulkarni 1994 Kulkarni, Raghunath P. Viśvakarmīya Rathalakṣaṇam: A Study of Ancient Indian Chariots, with a Historical Note, References, Sanskrit Text, and Translation in English. Delhi: Kanishka Publisher, 1994. Kulke 1993 Kulke, Hermann. Kings and Cults: State Formation and Legitimation in India and Southeast Asia. New Delhi: Ma- nohar, 1993. Kulke and Schnepel 2001 = Kulke, Hermann and Burkhard Schnepel. Jagannath Re- visited: Studying Society, Religion and the State in Orissa. New Delhi: Manohar, 2001. L’Hernaut 1984 L’Hernaut, F. “Subramaṇya as Supreme Deity.” Dis- courses on Śiva: Proceedings of a Symposium on the Na- ture of Religious Imagery. Ed. Michael W. Meister. Bom- bay: Vakils, Feffer and Simons, 1984. 258–270. Luchesi 2008 Luchesi, Brigitte. “Parading Hindu Gods in Public: New Festival Traditions of Tamil Hindus in Germany.” South Asian Religions on Display: Religious Processions in South Asia and in the Diaspora. Ed. K. A. Jacobsen. Lon- don: Routledge, 2008. 178–190. Nabokov 2000 Nabokov, Isabelle. Religion against the Self: An Ethnogra- phy of Tamil Rituals. Oxford: Oxford University Press, 2000. Nandakumar 2003 Nandakumar, Prema. “Hindu Temples and Festivals: Spirituality as Communal Participation.” Hindu Spiritu- ality: Postclassical and Modern. Ed. K. R. Sundararajan and Bithika Mukerji. Delhi: Motilal Banarsidass, 2003. 421–436. O’Malley 1908 O’Malley, L. S. S. Bengal District Gazetteer: Puri. (1st ed. 1908). New Delhi: Logos Press, 2007. Ramesan 1981 Ramesan, Natesa. The Tirumala Temple. Tirupati: Tiruma- la Tirupati Davasthanams, 1981. Schnepel 2002 Schnepel, Burkhard. The Jungle Kings: Ethnohistorical Aspects of Politics and Ritual in Orissa. Delhi: Manohar, 2002. Sharma 2008 Sharma, Usha. Festivals in Indian Society. 2 Vols. New Delhi: Mittal, 2008. Sinha 2011 Sinha, Vineeta. Religion-State Encounters in Hindu Do- mains: From the Straits Settlements to Singapore. Dor- drecht: Springer, 2011. Sparreboom 1985 Sparreboom, Marcus. Chariots in the Veda. Leiden: E. J. Brill, 1985. Starza 1993 Starza, O. M. The Jagannatha Temple at Puri: Its Archi- tecture, Art, and Cult. Leiden: E. J. Brill, 1993. Studies in South Asian culture 15.  Chariot Festivals 215 Stein B. 1984 Stein, Burton. All the Kings’ Mana: Papers on Medieval South Indian History. Madras: New Era Publications, 1984. Stein M. A. 1907 Stein, M. Aurel. Ancient Khotan. Detailed Report of Ar- chaeological Explorations in Chinese Turkestan. 2 Vols. Oxford: Clarendon Press, 1907. Thomas 2003 Thomas, Susanna. Hatshepsut: The First Woman Pharaoh. New York: Rosen Publishing Group, 2003. Trouillet 2008 Trouillet, Pierre-Yves. “Mapping the Management of Threatening Gods and Social Conflict: A Territorial Ap- proach to Processions in a South Indian Village (Tamil Nadu).” South Asian Religions on Display: Religious Pro- cessions in South Asia and in the Diaspora. Ed. K. A. Ja- cobsen. London, New York: Routledge, 2008. 45–62. Turner 1973 Turner, Victor. “The Centre Out There: Pilgrim’s Goal.” History of Religions 12.3 (1973): 191–230. Willford 2007 Willford, Andrew C. Cage of Freedom: Tamil Identity and the Ethnic Fetish in Malaysia. Singapore: National Uni- versity of Singapore Press, 2007. Witzel 2013 Witzel, Michael. Das alte Indien. München: C. H. Beck, 2003. Beck'sche Reihe 2304. Younger 2002 Younger, Paul. Playing Host to Deity: Festival Religion in the South Indian Tradition. Oxford: Oxford University Press 2002.  Section 3: Adaptive Reuse of Untraced and Virtual Texts  “This is Not a Quote”: Quotation Emplotment, Quotational Hoaxes and Other Unusual Cases of Textual Reuse in Sanskrit Poetics-cum-Dramaturgy* Daniele Cuneo This chapter will highlight some instances of textual reuse in selected works of alaṃkāraśāstra (poetics) and nāṭyaśāstra (dramaturgy). The material will be investigated in order to find a provisional rationale regarding the what, how and why of various kinds of quotation scenarios, specifically and pri- marily in connection with the issue of novelty and its relation to the self-un- derstanding of traditional knowledge systems. After illustrating the well- known standard view of the denigration of novelty as such in Sanskrit śāstra (Pollock 1985, 1989a, 1989b, 1989c), I will tackle the reverse in the field of alaṃkāraśāstra, as has been propounded by McCrea (2011). My cursory survey will therefore begin with a contrastive example regarding the origina- tion of rasas (“aesthetic emotions”) according to Abhinavagupta (10th–11th CE) and Bhoja (11th CE), each in dialogue with Bharata’s Nāṭyaśāstra (2nd– 4th CE), the seminal work of Sanskrit dramaturgy. The second case to be examined is the often-studied list of the views of earlier thinkers found in Abhinavagupta’s commentary on the rasasūtra of Bharata, a crucial aphor- ism in the sixth chapter of the Nāṭyaśāstra. In his commentary, Abhinava- gupta weaves, so to speak, a narrative of various authors’ opinions and refu- tations, at the end of which his own view is enthroned as the only correct endpoint of a history of progressively improving speculations. Borrowing from the thorough analysis in Cox (2013), the third case that will be ex- amined is Śāradātanaya’s Bhāvaprakāśana (13th CE), in which, on one hand, recognized citations and re-adaptations are employed to appropriate and do- mesticate the well-known Kashmirian version of literary theory in a South Indian theoretical milieu, and on the other, Śāradātanaya seems to attribute textual passages by both earlier authors and himself to texts and authors that * I am deeply grateful to Elisa Freschi, Philipp Maas, Elisa Ganser and Charles Li for their precious remarks and suggestions. All mistakes, of course, are mine alone. 220 Daniele Cuneo probably never existed, but are smuggled in as the “actual” ones. Arguably this has been done to legitimize his own cultural endeavor and sanction its theoretical and practical validity. The last work that will be considered is He- macandra’s Kāvyānuśāsana (11th–12th CE). In his own sub-commentary, called Viveka, Hemacandra repeats almost verbatim Abhinavagupta’s entire analysis of rasa found in the above-mentioned rasasūtra commentary. This act of sheer repetition, however, camouflages small but significant changes to the quoted portions, probably introduced to make the material better fit Hemacandra’s own theoretical agenda. It is worth noting that the case of Hemacandra represents a number of similar cases of extensive reuse of textual materials in alaṃkāraśāstra and nāṭyaśāstra in the second millen- nium. My tentative conclusion for assessing these very disparate and some- times unusual manners of textual reuse hinges on a partial acceptance of McCrea’s thesis of the peculiar nature of alaṃkāraśāstra as a laukika (this- worldly) system of knowledge, for which novelty and change are the norm and not the exception. My acceptance remains partial insofar as the evidence can be better interpreted by postulating two paradigms of textual authority whose fortunes alternate over the centuries. They are tightly linked both to the search for a foundational text in the specific folds of alaṃkāraśāśtra, and the overlapping of topics and the synthesis of theoretical notions across the two domains of alaṃkāra- and nāṭyaśāstra. Their intersection might be post- ulated as one of the main reasons for the fluctuation between meta-specula- tive stances regarding novelty and tradition. 1 Introduction: Reuse, novelty, and tradition purāṇam ity eva na sādhu sarvaṃ Kālidāsa’s Malavikāgnimitra (prastāvanā, verse 2a) Of the many approaches and perspectives on adaptive reuse outlined by Elisa Freschi and Philipp Maas in the introduction to this volume, my focus lies on the link of reuse to the question of originality and, more specifically, to spe- culative innovation. Among the numerous Sanskrit knowledge systems known as śāśtras, whose overall aim is the nomothetic establishment and sanction of a correct, regulated standard of human practice in the respective fields of application, my case studies are taken from the two interconnected fields of nāṭyaśāstra and alaṃkāraśāstra. Investigating their many patterns of textual reuse and their aberrancy and mutability through time, these two  “This is Not a Quote” 221 fields will be tentatively interpreted based on the emic self-perception of the two learned, traditional discourses along with their often implicit meta-theo- retical assumptions, as well as the etic stance of Pollock’s studies on the cat- egory of śāstra (1985, 1989a, 1989b, 1989c) and their recent sequel-cum- rejoinder by McCrea (2011). In particular, the close link between the two śāstras in question will be one of the central pivots for challenging, in part, earlier scholarship and gaining a more nuanced understanding of the link of innovation and textual reuse. The significance of this understanding depends on conceptualizing the idea of “tradition” as a modality of change. Similarly, the actual realities of “traditional societies” show that these are kinds of societies that understand their own transformations in terms of this modality, commonly in a non-self- reflexive manner in which being “traditional” seems to be considered a natu- ral fact and not a cultural construction.1 Seen from another perspective, the issue at stake is “the issue of tradition,” insofar as including cultural change within the boundaries of traditional knowledge must be seen as an inherent and constitutive aspect of any traditional discourse as such. An assessment, albeit limited, of textual reuse and its vagaries offers a privileged perspective on this issue; the skillful reuse of texts can be, and has been, wielded as a powerful weapon to bridge the tensions that arise when coping with the un- avoidable cultural antagonism between the introduction and legitimization of novelty and the reiteration and re-affirmation of bequeathed knowledge. 2 Śāstra as an ideological apparatus In the above-mentioned series of pioneering articles written in the 1980s, Pollock argued that the discursive technology of śāstra2 – arguably and em- blematically born in its almost classical argumentative form with the work of the grammarian Patañjali (2nd BCE), whereupon it occupied central stage in the two following millennia of the Sanskritic episteme3 – is based on the as- 1 See Hobsbawm and Ranger 1983, Guolo 1996 and Squarcini 2008. Moreover, for a multi-perspective study on tradition as a device and basis for both change and its vali- dation, see Squarcini 2005. 2 On the meaning of the term according to various emic understandings (“system of rules” and “revelation,” i.e., the Veda itself), see Pollock 1985: 501–502. 3 One might want to include in the genre the even earlier texts known as vedāṅgas (6th– 3rd BCE), ancillary disciplines chiefly conceived as means for preserving the Vedic corpus and properly performing the Vedic rites, and the dharmasūtras (from the 4th–3rd BCE), the first legal texts of the tradition that was later called dharmaśāstra (see be- 222 Daniele Cuneo sumption that truth is revealed and given once and for all in a timeless past. It is then merely repeated and explained in any given present time in any single traditional knowledge system. Specifically, Pollock’s main thesis pivots on how the authoritativeness of all śāstras (“cultural grammars,” “cultural software” or “knowledge systems,” as he cleverly dubs them) is rooted in a kind of transhistoric, trans-human and transcendent source, usually a lost or otherwise inaccessible Vedic or semi-Vedic (even divine) scripture. For in- stance, it is stated in the Nāṭyaśāstra that the various branches of the theater derive from the four Vedas.4 Or, in the case of āyurveda (medicine), the śāstra traces its origin back to the god Brahmā.5 In Pollock’s view, this ac- tively negates history and novelty as concrete possibilities in the development of knowledge. This meta-theoretical stance, Pollock argues, is borrowed from the Mīmāṃsā model of the textual authority of the Veda, whose unique status as an authorless and ahistorical text, i.e., its apauruṣeyatva, makes its legiti- macy intrinsically unquestionable.6 In other words, in the self-understanding of śāstra and its understanding of the world, theory must always precede practice. This nomological mechanism of warranty and validation determines and seals the supremacy of the proponents of a tradition (such as the Vedic or, potentially, any other) by making them in practice the only judges who can sanction any possible or actual novelty by means of an ex post reinter- pretation of theory that strives to include that new practice (or idea) in an older framework, in the reassuring womb of traditionally validated know- ledge. The naturalization and de-historicization of cultural practices actually hides the hand of elitist and dominant powers and, at least in theory, immu- nizes them from any critical attack or derogatory value judgment. low: p. 225). 4 Nāṭyaśāstra 1.17: jagrāha pāṭhyam ṛgvedāt sāmabhyo gītam eva ca | yajurvedād abhi- nayān rasān ātharvaṇād api ||. 5 In Pollock’s theory, the properly “Vedic” origin and the “divine” origin (such as the one of āyurveda) are thought to belong to the same general framework of understand- ing the śāstra, its genesis and role. Whether or how far this conflation might be re- garded as problematic lies beyond the limited boundaries of the present chapter. See, for instance, Pollock (1989c: 609): “Veda is the general rubric under which every sort of partial knowledge – the various individual śāstras – are ultimately subsumed. There are several routes to establishing this contiguity: through some formal convention em- bodied in the text – a śāstra will explicitly claim status as a Veda, or establish for itself a paraṃparā reverting to God, or present itself as the outcome of divine revelation.” 6 It is worth noting that early Buddhist and Jain works may well have contributed, at least indirectly, to this meta-theoretical idea of a timeless and unquestionable truth, since the founders of the non-brahmanical religions were regarded as omniscient and, to some extent, beyond the pale of criticism based on human reason. I thank the editors of this volume for their useful insights on this crucial point.  “This is Not a Quote” 223 Given the major role that this ideological apparatus played in shaping the culture of the Indian subcontinent for almost two millennia, concrete exam- ples are easy to find. A first one might be the dharmaśāstra, the corpus of social and cultural textualized norms aimed at regulating every aspect of human behavior. At its cornerstone is the very concept of dharma, especially insofar as it is considered to share the transhistoric and naturalized status of the Veda, in its turn the paramount source of dharma.7 However, in the his- tory of the Sanskrit cultural hegemony, practically all fields of human activity became the object of a śāstra, from the creation of art to sexual intercourse, from archery (or the science of weapons in general) to astronomy or astrol- ogy, from architecture to lexicography. Numerous traditional lists of sciences (śāstras or, more often vidyāsthānas) are well known; again Pollock (1985: 502–503) is a good guide through the dizzying lists and their variations.8 Anyway, our present concern is the realization that “virtually any organized activity known to a premodern society is amenable to treatment in śāstra” (Pollock 1985: 502). Having acknowledged the wide scope of the technology of śāstra, it is useful to note how its meta-speculative stance of the primacy of traditionality over novelty – of recovery over discovery – is not, for the most part, estab- lished textually on the foundational works of the various branches of know- ledge, but more aptly and commonly on the impressive number of commenta- ries and sub-commentaries on these works, which constitute the overwhelm- ing majority of Sanskrit texts tout court.9 Obviously, the textual genre of commentary highlights, both implicitly and explicitly, the pre-eminence of the principle of authority over individual originality, since the task of a “commentator” is completely different from that of an “author,” at least in theory.10 The theoretical-cum-practical technology of śāstra and the practical 7 For the latest review of the issue of the various sources of dharma, with a selected bibliography and references to relevant dharmaśāstra literature, see David 2015. 8 On the progressive opening up of these lists and the enduring restrictiveness of Mīmāṃsā with regard to conferring the transcendent legitimacy of the Veda to other fields of knowledge, see Pollock 1989b. 9 On the idea and analysis of numerous cultural traditions as “Commentary Cultures,” see Quisinsky and Walter 2007 and the workshop “Commentary Cultures. Technolo- gies of Medieval Reading” that was held within the framework of “Zukunftsphilolo- gie” 16–17 May 2013 in Berlin. On various aspects of the commentary culture of South Asia, see von Stietencron 1995, Chenet 1998, Hulin 2000, von Hinüber 2007, Slaje 2007, Tubb and Boose 2007 and Ganeri 2010. 10 The cautionary double quotes are meant to indicate that in actual practice there is no clear-cut divide between the roles of author and commentator, and that many com- mentators can be regarded as more original and “authorial” than authors, both in South 224 Daniele Cuneo technology of commentary represent, so to speak, two sides of the same ideological apparatus, an apparatus that is aimed – generally speaking and in Pollock’s parlance – at creating, preserving and naturalizing a set of norms conceived to assure the crystallization of the power structures in the social and cultural status quo. 3 The worldly śāstra, its fuzzy boundaries, and the derivation of rasas The general validity of this basic paradigm can be challenged using evidence drawn from the field of alaṃkāraśāstra. When compared to the others, this field is a quite extraordinary knowledge system.11 I’ll just briefly state three reasons for its distinctiveness.12 First, it is a latecomer, its first extant work dating only to the 7th century CE.13 Other śāstras have a significantly older pedigree. Second, its subject matter is thoroughly laukika, that is, this- Asia and elsewhere. 11 This section of the present chapter contains material I presented at the 14th World Sanskrit Conference in Kyoto in a paper entitled “Smuggling Novelty or Dismantling Tradition. Abhinavagupta and Bhoja on the derivation of rasas.” 12 The present remarks are developed along similar lines as in Tubb (2008: 173–176), where the “murkiness of the status of poetics as a śāstra,” concerning its subject mat- ter, its sources as well as its audience, is briefly dealt with. The peculiarity of alaṃkā- raśāstra as a knowledge system and the issue of novelty in its historical development are also dealt with in Tubb and Bronner 2008, but with a specific focus on the authors of the 16th and 17th centuries, the self-consciously navya (“new”) school (in this regard, see also Bronner 2002 and 2004). 13 Although they quote earlier authors by name, the first two extant works of alaṃkāra- śāstra, Bhāmaha’s Kāvyālaṃkāra and Daṇḍin’s Kāvyalakṣaṇa (or Kāvyādarśa), both date to the 7th century (with Bhāmaha prior to Daṇḍin, as has been convincingly argued in Bronner 2012). The Nāṭyaśāstra, on the contrary, is considered to date back to the first centuries of the Common Era, or maybe even earlier. The intersection of the two fields of knowledge (nāṭya- and ālaṃkāraśastra) is a cultural event whose beginning can be dated with reasonable certainty: The Kashmirian author Udbhaṭa (8th–9th CE) wrote, as probably the first, on both alaṃkāraśāstra (his Kāvyālaṃkārasaṃgraha and his mostly lost commentary on Bhāmaha; see Gnoli 1962) and nāṭyaśāstra (a lost commentary on Bharata’s work). A partial fusion of the two knowledge systems oc- curred with Ānandavardhana’s Dhvanyāloka (see McCrea 2008). And Abhinavagupta (10th–11th) masterfully attempted the complete convergence of the two cultural gram- mars in his twofold effort of commenting on and harmonizing the Nāṭyaśāstra and the Dhvanyāloka. As I will argue throughout this chapter, it is this convergence of know- ledge fields, along with their different styles and attitudes towards novelty and tra- dition, that might be regarded as one of the causes for the peculiar alternation between the two different methods of validating and norming authority (see below, section 7).  “This is Not a Quote” 225 worldly. The genre of kāvya (“belles-lettres”) even has a specific, albeit semi-mythical, beginning in “history,” so to speak, namely Vālmiki’s Rāmā- yaṇa, the first kāvya by the first poet (ādikavi). Third, alaṃkāraśāstra lacks a root text that could have been the object of a chain of commentaries and sub- commentaries, as have other śāstras such as nyāya, mīmāṃsā and so on. I was very happy to discover that Lawrence McCrea, in a contribution to the 2011 volume in honor of Sheldon Pollock, South Asian Texts in History, ex- plored the usefulness of Pollock’s understanding of the transcendent śāstraic model, limiting the scope of its application by showcasing the discourse on literary theory. Here, he argues, elements of historical consciousness and pride, as well as practice-driven, historically self-aware theoretical inno- vations are actually quite frequent and possibly the norm. In addition to focusing on alaṃkāraśāstra and nāṭyaśāstra, I would like to propose that there were two opposing methods for how bequeathed know- ledge was dealt with in the crafty hands of South Asian commentators.14 The two commentarial meta-techniques I am suggesting could also be regarded as two extremes in the spectrum of commentarial approaches. On one side, which I somewhat fancily dub the “novelty-smuggling” strategy, theoretical and practical changes were introduced to the framework of traditional lore by disguising transformations in the reassuring garb of the old system, thereby rejecting novelty per se as a legitimate cultural category. On the other side, which I call the “tradition-dismantling” strategy, bequeathed knowledge was de-legitimized and the novelty of change was invested as sovereign for building cultural discourse, thus setting new parameters for future develop- ment. I am presenting two case studies, intended as paradigmatic examples of this. The first, as an instance of the “novelty-smuggling” strategy, is a pas- sage from Abhinavagupta’s commentary (10th–11th century) on some verses of Bharata’s Nāṭyaśāstra; the second, as an instance of the “tradition-dis- mantling” strategy, a heated discussion in the Śṛṅgāraprakāśa of Bhoja (11th century) on the same portion of the Nāṭyaśāstra. The textual details in them- selves do not concern us directly, but given the brevity of the Nāṭyaśāstra passage and the marked difference between the two takes on it, it is a good example in this investigation of textual reuse in texts on dramaturgy and poetics. 14 One of the aims of the present chapter is also to show how these two commentarial approaches can be considered two general “authorial” attitudes towards novelty and tradition, independent of their use in commentaries proper or in any other work within the Sanskritic episteme. 226 Daniele Cuneo Verses 39–41 of the sixth chapter of Bharata’s Nāṭyaśāstra, the first and most important treatise on Indian dramaturgy available to us, deal with what I have tentatively called the “derivation of rasas.” To describe this briefly, without entering the centuries-long debate on their epistemological status and their definite locus,15 the rasas are the various possibilities of the audience’s emotional response to a theatrical performance elicited by an array of com- ponents and representing the performance’s ultimate aim. The standard text of Bharata’s Nāṭyaśāstra lists eight such emotional experiences.16 The threes verse we are concerned with here construct a derivational pat- tern among these eight rasas, whereby four of them are seen as originating from the other four. Consequently, the former are considered the “causes of origination” (utpattihetu) of the latter. One might say, in other words, that the former are primary – at least in this respect – and the latter are secondary. The introductory prose to verse 6.39 and the verse itself read as follows: teṣām utpattihetavaś catvāro rasāḥ | tad yathā – śṛṅgāro raudro vīro bībhatsa iti | atra śṛṅgārād dhi bhaved dhāsyo raudrāc ca karuṇo rasaḥ | vīrāc caivādbhutotpattir bībhatsāc ca bhayānakaḥ || Among those [eight rasas], four rasas are the causes of origination [of the other four]. Namely, śṛṅgāra, raudra, vīra and bībhatsa. In this re- spect: hāsya arises from śṛṅgāra, and from raudra [arises] the rasa karuṇa, then, adbhuta originates from vīra, and bhayānaka from bī- bhatsa. The following schematic table employs the rough but usually accepted translation of the names of the various rasas: 15 I have examined the two main interpretations of rasa in an earlier article (Cuneo 2013); its arguments do not need repetition here, but its tentative conclusions might be useful for framing the general problem, at least in a note. According to the interpretation of the “ancients” (theoretically including Bharata himself, although, in my reading of his text, there are some significant doubts and grey areas regarding this), rasas are nothing but heightened ordinary emotions, experienced by the characters in dramatic represen- tations and enjoyed secondarily by the audience (Bhoja shares this view, with his per- sonal accent on the singularity of rasa, as discussed below.) According to the new pa- radigm, championed by Abhinavagupta and followed by many other authors after him, rasa is the emotion directly savoured by the audience. It consists of a blissful aesthetic Erlebnis that is qualitatively different from ordinary experience, insofar as the felt emotion is distilled of any reference to personal identity, causality or spatio-temporal- ity. This distillation eliminates desire and, hence, allows the beatific savouring of the emotional experience itself, ultimately not different from the spectators’ own con- sciousness. 16 In the subsequent history of the nāṭyaśāstra, many authors recognized and argued for a different number of rasas. For an overview of this matter, see Raghavan 1967.  “This is Not a Quote” 227 Table 1: The origination of rasas “Originating” rasa → “Originated” rasa 1. the erotic (śṛṅgāra) → the comic (hāsya) 2. the furious (raudra) → the pathetic (karuṇa) 3. the heroic (vīra) → the wondrous (adbhuta) 4. the loathsome the fearful (bhayānaka) → (bībhatsa) The internal logic of this derivational pattern is very briefly outlined in the next two verses, Nāṭyaśāstra 6.40–41: śṛṅgārānukṛtir yā tu sa hāsyas tu prakīrtitaḥ | raudrasyaiva ca yat karma sa jñeyaḥ karuṇo rasaḥ || vīrasyāpi ca yat karma so ’dbhutaḥ parikīrtitaḥ | bībhatsadarśanaṃ yac ca jñeyaḥ sa tu bhayānakaḥ || Hāsya is well known as the imitation of śṛṅgāra, and the karuṇa rasa is known to be the result (lit. “action, activity”) of raudra. Moreover, the result (lit. “action, activity”) of vīra is well known to be adbhuta, whereas the vision of bībhatsa is to be known as bhayānaka. On first sight, this brief explanation of the relationship between various emotional states seems sound and comprehensible; for instance, looking at something disgusting can also engender fear. However, on closer inspection, this explanation is far from obvious and self-explanatory. There are a consi- derable number of possibilities regarding the status and locus of rasas as conceived in the text of Bharata that can change how the derivational pattern between the emotions is understood.17 Nonetheless, the present aim is not to clarify the contents of this passage in the Nāṭyaśāstra itself, but rather how Bhoja and Abhinavagupta approached its problematic nature. While the two authors were close in time, they were probably not only unaware of each other, but also unaware of each other’s account of this aspect of Bharata’s aesthetic theory, an aspect that was either outdated, somewhat underdeve- loped, or simply no longer fully understood. 17 See n. 15 and Cuneo 2013 for some speculations on the issue. 228 Daniele Cuneo In a passage of his Śṛṅgāraprakāśa, translated in part in a seminal essay by Pollock (1998),18 Bhoja, the king of Dhāra in the reign of Malwa, is quite adamant in rejecting Bharata’s proposal that some rasas derive from others, seeing this as completely inadequate, as well as in stating the superiority of his own theory of aesthetics. Bhoja’s theory recognizes, in fact, the existence of only one rasa, namely śṛṅgāra “passion,” also called abhimāna “sense of self,” ahaṃkāra “ego,” preman “love” and rasa (in the singular), representing the real and only origin of all other rasas. In Pollock’s words (1998: 126), Bhoja’s śṛṅgāra is “what enables a person to experience the world richly” and “the capacity of emotional intensity as such.”19 In his argumentation, Bhoja denies that one rasa might arise from another, stating that such a pattern of arising can be logically understood in only two ways and both are erroneous. According to the assumed under- standing of the production of psychological states in the nāṭyaśāstra (here with the lowercase I mean the knowledge system, not the foundational text), either the originating rasa is the “determinant as substratum” or the “concrete cause” of the originated rasa (ālambanavibhāva) – such as a hero and he- roine being considered the ālambanavibhāva of the rasa of love (śṛṅgāra- rasa) – or, in a manner reminiscent of Sāṃkhya thought, the originating rasa is the primordial state (prakṛti) from which the originated rasa develops in a kind of self-transformation. In the first explanation (treated in Śṛṅgāraprakāśa, pp. 684–685), Bha- rata’s theory does not hold because there would be an invariable concomit- ance between the originating rasa and the originated rasa, and this is not the case. For instance, the comic (hāsya) can be found arising from rasas other than the erotic (śṛṅgāra), and, moreover, it can also be found as not arising from the erotic. Furthermore, the pathetic rasa can arise from a rasa other than the furious, and not all instances of the furious govern the arising of the pathetic, since the furious can also produce the fearful or the loathsome. At this point, Bhoja offers numerous examples of possible breaks in Bharata’s pattern, such as the comic rasa arising from a rasa other than the erotic, a rasa other than the comic from the erotic, and so forth.20 18 The same passage has also been dealt with briefly in Raghavan (1978: 424–426). 19 To anticipate some of the conclusions of this digression: Bhoja’s “monistic” aesthetic philosophy is a complete novelty, a novelty that is consciously aimed at revolutioniz- ing its field of knowledge. 20 Just as an example of the many poetical examples, we can cite Kirātārjunīya 3.21, a verse quoted by Bhoja as an instance of a rasa other than the pathetic – in this case, the fearful – arising from the furious: “On seeing the son of Radhā (i.e., Karṇa), who by his fury made [his enemies] lose their composure and who had propitiated the son of  “This is Not a Quote” 229 In the second explanation (treated in Śṛṅgāraprakāśa, pp. 686–687), with the originating rasa being the prakṛti of the originated rasa, Bharata’s theory does not hold either. According to this understanding, the comic is a trans- formation of the erotic, indeed, in Bharata’s definition, an imitation of the erotic. However, the comic can also arise from imitations of other rasas, for instance, the heroic. Moreover, the same variability in derivational patterns can be identified in all of the other cases as well, since – and this is the real point at stake – there is no compulsory developmental relationship between any two rasas. Namely, it is impossible to establish that rasa x will con- stantly and invariably arise from rasa y. In other words, it is always possible to find examples in literature in which a given rasa derives from some other rasa, a given emotional situation is seen as deriving from another one. There- fore, Bharata’s theory of a one-to-one relationship between four rasas that are primary and four rasas that are secondary does not manage to pass the strict scrutiny of reason as is orchestrated in Bhoja’s criticism. In Kashmir, quite some distance from the homeland of Bhoja, Abhinava- gupta dealt with the same problem. His solution, however, is based on a very different meta-theoretical standpoint, in this case the need to integrate the innovation into the reassuring form of traditional knowledge. For this reason he can not refute the respected and authoritative text of Bharata; it must be commented upon and tacitly reinterpreted. The development of knowledge climbs, so to speak, up the rungs of the commentarial ladder. In four celebrated and oft-quoted verses of his Abhinavabhāratī (see the appendix to this chapter), found in the middle of his rasasūtra commentary as a kind of manifesto-like intermezzo, Abhinavagupta proclaims his view. The tentative understanding of these verses and of Abhinavagupta’s meta-theo- retical attitude towards change will be the focus of the conclusions to the present chapter. At the level of his commentarial practice, the attitude of respect towards the tenets of the traditional knowledge of the Nāṭyaśāstra influences both his treatment of the derivation of rasas and his strikingly in- novative interpretation, aimed at resolving the conundrum represented by the apparent inadequacy of Bharata’s view while creating a new paradigm for the Rasa theory (see n. 15). In his commentary on the verses of Bharata cited above, Abhinavagupta “explains” that what seems a rather rigid and schematic model of subdividing Jamadagni (i.e., Paraśu Rāma) [in order to acquire the knowledge of the missiles], even in the God of Death would forcibly arise an acquaintance with feelings of fear, un- known [to him before].” (nirīkṣya saṃrambhanirastadhairyaṃ rādheyam ārādhitajā- madagnyam | asaṃstuteṣu prasabhaṃ bhayeṣu jāyeta mṛtyor api pakṣapātaḥ ||). Translation modified from that of Roodbergen 1984: 170. 230 Daniele Cuneo rasas between janaka and janya, “producer” and “produced,” is nothing but a paradigmatic exemplification of four conceivable derivational patterns be- tween rasas, four in a wide array of possible combinations. Therefore, the comic (hāsya) comes not only from imitating the erotic (śṛṅgāra), it can also arise from the imitation of any other rasa.21 Note that this is exactly one of the arguments wielded by Bhoja against Bharata’s theory. Similarly, according to Abhinavagupta, who continues his thoughts on the matter, the origination of the pathetic (karuṇa) from the furious (rau- dra) as stated in the verse is simply an example of a possible relation between rasas, i.e., a relation in which an originated rasa is the result of the result of the originating rasa; in this case, the pathetic (karuṇa) is the result of impri- sonment and murder, as Abhinavagupta explains, that are the result of the furious (raudra). The same reasoning is to be considered valid for the two remaining derivational patterns. The origination of the wondrous from the heroic is simply an example of a relation between rasas in which the origi- nated rasa is the direct result of the originating rasa. Finally, the origination of the fearful from the loathsome is an example of a relation between rasas in which the originated rasa derives from the same vibhāvas, i.e., from the same “dramatic” causes, to put it briefly, of the originating rasa. Unfortunately, it would overextend the limits of this short chapter to discuss the several poetic examples quoted by Abhinavagupta to substantiate his interpretation of Bha- rata’s verses. To present Abhinava’s view schematically, an originated rasa can be: 1 a semblance of the originating rasa, such as the comic (hāsya) for the erotic (śṛṅgāra), 2 an indirect result of the originating rasa, such as the pathetic (karuṇa) for the furious (raudra), 3 a direct result of the originating rasa, such as the wondrous (adbhuta) for the heroic (vīra), or 4 a further result of the “dramatic” causes (vibhāva) of the originating rasa, such as the fearful (bhayānaka) for the loathsome (bībhatsa). 21 In Abhinavagupta’s words, in Abhinavabhāratī ad Nāṭyaśāstra 6.39, vol. 1, p. 294, “Along the same lines, the word ‘śṛṅgāra’ [in Nāṭyaśāstra 6.39] suggests a modality in which [hāsya] comes from the semblance of one [of the other rasas]. Therefore, hāsya must be recognized as also present in the semblances of all [the other rasas], such as karuṇa and the like. [This obtains] because being a determinant (i.e., a dramatic cause, to put it briefly) for hāsya is merely brought about by the activity of inappropriateness [of any kind].” (evaṃ tadābhāsatayā prakāraḥ śṛṅgāreṇa sūcitaḥ. tena karuṇādyābhā- seṣv api hāsyatvaṃ sarveṣu mantavyam. anaucityapravṛttikṛtam eva hi hāsyavibhāva- tvam.)  “This is Not a Quote” 231 To conclude this first bundle of evidence regarding reuse, traditionality and innovation, I have argued that in their respective aesthetic theories, both Bhoja and Abhinavagupta recognized the inadequacy of Bharata’s view – although this recognition is only implicit in the case of Abhinavagupta, whose reinterpretation of the text is concealed under the cloak of respect for an allegedly infallible tradition – and that both Bhoja’s and Abhinavagupta’s solutions to that inadequacy, as well as their overall aesthetic theories, are strikingly innovative in their treatment of Bharata’s Nāṭyaśāstra text. How- ever, to clearly state what was already hinted at, the different approaches of these two authors represent two extremes in the spectrum of strategies through which cultural change is brought about, accounted for and legiti- mated in an unending process of constructing, preserving and re-inventing any traditional discourse. Bhoja implements the method I call the “tradition-dismantling” strategy. He consciously and overtly takes apart and de-legitimizes the traditional dis- course on dramaturgy by refuting its tenets through both logical argumen- tation and phenomenological exemplification. On these ruins of traditional knowledge, he then constructs the new building of his own theory, trying thereby to set new parameters for future development. In contrast, Abhinavagupta implements the method I call the “novelty- smuggling” strategy. Accordingly, he does not directly challenge the norma- tive authority of the tradition represented by Bharata’s Nāṭyaśāstra. Rather, by commenting in his Abhinavabhāratī on the text of Bharata instead of composing an independent treatise as Bhoja did with his Śṛṅgāraprakāśa, he both implicitly and explicitly pays respect to the great semi-mythical master of dramaturgy, thereby also acknowledging the master’s authoritative and prescriptive status. Nevertheless, while interpreting Bharata’s text by means of the various hermeneutical devices common to the Sanskrit commentarial praxis, Abhinavagupta introduces crucial innovations. A noteworthy innova- tion is for instance Abhinavagupta’s famous conception of śāntarasa as the main rasa from which all other rasas develop and of which all other rasas ultimately consist.22 Reducing the different rasas in this way to a unique and supreme rasa is not far from Bhoja’s conception of śṛṅgāra as the true and only rasa. Nevertheless, Abhinavagupta’s meta-theoretical strategy entails an inclusion of theoretical and practical change within the seemingly unda- 22 Much has been written on the concept of śāntarasa. Without presuming exhaustive- ness, I will mention Pandey 1944, De 1960, Raghavan 1967, Masson and Patwardhan 1969 and 1970, Bhattacharya K. 1972, Gerow and Aklujkar 1972, Bhattacharya S. P. 1976 and Gerow 1994. I will also briefly express my take on the issue in Cuneo 2016: 59–60. 232 Daniele Cuneo maged, unaltered and inalterable framework of traditional lore. This means that in this case, the inherently disruptive nature of novelty remains disguised in the apparently harmless verses of the long-established and revered text of Bharata’s Nāṭyaśāstra.23 4 Quotation emplotment and the teleology of commentarial thought As is certainly clear, the first scenario presented above is not really about actual quotations.24 However, within the disputably central issue of commen- tary it allowed me to propose a theoretical background – two modalities for change and innovation – that can be used for exploring other quotation scena- rios within the fields of nāṭya- and alaṃkāraśāstra. A different and more general observation on quotations and their relation- ship to tradition is the following: Quoting a text or extensively reusing its material as “an authority” ‒ as a case of ipse dixit, so to speak ‒ can only be considered a hint at the self-perceived traditionality of a knowledge system. However, if an earlier text of the same tradition is quoted or mentioned as a rival to be refuted, this actually indicates a tolerance for novelty. Indeed, in the first few documented centuries of the development of alaṃkāraśāstra, it seems that the second case is the norm, either explicitly or implicitly.25 During this period a handful of authors follow one another, re- futing each other’s theories and attempting to build a coherent system that accounts for the poeticalness of poetry or properly describes the specific 23 As a postscript to this section, one might add that although Bhoja overtly employs what I have called “tradition-dismantling” and Abhinavagupta what I have called “novelty- smuggling,” of the two it is Abhinavagupta who arguably develops the more innova- tive theory regarding the epistemology of rasa. I am referring to what I have elsewhere called in Kuhnian terms the “second paradigm shift” (with respect to the first one pro- pounded by Ānandavardhana and identified as such by McCrea 2008). This second revolution “marks the change from a conception of aesthetic experience (rasa) that does not account for the ontological difference between the universe experienced in or- dinary reality and the universe created by, and experienced in, art to a conception of aesthetic experience (rasa, again) […] that does account for such a difference and makes it the crucial speculative argument justifying and legitimizing the intrinsically pleasurable, or even beatific, nature of the emotions aroused by art” (Cuneo 2013: 62). 24 One might argue, however, that in every commentary the entire text being commented upon is either explicitly or at least implicitly quoted, and that therefore the intellectual practice of commenting is inherently quotational. 25 See, for instance, Bronner 2012, in which various passages of Daṇḍin are convincingly identified and interpreted as rejoinders to tenets propounded by Bhāmaha.  “This is Not a Quote” 233 features of poetry.26 As the history of the discipline continues, the issue be- comes more and more complicated. A crucial complication in the picture – the factor that from the beginning shakes the assumption that the discipline is laukika (as maintained in the above-mentioned publication McCrea 2011) – is the ambiguous nature of the Nāṭyaśāstra. As already mentioned, it is supposedly the root text of the scholarly discipline of dramaturgy. However, Abhinavagupta’s Abhinavabhāratī is practically27 its only extant commen- tary, unlike the case of other mūlasūtras, which were commented upon many times and whose commentaries received sub-commentaries and so on. Fur- thermore, the Nāṭyaśāstra is attributed to a semi-mythical figure, Bharata, literally meaning the “actor,” and begins with a story of the descent of the creation of theater by Brahmā and its transmission to Bharata. It therefore possesses at least some of the crucial features that Pollock tried to identify as common to those śāstras that share the transhistoric character of the Vedic texts.28 However, as the above-mentioned passages of Bhoja testify, its status as an infallible source of epistemic authority was challenged just as often as it was resorted to, especially from the 9th century onwards, when the fields of dramaturgy and poetics gradually began to be integrated. Much more research is needed to settle the issue of tradition vs. innova- tion in these two interconnected fields. But despite the cursory nature of this survey of some specific textual material, a working hypothesis for outlining and explaining the major quotational trends in the discipline will be at- tempted. In particular, three quotation scenarios will be presented, cases that are quite unusual with regard to the interpretive grids of traditionality or to- lerance-for-novelty as outlined above. 26 For the history of poetics and dramaturgy, one can consult the classical De (1960), Kane (1961), and Gerow (1977). A very useful and more up-to-date discussion of the first couple of centuries, with a focus on the issue of rasa, is McCrea (2008: 30–54). 27 Another versified commentary on the Nāṭyaśāstra called Sarasvatīhṛdayālaṃkāra (although it is often referred to as Bharatabhāṣya, see Primary Sources) by Nānyadeva (11th–12th CE) also exists, but it only covers the sections on music, which do not con- cern us here. As an aside, as far as I know the only edition of this work, by Chaitanya P. Desai, seems to be based only on the manuscript found in the Bhandarkar Oriental Research Institute of Pune (MS no. 111 of 1869–70) and does not use the manuscript held in the Government Oriental Manuscript Library of Madras (MS no. R. 5598 – Vol. 1, S.R. 2981). Moreover, more manuscript material related to this work might well be unearthed by further research. 28 For the divine origin of both theater and the knowledge of theater, see the first chapter of the Nāṭyaśāśtra (in particular, the many verses where the expression nāṭyaveda ap- pears). For an analysis of the myth of origin, see Bansat-Boudon 2004. 234 Daniele Cuneo The first case I would like to examine is what I tentatively call “quotation emplotment.” I am referring to the famous commentary by Abhinavagupta on the much-quoted rasasūtra of Bharata: vibhāvānubhāvavyabhicārisaṃyogād rasaniṣpattiḥ. As mentioned above, the Abhinavabhāratī is practically the only extant commentary on the Nāṭyaśāstra. However, Abhinavagupta quotes a plethora of other authors who commented on the work or, at least, dealt specifically with the issue of rasa. Just to name the most important, we en- counter Bhaṭṭa Lollaṭa, Śrī Śaṅkuka, Bhaṭṭa Tauta and Bhaṭṭa Nāyaka. The range and content of their various opinions will not be dealt with in the present context, since they have already been the object of quite a lot of scholarship.29 Here I am interested in how Abhinavagupta quotes them and uses their theories in relation to one another. But first a caveat is necessary. Since Abhinavagupta is practically our only source, no precise details are presently known about the respective theo- ries of the aforementioned authorities on rasa. Indeed, we might reasonably doubt the accuracy of Abhinavagupta’s re-use of their words and their posi- tions.30 Even after quick reading the first passage of the commentary on the ra- sasūtra, it is possible to appreciate the power of rhetorical technique imple- mented by Abhinavagupta, both in using direct quotations and rephrasing his predecessors’ textual materials. He weaves a narrative, so to speak, of the 29 The commentary on the rasasūtra is the object of the pioneering translation and study by Gnoli (1968). The most recent treatment, albeit somewhat lacking in fresh ideas, of this seminal section is found in Gopalakrishnan 2006. I have attempted an improved translation of the text (within the context of the whole sixth chapter) in my unpublished PhD thesis (Cuneo 2008–2009). Several portions of the rasasūtra commentary have been newly translated and analyzed in a number of articles, as for example Pollock (2010b) and David (2016). Generally it is also worth reading Ingalls (1990), since the same discussion on the nature of rasa is contained in an abbreviated form in Abhinavagupta’s Locana on Ānandavardhana’s Dhvanyāloka (especially 2.4). Many arguments on rasa found in the poorly studied Kalpalātāviveka can be traced back to those of Abhinavagupta (although most are probably from his Locana). In the history of alaṃkāraśāstra, the rasasūtra commentary has been often taken as a model and also quoted en bloc, sometimes in an abridged form but sometimes expanded upon (see section 6 of this chapter for a brief examination of such re-use). 30 An extreme stance, if we want to give in to scepticism, would consist in doubting even the very existence of these authors and postulating that they are fictional characters in a dialectic drama enacted by Abhinavagupta himself, who is both director and the only actor impersonating different roles. I personally do not hold this view, since I am con- vinced, at least in general terms, by the arguments provided by Pollock (2010b) that identify many of the ideas of Bhaṭṭa Nāyaka in the fourth chapter of Dhanañjaya’s Daśarūpaka and, especially, Dhanika’s Avaloka commentary on it.  “This is Not a Quote” 235 opinions, arguments and refutations of the various authors who preceded him, thus building something between a sort of historical report, a fictional narra- tive and a doxographical account, in which every theoretician’s viewpoint is refuted by the arguments of the next. In this way, Bhaṭṭa Lollaṭa is refuted by Śrī Śaṅkuka, Śrī Śaṅkuka refuted by Bhaṭṭa Tauta and so on, in a crescendo of speculative acumen as well as the sheer number of lines devoted to each thinker. At the end of this “history” or “story” of progressively improving theories, it is Abhinavagupta’s own view that is established as the only correct one, a final view of the ontology and epistemology of the rasa expe- rience. If we take Abhinavagupta’s account at face value, that is, as a kind of do- xographical or, one might even say, historical report, then the model of tex- tual authority being implicitly called upon is clearly the one that McCrea postulates for alaṃkāraśāstra as a whole. In this model, theoretical and even historical novelty is both praised as such and expected as the norm for a knowledge system dealing with a laukika topic. However, as mentioned above, it is legitimate to doubt, at least in principle, the accuracy of Abhina- vagupta’s quotes, or quotation emplotment as I have called it, exactly because it is a bit too neat in its gradual, progressive and almost teleological devel- opment of the argumentation, an argumentation that ends in a kind of specul- ative apotheosis of Abhinavagupta’s own conclusions (siddhānta). It is there- fore fair, although this judgment amounts to nothing more than mere edu- cated guesswork, to assume that Abhinavagupta undertook a certain amount of tweaking and tampering with the material he had at his disposal.31 What is more important, however, the four elegant verses at the end of this quota- tional narrative seem to represent Abhinavagupta’s own judgment of the rationale of his argumentation, from both a historical and theoretical view- point. And they seem to reflect questions of originality, innovativeness and sources of knowledge. However, in order to complicate the discussion further and for the sake of the larger picture drawn at the onset of this chapter, before tackling these verses and trying to gauge their significance I would like to describe two other unusual quotation scenarios that a reader of alaṃkāra texts might stum- ble upon. 31 I am not accusing Abhinava of malignity or having a bad conscience, but I simply accept that theories and arguments are inevitably transformed when reconstructed in any narrative account, especially if the account is aimed at becoming some kind of te- leological narrative. In the words of Tubb and Bronner (2008: 626), “Abhinavagupta’s real purpose in retelling the history of the rasa discussion is to impose upon it a linear narrative in which his own view is the triumphant culmination.” 236 Daniele Cuneo 5 Quotational hoaxes and novelty under siege In order to illustrate what I tentatively call “quotational hoaxes,” as men- tioned above I will briefly examine the Bhāvaprakāśana of Śāradātanaya, a lengthy 13th-century South Indian versified text on both literary and dramatic theory.32 In the words of a recent article by Whitney Cox (2013: 136–137), from which I am heavily borrowing here, Śāradātanaya’s work is “rife with quotations and recastings both acknowledged and unacknowledged, begin- ning with the Nāṭyaśāstra and extending up to Mammaṭa’s Kāvyaprakāśa,” including works “from the literary salon of the Paramāra court at Dhāra (es- pecially the Daśarūpaka and Bhoja’s Śṛṅgāraprakāśa).” Of interest in the present context, within the multi-layered and inherently quotational nature of Śāradātanaya’s work, are a number of pseudo-quotations or pseudepigraphi- cal quotations, as Cox calls them. These are passages from works of known authors such as Bhoja or Mammaṭa whose authorship is however attributed by Śāradātanaya to another source, a source that in turn probably never ex- isted and often seems to have a mythical, semi-Vedic or some kind of au- thoritative authorship. For instance, one of these sources is a certain Yoga- mālāsaṃhitā, attributed to Vivasvat who was instructed by Śiva himself. An- other is a certain Kalpavallī, the supposedly original source of Mammaṭa’s Kāvyaprakāśa. The most remarkable example of a pseudepigraphical quota- tion in the Bhāvaprakāśana among those cited by Cox is a prose passage describing the derivation of rasas from ordinary emotions and their savoring on the part of spectators. The passage is attributed to Bharatavṛddha, “the elder Bharata,” a mythical figure who supposedly predated the Bharata of the Nāṭyaśāstra. As brilliantly recognized by Cox, it is in this very passage that we find Śāradātanaya’s most striking innovation in the theory of rasas.33 In 32 A better known and even more complex example of quotational hoax are the consi- derable number of untraced scriptural quotations found in the works of the 13th-century Dvaita Vedānta philosopher Madhva, usually considered forgeries by the author. For an analysis of this extremely interesting case, see Mesquita 2000 and 2008, as well as the contribution of Okita to the present volume. 33 Although not specifically relevant to our present concern, I will repeat the innovation for its sheer interest: “the idea that the rasa-experience varies depending upon the mental states of the spectator at the moment of reception (i.e., that the rasas are expe- rienced tādātvikamanovṛttibhedabhinnāḥ)” (Cox 2013: 144). The accent on the muta- bility of the aesthetic experience in its singular, personal instances and on account of individual variables is very far from the standard concept of sahṛdaya “ideal connois- seur” and its normative character in both Ānandavardhana and Abhinavagupta (see Masson 1979, Hardikar 1994, Kunjunni Raja 1997 and McCrea 2008: 114–117). Al- though any comparative attempt must be undertaken with due caution, it could be li-  “This is Not a Quote” 237 Cox’s words, Śāradātanaya is “prepared to conceal or to downplay his own innovations and unprecedented combinations by displacing these onto other, invented works,” these often ascribed to supernatural figures (like Śiva) or mythical ur-authors (like Bharatavṛddha). Cox describes this confectioning of textual authorities in the light of the text-historical panorama of scholarship in South India around the 12th century, a period that witnessed a creative explosion of literary works in Tamil as well as the production of numerous anonymous Sanskrit texts, “resulting in whole new canons for different Śaiva, Vaiṣṇava and Śākta religious orders” (Cox 2013: 153). From a more general perspective, however, the abundance of pseudo-quotations in the Bhāvaprakāśana can certainly be considered implicit approval of the model of textual authoritativeness that grants the greatest value to what is transhis- toric and transcendent. One might well argue that the mere mention of human authors would not be enough to empower the quotations with validifying Śāradātanaya’s work, not to mention the case of an innovation championed by the author himself. Thus they were craftily attributed to the mythical Bharatavṛddha. Such a meta-theoretical ascription points in the opposite di- rection from the model postulated by McCrea for alaṃkāraśāstra (pace what Cox seems to argue for at the end of his article), a model nevertheless well supported by many cases from the first centuries of the discipline, examples offered by McCrea himself (especially the self-conscious attitude of being an innovator taken up by Ānandavardhana) and possibly by Abhinvagupta’s quotation emplotment described above. But before attempting to draw a con- clusion, let’s review one last piece of quotational evidence. 6 Unabashed repetition and authorial sleight of hand The last quotational scenario that will be discussed here might be better un- derstood as large-scale borrowing, also dubbed as the phenomenon of “re- peat” (Hugon 2015) ‒ the acknowledged or unacknowledged appropriation of large chunks of earlier textual material in one’s own work. The object of the “repeat” under question is again the core discussion on the ontology, episte- mology and psychology of rasa in Abhinavagupta’s commentary on the rasasūtra. This discussion is borrowed and heavily summarized in the fourth ullāsa of Mammaṭa’s Kāvyaprakāśa (second half of the 11th century), which kened to certain strands of contemporary hermeneutics and aesthetics of reception. (Although unaware of Śāradātanaya’s position at the time, I briefly touched on this is- sue in Cuneo 2006: 156–157.) 238 Daniele Cuneo became the standard manual for poetics in the second millennium. The same passage is then generously quoted, rearranged and rephrased in numerous commentaries on the Kāvyaprakāśa itself, starting with the commentary by Māṇikyacandra, the Saṃketa (late 12th century).34 Similarly, the same passage of Abhinavagupta is appropriated in its entirety in a section of Hemacandra’s sub-commentary, the Viveka, on the second chapter of his own Kāvyānu- śāsana (first half of the 12th century).35 All of these cases of appropriation are marked by more or less minor reworking,36 something worth studying in itself to gauge the theoretical differences between these authors. But as one example, a passage from the auto-sub-commentary on the Kāvyānuśāsana might be briefly analyzed. In order to clarify the often very terse Sanskrit of Abhinavagupta, Hemacandra expands on the text in several sections,37 especially by re-stating the views of the authors at the end of the parts dedicated to them (as in the case of Lollaṭa’s view on rasa) or making obscure or elliptic lines of reasoning more explicit (as in the case of the seven reasons why Śaṅkuka cannot accept Lollaṭa’s view). In doing this, Hemacan- dra expands the text considerably, from a handful of lines to a full page, of- fering examples and textual authorities in support of the reasoning. In re- writing this long passage, Hemacandra also relies on the version of the text as it was summarized and re-elaborated by Mammaṭa less than a century earlier. However, independently from Mammaṭa, Hemacandra also provides addi- tional material on some of the authors Abhinavagupta is allegedly quoting. 34 The passage as summarized and re-elaborated by Mammaṭa is also quoted (“repeated”) or reworked in several other independent works of alaṃkāraśāstra (for instance, in Śiṅgabhūpāla’s Rasārṇavasudhākara, vilāsa II, vṛtti ad 168ab, pp. 251–252) and in various commentaries on poetical texts, as for instance in Sūryadāsa’s Śṛṅgārataraṃ- giṇī commentary on the Amaruśataka (Pintucci 2014: 83–85). I sincerely thank Gaia Pintucci for these useful references. The present survey of the quotational Wirkungs- geschichte of the rasasūtra commentary is quite cursory and very personal. A more comprehensive examination of the issue would be an ideal subject for continuing this study. 35 On the aesthetic thought of Hemacandra in general, see Upadhyay 1987. For a German translation of the first two chapters of the Kāvyānuśāsana (therefore including the pas- sage under discussion), see Both 2003. 36 The lack of trustworthy critical editions for practically all of these texts might be a reason for minor differences in their textual reuse. See, in this regard, Freschi 2015, section 3.1. 37 An interesting counterexample is the above-mentioned (in section 3, as well as section 7) intermezzo in which Abhinavagupta presents his view on the issue of traditionality and novelty – not only does Hemacandra not expand on this, he actually expunges it. However, it is not clear at this stage of my research whether Hemacandra’s omission is due to his disagreeing with Abhinavagupta.  “This is Not a Quote” 239 One example of this is the explicit differentiation between everyday infe- rences and “aesthetic” inferences, reported with regard to Śaṅkuka’s opinion of rasa as a kind of imitation (anukaraṇa). Rasa is an object of an inferential process that is actually based on artificial (kṛtrima) or unreal premises and reasons, i.e., a dramatic performance. But this inferable object is different from those cognized in common inferences, insofar as its nature is enjoyable, just like saliva forming in the mouth due to the sight of someone else tasting an astringent fruit (kaṣāyaphalacarvaṇaparapuruṣadarśanaprabhavamukha- prasekakalanākalpayā). This graphic simile, useful for understanding the proxy-like nature of rasa in Śaṅkuka’s view, is absent in the texts of both Abhinava and Mammaṭa. It is currently impossible to determine whether He- macandra had direct access to the work of Śaṅkuka or other authors, or whether here he was creatively elaborating on the text. To conclude this brief analysis of an episode of quotational “repeat,”38 one might presume that Hemacandra attempted to improve the text he was quoting: while he clearly considers this large textual chunk to be authoritative with regard to the nature of rasa, in no way does he see it as untouchable, as is proved by his active tampering with it. To a contemporary scholar, the text as reported by Hemacandra is considerably clearer than the original by Abhi- navagupta; thus Hemacandra’s reworking seems quite effective and achieves the aim I have implicitly assumed it had. In all its various degrees of liberty towards the repeated texts and in the wide variety of aims one might postulate for it, the phenomenon of “repeat” is decidedly in favor of the “traditionalistic” model of textual authority, usually common to all śāstras, insofar as long quotations from well-estab- lished works can be regarded as the mark of an understanding of history and novelty in line with the well-known Mīmāṃsā model of their theoretical negation (Pollock 1989c). 7 Conclusions: The alternate fortunes of the two paradigms of textual authoritativeness To close this short survey of quotational scenarios, an analysis of the verses by Abhinavagupta mentioned in section 3 might be fruitful for illustrating 38 Cases in which Hemacandra expands, changes or reduces the text of the rasasūtra commentary could be analyzed from various perspectives, but this will have to be the focus of a different study. 240 Daniele Cuneo various attitudes toward textual reuse as well as toward tradition and innova- tion: āmnāyasiddhe kim apūrvam etat saṃvidvikāse ’dhigatāgamitvam | it- thaṃ svayaṃgrāhyamahārhahetudvandvena kiṃ dūṣayitā na lokaḥ || ūrdhvordhvam āruhya yad arthatattvaṃ dhīḥ paśyati śrāntim aveda- yantī | phalaṃ tad ādyaiḥ parikalpitānāṃ vivekasopānaparamparā- ṇām || citraṃ nirālambanam eva manye prameyasiddhau prathamāva- tāram | sanmārgalābhe sati setubandhapurapratiṣṭhādi na vismayāya || tasmāt satām atra na dūṣitāni matāni tāny eva tu śodhitāni | pūrva- pratiṣṭhāpitayojanāsu mūlapratiṣṭhāphalam āmananti || (Abhinava- bhāratī ad Nāṭyaśāstra 6, prose after 31, rasasūtra, vol. 1, p. 277)39 If it is established by tradition, how can it be new? If there is an ex- pansion in knowledge, it already belongs to a received tradition. How can the world not make such critiques by means of a hostile argument about what is knowable by oneself and determined by the highly hon- ored [predecessors]?40 39 To my knowledge, these complex verses have been translated four other times, always differently with regard to the problematic issue of traditionality and novelty (Gnoli 1968: 51–52, Kaviraj 2005: 127, Visuvalingam 2006: 8 and Graheli 2008: 24). For the sake of completeness, these other translations are listed in the appendix. In an earlier paper I already offered a translation of these verses (Cuneo 2013: 50, n. 1); the present rendering is a new attempt at making sense of them. I thank Philipp Mass, Elisa Fre- schi and Vincenzo Vergiani for their help and sagacious suggestions in improving my understanding of this passage, especially the first line. 40 The first verse is the most problematic, since both the meaning of certain words and the subdivisions in the sentences and clauses are far from clear. What seems certain is that common people (lokaḥ) will criticize (dūṣayitā) a behavior, as is laid down implicitly in the first line. Gnoli translates: “Why repeat truths disclosed already in the thought of our predecessor [sic] and thus behave as no one has behaved before? This double, se- rious and evident error will certainly be imputed to me by audience.” He understands the entire first line as a single question, and seems to conflate āmnayasiddhe and saṃ- vidvikāse in “disclosed already in the thought of our predecessors.” Then he renders [a]dhigatāgamitvam as “behave as no one has behaved before,” which is not convin- cing either for the meaning “to behave” or for the insertion of a negation, as he says “no one” (possibly he wants to read adhigata-agamitvam with a short a in the sandhi). For the second line, Gnoli implicitly adds the object “to me,” i.e., Abhinavagupta, which is not in the Sanskrit. Moreover, he renders the difficult compound svayaṃ- grāhyamahārhahetudvandvena as “double (dvandva), serious (mahārha) and evident (svayaṃgrāhya) error (hetu, probably, taken together with the verb dūṣayitā).” In a clearly innovation-oriented interpretation, Kaviraj understands the first line as three separate sentences: “What is new [in this idea]? It is already established in the tra- dition. With the development of the intellect/understanding people grasp [better] what  “This is Not a Quote” 241 they understood earlier.” He does not connect āmnayasiddhe and saṃvidvikāse, and he renders [a]dhigatāgamitvam in cognitive terms. For the second line, he seems to understand svayaṃgrāhyamahārhahetudvandvena as the object of blame: “[the one] who seeks to contradict the precious self-justifying ideas of the tradition?” However it is difficult to find a one-to-one correspondence in the original. Visuvalingam is clearly tradition-oriented in his translation and understands the first line as two pieces of criti- que. I am again not convinced of the rendering of [a]dhigatāgamitvam, for which he gives: “why bother to cram down these stifling canons?” With this he is forced to use the kiṃ twice to introduce both rhetorical questions (see Appendix). Graheli translates: “How can this be anything new, if it was established by tradition? It is just the appre- hension of something already known, albeit within an expanded awareness. Isn’t be- cause of such a conflict, between something readily available and something of great value, that people find faults?” Again [a]dhigatāgamitvam is rendered in cognitive terms, but he is forced to add “albeit” to make better sense of the sentence, which is also possible although not obvious from the Sanskrit. Extremely interesting is his ren- dering of the problematic compound in the second line, but I am not sure what the two elements of the conflict are supposed to be. Elisa Freschi (personal communication) also proposes to understand -dvandva as conflict, but a conflict between the two op- tions given in the first line. Namely, if it is new, it cannot be part of tradition and if there is a development, it already belongs to tradition. I am sympathetic to this reading, although I am not sure in this case how to make good sense of the rest of the com- pound (svayaṃgrāhyamahārhahetu-). In my 2013 paper, I tried to understand the first line as a single sentence “There should be no wonder (kiṃ āpūrvam etad) in following what is already known ([a]dhigatāgamitvam) in the disclosure of knowledge (saṃ- vidvikāse) as established by tradition (āmnayasiddhe).” I am less convinced of this now, and thus have opted here for a critique of the introduction of novelty in the first pāda and, in the second, the reason for that, i.e., the fact that any development in understanding can only obtain for those who already belong to a tradition. However, I also find Graheli’s translation here possibly convincing. For the second line, I unders- tood the term dvandva as “quarrel,” “strife,” as I still do, but I am less convinced now of my previous understanding of the passage as an implicit critique of the world in its hypercritical attitude towards novelty (I had understood an implied object and trans- lated “does not the world spoil [everything]”). As Gnoli and Visuvalingam seem to do, it is also possible to understand dvandva as simply “two,” “a couple,” and thus to trans- late “by means of two (dvandva) reasons (hetu) that are self-evident and valuable,” possibly referring to the claim to novelty per se and the claim to novelty without pre- viously belonging to a tradition. Philipp Maas (personal communication) suggests un- derstanding the first line as two pieces of critique: “If it is established by tradition, how can it be new? If there is an expansion of knowledge, how can it be found (adhigata) to belong to the tradition (-āgamitvam)?” These rhetorical questions would be meant to criticize Abhinava’s enterprise from two perspectives. If he establishes what is already established by tradition, he does not achieve anything new. If he achieves something new, this would be not part of tradition. In this interpretation, the -dvandvena of the second line would refer to this very “pair of opposing arguments.” Even more clearly in this interpretation, the following verse would represent a reply to these critiques. I am quite sympathetic to this reading. However, I still prefer the understanding I chose because it allows for the actual development of knowledge within a tradition, even in 242 Daniele Cuneo Ascending ever higher, the unwearied intellect beholds the truth, which is nothing but the fruit of the succession [of steps] on the ladder of discrimination, as conceived by the ancients.41 Wonderful is, I believe, the first manifestation in the establishment of the knowable, as it is completely supportless. [But], once the right path has been taken, it is no cause for wonder that bridges are built, ci- ties are founded and so on and so forth.42 the mouth of an adversary, which I consider a more plausible stance. Clearly I do not claim to have resolved the difficulties of this passage. But, as I argue below, I find that the complexity, polysemy and ambiguity of this and the following verses have a pro- grammatic and purposeful nature. 41 This verse, representing Abhinavagupta’s reply to the critiques of the first verse, seems to be claiming that knowledge develops in an almost Enlightenment-like way, with the intellect ascending progressively ever higher until it beholds the truth. However, its progression is enabled only by a ladder that consists of the previous doctrines. The ex- act meaning or at least an appropriate rendering of the word paramparā is not clear, as is reflected in the various translations. Gnoli reads: “the doctrines which have suc- ceeded each other on the ladder of thought.” Visuvalingam has: “treading the rungs of discrimination, the conceptual ladder built up by generations of forerunners.” Kaviraj translates: “the succession of intellectual/theoretical steps of the staircase prepared by the scholars of old.” Moreover, he adds an interesting note on the varia lectio “alam” for “phalam”: “This would alter the meaning of the assertion dramatically and suggest a Wittgenstein-like point that once the results have been reached, we can throw away the ladder” (Kaviraj 2005: 140, n. 23). Graheli renders “the fruit of the many theories conceived by former thinkers on the ladder of discrimination.” In any case, the tension between the development of novelty and the centrality of tradition remains part and parcel of this elegant stanza. 42 This verse does not pose too many problems. It simply seems to argue that, after the difficulties of inception, anything becomes easier. As cleverly pointed out by Philipp Maas (personal communication), the use of the term avatāra probably refers to the descent of the knowledge about the theater from Brahmā to mankind, which received it with a fair degree of astonishment. However, both the translation and the relation be- tween the words citra and nirālambana are not obvious. Gnoli understands them on the same level and renders them as “doubtful and vacillating.” Kaviraj relates the one to the other and translates: “It is fascinating [citra], I think, that the first appearance of things seems to be without a prior supporting cause [nirālambana].” Visuvalingam un- derstands them on the same level, as Gnoli does, and freely paraphrases: “Groping in so many directions and, indeed, without a firm foothold …” Graheli, like Kaviraj, un- derstands them as connected (the fact of being nirālambana is citra) and renders citra with “strange.” I definitely opt for correlating and subordinating the two adjectives, but translate citra as “wonderful.” The word citra does have different and opposite nu- ances, and I am not at all adamant in my choice of the positive “wonderful” in the face of the more cautions “strange.” As Elisa Ganser has suggested to me (personal commu- nication), one more meaning of citra is “varied,” “multifarious,” and the phrase might  “This is Not a Quote” 243 Therefore, here, I do not refute but refine the views of the wise, which hand down the result of the root-foundation in constructions that were erected in the past.43 My understanding of these verses is far from final, and my feeling is that their meaning is bound to remain at least partly ambiguous. Moreover, my contention is that this ambiguity might be intentional and programmatic. I believe that Abhinavagupta was possibly trying to find a viable in-between path that could satisfy both those who upheld a “traditional” view and who upheld an innovation-oriented view. In other words, he was possibly trying to reconcile two very different models, one of textual authority and the other of positive evaluation of novel knowledge. Along these lines of interpretation, he seems to be, on one hand, following the idea that the only source of under- standing is in the transhistoric past of the onset of traditional lore, represented in this case by the work of Bharata and its infallible description-cum-pre- scription of the dramatic arts and their workings. However, on the other hand, therefore refer to the “varied first crossing in the ascertainment of the knowable,” i.e., the multiplicity of the opinions of previous commentators. 43 The last verse is comparatively easier. Abhinavagupta claims that he is only refining the theories of his predecessors. However, the close of the construction metaphor is not crystal clear. Gnoli changes the metaphor and underlines the traditionality of Abhina- va’s enterprise by speaking of “the harvest” of thought that the predecessors have left us as a legacy. In contrast, in an interpretation that favours novelty over tradition, Ka- viraj seems to understand the term yojanā as “bringing coherence” in reference to what has been previously established and posits “the establishment of entirely new truths/ foundations” – I am not sure what Sanskrit terms he is translating – as its result. In a clearly tradition-oriented but markedly interpretive and free rendering, Visuvalingam speaks of “the blueprints bequeathed by our predecessors,” in which it is possible to re- cognize “the foundations of this crowning achievement” of Abhinavagupta. But he does not respect the syntax of the verse (āmananti is rendered as a first person singular, it seems). Graheli respects the architectural metaphor and follows the text very closely. Thus the views of the wise “pass down a fruit whose support is rooted in formerly sup- ported theories.” There might be some difficulty in connecting the word mūla in the compound with the preceding pūrvapratiṣṭhāpitayojanāsu, although sāpekṣasamāsas are commonly accepted, and I would rather stick even more closely to the metaphor and understand the word yojanā as some kind of building. In any case, Graheli’s trans- lation is the one closest to mine, in which I have tried to give a very plain rendering. Moreover, I would argue that the word mūla is a clear reference to the Nāṭyaśāstra and that the constructions are the various commentaries. Therefore, Abhinavagupta might be suggesting a quasi-archeological feat, as he is going back to the “authentic” fruit of the original foundation preserved as a part of later constructions, like Roman capitals in medieval churches. I am arguing for an interpretation that mediates between the novel- ty-oriented one and the tradition-oriented one, since I believe Abhinavagupta was trying to find a sort of perfect middle ground. 244 Daniele Cuneo he seems also to accept a more novelty-oriented meta-theory that permits and possibly values development and increase in knowledge and understanding thanks to a chain of progressively more refined thinkers. To return to the main issue at stake ‒ the cultural history of alaṃkāra- and nāṭyaśāstra and their relation to novelty and tradition ‒ I argue that we should examine this question with the same flexibility, cautiousness and attention to multivocality that I tentatively attribute to Abhinavagupta in these verses. As far as the evidence goes, it seems that in both the self-repre- sentation and the practice of literary and dramatic theory two competing un- derstandings of tradition and innovation and two parallel ways of dealing with the reuse of earlier materials coexisted and were therefore in a conti- nuous, dynamic interrelation. It is thus not at all surprising to find numerous examples of both approaches, both traditionalistic and non-traditionalistic. Also unsurprising is Abhinavagupta’s attempt, at least in my interpretation, to bridge the gap between the two tendencies and to reconcile them in a unified semi-historical narrative. Moreover, contrary to the early centuries of alaṃkāraśāstra when the lau- kikatva model argued by McCrea seems to have been common within the conflicts between the various authors, and contrary to the centuries between the 9th and the 11th, when the mostly Kashmirian thinkers vied fiercely with each other for establishing the correctness of ever newer theories in accor- dance with what I call the “dismantling tradition” strategy,44 I would argue that after Abhinavagupta’s momentous contribution and after the composition of Mammaṭa’s Kāvyaprakāśa, the traditionalistic tendency seems to have gained the upper hand. Indeed, Mammaṭa’s work virtually assumed the role of a root text (mūla) – for which the incredible number of commentaries45 on this text are ample evidence – and Abhinavagupta’s understanding of artistic epistemology and ontology often became the norm, albeit with numerous 44 It is interesting to note that it is exactly in 9th-century Kashmir (and with Udbhaṭa) that the two fields of knowledge start to intersect and merge into a single system of knowl- edge, i.e., poetics-cum-dramaturgy (alaṃkāra- and nāṭyaśāstra taken together, which might be called sāhityaśāstra, although this term has not been used emically very of- ten). It seems safe to postulate that the convergence of theories and practices coming from the different domains of poetry and drama was one of the theoretical causes for the blossoming of philosophical speculation. A possible avenue of research is the hy- pothesis that authors hailing from Kashmir had a more novelty-oriented stance and the rest of South Asian authors, a more tradition-oriented one, but this geographical typol- ogy is currently only educated guesswork. 45 It is “the most often commented upon śāstra text in Sanskrit literature” (Cahill 2001: 23). For an idea of the number of commentaries, see the indeed long but still non-ex- haustive list in Cahill 2001: 23–37.  “This is Not a Quote” 245 exceptions.46 But then again, a general caveat must be stated. Many theoreti- cians of the second millennium kept writing independent treatises that chal- lenged any strictly univocal interpretation of the tradition, although there were those who did give in to the traditionalistic, transhistoric model of tex- tual authority, as for instance the aforementioned case of Śāradātanaya’s Bhāvaprakāśana. Therefore, the two models of coping with novelty, the “no- velty-smuggling strategy” and the “tradition-dismantling strategy,” seem to have enjoyed alternating fortunes. This followed, as evinced above, recogniz- able patterns – for instance, the prevalence of the traditionalistic view in the second millennium and in texts more strictly related to the tradition of nāṭya- śāstra. But there were also unpredictable cases. As I have shown, a potential reason, although probably not the only one, for the alternating in history of two epistemic modes and models of innovation and preservation of cultural legacy is the fusion of the two śāstras in question, with their very different meta-theoretical pedigrees. Further research in this direction remains a desideratum. As a last remark, to offset the meta-theoretical tendency prevalent in our contemporary world, i.e., the anti-traditionalistic model that treasures novel- ty, originality and authoriality over anything else, I would like to draw atten- tion to what I consider an evident rhetorical advantage – or maybe the ad- vantage – of the “novelty-smuggling” strategy over the “tradition-dismantl- ing” one. In the agonistic realm of any cultural discourse, the “novelty- smuggling” strategy offers a possibility for exploiting the accepted authorita- tiveness and trustworthiness of traditional knowledge – that is, its status as a paradigmatic abode of truth – as a device for validating and legitimizing something new. Furthermore, to disguise innovative theories, “traditionality” can be employed to advantage as a defense against the allegedly baseless and dangerous nature of novelty as such. Traditionalistic or tradition-oriented views have often, and often correctly, been regarded as conservative and prone to cultural and political fundamen- talism. Moreover, any claim of truth is also a claim of power, and therefore any meta-theory of validation – be it traditionalistic or not – runs the risk of being used to shut the intellectual field and freeze both the cultural discourse and the power structures that inform it, to the deep detriment of the subal- tern.47 46 For instance, consider the famous works of Bhānudatta (15th century), recently trans- lated in Pollock 2009 for the Clay Sanskrit Library. Otherwise, consider the new school of alaṃkāraśāstra as described in Bronner 2002 and 2004, in which self-con- scious innovation comes back in fashion. 47 This politically flavored conclusion might seem out of place at the end of a discussion 246 Daniele Cuneo Appendix: Four translations of Abhinavagupta’s intermezzo Gnoli (1968: 51–52): “Why repeat truths disclosed already in the thought of our predecessor [sic] and thus behave as no one has behaved before? This double, serious and evident error will certainly be imputed to me by au- dience. Tireless, the mind of man climbs ever higher to gaze on truth. This is just the fruit of the doctrines which have succeeded each other on the ladder of thought. In the beginning, the crossing of the river of the knowable is, I know, agitated and supportless: but as we advance doggedly along this road, we cease to be amazed by built bridges, city foundations, or anything else. A rich and fruitful harvest may be culled by posterity from the inheritance of thought left to it by predecessors. Thus the doctrines of the sages of antiquity will only be refined by us here and not refuted.” Kaviraj (2005: 127): “What is new [in this idea]? It is already established in the tradition. With the development of the intellect/understanding people grasp [better] what they understood earlier. Otherwise, would people not blame who seeks to contradict the precious self-justifying ideas of the tradi- tion? That the intellect, never flagging, constantly rises upwards and under- stands the meaning of theories/truths – is not that the very fruit of the succes- sion of intellectual/theoretical steps of the staircase prepared by the scholars of old? It is fascinating, I think, that the first appearance of things seems to be without a prior supporting cause, yet once the proper way is found, it is not surprising that bridges can be built and cities constructed. Therefore, I have here not found fault with the ideas of these good (earlier) thinkers, but only refined them. They say that in bringing coherence to the views established earlier, the result is similar to the establishment of entirely new truths/foun- dations.” Visuvaligam (2006: 8): “When it has been already established by tradi- tion, why these pretentious claims to originality? When self-conscious thought blossoms so freely on its own, why bother to cram down these stifl- ing canons? With these two objections, ever so precious and within easy reach, what’s then left that this world has not turned to derision? Climbing ever higher and higher, knowing no repose, the intellect finally perceives the on a seemingly non-political field of knowledge, poetics-cum-dramaturgy, and the shifts in epistemic stances on normative validation in its cultural history. However, the highly social, moral and political nature inherent in the normative nature of discussions on matters of aesthetic taste has often been shown in contemporary theory (see, for in- stance, Bourdieu 1996 and 2003, and Rancière 2004) as well as in contemporary scho- larship on alaṃkāraśāstra and nāṭyaśāstra (see Pollock 2001, Leavitt 2011, and Ganser and Cuneo 2012).  “This is Not a Quote” 247 truth of things. This is the reward of treading the rungs of discrimination, the conceptual ladder built up by generations of forerunners. Groping in so many directions and, indeed, without a firm foothold, such I say, is our first plunge into the ocean of certain knowledge. Once the right path has been found and cleared, building bridges and founding entire cities, such architectural feats are no cause for wonder. Therefore, far from having been overturned and demolished here, the views of fellow truth-seekers have been merely refined. In the blueprints bequeathed by our predecessors, we recognize the founda- tions of this crowning achievement of our own labors!” Graheli (2008: 24): “How can this be anything new, if it was established by tradition? It is just the apprehension of something already known, albeit within an expanded awareness. Isn’t because of such a conflict, between something readily available and something of great value, that people find faults? Climbing higher and higher, the restless intellect observes reality, which is the fruit of many theories conceived by former thinkers on the lad- der of discrimination. Indeed, what I find strange is that the first approach in the ascertainment of the object of knowledge can be groundless, while to build bridges and cities – once the right path has been determined – is not a reason of surprise. Therefore, here the opinions of wise people have not been censured, but rather improved, because they pass down a fruit whose support is rooted in formerly supported theories.” References Primary Sources Bharatabhāṣya of Nānyabhūpāla, Part I, ed. Chaitanya P. Desai. Khairagarh: Indira Kala Sangit Vishwavidyalaya, 1961. Bharatabhāṣya of Nānyabhūpāla, Part II, ed. Chaitanya P. Desai. Khairagarh: Indira Kala Sangit Vishwavidyalaya, 1976. Bhāvaprakāśana by Śāradātanaya, ed. K. S. Ramaswamy. Baroda: Oriental Institute, 1930. Gaekwad’s Oriental Ser. 45. Daśarūpaka by Dhanañjaya, with the Commentary Avaloka by Dhanika, and the Sub-commentary Laghuṭīkā by Bhaṭṭanṛsiṃha, ed. T. Venkatacharya. Madras: Adyar Library and Research Centre, 1969. Dhvanyālokalocana by Abhinavagupta, ed. Pattābhirāma Śāstrī. Benares: Chowkhamba Sanskrit Ser. Office, 1940. Kashi Sanskrit Ser. 135. Kalpalātāviveka, ed. M. L. Nagar and H. Shastri, with an introduction by P. R. Vora. Ahmedabad: Lalbhai Dalpatbhai Institute of Indology, 1968. Kāvyānuśāsana of Hemacandra, ed. R. C. Parikh and V. M. Kulkarni. 2nd rev. ed. Bombay: Śrī Mahāvīra Jaina Vidyālaya, 1964. 248 Daniele Cuneo Kirātārjunīya by Bhāravi with the Commentary of Mallinātha, ed. Jibananda Vidya- sagara. Calcutta: Jibananda Vidyasagara, 1875. Nāṭyaśāstra of Bharatamuni with the Commentary Abhinavabhāratī by Abhinava- guptācārya, ed. M. R. Kavi. Baroda: Oriental Institute, 1926–1960. Gaekwad’s Oriental Ser. Rasārṇavasudhākara of Śiṅgabhūpāla, Crit. Ed. with Introduction and Notes by T. Venkatacharya. Adyar, Madras: Adyar Library and Research Centre, 1979. Śṛṅgāraprakāśa of Bhoja, ed. V. Raghavan. Cambridge: Harvard University Press, 1998. Secondary Sources Assmann and Gladigow 1995 = Assmann, Jan and Burkhard Gladigow, eds. Text und Kommentar: Archäologie der literarischen Kommunika- tion IV. München: Wilhelm Fink, 1995. Bansat-Boudon 2004 Bansat-Boudon, Lyne. Pourquoi le Théâtre? La réponse indienne. Paris: Summulae, 2004. Bhattacarya K. 1972 Bhattacarya, Kamaleswar. “Śāntarasa et Advaita.” Journal Asiatique 160 (1972): 89–105. Bhattacharya S. P. 1976 Bhattacharya, Shashthi Prasad. Śānta Rasa and Its Scope in Literature. Calcutta: Sanskrit College, 1976. Both 2003 Both, Leo. Hemacandras Kāvyānuśāsana. Kapitel 1 und 2. Wiesbaden: Harrassowitz, 2003. Bourdieu 1996 Bourdieu, Pierre. The Rules of Art: Genesis and Structure of the Literary Field, trans. Susan Emmanuel. (French edition: Les règles de l’art: genèse et structure du champ littéraire. Paris: Seuil, 1992.) Cambridge: Polity Press, 1996. Bourdieu 2003 — Méditations pascaliennes. 3rd ed. (English edition: Pas- calian Meditations. Stanford: Stanford University Press, 2000). Paris: Éditions du Seuil, 2003. Bronner 2002 Bronner, Yigal. “What is New and What is Navya: Sans- krit Poetics on the Eve of Colonialism.” Journal of Indian Philosophy 30 (2002): 441–462. Bronner 2004 — “Back to the Future: Appaya Dīkṣita’s Kuvalayānanda and the Rewriting of Sanskrit poetics.” Wiener Zeitschrift für die Kunde Südasiens 48 (2004): 47–79. Bronner 2012 — “A Question of Priority: Revisiting the Bhāmaha- Daṇḍin Debate.” Journal of Indian Philosophy 40 (2012): 67–118. Bronner, Cox and McCrea 2011 = Bronner, Yigal, Whitney Cox and Lawrence McCrea, eds. South Asian Texts in History. Critical En- gagements with Sheldon Pollock. Ann Arbor, MI: Associa- tion for Asian Studies, 2011.  “This is Not a Quote” 249 Cahill 2001 Cahill, Timothy. An Annotated Bibliography of the Alaṃ- kāraśāstra. Leiden: Brill, 2001. Chenet 1998 Chenet, François. “Le commentaire en Inde.” In: Mattéi 1998: 1656–1664. Cox 2013 Cox, Whitney. “From Source-criticism to Intellectual His- tory in the Poetics of the Medieval Tamil Country.” In: Cox and Vergiani 2013: 115–160. Cox and Vergiani 2013 Cox, Whitney and Vincenzo Vergiani, eds. Bilingual Dis- course and Cross-Cultural Fertilisation: Sanskrit and Tamil in Medieval India. Pondicherry: Institut français de Pondichéry / École française d'Extrême-Orient, 2013. Cuneo 2006 Cuneo, Daniele. “Linguisticità ed esperienza estetica in Abhinavagupta e Gadamer.” Rivista di Studi Sudasiatici 1 (2006): 139–160. Cuneo 2008–2009 — “Emotions without Desire. An Interpretive Appraisal of Abhinavagupta’s Rasa Theory. Annotated Translation of the First, Sixth and Seventh Chapters of Abhinavagupta’s Abhinavabhāratī.” Diss. “Sapienza” University of Rome 2008–2009. Cuneo 2013 — “Unfuzzying the Fuzzy. The Distinction between rasas and bhāvas in Abhinavagupta and Bharata.” In: Mirnig, Szántó and Williams 2013: 49–75. Cuneo 2016 — “Detonating or Defusing Desire. From Utpaladeva’s Ecstatic Aesthetics to Abhinavagupta’s Ecumenical Art Theory.” In: Torella and Bäumer 2016: 31–76. Dallapiccola 1989 Dallapiccola, Anna L., ed. Shastric Traditions in Indian Arts. Wiesbaden: Harrassowitz, 1989. David 2015 David, Hugo. “Human Action as Seen in Early Medieval Indian Philosophical texts (600–1000): Activity, Speech and Desire.” The Journal of Value Enquiry 49.4 (2015): 567–595. David 2016 — “Time, Action and Narration. On Some Exegetical Sources of Abhinavagupta’s Aesthetic Theory.” Journal of Indian Philosophy 44.1 (2016): 125–154. De 1960 De, Sushil Kumar. History of Sanskrit Poetics. 2nd rev. ed. Calcutta: Firma K. L. M., 1960. Freschi 2015 Freschi, Elisa. “The Reuse of Texts in Indian Philosophy: Introduction.” Journal of Indian Philosophy 43.2–3 (2015): 219‒255. Ganeri 2010 Ganeri, Jonardon. “Sanskrit Philosophical Commentary.” Journal of the Indian Council of Philosophical Research 25.1 (2010): 187–207. 250 Daniele Cuneo Ganser and Cuneo 2012 Ganser, Elisa and Daniele Cuneo. “The Actor’s Social Sta- tus and Agency. Fame or Misery?” Cracow Indological Studies 14 (2012): 87–131. Gerow 1977 Gerow, Edwin. Indian Poetics. Wiesbaden: Harrassowitz, 1977. A History of Indian Literature 5.2.3. Gerow 1994 — “Abhinavagupta’s Aesthetics as a Speculative Para- digm.” Journal of the American Oriental Society 114 (1994): 186–208. Gerow and Aklujkar 1972 Gerow, Edwin and Ashok Aklujkar. “On Śānta Rasa in Sanskrit Poetics.” Journal of the American Oriental Society 92 (1972): 80–87. Gnoli 1962 Gnoli, Raniero. Udbhaṭa’s Commentary on the Kāvyālaṃ- kāra of Bhāmaha. Roma: Istituto italiano per il Medio ed Estremo Oriente, 1962. Gnoli 1968 — The Aesthetic Experience According to Abhinavagupta. 2nd ed. Varanasi: Chowkhamba Sanskrit Ser. Office, 1968. Chowkhamba Sanskrit Studies 62. Gopalakrishnan 2006 Gopalakrishnan, Sudha. “Abhinavagupta’s Interpretation of Bharata’s Rasa-Sutra: A Translation.” In: Paranjape and Visuvalingam 2006: 130–143. Goulet-Cazé 2000 Goulet-Cazé, Marie-Odile, ed. Le commentaire entre tra- dition et innovation. Actes du Colloque International de l’Institut des Traditions Textuelles. Paris et Villejuif 22–25 septembre 1999. Paris: Vrin, 2000. Bibliothèque d’histoire de la philosophie. Nouvelle série. Graheli 2008 Graheli, Alessandro. “In Praise of Repetition.” IIAS News- letter 48 (Aug. 2008). Guolo 1996 Guolo, Renzo. Il paradosso della tradizione. Religioni e modernità. Milano: Guerini e Associati, 1996. Hardikar 1994 Hardikar, Avadoot R. “The Aesthetic Appreciator or sahṛ- daya.” Annals of the Bhandarkar Oriental Research Insti- tute 75 (1994): 265–272. von Hinüber 2007 von Hinüber, Oskar. “Buddhistische Kommentare aus dem alten Indien. Die Erklärung des Theravāda-Kanons.” In: Quisinsky and Walter 2007: 99–114. Hobsbawm and Ranger 1983 = Hobsbawm, Eric and Terence Ranger. The Invention of Tradition. Cambridge: Cambridge University Press, 1983. Hugon 2015 Pascale Hugon. “Text Re-use in Early Tibetan Episte- mological Treatises.” Journal of Indian Philosophy 43.2–3 (2015): 453–91. Hulin 2000 Hulin, Michel. “Le commentaire dans la littérature philo- sophique de l’Inde ancienne.” In: Goulet-Cazé 2000: 425– 434.  “This is Not a Quote” 251 Ingalls 1990 Ingalls, Daniel H. H., ed. The Dhvanyāloka of Ānandavar- dhana with the Locana of Abhinavagupta, transl. D. H. H. Ingalls, J. M. Masson, and M. V. Patwardhan, London: Harvard University Press, 1990. Harvard Oriental Ser. 49. Kane 1961 Kane, Pandurang Vaman. History of Sanskrit Poetics. 3rd rev. ed. Delhi: Motilal Banarsidass, 1961. Kaviraj 2005 Kaviraj, Sudipta. “The Sudden Death of Sanskrit Knowl- edge.” Journal of Indian Philosophy 33 (2005): 119–142. Kuhn 1970 Kuhn, Thomas Samuel. The Structure of Scientific Revo- lution. 2nd enl. ed. Chicago: University of Chicago Press, 1970. Kunjunni Raja 1997 Raja, K. Kunjunni. “Sahṛdaya: The Ideal Art-Connois- seur.” In Lienhard and Piovano 1997: 419–423. Leavitt 2011 Leavitt, Guy. “The Social in Kashmiri Aesthetics: Sug- gesting and Speciously Savoring Rasa in Ānandavardhana and Abhinavagupta.” In Bronner, Cox and McCrea 2011: 267–289. Lienhard and Piovano 1997 = Lienhard, Siegfried, and Irma Piovano, eds. Lex et Litterae. Studies in Honour of Professor Oscar Botto. To- rino: Edizioni dell’Orso, 1997. Masson 1979 Masson, Jeffrey Moussaieff. “The Sahṛdaya.” In: Sinha 1979: 313–316. Masson and Patwardhan 1969 = Masson, Jeffrey Moussaieff and M. V. Patwardhan. Śāntarasa and Abhinavagupta’s Philosophy of Aesthetics. Poona: Bhandarkar Oriental Research Institute, 1969. Bhandarkar Oriental Ser. 9. Masson and Patwardhan 1970 = — Aesthetic Rapture. 2 vols. Poona: Deccan College, 1970. Mattéi 1998 Mattéi, Jean-François, ed. Encyclopédie philosophique universelle, IV: Le discours philosophique. Paris: Presses Universitaires de France, 1998. McCrea 2008 McCrea, Lawrence J. The Teleology of Poetics in Medie- val Kashmir. Cambridge, MA: Harvard University Press, 2008. Harvard Oriental Ser. 71. McCrea 2011 — “Standards and Practices: Following, Making, and Breaking the Rules of Śāstra.” In: Bronner, Cox and McCrea 2011: 229–239. Mesquita 2000 Mesquita, Roque. Madhva’s Unknown Literary Sources: Some Observations. New Delhi: Aditya Prakashan, 2000. Mesquita 2008 — Madhva’s Quotes from the Purāṇas and the Mahā- bhārata: An Analytical Compilation of Untraceable Source-Quotations in Madhva’s Works along with footnotes. New Delhi: Aditya Prakashan, 2008. 252 Daniele Cuneo Mirnig, Szántó and Williams 2013 = Mirnig, Nina, Péter-Dániel Szántó and Michael Williams, eds. Puṣpikā: Tracing Ancient India through Texts and Traditions. Contributions to Current Research in Indology. Vol. I. Oxford: Oxbow Books, 2013. Pandey 1944 Pandey, Kanti Candra. “Dhanañjaya and Abhinavagupta on Śāntarasa.” Proceedings of the All India Oriental Con- ference 12 (1994): 326–330. Paranjape and Visuvalingam 2006 = Paranjape, Makarand and Sunthar Visuvalingam, eds. Abhinavagupta: Reconsiderations, New Delhi: Sam- vad India Foundation, 2006. Pintucci 2014 Pintucci, Gaia. “The Western Recension of the Amaruśa- taka. A Critical Edition and Translation of Sūryadāsa’s Śṛṅgārataraṃgiṇī (stanzas 1 to 3) in Comparison to Arjunavarman’s Rasikasaṃjīvanī.” MA thesis. University of Hamburg 2014. Pollock 1985 Pollock, Sheldon. “The Theory of Practice and the Prac- tice of Theory in Indian Intellectual History.” Journal of the American Oriental Society 105.3 (1985): 499–519. Pollock 1989a — “The Idea of Śāstra in Traditional India.” In: Dalla- piccola 1989: 17–26. Pollock 1989b — “Playing by the Rules: Śāstra and Sanskrit Literature.” In: Dallapiccola 1989: 301–312. Pollock 1989c — “Mīmāṃsā and the Problem of History in Traditional India.” Journal of the American Oriental Society 109.4 (1989): 603–610. Pollock 1998 — “Bhoja’s Śṛṅgāraprakāśa and the Problem of rasa. A Historical Introduction and Annotated Translation.” Asia- tische Studien/Études Asiatiques 52.1 (1998): 117–192. Pollock 2001 — “The Social Aesthetic and Sanskrit Literary Theory.” Journal of Indian Philosophy 29 (2001): 197–229. Pollock 2009 — ed., trans. “Bouquet of Rasa” and “River of Rasa” by Bhānudatta. New York: New York University Press, JJC Foundation, 2009. Clay Sanskrit Library. Pollock 2010a —, ed. Epic and Argument in Sanskrit Literary History. Essays in Honor of Robert P. Goldman. New Delhi: Manohar, 2010. Pollock 2010b — “What was Bhaṭṭa Nāyaka Saying? The Hermeneutical Transformation of Indian Aesthetics.” In: Pollock 2010a: 143–184. Quisinsky and Walter 2007 = Quisinsky, Michael and Peter Walter, eds. Kommentar- kulturen: Die Auslegung zentraler Texte der Weltreligio- nen. Ein vergleichender Überblick. Köln: Böhlau, 2007. Raghavan 1967 Raghavan, Venkataraman. The Number of Rasas. Madras: Adyar Library and Research Centre, 1967.  “This is Not a Quote” 253 Raghavan 1978 — Bhoja’s Śṛṅgāraprakāśa. Madras: Punarvasu, 1978. Rancière 2004 Rancière, Jacques. The Politics of Aesthetics. London: Continuum, 2004. Roodbergen 1984 Roodbergen, Jouthe Anthon Fokko. Mallinātha’s Ghaṇṭā- patha on the Kirātārjunīya, I–VI. Part 1: Introduction, Translation and Notes. Leiden: E. J. Brill, 1984. Sinha 1979 Sinha, J. P., ed. Ludwik Sternbach Felicitation Volume. Lucknow: Akhila Bharatiya Sanskrit Parishad, 1979. Slaje 2007 Slaje, Walter. “Der Sanskrit-Kommentar.” In: Quisinsky and Walter 2007: 69–97. Slaje 2008 —, ed. Śāstrārambha. Inquiries into the Preamble in San- skrit. Wiesbaden: Otto Harrassowitz, 2008. Squarcini 2005 Squarcini, Federico, ed. Boundaries, Dynamics and Con- struction of Traditions in South Asia. Firenze: Firenze University Press, 2005. Squarcini 2008 — Tradens, Traditum, Recipiens. Studi storici e sociali sull’istituto della tradizione nell’antichità sudasiatica. Fi- renze: Società Editrice Fiorentina, 2008. von Stietencron 1995 Stietencron, Heinrich von. “Typisierung und Sitz im Le- ben: Anmerkungen zum Kommentar in Indien.” In Ass- mann and Gladigow 1995: 249–255. Torella and Bäumer 2016 = Torella, Raffaele and Bettina Bäumer, eds. Utpaladeva: Philosopher of Recognition. Shimla: Indian Institute of Advanced Studies. Tubb 2008 Tubb, Gary A. “Philosophical Beginnings in Sanskrit Treatises on Poetics.” In: Slaje 2008: 171–182. Tubb and Boose 2007 Tubb, Gary A. and Emery R. Boose. Scholastic Sanskrit. A Manual for Students. New York: The American Institute of Buddhist Studies, 2007. Tubb and Bronner 2008 Tubb, Gary A. and Yigal Bronner. “Vastutas tu: Method- ology and the New School of Sanskrit Poetics.” Journal of Indian Philosophy 36 (2008): 619–632. Visuvalingam 2006 Visuvalingam, Sunthar. “Towards an Integral Appre- ciation of Abhinavagupta’s Aesthetics of Rasa.” In: Paran- jape and Visuvalingam 2006: 7–55. Upadhyay 1987 Upadhyay, Amrut M. The Kāvyānuśāsana of Acharya He- macandra: A Critical Study. Ahmedabad: Amrut M. Upa- dhyay, 1987.  Quotation, Quarrel and Controversy in Early Modern South Asia: Appayya Dīkṣita and Jīva Gosvāmī on Madhva’s Untraceable Citations* Kiyokazu Okita Introduction In an important paper published in 2012, Elisa Freschi effectively establishes the significance of what we might call “Quotation Studies,” an area of re- search that is still quite underexplored. Freschi points out that among the many benefits the study of quotations can yield, it can reveal the way in which authors understood authority in their traditions. In this context Freschi mentions an exciting case of the citations used by Madhva to validate his own view, citations that are however untraceable. As Freschi points out, Madhva’s quotes are quite controversial because many of the works and passages he quotes are found exclusively in his own works. For example, he cites a passage and attributes it to a certain text such as the Caturvedaśikhā, a text of which no one has ever heard. Or after citing a verse, he states iti varāhe “thus it is said in the Varāha-Purāṇa,” but the ∗ I thank Prof. Harunaga Isaacson (University of Hamburg) for reading a section of Appayya Dīkṣita’s Madhvatantramukhamardana and its commentary with me. I am also greatly indebted to the detailed and insightful comments from the editors Dr. Elisa Freschi (IKGA, Austrian Academy of Sciences) and Dr. Philipp Maas (University of Leipzig). Earlier versions of this chapter were first read at the Deutscher Orientalis- tentag 2013 on 24 September 2013, then at the Oxford Centre of Hindu Studies on 22 May 2014, as well as at the Faculty of Indological Studies, Kyoto University on 3 June 2014. The current chapter was revised based on the comments I received on these occasions. I acknowledge contributions from the following colleagues: Dr. Rem- bert Lutjeharms (University of Oxford), Lucian Wong (University of Oxford), Dr. Jes- sica Frazier (University of Kent), Prof. Diwakar Acharya (Kyoto University), Prof. Somdev Vasudeva (Kyoto University) and Prof. Yuko Yokochi (Kyoto University). 256 Kiyokazu Okita verse cannot be found in any of the editions of the Purāṇa currently availa- ble. Or he simply writes iti ca, without mentioning a source for his citation. The topic of Madhva’s untraceable quotes has been systematically and extensively explored by the Indologist Roque Mesquita, who argues that most of Madhva’s untraceable quotes are not actual quotes but his own creation. This claim has received considerable criticism from scholars belonging to the Mādhva tradition. In the present chapter, I shall first briefly describe the his- tory and the nature of the modern controversy concerning Madhva’s quotes. Then I will explore the relevance of this controversy for Purāṇic studies and the study of Vedānta as Hindu Theology. The main part of this article, how- ever, consists of an exploration of the writings of two important Hindu theo- logians of the 16th century, namely Appayya Dīkṣita and Jīva Gosvāmī, who held opposing views with regard to Madhva’s quotes. While Appayya rejects the validity of Madhva’s untraceable quotes to refute the latter’s Dvaita posi- tion, Jīva refers to the same quotes to validate his own Gauḍīya viewpoint. By examining the respective positions of these two authors, I hope to present the complexity of this topic, which in my view has not been fully addressed in Mesquita’s works. 1 The modern controversy: Mesquita vs. Sharma As Ludo Rocher (2008: 603f.) points out, Mesquita was not the first modern scholar to discuss the issue of Madhva’s untraceable quotes. Starting with Suzanne Siauve, a number of scholars, including Betty Heimann, Helmuth von Glasenapp, R. G. Bhandarkar, V. S. Ghate and A. Venkatasubbiah, no- ticed and addressed the intriguing nature of Madhva’s quotes. Mesquita, however, was the first scholar to examine the issue in a systematic manner. In his monograph, first published in 1997,1 Mesquita discusses Madhva’s untra- ceable quotes in detail and points out that these quotes are not only untracea- ble but that they also tend to support Madhva’s unique teachings such as the doctrines of the five-fold difference (pañcabheda), the differentiating capac- ity (viśeṣa) and the ontological hierarchy (tāratamya). On the basis of Ap- payya Dīkṣita’s texts, Mesquita also examines Madhva’s claim of being the third avatāra of Vāyu, which enabled him to justify his citation of untracea- ble quotations. Based on these observations, Mesquita concludes that Madhva himself authored these untraceable quotes to support his arguments. 1 The monograph was subsequently translated into English. See Mesquita (2000).  Quotation, Quarrel and Controversy 257 Mesquita’s work was severely criticized by a number of Mādhva scholars who believe that Madhva’s untraceable quotes indeed are citations of real works, works that are presently simply not available or lost. B. N. K. Sharma has been most vocal on this. In a paper published in 2001, Sharma advises Mesquita: To err is human. Even supposing that the Professor’s [i.e., Mesquita’s] charges against Madhva are due purely to errors of judgment, their cumulative effect may well have its own adverse repercussions. It would therefore be advisable for the Professor to withdraw his char- ges, apologize to Madhva and close the chapter, for good once for all (34). One may note the harsh overtone in Sharma’s charge. In response, Mesquita points out in his publication in 2007 that Sharma’s defense of Madhva’s untraceable quotes is weak. For example, Sharma writes: The Gita says God descends on earth in all Yugas and surely some of the gods too do so with Him […] [W]hy disbelieve the ability of a great thinker like M [i.e., Madhva] to be able to recapture lost sakhas by his Yogic spiritual power, centuries ago, for the benefit of poster- ity? (Sharma 2001: 21, quoted in Mesquita 2007: 27) This passage shows that Sharma bases his academic arguments on religious beliefs. Regardless of whether one believes in Madhva’s supernatural power, the very fact that Madhva quotes thousands of verses supporting his unique teachings, verses that occur only in his own works, appears to indicate that Madhva authored these verses himself.2 2 Untraceable quotes and Purāṇic studies While Mesquita’s conclusions concerning Madhva’s untraceable quotes are generally convincing, I believe there are important issues that Mesquita has not fully addressed. For example, the issue of untraceable quotes is related to the lack of reliability in the currently available printed editions of Sanskrit texts, of which many have not been critically edited on the basis of the va- 2 Govindacharya (1997) says he discovered a text called the Maitreyeni Upaniṣad, which authors such as Vidyāraṇya considered a fabrication of Madhva. If this is the case, it indicates that at least some of Madhva’s quotes that are currently untraceable might be traced in future. Unfortunately, I have not been able to examine the authenticity of this Upaniṣad. I thank Dr. Ravi Gupta for having drawn my attention to this reference. 258 Kiyokazu Okita riety of manuscripts available in diverse geographical locations in South Asia. In this regard, it is particularly challenging to deal with Madhva’s un- traceable quotes attributed to the Purāṇas since the Purāṇic text corpora are generally known for their fluidity (Rocher 1986: 37–45). This means that there is a genuine possibility that quotes attributed to the Purāṇas are un- traceable not because they are an author’s concoction but because the texts in the currently available editions differ from the versions that were available to Madhva. Mesquita is aware of this difficulty and points out that there are a number of verses in Vedānta Deśika’s works that he attributes to the Varāha Purāṇa but that are not traceable in the current printed Purāṇa editions (Mesquita 2000: 155). In this context, the case of the Skanda Purāṇa verses cited by the Dhar- manibandha authors is worth considering, which exemplifies the poor state of the textual basis of current Purāṇic studies. The Dharmanibandha authors such as Lakṣmīdhara cite verses from the Skanda Purāṇa that are not found in most printed editions. However, a study by Adriaensen, Bakker and Isaac- son in the 1990s based on the oldest surviving manuscript of the Purāṇa dating from 810 CE led to the discovery that many hitherto “untraceable” verses can be found in what seems to be the oldest text bearing the name Skanda Purāṇa.3 As Bakker (2004: 2) points out, this old Skanda Purāṇa turned out to have hardly anything in common with the well-known edition of the Purāṇa edited by Khemarāja Śrīkṛṣṇadāsa. This means that there is at least a theoretical possibility that some of the verses quoted by Madhva, which Mesquita identifies as untraceable, may turn out to be existing in the Purāṇa.4 The fundamental issue is that most of the Purāṇas now available in print are not critically edited, and therefore, they are not necessarily reliable.5 Any discussion regarding Madhva’s untraceable quotes will remain inconclusive as long as considerable progress in research in the Purāṇic textual history has not been achieved. 3 See Adriaensen, Bakker and Isaacson (1998) Prolegomena and Bakker (2004). I thank Dr. Kengo Harimoto for kindly drawing my attention to this case. 4 I checked the 129 verses attributed to the Skandapurāṇa, which Mesquita lists as un- traceable (2008: 263–299), against the first 31 chapters of the Skandapurāṇa critically edited by Adriaensen, Bakker and Isaacson (www.rug.nl/ggw/onderzoek/ onderzoeksinstituten/indian/skandapurana/indianText). However, I could not find any matches. 5 The challenge is also that there are several versions of the same text existing in parallel.  Quotation, Quarrel and Controversy 259 3 Untraceable quotes and Vedānta as Hindu theology As discussed, the issue of untraceable quotes is complex and does not allow us to reach a simple conclusion. At the same time, we should recognize that the list of Madhva’s untraceable quotes compiled by Mesquita in his publica- tion in 2007 is more than 250 pages long. The sheer number of untraceable quotes found in Madhva’s works seems to indicate that the majority of them, if not all, were his own creation. Mesquita’s observation that these quotations tend to support Madhva’s unique doctrines further supports this conclusion.6 If we accept that Madhva’s untraceable quotations, at least some of them, were indeed written by Madhva himself to support his own teaching, then this seems to have significant implications for our understanding of Vedānta. In recent years, scholars such as Francis Clooney (2003) and Jonathan Edel- mann (2013) have argued that Vedānta needs to be understood as a discipline of scriptural exegesis, and therefore, it is more appropriate to label Vedānta a theology than to regard it as philosophy. One of the distinctive characteristics 6 To what extent this is unique to Madhva is another question. Similar phenomena might have existed. For example, Prof. Alexis Sanderson suggests that not only the Śvetāśvatara Upaniṣad was written in a much later period (i.e., sometime after the 2nd century CE) than normally assumed, but also the text embodies the aspiration of a group of Rudrabhaktas / Śivabhaktas to vedicize their tradition, thus making a bid to elevate its status in the eyes of the orthoprax, by producing a scripture that embeds new material drawn from a Pāśupata environment into a matrix of ancient Vedic verses: “My hypothesis starts from the observation that there are some echoes in that [Śvetāśvatara] Upaniṣad of the Pañcārtha/Pāśupatasūtra and its terminology. Evidence of these echoes is given in n. 124 of my 2007 publication ‘Atharvavedins in Tantric Territory …’ It has been assumed that since the ŚvUp [Śvetāśvatara Upaniṣad] is an Upaniṣad it should be seen as marking an early stage in the evolution from Vaidika re- ligion to Śaivism proper. This evolutionist model of progressive ‘hinduisation’ seems to me to be naive. It fails to consider that when a scriptural text is of mixed character this may well be not because two phenomena which would later be distinct are not yet distinct but because an author or compiler redacted elements of both traditions into a single text to serve the purpose of diminishing this distinctness. This author articulates, I propose, the aspiration of a group of Rudrabhaktas/Śivabhaktas to vedicise their tra- dition, so making a bid to elevate its status in the eyes of the orthoprax, by producing a scripture that embeds new material drawn from a Pāśupata environment into a matrix of ancient Vedic verses. It is with this in mind that I have called the ŚvUp Vaidika- Pāśupata in n. 247 of my 2003 study ‘The Śaiva Religion among the Khmers (Part I)’” i.e., in Sanderson 2003. Personal correspondence, 3 June 2014. The content in square brackets has been inserted by the current author. As for a commonly accepted date for the Śvetāśvatara Upaniṣad, Thomas Oberlies for example suggests that it was written between the beginning of the Christian Era and 200 CE (1995: 66‒67). I thank Dr. Philipp Maas for this reference. 260 Kiyokazu Okita of theology is its commitment to texts that the tradition considers to be di- vinely revealed. Theology in this sense is contrasted with philosophy in the post-Cartesian sense of the term, which maintains an intellectual space to critically inquire the validity of any type of commitment, and consequently does not require commitment to any particular set of texts.7 The goal of Vedānta as it is traditionally understood is to articulate a co- herent system of thought based on a core of revealed texts such as the Upaniṣads, the Brahmasūtras and the Bhagavadgītā. Therefore, Vedānta is fundamentally rooted in these works, and the commitment of any exegete to these works should have priority over their individual ideas. If, however, Madhva had his ideas first and then created fictional quotes to support them, can his enterprise still be called Vedānta? Or do we need to modify our un- derstanding of Vedānta as Hindu theology? These are some of the questions we need to address when dealing with Madhva’s untraceable quotes. 4 Early modern controversy: Appayya Dīkṣita vs. Jīva Gosvāmī The previous part of this chapter dealt with present-day discussions on the issue of Madhva’s untraceable quotes, and its relations to Purāṇic studies and Hindu theology. In order to further demonstrate the complexity involved, I shall now discuss how two prominent Hindu theologians saw the issue. 4.1 Appayya Dīkṣita One of these theologians is Appayya Dīkṣita (1520–1592) (Bronner 2007: 1), a prominent Advaita author from northern Tamil who is remembered inter alia for his Śivārkamaṇidīpikā, a sub-commentary on Śrīkaṇṭha’s Brahma- sūtra commentary. The other is Jīva Gosvāmī (1517 terminus ad quem–1608 terminus a quo), one of the six Gosvāmīs of Vṛndāvana, who was the syste- matizer of the Gauḍīya Vaiṣṇava theology. Appayya Dīkṣita squarely re- jected Madhva’s untraceable quotes while Jīva Gosvāmī accepted and uti- lized some of them to build his own theological system. In this section, I will first examine the first two verses of Appayya’s Madhvatantramukhamardana, or Crushing the Face of Madhva’s Teaching and his auto-commentary Madhvamatavidhvaṃsanā, or Destroying Madh- va’s Doctrine on these two verses. While Mesquita deals with these works (2000: 29–33) to some extent, I shall examine them in more detail, incor- 7 For a further discussion on this point by Edelmann and responses to his argument, see http://indianphilosophyblog.org/2014/03/07/philosophy-and-theology-lets-be-clearer/.  Quotation, Quarrel and Controversy 261 porating sections from Appayya’s auto-commentary that Mesquita does not discuss. Appayya’s works are significant, because he was the first author who explicitly addressed the issue of Madhva’s untraceable quotes. Although Mesquita (2000: 27–29) claims that in the 13th century Varadaguru and Vedānta Deśika already criticized Madhva’s quotes, it is not clear whether their criticism was indeed directed against Madhva. The two authors criti- cized the practice of fabricating citations for one’s own purposes, but they did not specify, as Jan Willem de Jong (1999: 63–64) and Rocher (2008: 603) point out, to whom their criticism was directed. Going back to Appayya, his Madhvatantramukhamardana is a metrical composition refuting Madhva’s commentary on the first five Adhikaraṇas.8 He starts the work in the following way: If one names the topic of scripture to be Śiva or Viṣṇu, that is ac- cepted. [For, such an interpretation] should be accepted by us who worship that Brahman, even if it is with qualities. The contradiction does not overly manifest itself for those who know the principle of Nahinindā, nor is it appropriate to reject even a divergent meaning of the Sūtras, if [such a meaning] is [somehow] acceptable.9 As an Advaitin, Appayya believes that Brahman without qualities (nirguṇa) is the highest reality. Nevertheless, in his opening verse he demonstrates a concessive attitude towards the worship of saguṇa Brahman as Śiva or Viṣṇu. The principle of Nahinindā, to which Appayya refers in the above cited passage, is first mentioned in the Śābarabhāṣya: na hi nindā nindyaṃ nindituṃ prayujyate kiṃ tarhi ninditād itarat praśaṃsitum. 8 That this work is concerning the first five Adhikaraṇas of the Brahmasūtras is ex- pressed in the third verse: ataḥ pañcādhikaraṇīṃ lakṣyīkṛtyaiva tanmate / dūṣyastha- lāni sarvatra sūcyante sudhiyāṃ mude // “Therefore, targeting only the [first] five Adhikaraṇas [in his Brahmasūtra commentary], I point out in all cases those places to be condemned in his teachings, for the pleasure of people with good intelligence.” Mesquita understands the expression pañcādhikaraṇīṃ to refer to the five-fold divi- sions in the Vedas (2000: 30). However, this is incorrect since in his auto-commentary Appayya explains that the term refers to the first five Adhikaraṇas in Madhva’s Brahmasūtra commentary: tanmatasiddhām īkṣatyadhikaraṇāntāṃ pañcādhikaraṇīṃ “pañcādhikaraṇīm means what is established in his system ending with the Īkṣati Adhikaraṇa i.e., the commentary on the fifth sūtra, ‘ikṣater nāśabdam’ to the eleventh sūtra, ‘śrutatvāc ca’.” 9 śivaṃ viṣṇuṃ vā yady abhidadhati śāstrasya viṣayaṃ tad iṣṭaṃ grāhyaṃ naḥ saguṇam api tad brahma bhajatām / virodho nātīva sphurati nahinindānayavidāṃ na sūtrāṇām arthāntaram api bhavad vāryam ucitam // (Appaya 1941:1). 262 Kiyokazu Okita Blame is not employed in order to blame what is to be blamed but rather to praise what is different from that which is blamed.10 By referring to this principle, Appayya suggests that when the Advaitins blame the saguṇa interpretation of the Brahmasūtras, we should not take this literally. Rather, we should understand this blame as praise of the alternative, namely the nirguṇa interpretation. Appayya also demonstrates his concessive attitude toward the worship of saguṇa Brahman when he begins his auto-commentary with a maṅgala verse dedicated to Kṛṣṇa: May Mukunda always bestow good on me, [Mukunda] who appears constantly in a complete form, [Mukunda who is] grasped by the for- tunate ones according to their taste though after a long time, after [they] have opened up the lotus bud of [their] heart by the practice of yoga.11 Since Appayya is the author of works with a Śaiva leaning such as the Śi- vārkamaṇidīpikā and Śivārcanacandrikā, one might expect him to dedicate his maṅgala verse to Śiva. That he offers it to Mukunda or Kṛṣṇa suggests that for him it does not really matter whether one worships Śiva or Viṣṇu, because in any case the highest reality is nirguṇa Brahman. This is also the interpretation of Pālghāṭ Nārāyaṇa Śāstrī’s Ṭippaṇī on Appayya’s auto-com- mentary. According to Śāstrī, yathāruci or “according to their taste” means “in the form of Rāma, Kṛṣṇa or Naṭarāja.”12 In his auto-commentary Appayya explains that the Advaitins, too, accept the worship of saguṇa Brahman since it helps them to understand Brahman’s nature, quality, greatness and so on.13 He further says that even for the Advai- tins the worship of saguṇa Brahman is necessary for attaining liberation and other desirable aims: 10 My translation is based on Apte 1992: 64. This is Śabara’s commentary on Mīmāṃ- sāsūtra 2.4.20: ekatve ’pi parāṇi nindāśaktisamāptivacanāni //. 11 Appayya, Madhvamatavidhvaṃsanā: udghāṭya yogakalayā hṛdayābjakośaṃ dhanyaiś cirād api yathāruci gṛhyamāṇaḥ / yaḥ prasphuraty avirataṃ paripūrṇarūpaḥ śreyaḥ sa me diśatu śāśvatikaṃ mukundaḥ // 1 // (Appayya 1941: 1). 12 Makhivaryabhūṣaṇā: yathāruci rāmakṛṣṇanaṭarājarūpena (Pālaghāṭ Nārāyaṇa Śāstri 1940: 1). 13 Madhvamatavidhvaṃsanā: saguṇaṃ brahmopāsīnānām asmākam advaitavādinām upāsyasvarūpaguṇamahimāvadhāraṇāya tatpratipādanapravṛttaṃ tadīyaṃ śāstrasya yojanaṃ grāhyam / (Appayya 1941: 2).  Quotation, Quarrel and Controversy 263 Indeed, we [Advaitins] also require the worship of these two [deities], because we need the controlling of our mind which is conducive to meditation, the sequence of experiencing celestial enjoyment in the [heavenly] places such as Kailāsa and Vaikuṇṭha, and its result in the form of liberation.14 Appayya makes it clear that the worship of saguṇa Brahman is something required (apekṣita) for the Advaitins. From Appayya’s perspective, the Vaiṣ- ṇavas such as Rāmānuja who worship Viṣṇu or saguṇa Brahman as the ulti- mate reality are not entirely correct. However, their worship is not entirely meaningless, since it leads to a good result. In fact, Appayya even accepts that the Brahmasūtras can be interpreted as teaching saguṇa Brahman, since in his view “the Sūtras are multifaceted.”15 Appayya ends his auto-commentary on the first verse of the Madhvatan- tramukhamardana with the following rhetorical question: Why should we not also accept the doctrine of Ānanda Tīrtha, which aims at investigating by means of Śrutis, Smṛtis and reasoning the na- ture of the Lord Nārāyaṇa, his inconceivable and astonishing qualities, greatness, wealth and power, the procedure for his worship and so on, and the sequence of enjoyments that are its (i.e., the worship’s) re- sult?16 If Madhva teaches saguṇa Brahman, and if Appayya thinks its worship is required for the Advaitins, then why does Appayya reject Madhva’s teach- ings? Appayya answers his own question in the second verse of the Madh- vatantramukhamardana: Nevertheless, we cannot accept the teachings of Ānanda Tīrtha in which the important norm of those who follow Vedic passages be- comes confused.17 14 Madhvamatavidhvaṃsanā: apekṣitaṃ khalv asmākam api nididhyāsanopayogicittava- śīkaraṇakailāsavaikuṇṭhādisthānagatadivyabhogānubhavakramamuktirūpaphalāpekṣi- ṇāṃ tayor upāsanam / (Appayya 1941: 2). 15 Madhvamatavidhvaṃsanā: sūtrāṇāṃ viśvatomukhatvād iti brūmaḥ / (Appayya 1941: 2). 16 Madhvamatavidhvaṃsanā: bhagavato nārāyaṇasya svarūpaṃ tasyācintyādbhutaguṇa- mahimavibhūtivaibhavaṃ tadupāsanādiprakāraṃ tatphalabhogakramaṃ ca śrutismṛ- tinyāyair nirūpayituṃ pravṛttam ānandatīrthīyaṃ matam api kuto na grāhyam? (Ap- payya 1941: 3). 17 Madhvatantramukhamardana: tathāpy ānandatīrthīyaṃ matam agrāhyam eva naḥ / yatra vaidikamaryādā bhūyasy ākulatāṃ gatā // 2 // (Appayya 1941: 3). 264 Kiyokazu Okita The important expression in this verse is vaidikamaryādā. The word maryādā literally means “limit” or “boundary.” Therefore, the compound literally means “the limit or the boundary of those who follow the Vedas.” Appayya says that in Madhva’s system this limit or boundary becomes confused (āku- latāṃ gatā). This means that Madhva’s teaching goes beyond what is accept- able to people who follow the authority of the Vedas. In Appayya’s view, Madhva’s teaching should be rejected not because it teaches saguṇa Brahman, but because the way he teaches does not comply with the accepted norm in the Vedāntic discourse. Appayya spells this out in his auto-commentary on the second verse just discussed. It is here where Appayya addresses the issue of Madhva’s untraceable quotes. Appayya’s cri- ticism against Madhva is both theological and epistemological. He voices his critique as follows: Even if we should accept the interpretations of others concerning [Brahman] with qualities in the manner mentioned, even then we can never accept the interpretation of Ānanda Tīrtha. For, in the interpre- tations of others there is a difference in the mode of interpretation in a few Sūtras only, as having as their object [Brahman] with qualities, without qualities, distinction, non-distinction and so on. Concerning other Sūtras, however, there is no disagreement regarding the outcome even if there are sometimes different modes of interpretations arrang- ing the prima facie view and the conclusive view. In Ānanda Tīrtha’s interpretation however, different modes of interpretations are found almost everywhere.18 Appayya accepts that there are Vedāntic views differing from Advaita, and acknowledges that these provide legitimate interpretations of the Brahma- sūtras. However, he rejects Madhva’s teaching because it is based on differ- ent modes of interpretations (prakārabheda). The point seems to be that Madhva often has an entirely different scheme for understanding the Brah- masūtras. For example, according to Śaṅkara and Rāmānuja, the Brahma- sūtras 1.4.23–27 constitute a section dealing with the topic of prakṛti, the primordial matter that is the source of the material universe. According to 18 Madhvamatavidhvaṃsanā: yady apy evam uktarītyānyadīyāni saguṇayojanāny upā- deyāni tathāpy ānandatīrthīyaṃ yojanam anupādeyam eva / anyadīyeṣu hi yojaneṣu saguṇanirguṇabhedābhedādiviṣayatayā katipayeṣv eva sūtreṣu prakārabhedaḥ / anye- ṣu tu sūtreṣu kvacit kvacit pūrvapakṣasiddhāntaracanābhaṅgībhede ’pi phalato na vi- vādaḥ / ānandatīrthīye tu yojane prāyaḥ sarvatraiva prakārabhedaḥ / (Appayya 1941: 3).  Quotation, Quarrel and Controversy 265 Madhva, however, this section deals with the principle of samanvaya, which explains how all the words in the Śruti ultimately refer to Brahman. Appayya further says that this difference “concerns the contents which are not agreed on by others, which are concocted by him alone.”19 What Appayya means is that while other Vedāntins build their teachings on existing Śruti and Smṛti texts, Madhva’s teaching is often based on texts that no one has ever heard of. As Mesquita points out, Appayya then gives a list of the ob- scure texts Madhva cites from.20 The list names twenty-nine texts, including the Agniveśya and the Bāllaveya, which according to Madhva are Śruti texts. It also names another ten texts such as the Brahmatarka, which from Madhva’s perspective are not the Śrutis but authoritative.21 After listing the obscure texts cited by Madhva, Appayya analyzes Madhva’s strategy for counterbalancing their obscurity. As Mesquita points out, this is where Appayya addresses Madhva’s claim of being the third avatāra of Vāyu: In order to remove suspicion concerning his own untrustworthiness, which arises as a result of citing these [works], [Madhva] proclaims that he himself is the third Avatāra of the god Vāyu after Hanuman and Bhīmasena.22 Then, Appayya provides two passages that Madhva uses to justify his claim. The first one reads as follows: The first [incarnation of Vāyu] is named Hanuman, the second Bhīmasena. As for the third, it is Pūrṇaprajña who accomplishes the work of the Lord.23 19 so ’py anyeṣām asampratipanneṣu svamātrakalpiteṣv artheṣu / (Appayya 1941: 3). 20 Appayya’s list of obscure Śruti texts cited by Madhva: Catura, Kamaṭha, Māṭhara, Kauṇṭharavya, Kauṇḍinya, Māṇḍavya, Mārkaṇḍeya, Maudgalya, Pauṣyāyaṇa, Pau- trāyaṇa, Sautrāyaṇa, Pārāśaryāyaṇa, Mādhyandināyana, Kauṣārava, Kāṣāyaṇa, Bṛ- had, Uddālaka Auddālakāyaṇa, Kauśika, Sauparṇa, Śāṇḍilya, Vatsa, Gaupavana, Bāl- laveya Agniveśya, Caturvedasaṃhitā, Caturvedaśikā, Indradyumna, Paramaśruti (Ap- payya 1941: 3–4). 21 Appayya’s list of other obscure texts cited by Madhva: Adhyātmanārāyaṇa (or Adhyāt- ma and Nārāyaṇa?), Adhyātmasaṃhitā, Bṛhatsaṃhitā, Mahāsaṃhitā, Nārāyaṇatantra, Bṛhattantra, Puruṣottamatantra, Māyāvaibhava, Brahmatarka, Bhaviṣyatparva (Ap- payya 1941: 4). 22 Madhvatantramukhamardana: tadupanyāsaprasaktasvānāptatvaśaṅkāparihārāya ha- numadbhīmasenakrameṇa vāyoḥ svayaṃ tṛtīyāvatāra ity udghoṣaḥ / (Appayya 1941: 4). 23 Madhvatantramukhamardana: prathamo hanumānnāma dvitīyo bhīmasenakaḥ / pūr- 266 Kiyokazu Okita The second verse is from the Ṛg Veda, which says: “In this way that which is visible is carried for the body […].”24 Although Appayya apparently holds the first passage to be a Smṛti text, it is actually Madhva’s own composition that can be found in his Mahābhāra- tatātparyanirṇaya 2.118 (Madhva 1971: 24). The second passage is identifi- able as Ṛg Veda 1.141.1a. However, as pointed out by Pālghāṭ Nārāyaṇa Śāstrī in his Ṭippanī, this Vedic passage praises Agni and has nothing to do with Vāyu: By this praise of Agni, the excellence of the offering to Agni is made clear. By this it is pointed out that [Madhva’s] explanation of this Ṛg Veda passage as being intent on the triad of Vāyu Avatāras indicates heavily that he transgresses the limit of those who follow the Vedas.25 This is an example of Madhva twisting the meaning of an existing passage for his own purpose.26 In short, Madhva’s claim for being the third avatāra of Vāyu is supported only by his own composition and a deliberate misrepresentation of an existing text. Appayya says Madhva’s claim is noth- ing more than his imagination,27 and concludes his analysis by stating, “We repeatedly observe that [Madhva] transgresses the limit of those who accept scriptural authority.”28 It is worth reflecting on Appayya’s charge that Madhva transgresses the limit of those who accept scriptural authority (prāmāṇikamaryādālaṅ- ghanaṃ). In this expression, the term prāmāṇika literally means “those who accept the reliable means of knowledge (pramāṇa).” As is well known, pra- māṇa in Vedānta includes not only scriptural authority (śabda) but also other ṇaprajñas tṛtīyas tu bhagavatkāryasādhakaḥ // (Appayya 1941: 5). 24 báḷ itthā́ tád vápuṣe dhāyi darśatáṃ. 25 Makhivaryabhūṣaṇā: anenāgniśaṃsanenāgneyayāgasya prāśastyaṃ dyotyate / etena vāyvavatāratrayaparatayāsyā ṛco varṇanaṃ vaidikamaryādollaṅghanasya mahal liṅgam iti sūcitam / (Pālghāṭ Nārāyaṇa Śāstrī 1940: 11). Pālghāṭ Nārāyaṇa Śāstrī was a modern scholar who specialized in Advaita Vedānta. As his name suggests, he hailed from the Palakkad (Pālghāṭ) region in Kerala. This is also confirmed in the concluding verse Śāstrī gives at the end of his Ṭippaṇī (Pālghāṭ Nārāyaṇa Śāstrī 1940: 129): kṣīrāraṇyā(pālaghāṭ)khyadeśo guṇagaṇamahito darśanāt pāvanātmā / He taught at Maharaja’s college in Mysore (http://profskr.com/?page_id=9). Although his exact dates are not known to the current author, he seems to have lived from the late 19th century to the mid-20th century. 26 See Mesquita (2000: 55–58, 2003: 114) for an analysis of how Madhva uses this Ṛg Veda verse. 27 ṛgvedamantrasya svakalpitavāyvavatāratrayaparatayā pradarśakam (Appayya 1941: 11). 28 prāmāṇikamaryādālaṅghanaṃ bhūyasā dṛśyate / (Appayya 1941: 5).  Quotation, Quarrel and Controversy 267 means of knowledge such as perception (pratyakṣa) and inference (anumāṇa). Nevertheless, I believe what Appayya has in mind is scriptural authority (śabdapramāṇa), because this expression appears in the context of Appayya’s criticism of Madhva’s use of scripture. This is further supported by the fact that Pālghāṭ Nārāyaṇa Śāstrī rephrases prāmāṇika as “those who follow the Vedas (vaidika).” As an Advaitin, Appayya ultimately does not accept Madhva’s ontology, which is dualistic and theistic. At the same time, Appayya’s fundamental problem with Madhva is not his ontology as such. Appayya acknowledges that there are valid Vedāntic views that are not Advaita. He even praises the benefit of worshipping saguṇa Brahman. However, what Appayya finds problematic and unacceptable is the way in which Madhva builds his ontology. From Appayya’s perspective, Madhva’s teaching is unacceptable, because he does not follow the rules of the game. What is implicit in Appayya’s criticism is his understanding of the nature of Vedānta, according to which one should formulate one’s opinion based on commonly available textual sources. When Madhva cites texts that nobody knows, this is a threat to the discipline itself, because this challenges the very notion of scriptural authority (śabdapramāṇa), which is the basis of Vedānta. What is Appayya’s contribution to the study of citations? In their intro- duction to the present volume, Freschi and Maas discuss the reuse of parts of an old building to construct a new building. In the context of Vedānta, ac- cording to Appayya, it is allowed to formulate new systems of thought, but the ingredients for such a construction are pre-determined. The most widely accepted ingredients are the Upaniṣads, the Bhagavadgītā and the Brahma- sūtras. Smṛti texts such as the Purāṇas are accepted, but they have to be commonly available. Thus, Appayya emphasizes the necessity of verifiable scriptural evidence and rejects the use of apocryphal scriptural citations. 4.2 Jīva Gosvāmī In the previous section, I discussed Appayya’s criticism of Madhva’s use of untraceable citations. Appayya’s argument is clear and forceful. However, the matter is complicated by Jīva Gosvāmī, who, unlike Appayya, accepts some of Madhva’s untraceable quotes as authentic. A contemporary of Appayya, Jīva lived in North India in the 16th century. He was one of the so-called Ṣaḍgosvāmī of Vṛndāvana, who founded the Gauḍīya Vaiṣṇava school that was initiated by Caitanya (1486–1534).29 It is 29 According to the Caitanyacaritāmṛta of Kṛṣṇadāsa Kavirāja (1496–??), Caitanya was initiated by Īśvara Purī into the path of Kṛṣṇa devotionalism. Some of the later Gauḍīya authors, such as Baladeva Vidyābhūṣaṇa (??–1793), claim that Caitanya belonged to 268 Kiyokazu Okita not clear to me whether Jīva knew Appayya’s works. As far as his discussion of Madhva’s quotes is concerned, Jīva does not address Appayya’s criticism. Jīva’s view on Madhva’s untraceable quotes is significant in our discussion because it shows that the authenticity of his quotes was accepted not only by Madhva’s immediate followers but also by Vedāntins outside the Mādhva school. Before we analyze Jīva’s view, however, we need to clarify the relation between the Gauḍīyas and the Mādhva school. This is necessary, because there is a controversial claim that the Gauḍīyas actually belong to the Mādh- va school and form the Brahmā-Mādhva-Gauḍīya Sampradāya. The idea of the Gauḍīya affiliation with the Mādhvas possibly goes back to Kavi Karṇa- pura’s Gauragaṇoddeśadīpikā, which was written in 1576 (see n. 29). Even if the passage of this work that mentions the Sampradāya affiliation is an interpolation, as John Hawley (2013) argued recently, the Mādhva-Gauḍīya connection was widely accepted in North India by the 18th century. As David Buchta has argued and as I have discussed elsewhere,30 how- ever, it is best to see this affiliation as something formal and external. In other words, in spite of the fact that some Gauḍīyas claim to be a part of the Mādhva tradition, we must understand the two traditions as different from one another. It is clear that the Gauḍīya tradition is distinct from the Mādhva tradition if, for example, we examine Jīva Gosvāmī’s magnum opus, the Bhā- gavatasandarbhas, which are arguably the most foundational texts in the Gauḍīya theology. In the beginning of the Tattvasandarbha, which is the first volume in the Sandarbhas, Jīva does not mention any of the Mādhva teach- ers. Instead, he offers the maṅgala verses to Kṛṣṇa, to Caitanya and to his uncles Rūpa and Sanātana.31 Caitanya, Rūpa and Sanātana are all prominent figures in the Gauḍīya tradition but are not recognized as such in the Mādhva tradition. As Alexis Sanderson (2005: 89) and Christopher Minkowski (2008: 1) have pointed out, the maṅgala or the opening verses are the “face” of the work, where the author’s personal affiliation can be observed. Therefore, the the Mādhva sampradāya because Mādhavendra Purī, Īśvara Purī’s guru, was a Mādhva sannyāsī. However, whether Mādhavendra belonged to the Mādhva sampradāya or not is by no means self-evident. In fact, scholars such as S. K. De (1961) and Sharma (2000) reject Mādhavendra’s affiliation with the Mādhva school. See Okita 2014b: 44– 55 for a detailed discussion of this point. 30 See for example, Buchta 2003, 2005, 2007 and Okita 2008, 2009, 2014a, 2014b. 31 Tattvasandarbha: antaḥ kṛṣṇaṃ bahir gauraṃ darśitāṅgādivaibhavam / kalau saṅ- kīrtanādyaiḥ smaḥ kṛṣṇacaitanyam āśritāḥ // 2 // jayatāṃ mathurābhūmau śrīlarūpa- sanātanau / yau vilekhayatas tattvaṃ jñāpakau pustikām imām // 3 // (Jīva 1983: 5).  Quotation, Quarrel and Controversy 269 opening verses of the Tattvasandarbha indicate that Jīva does not see his Gauḍīya tradition as a part of the Mādhva school. Moreover, Jīva mentions that his Sandarbhas are based on Gopāla Bhaṭṭa Gosvāmī’s composition, which Gopāla Bhaṭṭa himself wrote following what was written by great Vaiṣṇavas.32 In his auto-commentary on this verse, Jīva explains that the expression “great Vaiṣṇavas” refers to authors such as Rāmānuja, Madhva and Śrīdhara Svāmī.33 In other words, Jīva acknowledges his indebtedness to Madhva but not in an exclusive way. In fact, as discussed by Ravi Gupta and others, Rāmānuja’s and Śrīdhara’s influence on Jīva’s theology is at least equally strong as that of Madhva, and there are a number of significant theological disagreements between Madhva and Jīva.34 For example, Jīva accepts the bhedābheda position in which Brahman is considered the material cause (upādānakāraṇa) of the manifest world; that the living entities and the manifest world are Brahman’s energies (śaktis); that Kṛṣṇa is the highest manifestation of the ultimate reality and the source of all Viṣṇu avatāras; that Rādhā is Kṛṣṇa’s pleasure-giving energy (hlādinī śakti) and his best consort; and that Caitanya is non-different from Kṛṣṇa. While Jīva promotes these doctrines, Madhva does not accept them. In con- trast, Madhva’s followers accept his teachings rigidly; for them Madhva is the supreme teacher. Complicated discussions notwithstanding, the only point I wish to make clear is that the Gauḍīyas are theologically distinct from the Mādhvas, and therefore they should be considered as forming a separate school. Provided that this point is acceptable, we can now go back to address Jīva’s view of Madhva’s untraceable quotes. From Jīva’s perspective, the untraceability of certain texts does not pose such a great challenge, since he believes there are many texts that are lost and hence not available to us. In his Tattvasandarbha Anuccheda 12, Jīva states that at the present mo- ment (samprati), scriptural authority (śabda) means the Itihāsas and the Purāṇas, because the Vedas are difficult to go through in their entirety (duṣpāratvāt).35 In his auto-commentary, Jīva explains that the expression 32 Tattvasandarbha: ko ’pi tadbāndhavo bhaṭṭo dakṣiṇadvijavaṃśajaḥ / vivicya vyalikhad granthaṃ likhitād vṛddhavaiṣṇavaiḥ // 4 // tasyādyaṃ granthanālekhaṃ krāntam ut- krāntakhaṇḍitam / paryālocyātha paryāyaṃ kṛtvā likhati jīvakaḥ // 5 // (Jīva 1983: 6). 33 Sarvasaṃvādinī: vṛddhavaiṣṇavaiḥ śrīrāmānujamadhvācāryaśrīdharasvāmyādibhir yal likhitam tasmād uddhṛtasyety arthaḥ / (Jīva 1983a: 6). 34 See for example Gupta 2007, Kapoor 1976. 35 Tattvasandarbha: vedaśabdasya samprati duṣpāratvād duradhigamārthatvāc ca tad- arthanirṇāyakānāṃ munīnām api parasparavirodhād vedarūpo vedārthanirṇāyakaś cetihāsapurāṇātmakaḥ śabda eva vicāraṇīyaḥ / (Jīva 1983: 29). 270 Kiyokazu Okita “the present moment” refers to the age of Kali, and that the Vedas are diffi- cult to go through because some texts are not available and because we lack the intelligence to comprehend the works that have come down to us.36 Rādhā Mohana Gosvāmī (late 18th century)37 further clarifies that the Vedas are difficult to go through in their entirety “because some Vedas are destroyed and some concealed.”38 Appayya’s main criticism against Madhva was that Madhva transgresses the limits of those who accept scriptural authority. If we accept Jīva’s view of scripture, however, the convention of what constitutes authoritative texts becomes much more ambiguous, because in Jīva’s view there is always a possibility that certain texts are untraceable not because they were fabricated, but because they used to exist but are now lost or unavaila- ble. Jīva was not the first authority who discussed the problem of lost scrip- tures, although he was possibly among the few who defended the validity of extant passages from lost texts. A defense of the authority of lost texts can be found, for example, in the tradition of the Dharmaśāstra. According to this tradition, the rules enjoined in the Dharmaśāstras are authoritative because they are based on the Vedas. A problem, however, is that many Dharma- śāstric rules are not found in the revealed scriptures. One way in which the tradition deals with this is to appeal to the Mīmāṃsaka concept of “inferred Vedic text” (anumitaśruti). If certain rules are mentioned in the Smṛti texts or observed in the conduct of virtuous people but not mentioned in the available Śruti texts (pratyakṣaśruti), then we must infer the existence of lost Śruti texts that mention those rules (Olivelle 1993: 84–85, 1999: xli, 2006: 173). As early as the 3rd century BCE, Āpastamba says, All injunctions were (originally) taught in the Brāhmaṇas. Those sec- tions of theirs that have been lost are inferred on the basis of custom (trans. Olivelle 1993: 85).39 A critical difference, however, between Jīva and those Dharmaśāstric authors is that the former uses available passages attributed to a lost text while the latter do not claim to have such a direct access to “inferred Vedic text.”40 36 Sarvasaṃvādinī: apracaradrūpatvena durmedhastvena ca ‘duṣpāratvāt’ (Jīva 1983a: 29). 37 On Rādhā Mohana Gosvāmī, see Elkman 1986: 51–55. 38 Ṭīkā: duṣpāratvād iti keṣāṃcid vedānām ucchannatvāt keṣāṃcit pracchannatvāc ca iti bhāvaḥ (Jīva 1983: 29). For a detailed discussion on Jīva’s epistemology, see Broo 2006, Buchta 2006, Uskokov 2009. 39 Āpastambadharmasūtra 1.12.10: brāhmaṇoktā vidhayas teṣām utsannāḥ pāṭhāḥ pra- yogād anumīyante /  Quotation, Quarrel and Controversy 271 The idea of “lost text” is found in the Vaiṣṇava tradition as well. Yāmunācārya, a well-known teacher of the Śrī Vaiṣṇava school in the 10th century in South India, wrote a text called the Āgamapramāṇya. In this work, he argues that the lost Vaiṣṇava texts in Kashmir such as the Ekāyanaśākhā are revealed (apauruṣeya) scriptures and therefore authorita- tive (Sanderson 2009: 108). Returning to Jīva’s discussion, in Anuccheda 28 he addresses Madhva’s untraceable citations. This is in the context of Jīva’s discussion of scriptural means of knowledge (śabdapramāṇa). Like many other Vedāntins, Jīva says that scriptural knowledge is the only means to understand Brahman. Con- cerning the use of scripture in his own work, Jīva first says that he will cite the Śrutis and the Smṛtis exactly as he saw (yathādṛṣṭam eva) them.41 How- ever, to this general principle Jīva adds an exception, namely, Madhva’s untraceable citations: And in some cases, I humbly accepted sources that I have not seen myself from glorious texts such as the Bhāgavatatātparya, the Bhāra- tatātparya and the Brahmasūtrabhāṣya by the glorious Madhvācārya, who is the ancient teacher of realism, who greatly propagated a particular Vaiṣṇava thought, [...] who has as his disciples and grand- disciples the best among the knowers of the Vedas and their meaning, such as Vijayadhvaja, Brahmatīrtha and Vyāsatīrtha, who were well known in southern countries and so on.42 The key expression in the above passage is “the sources that I have not seen myself (svayam adṛṣṭākarāṇi).” When Jīva realizes that certain quotations are found only in Madhva’s works, he accepts these passages as authoritative although he cannot trace the original, and does not discuss their authenticity. He has a very positive attitude toward Madhva and his followers. This is 40 I thank Dr. Elisa Freschi for her perceptive comment on this point. For her reflections on this, see her blog post: http://elisafreschi.com/2014/07/01/forging-indian- philosophical-texts/. 41 Tattvasandarbha: pramāṇāni śrutipurāṇādivacanāni yathādṛṣṭam evodāharaṇīyāni / (Jīva 1983: 83). 42 Tattvasandarbha: kvacit svayam adṛṣṭākarāṇi ca tattvavādagurūṇām anādhunikānāṃ […] pracurapracāritavaiṣṇavamataviśeṣāṇāṃ dakṣiṇādideśavikhyātaśiṣyopaśiṣyībhū- tavijayadhvajabrahmatīrthavyāsatīrthādivedavedārthavidvarāṇāṃ śrīmadhvācāryaca- raṇānāṃ śrībhāgavatatātparyabhāratatātparyabrahmasūtrabhāṣyādibhyaḥ saṃgṛhī- tāni / (Jīva 1983: 83). – According to Baladeva’s commentary, “and so on (ādi)” in the expression “southern countries and so on (dakṣinādideśa)” indicates that the Mādhva sampradāya was well known in the eastern areas (gauḍe ’pi) as well: dakṣiṇādideśeti tena gauḍe ’pi mādhavendrādayas tadupaśiṣyāḥ katicid babhūvur ity arthaḥ /. 272 Kiyokazu Okita expressed, for example, in his description of Madhva as the teacher of realism (tattvavādaguru). “tattvavāda” was the term Madhva used to describe his teaching, in contrast to māyāvāda, which refers to the Advaita teaching. While Jīva teaches many Gauḍīya doctrines that Mādhva followers did not accept, Jīva agrees with Madhva in his severe critique of the Advaita Vedānta theory of non-dualism. In this regard, Jīva’s respect for Madhva’s teaching appears as an understandable consequence of their doctrinal agreement. Another place where we observe Jīva’s reverential attitude toward the Mādhva tradition is in his description of Madhva’s followers being well known in the southern country and so on (dakṣiṇādideśavikhyāta). For exam- ple, Vijayadhvaja (1410–1450)43 was well known as the author of the Padaratnāvalī, which is the first full commentary on the Bhāgavata Purāṇa within the Mādhva tradition. As for Vyāsatīrtha (1478–1539), B. N. K. Sharma calls him “the Prince of Dialecticians of the Dvaita school” (1994: viii). He was the famous author of the Nyāyāmṛta, which later prompted Madhusūdhana Sarasvatī, a famous Advaitin of the 16th century, to write his refutation called Advaitasiddhi.44 All of these points indicate that by Jīva’s lifetime in the 16th century, the Mādhva sampradāya had become a well-es- tablished and well-respected Vaiṣṇava tradition like that of Rāmānuja’s Śrī sampradāya. Thus for Jīva, who was establishing a new tradition of Gauḍīya Vaiṣṇavism, referring to the older tradition of Madhva was part of a strategy for presenting his theology in a respectable manner. After glorifying Madhva and his followers, Jīva quotes passages from Madhva’s Mahābhāratatātparyanirṇaya: Just as Lord Vyāsa, who is Lord Nārāyaṇa in person, spoke in the Mahābhārata and so on, knowing other scriptures in the light of Vedānta, and after seeing different types of books in various countries, in the same manner I [i.e., Madhva] speak, by examining these [books].45 In these passages, Madhva claims his Vyāsa-like authority in the Mahābhā- ratatātparyanirṇaya, which is his summary of the Mahābhārata. By referring 43 Sharma 2000: 456. 44 For a detailed discussion of this debate, see Sharma 1994. Also, for the indebtedness of Madhusūdana Sarasvatī to Vyāsatīrtha, see Pellegrini 2015. 45 Tattvasandarbha, Anuccheda 28, Mahābhāratatātparyanirṇaya 2.6cd~2.8ab: śāstrān- tarāṇi saṃjānan vedāntasya prasādataḥ / deśe deśe tathā granthān dṛṣṭvā caiva pṛ- thagvidhān // yathā sa bhagavān vyāsaḥ sākṣān nārāyaṇaḥ prabhuḥ / jagāda bhāra- tādyeṣu tathā vakṣye tadīkṣayā // iti / (Jīva 1983: 84).  Quotation, Quarrel and Controversy 273 to these passages, Jīva expresses his high regard for Madhva and emphasizes the scriptural authority of his writings. Then, at the end of the Anuccheda 28, Jīva gives a list of the texts that ap- pear only in Madhva’s works: Among them, the Śrutis are the Caturvedaśikhā and so on, which are selected by him, and the Purāṇas are the sections of the Garuḍa Purāṇa and so on, which are not available anywhere at present. Fur- thermore, the Saṃhitās mean the Mahāsaṃhitā and so on, and the Tantras are the Tantrabhāgavata, the Brahmatarka and so on.46 In his Madhvamatavidhvaṃsanā, Appayya mentions the Caturvedaśikhā, the Mahāsaṃhitā and the Brahmatarka as absolutely unknown and therefore untrustworthy. Jīva thus accepts as authoritative precisely the same works that Appayya rejects. Jīva’s argument is simple, but its implications are profound. Many texts became lost or inaccessible in the course of time. However, Madhva’s untra- ceable quotes must be authentic because illustrious teachers of the Mādhva school quote them. Therefore, even though he cannot trace the original, he can use these untraceable quotes to build his Gauḍīya theology. As noted earlier, the Dharmaśāstric tradition appealed to the idea of lost texts. These texts are inferred partly based on the conduct of virtuous people. If we ac- knowledge the cultural background in which this type of argument appears to have been well accepted, Jīva’s argument for the validity of Madhva’s untra- ceable quotes based on the authority of the Mādhva followers must have sounded fairly reasonable to his contemporary audience. Going back to the analogy of architecture, using the materials from an older building can give a new building an aura of tradition and being estab- lished. Thus Jīva’s reuse of Madhva’s untraceable quotes has the effect of reminding experienced readers of the legacy of the Mādhva tradition and its fierce and forceful attack on Advaita Vedānta. 46 Tattvasandarbha, Anuccheda 28: tatra taduddhṛtā śrutiś caturvedaśikhādyā, purāṇaṃ ca gāruḍādīnāṃ samprati sarvatrāpracaradrūpam aṃśādikam / saṃhitā ca mahāsaṃ- hitādikā tantraṃ ca tantrabhāgavatādikaṃ brahmatarkādikam iti jñeyam // (Jīva 1983: 84). 274 Kiyokazu Okita Conclusion In this chapter, I have explored the issue of Madhva’s untraceable quotations and its implications. As I described, the modern controversy between Mes- quita and Sharma has a prelude that goes back at least to the 16th century. Appayya criticizes Madhva for citing unknown texts because this goes against the Vedic norm. Appayya’s presupposition is that Vedānta is funda- mentally a discipline of scriptural exegesis and one’s opinions should be based on texts that are commonly available and accepted as authoritative.47 Clooney’s view that Vedānta is better understood as theology resonates with Appayya’s argument. On the one hand, Appayya criticizes Madhva for transgressing the limits of those who follow the Vedas (vaidikamaryādollaṅghana). On the other hand, according to Jīva, many works that were once existent are lost and therefore no longer available. If we embrace this view, then what exactly constitutes authoritative Vedic texts suddenly becomes ambiguous. The phe- nomenon of lost texts does in fact occur, as I discussed above (section 2), in the case of citations from the Skanda Purāṇa by Dharmanibandha authors. For a long time, these citations were regarded as pseudo-citations, because they could not be traced in the vulgate edition of the Skanda Purāṇa (see Khemarāja Śrīkṛṣṇadāsa 1910). More recently, however, they have been found in a text bearing the same title in a recension edited by Andriaensen, Bakker and Isaacson. For Jīva therefore, the obscure origin of texts does not necessarily indicate that they lack scriptural authority. Certainly Mesquita’s contribution is groundbreaking and significant. At the same time, I hope to have shown that the issue is more complex than he presented it. In relation to the architectural analogy, Appayya’s view is that only those building materials can be used whose origin is readily known. In contrast, Jīva argues that building blocks can be used effectively simply because of the fact that an established tradition uses them. Furthermore, since Jīva cites Madhva’s untraceable quotes, later Gauḍīya followers such as Viśvanātha Cakravartī in the 17th century and Baladeva Vidyābhūṣaṇa in the 18th century also refer to quotes of Madhva cited by Jīva.48 In other words, since Jīva 47 While I believe violating scriptural evidence (śabdapramāṇa) has a particularly signifi- cant implication in exegetical traditions such as Mīmāṃsā and Vedānta, such a viola- tion would also be problematic in other schools such as Nyāya, even though they are not exclusively dependent on scriptural evidence. This topic is unfortunately beyond the scope of the current chapter, but I thank Dr. Elisa Freschi for having pointed this out. 48 See Okita 2011 on this point.  Quotation, Quarrel and Controversy 275 became the authority in the Gauḍīya tradition, the fact in itself that Jīva used Madhva’s citations added authority to them. Thus, the reuse of Madhva’s un- traceable quotes in the Gauḍīya tradition presents an intriguing phenomenon of reuse in which an influential writer enhanced the authority of (alleged) ci- tations simply by citing them again. References Adriaensen, Bakker and Isaacson 1998 = Adriaensen, R. H., T. Bakker and H. Isaacson, eds. The Skandapurāṇa. Volume 1. Adhyāyas 1– 25. Groningen: Forsten, 1998. Groningen Oriental Studies, Supplement. Appayya 1940 Appayya Dīkṣita. Madhvatantramukhamardanam. Ed. Rā- macandra Śāstrī. Puṇyākhyapattane: Vināyak Gaṇesh Āpṭe, 1940. Ānandāśramasaṃskṛtagranthāvaliḥ 113. Appayya 1941 — The Madhvatantramukhamardanam. With his Auto- commentary Madhvamatavidhvaṃsanā and Cinnasvāmi Śāstrī’s Sandarbhavivaraṇa. Ed. Rāmanātha Dīkṣita. Kāśī: Hitacintakamudraṇālaya, 1941. Apte 1992 (1957) Apte, V. S. “Appendix E: A Collection of Popular Sanskrit Maxims (Nyāyasaṃgraha).” The Practical Sanskrit- English Dictionary, Revised & Enlarged Edition. Kyoto: Rinsen Book Company, 1992. Bakker 2004 Bakker, Hans T., ed. Origin and Growth of the Purāṇic Text Corpus: With Special Reference to the Skandapu- rāṇa. Delhi: Motilal Banarsidass, 2004. Papers of the 12th World Sanskrit Conference Held in Helsinki, Finland, 13– 18 July, 2003 3.2. Bronner 2007 Bronner, Yigal. “Singing to God, Educating the People: Appayya Dīkṣita and the Function of Stotras.” Journal of the American Oriental Society 127.2 (2007): 113‒130 (1– 18). Broo 2006 Mans Broo. “Jīva Gosvāmin and the Extent of Vedic Para- digm.” Journal of Vaishnava Studies 15.1 (2006): 5–30. Buchta 2003 Buchta, David. “Gems from the Gītā-Bhūṣaṇa.” Journal of Vaishnava Studies 12.1 (2003): 127–148. Buchta 2005 — “Baladeva Vidyābhūṣaṇa and the Vedāntic Refutation of Yoga.” Journal of Vaishnava Studies 14.1 (2005): 181– 208. Buchta 2006 — “Purāṇas as Śruti.” Journal of Vaishnava Studies 15.1 (2006): 87–108. 276 Kiyokazu Okita Buchta 2007 — “Baladeva’s Multi-Regional Influences.” Journal of Vaishnava Studies 15.2 (2007): 81–94. Clooney 2003 Clooney, Francis: “Restoring ‘Hindu Theology’ as a Cate- gory in Indian Intellectual Discourse.” The Blackwell Companion to Hinduism. Ed. Gavin Flood. Oxford: Blackwell Publishing, 2003. 447–477. de Jong 1999 de Jong, J. W. “Review of Madhva und seine unbekannten literarischen Quellen, by Roque Mesquita.” Indo-Iranian Journal 4.2 (1999): 63–64. De 1961 De, Sushil Kumar. Early History of the Vaisnava Faith and Movement in Bengal, from Sanskrit and Bengali Sources. 2nd ed. Calcutta: Firma K. L. Mukhopadhyay, 1961. Edelmann 2013 Edelmann, Jonathan. “Hindu Theology as Churning the Latent.” Journal of the American Academy of Religion 81.2 (2013): 427‒466 (1–40). Elkman 1986 Elkman, Stuart Mark. Jīva Gosvāmin’s Tattvasandarbha: A Study on the Philosophical and Sectarian Development of the Gaudīya Vaiṣṇava Movement. Delhi: Motilal Banarsi- dass, 1986. Freschi 2012 Freschi, Elisa. “Proposals for the Study of Quotations in Indian Philosophical Texts.” Religions of South Asia 6.2 (2012): 161–189. Govindacharya 1997 Govindacharya, Bannanje. “An Interview with Professor Bannanje Govindacharya on the Life of Madhvacarya.” Journal of Vaishnava Studies 5.3 (1997): 7–16. Gupta 2007 Gupta, Ravi. The Caitanya Vaiṣṇava Vedānta of Jīva Go- svāmī: When Knowledge Meets Devotion. London, New York: Routledge, 2007. Routledge Hindu Studies Ser. Hawley 2013 Hawley, John S. “How Do the Gauḍīyas Belong? Kavikar- ṇapūra, Jaisingh II, and the Question of Sampradāy.” Jour- nal of Hindu Studies 6.1 (2013): 114–130. Jīva 1983 Jīva Gosvāmī. Śrī Tattvasandarbhaḥ Śrīla-Jīva-gosvāmī- prādapraṇītaḥ. Śrīla-Jīva-gosvāmi-kṛtayā Sarva-samvā- dinī-vyākhyayā, Śrīmad-Baladevavidyābhūṣaṇa-kṛtayā ṭī- kayā, Śrī-Rādhāmohana-gosvāmi-viracitayā ṭīkayā, Śrī- Gaurakiśora-gosvāmi-kṛtayā Svarṇaletety-ākhyayā ṭīkayā, Śrī-Haridāsa-śāstri-praṇītayā vyākhyayā ca sambalitaḥ. Vṛndāvana: Śrī Gadādharagaurahari Press, 1983. Jīva 1983a Sarvasaṃvādinī in Jīva 1983. Kapoor 1976 Kapoor, O. B. L. The Philosophy and Religion of Sri Cai- tanya. Delhi: Munshiram Manoharlal, 1976. Khemarāja Śrīkṛṣṇadāsa 1910 = Khemarāja Śrīkṛṣṇadāsa, ed. Śrī-Skāndaṃ Mahā- purāṇam. 7 vols. Bombay: Veṅkaṭeśvara Press, 1910.  Quotation, Quarrel and Controversy 277 Madhva 1971 Madhva. Mahābhāratatātparyanirṇaya and Mahābhārata- tātparyam. Vol. 2 of Sarvamūlagranthāḥ. Ed. Bannanje Govindacharya. Udupi: Akhila Bhārata Mādhva Mahā Maṇḍala, 1971. Mesquita 1997 Mesquita, Roque. Madhva und seine unbekannten literari- schen Quellen. Einige Beobachtungen. Wien: Sammlung de Nobili, 1997. Publications of the De Nobili Research Library 24. Mesquita 2000 — Madhva’s Unknown Literary Sources: Some Observa- tions. Delhi: Aditya Prakashan, 2000. Mesquita 2003 — “The Rank and Function of God Vāyu in the Philo- sophy of Madhva.” Indo-Iranian Journal 46 (2003): 97– 117. Mesquita 2007 — Madhvas Zitate aus den Purāṇas und dem Mahābhā- rata: Eine analytische Zusammenstellung nicht identifi- zierbarer Quellenzitate in Madhvas Werken nebst Über- setzung und Anmerkungen. Wien: Sammlung de Nobili, 2007. Publications of the De Nobili Research Library 34. Mesquita 2008 — Madhva’s Quotes from the Purāṇas and the Mahābhā- rata: An Analytical Compilation of Untraceable Source- Quotations in Madhva’s Works along with Footnotes. Delhi: Aditya Prakashan, 2008. Minkowski 2008 Minkowski, Christopher. “Why Should We Read the Maṅ- gala Verses?” Śāstrārambha: Inquiries into the Preamble in Sanskrit. Ed. Walter Slaje. Wiesbaden: Harrassowitz, 2008. 1–24. Abhandlungen für die Kunde des Morgen- landes 62. Oberlies 1995 Oberlies, Thomas. “Die Śvetāśvatara Upaniṣad: Einleitung – Edition und Übersetzung von Adhyāya I.” Wiener Zeit- schrift für die Kunde Südasiens 39 (1995): 61‒102. Okita 2008 Okita, Kiyokazu. “Mādhva or Gauḍīya? The Philosophy of Baladeva Vidyābhūṣaṇa’s Prameyaratnāvalī.” Journal of Vaishnava Studies 16.2 (2008): 33–48. Okita 2009 — “A Bengali Vaiṣṇava Contribution to Vedānta: Bala- deva Vidyābhūṣaṇa on the Word atha in the Brahmasūtra 1.1.1.” Journal of Vaishnava Studies 18.1 (2009): 87–100. Okita 2011 — “Review of Madhva’s Quotes From the Purāṇas and the Mahābhārata: An Analytical Compilation of Untrace- able Source-Quotations in Madhva’s Works along with Footnotes, by Roque Mesquita.” Indo-Iranian Journal 54 (2011): 185–192. Okita 2014a — “Caitanya Vaiṣṇavism on Trial: Continuity and Trans- formation in the Eighteenth Century.” Caitanya Vaiṣṇava 278 Kiyokazu Okita Philosophy: Tradition, Reason, and Devotion. Ed. Ravi M. Gupta. Farnham, Surrey: Ashgate, 2014. 75–112. Okita 2014b — Hindu Theology in Early Modern South Asia: The Rise of Devotionalism and the Politics of Genealogy. Oxford: Oxford University Press, 2014. Olivelle 1993 Olivelle, Patrick. The Āśrama System: The History and Hermeneutics of a Religious Institution. New York: Oxford University Press, 1993. Olivelle 1999 — Dharmasūtras: The Law Codes of Āpastamba, Gautama, Baudhāyana, and Vasiṣṭha. New York: Oxford University Press, 1999. Olivelle 2006 — “Explorations in the Early History of Dharmaśāstra.” Between the Empires: Society in India 300 BCE to 400 CE. Ed. Patrick Olivelle. New York: Oxford University Press, 2006. 169–190. Pālghāṭ Nārāyaṇa Śāstri 1940 = Pālghāṭ Nārāyaṇa Śāstri. Makhivaryabhūṣaṇā in Appayya 1940. Pellegrini 2015 Pellegrini, Gianni. “‘Old is Gold!’ Madhusūdana Saras- vatī’s Way of Referring to Earlier Textual Tradition.” Journal of Indian Philosophy 43.4–5 (2015): 1–58. Rādhā Mohana Gosvāmī 1983 = Rādhā Mohana Gosvāmī. Ṭīkā in Jīva 1983. Rocher 1986 Rocher, Ludo. The Purāṇas. Wiesbaden: Harrassowitz, 1986. A history of Indian literature 2.3. Rocher 2008 — “Review of Madhvas Zitate aus den Purāṇas und dem Mahābhārata: Eine analytische Zusammenstellung nicht identifizierbarer Quellenzitate in Madhvas Werken nebst Übersetzung und Anmerkungen.” Journal of the American Oriental Society 128.3 (2008): 603–606. Śabarasvāmin and Yudhiṣṭhira Mīmāṃsaka 1977 = Śabarasvāmin and Yudhiṣṭhira Mīmāṃsaka, ed. Jaiminīyamīmāṃsābhāṣyam: Ārṣamata- vimarśinyā Hindīvyākhyayā sahitam. Bahālagadha: Yudhi- ṣṭhira Mīmāṃsaka, 1977. Sanderson 2003 Sanderson, Alexis. “The Śaiva Religion among the Khmers, Part 1.” Bulletin de l’Ecole française d’Extrême- Orient 90–91 (2003–2004): 349–463. Sanderson 2005 — “A Commentary on the Opening Verses of the Tantra- sāra of Abhinavagupta.” Sāmarasya: Studies in Indian Arts, Philosophy, and Interreligious Dialogue – in Honour of Bettina Bäumer. Ed. Sadananda Das and Ernst Für- linger. New Delhi: D. K. Printworld, 2005. 89–148. Sanderson 2007 — “Atharvavedins in Tantric Territory: The Āngirasakal- pa Texts of the Oriya Paippalādins and their Connection with the Trika and the Kālīkula, with Critical Editions of the Parājapavidhi, the Parāmantravidhi, and the Bhadra-  Quotation, Quarrel and Controversy 279 kālīmantravidhiprakarana.” The Atharvaveda and its Paip- palāda Śākhā: Historical and Philological Papers on a Vedic Tradition. Ed. Arlo Griffiths and Annette Schmied- chen. Aachen: Shaker Verlag, 2007. 195–311. Sanderson 2009 — “Kashmir.” Brill’s Encyclopedia of Hinduism, Volume One: Regions, Pilgrimage, Deities. Ed. Knut A. Jacobsen. Leiden: Brill, 2009. 99–126. Sharma 1994 Sharma, B. N. K. Advaitasiddhi vs Nyāyāmṛta: An Up To Date Critical Re-Appraisal. Part 1. Bangalore: Ānanda- tīrtha Pratiṣṭhāna, 1994. Sharma 2000 — History of the Dvaita School of Vedānta and Its Litera- ture: From the Earliest Beginnings to Our Own Times. 3rd rev. ed. Delhi: Motilal Banarsidass, 2000. Sharma 2001 — My Latest Four Research Papers. Mumbai: Navin Printers, 2001. Uskokov 2009 Uskokov, Aleksandar. “Bengali Vaishnava Epistemolo- gy.” Journal of Vaishnava Studies 18.1 (2009): 53–86.  Reusing, Adapting, Distorting? Veṅkaṭanātha’s Reuse of Rāmānuja, Yāmuna (and the Vṛttikāra) in his Commentary ad Pūrvamīmāṃsāsūtra 1.1.1∗ Elisa Freschi 1 Early Vaiṣṇava synthesizing philosophies Vaiṣṇavism has a long history in India in general and in South India in par- ticular. At the time of Rāmānuja (traditional dates: 1017‒1137), some of the texts later recognized as milestones of Śrī Vaiṣṇavism had already been com- posed. These are the texts (and practices) belonging to the Pāñcarātra corpus, whose origin can be placed in 9th-century Kashmir (Sanderson 2009b: 61–62, Leach 2012: 7), the hymns of the Āḻvārs and the treatises of certain theolo- gians such as Nāthamuni (traditional date of birth: 824) and Yāmunācārya (traditional dates: 918–1038).1 It is still unclear – as the following pages will show – who first attempted a synthesis of these three elements,2 not to men- tion the various philosophical currents (from Nyāya to Pūrva Mīmāṃsā and especially Vedānta) that have been part of the construction of later Śrī Vaiṣ- ∗ I am grateful to Marion Rastelli for several enlightening and enjoyable conversations about Pāñcarātra and related topics. I am also very much obliged to Philipp Maas for his careful reading of an earlier draft of this article. Research for this article was fi- nanced by the Austrian Science Fund (FWF) in the context of the FWF-project V-400. 1 See Neevel 1977: 14–16 for an explanation of these dates through the hypothesis that the life spans of the great masters of Śrī Vaiṣṇavism were lengthened in order to con- nect them directly to each other. In his 1977 book Neevel announces a forthcoming study focusing only on the issue of dates, but I have been unable to locate it (if it was ever written). 2 Neevel writes that Nāthamuni, whose two works have not been preserved, may have been the first to “discover” the hymns within a philosophical context (Neevel 1977: 88–89) but his main testimonies in this regard are later hagiographical biographies. Yāmuna has also been traditionally connected to the Āḻvārs (Neevel 1977: 88); the to- pos of being the worst human being and thus needing God’s compassion can be found in the Āḻvārs’ hymns as well as in the maṅgala verses of Yāmuna’s Stotraratna. 282 Elisa Freschi ṇava philosophy. I have, however, the strong impression that the synthesis of these elements as we know it was created by Veṅkaṭanātha (also known by the title of Vedānta Deśika, traditional dates 1269–1370). In his Āgamaprāmāṇya, Yāmuna defended the epistemological validity of the Pāñcarātra and engaged in the philosophical debate of his time. Nonethe- less, not Yāmuna but Rāmānuja is regarded as the founder of what is usually known as Viśiṣṭādvaita Vedānta (i.e., the philosophical counterpart of Śrī Vaiṣṇavism).3 Why? Perhaps because he offered the new-born system a well- developed number of works dealing with its principal tenets. Further, he wrote a commentary on the Brahmasūtra (henceforth BS), a practice that was probably already becoming the standard way of establishing a distinct Vedānta tradition.4 However, notwithstanding the status Rāmānuja gained for Viśiṣṭādvaita Vedāntins, his Śrībhāṣya (henceforth ŚrīBh) does not “feel” like a ground- breaking work, but rather like one that is part of a paramparā, a ‘teacher- pupil lineage.’5 In fact, Rāmānuja apparently based his work on already established interpretations, vocabulary and approaches. Apart from the au- thors mentioned above who were later recognized as the precursors of Viśi- ṣṭādvaita Vedānta, one might also recall among the forerunners of Rāmānuja an elusive Vṛttikāra, who may have commented on both the PMS and the BS6 3 Even Gerhard Oberhammer, who attempted a careful historical and philosophical investigation of the first stages of Viśiṣṭādvaita Vedānta, calls it the “Rāmānuja Schule”; see Oberhammer 1979 and the volumes following this publication. A simi- lar depiction is made in the second foreword of Narasimhachary 1998. 4 The practice had in fact already been adopted by Śaṅkara and Bhāskara and would be adopted by Madhva and Vallabha. The Gauḍīya Vaiṣṇavas, whose early figures (Cai- tanya and the three Gosvāmins Jīva, Sanātana and Rūpa) had not written a commentary on the BS, had to produce one in order to be accepted as a distinct sampradāya (see Okita’s chapter in the present volume, section 4.2). The fact that Śrīvatsaṅka Miśra’s (now lost) commentary on the BS was not accepted as the standard commentary among Śrī Vaiṣṇavas-Viśiṣṭādvaita Vedāntins seems to indirectly prove that it did not conform to what later came to be recognized as the basic tenets of the school. On Śrīvatsaṅka Miśra, see Oberhammer 1971: 118–119 and Neevel 1977: 69–75. Oberhammer high- lights the fact that Śrīvatsaṅka’s commentary was extraneous to Yāmuna’s tradition (and thus, I may add, also extraneous to Veṅkaṭanātha, who in fact never mentions it): “Vor Yāmuna hat es kaum eine Tradition von Brahmasūtren-Kommentaren des Pāñc- arātra gegeben, und Yāmuna mußte daher, wollte er von der von ihm vertretenen Auf- fassung der Vedānta-Lehren im Sinne einer Tradition sprechen, schulfremde Kommen- tatoren nennen […]” (Oberhammer 1971: 119). 5 The usage of single quotation marks for words’ meanings and of double quotation marks for the interpretation of sentences is the standard in linguistics, see Haspelmath 2014. 6 Against this view, see Bronkhorst 2007, end of section 1. Given that the present  Reusing, Adapting, Distorting? 283 and remained (see section 4.2) very influential for centuries, certainly until the time of Veṅkaṭanātha.7 In the opening of his ŚrīBh, Rāmānuja possibly refers to him as Bodhāyana.8 In this sense, Rāmānuja was a systematizer. He turned Śrī Vaiṣṇavism into a Vedāntic school, acknowledging the Upaniṣads, the Bhagavadgītā and the BS (the three texts acknowledged as the three foundations – prasthā- natrayī – of all Vedānta schools) as its foundations. Veṅkaṭanātha, however, went one step further. He tried to integrate the other schools of Indian philo- sophy into Viśiṣṭādvaita Vedānta, from Nyāya (Veṅkaṭanātha dedicated his Nyāyapariśuddhi to the goal of including Nyāya in Viśiṣṭādvaita Vedānta) to Pūrva Mīmāṃsā (see his Seśvaramīmāṃsā and Mīmāṃsāpādukā) and Pāñca- rātra’s apology (see his Pāñcarātrarakṣā). He wrote in Prakrit, Tamil, Maṇi- pravāḷa and Sanskrit, in the form both of śāstras and beautiful poems. His po- ems are full of the Āḻvārs’ poetical and mystical images, whereas his śāstras abound with scholarship. In this sense, Veṅkaṭanātha can be seen as a con- tinuator of the intellectual legacy of both Rāmānuja (especially as far as Ve- dānta is concerned, but [see section 4] also with regard to Pūrva Mīmāṃsā) and Yāmuna (as far as the Pāñcarātra and perhaps the Āḻvārs’ heritage are concerned). 2 Veṅkaṭanātha as a continuator of Rāmānuja (and of Yāmuna) In this chapter, I shall focus on Rāmānuja’s ŚrīBh and Veṅkaṭanātha’s Seśva- ramīmāṃsā (SM). In many ways, the two texts resemble each other. The first is a commentary on the BS (also called Uttaramīmāṃsāsūtra, UMS), whereas the second comments on the Pūrvamīmāṃsāsūtra (PMS). Further, while the first implicitly aims at appropriating the BS, the latter explicitly aims at gaining the PMS as part of the system Rāmānuja had established. More importantly, the Seśvaramīmāṃsā (henceforth SM) explicitly adopts the ŚrīBh’s agenda of showing how a (somewhat) different school was, in fact, part of one’s own and propounding the idea of the existence of an aikaśāstrya ‘unity of the teaching.’ However, Veṅkaṭanātha’s undestanding chapter focuses on Veṅkaṭanātha’s perspective on the unity of Pūrva and Uttara Mīmāṃsā, I will leave aside Bronkhorst’s points against the historicity of this hypothesis. 7 For an insightful excursus on the Vṛttikāra, see the contribution by Andrijanić to this volume. 8 bhagavadbodhāyanakṛtāṃ vistīrṇāṃ brahmasūtravṛttiṃ pūrvācāryāḥ saṃcikṣipuḥ, tanmatānusāreṇa sūtrākṣarāṇi vyākhyāsyante (Krishnamacharya 1938: 78). 284 Elisa Freschi of this slogan is in part different from and extends how it was used by Rā- mānuja. At the end of this chapter, I shall show how Veṅkaṭanātha also shares much of Yāmuna’s agenda. In fact, in several cases it is possible to detect whole passages from Yāmuna’s work being reused by Veṅkaṭanātha. How- ever, these passages are often silently embedded in Veṅkaṭanātha’s text. By contrast, Rāmānuja’s texts are also frequently embedded, but mentioning Rāmānuja is used as a philosopher’s stone, i.e., as the final argument that turns every argument into knowledge. Rāmānuja is thus the constant refer- ence point that settles any discussion. Why so? Perhaps because Yāmuna’s works were superseded by Rāmānuja’s, just as Yāmuna’s works had super- seded the works of his grandfather Nāthamuni.9 However, a further reason might be that Yāmuna’s agenda displays a major difference from that of Veṅkaṭanātha: Yāmuna’s Āgamaprāmāṇya is usually interpreted as having been devised against Pūrva Mīmāṃsakas (see Mesquita 1971, Neevel 1977, Mesquita 1980), whereas Veṅkaṭanātha went so far as to write genuine Pūrva Mīmāṃsā texts. Although Mesquita and Neevel may have exaggerated the anti-Pūrva Mīmāṃsā component of Yāmuna’s agenda (see below, section 5, for his respectful mention of Jaimini), he clearly tried to achieve what the Mīmāṃsaka Kumārila had tried to avoid some centuries earlier, namely the recognition of Pāñcarātra texts at the same level as the Vedas. Moreover, he used Nyāya arguments (e.g., the validity of the Pāñcarātra is said to depend on their author, God),10 and his general attitude is not yet clearly (Uttara) Mīmāṃsaka, unlike in the case of Rāmānuja.11 Using an analogy from another Indian school, the situation resembles that of Somānanda and Utpaladeva, with the former fighting against Bhartṛhari and the latter embedding Bhartṛhari’s ideas in a fruitful way in his own sys- tem in order to make a sounder philosophical school out of it (see Torella 2008). To sum up, Veṅkaṭanātha took on much of Yāmuna’s agenda, from the appraisal of Pāñcarātra to theism, but ultimately could not agree with his philosophical positions. As for the form of the general frame to be adopted, 9 “One of the reasons why the Nyāyatattva of Nāthamuni’s sank into oblivion and got lost, could be due to it being eclipsed by Yāmuna’s works. The same could be said of Yāmuna’s philosophical works ‒ the three Siddhis ‒ which are only fragmentarily pre- served. They themselves were also eclipsed by Rāmānuja’s Śrībhāṣya, as Neevel rightly observes” (Mesquita 1980: 211). 10 For a distinction between Mīmāṃsā- and Nyāya-based arguments concerning the valid- ity of sacred texts, see the introductory study in Freschi and Kataoka 2012. 11 My paper on Yāmuna is currently under preparation. Some of its conclusions are sum- marized on The Indian Philosophy Blog (http://wp.me/p486Wp-7C).  Reusing, Adapting, Distorting? 285 Veṅkaṭanātha agreed more with Rāmānuja’s aikaśāstrya suggestion – and pushed it much further. 3 The Śrībhāṣya and the Seśvaramīmāṃsā: Shared textual material It is, accordingly, not surprising that Veṅkaṭanātha reused the ŚrīBh to a large extent. After all, Veṅkaṭanātha’s attempt was risky, since the Pūrva Mīmāṃsā was known as an atheist system and he needed to be backed by his prede- cessor’s implicit support.12 Indeed, the SM is replete with quotations from the ŚrīBh. For our present purposes, however, it is perhaps more relevant that the SM reused a great deal of textual material from the ŚrīBh without acknowl- edging it. 3.1 Examples The following instances exemplify cases of reuse. Textual passages which are formally identical or almost identical (Ce/Ce’ or Cee/Ce’e in Steinkellner’s terminology, see Steinkellner 1988, Freschi 2012, Freschi 2015) are printed in grey. Textual passages which display only correspondences in the content (Re in Steinkellner’s terminology) have a wavy underline. The translations are purely indicative; the specificities of each text will need to be the object of a separate study. 3.1.1 The beginning of the commentary athātho brahmajijñāsā || 1.1.1 || atrāyam athaśabda ānantarye bhavati; ataśśabdo vṛttasya hetubhāve. (ŚrīBh ad 1.1.1)13 12 A similar device was implemented in Veṅkaṭanātha’s Pāñcarātrarakṣā, where an ac- knowledged quote from the BS was used at a critical point in order to show how the BS (in fact, the BS according to Rāmānuja’s interpretation, but Veṅkaṭanātha consis- tently refers to the unity of BS and ŚrīBh) already reconciled different Upaniṣadic au- thorities – and, thus, perhaps indirectly also in order to justify Veṅkaṭanātha’s attempt to harmonize different Pāñcarātra authorities. The reused passage is from the ŚrīBh ad BS 3.2.7 and is found in the Pāñcarātrarakṣā, Duraiswami Aiyangar and Venugo- palacharya 1942: 166. For more details, see Rastelli forthcoming: 7. 13 “Now, because of that, the desire to know the brahman starts || 1.1.1 || In this [sūtra] this word atha is [used] in the sense of [designating] an immediate succession. The word ataḥ is [used] in the sense of [stating that] something which has already occurred is the cause [of the desire to know].” (ŚrīBh ad BS 1.1.1). Please note that jijñāsā liter- ally means ‘desire to know,’ but it can be more smoothly translated with “investiga- tion.” 286 Elisa Freschi athātho dharmajijñāsā || 1.1.1 || athaśabdo ’trāyam ānantaryārtha eva prasiddhiprakarṣasyānapavāda- tvāt. maṅgalārthatvaṃ tv arthadharmaśabdayor anyārthatve ’pi pradī- pādivat syāt. yady apy atra pūrvaṃ na kiṃcid api nirdiṣṭam, tathāpy ataḥpadopādānena svarasataḥ sāmānyena kasyacid vṛttasya hetutokteḥ tasya dharmajijñāsāpadasamabhivyāhārasāmarthyāt tadapekṣitatayā viśeṣyamāṇatvāc ca sāṅgavedādhyayanarūpavṛttaviśeṣasiddhiḥ. na hi vedādhyayanāt pūrvam anyasmād vā niyatapūrvavṛttād anantaraṃ vedārthavicāraḥ taduktir vāvakalpate. ata evātra ataśśabdo ’pi vṛttasya hetubhāvārtha eva. (SM ad PMS 1.1.1)14 It is easy to see how Veṅkaṭanātha reused Rāmānuja’s text but extended it further with an interposed excursus. Where does this excursus come from? It was probably inspired by Śaba- ra’s commentary on PMS 1.1.1: tatra loke ’yam athaśabdo vṛttād anantarasya prakriyārtho dṛṣṭaḥ. na ceha kiñcid vṛttam upalabhyate. bhavitavyaṃ tu tena, yasmin saty an- antaraṃ dharmajijñāsāvakalpate. tathā hi prasiddhapadārthaka sa kal- pito bhavati. tat tu vedādhyayanam. tasmin hi sati sāvakalpate. naitad evam. anyasyāpi karmaṇo ’nantaraṃ dharmajijñāsā yuktā, prāg api ca vedādhyayanāt. ucyate […] tasmād vedādhyayanam eva pūrvam abhi- nirvartyānanataraṃ dharmo jijñāsitavya ity athaśabdasya sāmarthyam. na brūmo ’nyasyāpi karmaṇo ’nantaraṃ dharmajijñāsā na kartavyā iti, kiṃ tu vedam adhītya tvaritena na snātavyam, anantaraṃ dharmo jijñāsitavya ity athaśabdasyārthaḥ. ataḥśabdo vṛttasyāpadeśako hetvar- 14 “Now, because of that, the desire to know the dharma starts || 1.1.1 || In this [sūtra] this word atha has the meaning of an immediate succession, because a contact with some- thing very well known is not refuted (apavāda). But, although the words atha and dharma have a different meaning, they can have also indicate auspiciousness (maṅ- gala), like a lamp [which illuminates at the same time more than one thing]. Although here nothing had been made explicit before, nonetheless with the usage of the word ataḥ one automatically establishes a specific event, being the study of the Veda to- gether with its auxiliaries. It is so, because in general something which occurred before is mentioned as the cause, because this preceding [event] can possibly occur together with the word ‘investigation into the dharma’ (whose meaning is well known), and be- cause [the thing that occurred before] is specified by the fact that it (the investigation) depends on it. In fact, it is not the case that the investigation into the meaning of the Veda or its proclamation is possible before the study of the Veda or immediately after another preceding event that necessarily happened before. For this very reason, in this passage also the word ataḥ has the meaning of [stating that] something which has al- ready occurred is the cause [of the desire to know].” (SM ad PMS 1.1.1)  Reusing, Adapting, Distorting? 287 thena, yathā kṣemasubhikṣo ’yaṃ deśaḥ, ato ’ham asmin deśe prati- vasāmīti. evam adhīto vedo dharmajijñāsāyāṃ hetur jñātaḥ, anantaraṃ dharmo jijñāsitavyaḥ ity ataḥśabdasya sāmarthyam. (ŚBh ad PMS 1.1.1, Frauwallner 1968: 10–14)15 Note that the ŚBh on PMS 1.1.1 starts with a legitimation of the reference to worldly usages of words and that the term loka, thus, plays a key role there. In the ŚrīBh and the SM, by contrast, the issue does not even arise, perhaps because Śabara had apparently settled it once and for all. 3.1.2 Commentary on jijñāsā brahmaṇo jijñāsā brahmajijñāsā. brahmaṇa iti karmaṇi ṣaṣṭhī; kartṛkar- maṇoḥ kṛti iti viśeṣavidhānāt. yady api saṃbandhasāmānyaparigrahe ’pi jijñāsāyāḥ karmāpekṣatvena karmārthatvasiddhiḥ – tathāpy ākṣepa- taḥ prāptād ābhidhānikasyaiva grāhyatvāt karmaṇi ṣaṣṭhī gṛhyate. (ŚrīBh ad BS 1.1.1)16 15 “In this regard, we commonly experience that this word atha is used in the meaning of [something occurring] immediately after something has already occured. However, in the context of this [sūtra] one does not grasp anything which has already occurred. But there should be something [which has already occurred], so that once this occurs, im- mediately after it the desire to know dharma is adequate. To elaborate: it is postulated that this [word atha] has its [usual] well-known meaning. And, this [thing, which has already occurred] is the study of the Veda, for, once it has occurred, that (desire to know dharma) is adequate. [Obj.:] It is not so. It is suitable that one desires to know dharma also after any other action, and also before the study of the Veda. [R.:] […] Therefore the word atha means that immediately after having accomplished the study of the Veda the dharma must be investigated (lit. “desired to be known,” see n. 13). We (Mīmāṃsakas) do not claim that the investigation of dharma should not be done after another action. Rather, the meaning of the word atha is that after having studied the Veda one should not hurry up to take a bath, but rather one should immediately after it (the study of the Veda) investigate dharma. The word ataḥ points to something which has already occurred as the cause, as when one says ‘this region is prosperous and has abundant food, therefore (ataḥ) I live in this region.’ In this way, the word ataḥ means that the Veda, which has been studied, is understood as the cause for the desire to know dharma [and] that immediately after [the study of the Veda] one should investigate dharma.” The text of ŚBh ad PMS 1.1.1 has already been translated into German by Frauwallner (Frauwallner 1968: 11–15). The expression prakriyārtha is unusual; I have not seen it in any other text. The context and the fact that the sub-commentators do not dwell on it make me think that the compound only emphasized the ordinariness of the meaning suggested. Cf. Frauwallner’s translation: “Wir sehen doch, daß das Wort „danach“ im täglichen Leben normalerweise etwas bedeutet, das […]” (Frau- wallner 1968: 11). 16 “The brahman desire (brahmajijñāsā) is the desire to know the brahman. The genitive 288 Elisa Freschi dharmasya jijñāsā dharmajijñāsā. karmaṇy evātra ṣaṣṭhī samban- dhyasāmānyaṣaṣṭhīm aṅgīkṛtya tena tatra karmasāpekṣajijñāsāsā- marthyataḥ karmārthatvakḷpter vilambitatvāt. kartṛkarmaṇoḥ kṛti iti viśeṣavidhānāt. (SM ad PMS 1.1.1) 17 In this second passage, the strategy is different: no excursus is inserted, just Rāmānuja’s points, but they are reorganized in a different sequence. This is followed by a more detailed discussion of the grammatical issue at stake. By contrast, Śabara does not analyze the compound dharmajijñāsā as pre- supposing a ṣaṣṭhī ‘genitive case’: dharmāya jijñāsā dharmajijñāsā. sā hi tasya jñātum icchā. sa punaḥ kathaṃ jijñāsitavyaḥ? ko dharmaḥ, kathaṃlakṣaṇaḥ, kāny asya sādhanāni, kāni sādhanābhāsāni, kiṃparaś ceti. (ŚBh ad PMS 1.1.1)18 3.1.3 vyatireka cases For the sake of comparison it might be useful to take a look at Śaṅkara’s and Bhāskara’s commentaries on the same sūtra, which are the only ones predat- ing the ŚrīBh that are extant. 3.1.3.1 Śaṅkara’s commentary on the same sūtra Apart from some elements common to all of the commentaries on the words included in UMS 1.1.1 and PMS 1.1.1, Śaṅkara elaborates on altogether ending in brahman denotes the object (genitivus objectivus) because of the specific [grammatical] rule ‘[The sixth case ending is employed after a stem] meaning agent or object, when [used] along with a kṛt affix’ (Aṣṭādhyāyī 2.3.65). [Obj.:] Once the gen- eral connection has been seized, the meaning of [the genitive in brahman] is ascer- tained as the object because the desire to know requires an object (that is, even if the genitive did not indicate brahman as object, one would understand it as such because an object would be needed). [R.:] Notwithstanding, the genitive is understood as meaning the object because only lexical items are obtained through implication.” 17 “The dharma desire (dharmajijñāsā) is the desire to know dharma. Here the genitive termination denotes the object (genitivus obiectivus), 1. because, once one has accepted the genitive termination [as meaning] a general connection (‘an investigation related to dharma’), because of that, given that in the [sūtra] an investigation which [still] re- quires an object would be impossible, one would be delayed by the fact of [having to] imagine which [referent is implicitly] meant as the object, 2. because of the specific [grammatical] rule ‘[The sixth case ending is employed after a stem] meaning agent or object, when [used] along with a kṛt affix’ (Aṣṭādhyāyī 2.3.65).” (SM ad PMS 1.1.1) 18 “The dharma investigation is the investigation for the sake of dharma. In fact, this is the resolution to know it. And how, in turn, can this be investigated? What is dharma? How is it defined? What are the means to realize it? What are the erroneous means to realize it? What does it aim at?”  Reusing, Adapting, Distorting? 289 different topics. When he mentions, like Veṅkaṭanātha, the possibility of interpreting atha as having an auspicious meaning, he, unlike Veṅkaṭanātha, explicitly denies this. athāto brahmajijñāsā || 1.1.1 || tatrāthaśabda ānantaryārthaḥ parigṛhyate nādhikārārthaḥ, brahmajijñā- sāyāḥ anadhikāryatvāt. maṅgalasya ca vākyārthe samanvayābhāvāt. arthāntaraprayukta eva śrutyā maṅgalaprayojano bhavati. sati cānan- taryārthatve yathā dharmajijñāsā pūrvavṛttaṃ vedādhyayanaṃ niya- menāpekṣata evaṃ brahmajijñāsāpi yat pūrvavṛttaṃ niyamenāpekṣate tad vaktavyam. svādhyāyānantaryaṃ tu samānam. nanv iha karmāva- bodhanānantaryaṃ viśeṣaḥ. na. dharmajijñāsāyāḥ prāg apy adhītave- dāntasya brahmajijñāsopapatteḥ.[…]19 ataḥśabdo hetvarthaḥ. yasmād veda evāgnihotrādīnāṃ śreyaḥsādha- nānām anityaphalatāṃ darśayati – “tad yatheha karmacito lokaḥ kṣī- yata evam evāmutra puṇyacito lokaḥ kṣīyate” (ChUp 8.1.6) ityādiḥ. […] brahmaṇo jijñāsā brahmajijñāsā. brahma ca vakṣyamāṇalakṣaṇaṃ “janmād yasya yataḥ” iti. ata eva na brahmaśabdasya jātyādyarthānta- ram āśaṅkitavyam. brahmaṇa iti karmaṇi ṣaṣṭhī na śeṣe, jijñāsyāpe- kṣyatvāj jijñāsāyāḥ, jijñāsyāntarānirdeśāc ca. […] (BSBh ad BS 1.1.1)20 19 In the omitted sections the discussion on whether the knowledge of the Brāhmaṇa part of the Vedas is a presupposition of brahmajijñāsā is developed further. 20 “Now, because of that, the desire to know the brahman starts || 1.1.1 || In this [sutra] the word atha [must] be taken as meaning an immediate succesion, not as meaning the introduction [of a topic], because the desire to know the brahman does not need to be introduced. And [the word atha also does not have its common auspicious meaning] because auspiciousness cannot be connected within the meaning of the sentence. The [word atha], exactly while it is used for another purpose, can have the purpose of [con- veying] auspiciousness insofar as it is heard [and not because of its meaning]. And since [the word atha] means immediate succession, just as the desire to know the dharma necessarily requires a preceding study of the Veda, in the same way one must also say what preceding event the desire to know the brahman necessarily requires. The immediate succession after the study of one’s portion of the Veda is, however, common [to both Mīmāṃsās]. [Obj.:] But here, the specific element [which is required in order for the desire to know the brahman to take place] is the immediate succession after the ritual act. [R.:] No, because the desire to know the brahman is fit also before the desire to know the dharma for one who has studied the Upaniṣads. […] The word ataḥ means a reason, since the Veda itself shows that sacrifices such as the Agnihotra, which are means to realize one’s welfare, have no permanent result, e.g., in ‘therefore, as in this world a territory (loka) which has been gained through ritual actions comes to an end, so in the other a world gained through merit** comes to an end’. […] The 290 Elisa Freschi Śaṅkara then goes on to discuss whether the brahman is prasiddha ‘well- established’ or not and how in this sense it can be the object of an enquiry (if it is already well established it seems that no further investigation is needed). The same topic is present in the ŚBh but absent in the ŚrīBh and the SM. 3.1.3.2 Bhāskara’s commentary on the same sūtra Bhāskara’s commentary on BS 1.1.1 starts on a different note, that is, with the four different meanings of atha.21 Among them ānantarya is also in- cluded in Bhāskara’s interpretation of BS 1.1.1, whereas the meaning of auspicious (maṅgala) is absent. Overall, the readers will notice that the com- mentary is much closer to the one by Śaṅkara than to the ŚrīBh or the SM: athāto brahmajijñāsā || 1.1.1 || caturṣv artheṣv athaśabdaprayogo dṛśyate. yathā ānantarye bhuktvātha vrajati. yathā pūrvavṛttāpekṣāyām pūrvaṃ kiñcid uktvā vikalpāntaraṃ kartum icchann āha athāyam abhiprāya iti. yathā māṅgalye. […] ya- thādhikāre “atha śabdānuśānam”22 iti. tatra nādhikārārtho brahmaji- jñāsāyā anadhikāryatvāt. nāpi maṅgalārthaḥ, maṅgalasya vākyārthe ’nvayābhāvāt paṭahadhvanivat śravaṇamātragamyatvāt, pūrvavṛttāpe- kṣāyāś cānantaryāvyatirekāt ānantaryārtha iha gṛhyate. (Dvivedin 1903: 1; Kato 2011: 1–2)23 brahman desire is the desire to know the brahman. And brahman is that of which the definition will be provided in janmād yasya yataḥ [UMS 1.1.2]. Therefore, one should not suspect that the word brahman has a different meaning here, for instance that of [Brahmanical] caste. The genitive case in brahman denotes the object (genitivus objec- tivus); it is not [used] in the sense of [designating] any other [relation],* because the desire to know requires something to be known and no other things to be known are stated.” *This evokes Aṣṭādhyāyī 2.3.50, ṣaṣṭhī śeṣe “The genitive case is used in the case of any other [relation, not covered in the previous sutras].” **I have consulted Thibaut’s translation of Śaṅkara’s commentary (Thibaut 1890: 9–13) and Olivelle’s edition and translation of the Chāndogya Upaniṣad (Olivelle 1998: 275), which reports karmajito and puṇyajito for karmacito and puṇyacito. 21 Many thanks are due to Andrew Nicholson, who kindly sent me a copy of the editio princeps of Bhāskara’s commentary on the BS (Dvivedin 1903). I was later able to improve the text of the editio princeps by comparing it to the unpublished critical edi- tion of the same text, which Takahiro Kato kindly agreed to share with me. This criti- cal edition is an improved version of the critical edition Kato produced for his PhD (Kato 2011). Minor divergences between the two editions will not be noted, since they are beyond the scope of this chapter. 22 This is the beginning of the Mahābhāṣya. 23 “Now, because of that, the desire to know the brahman starts. The word atha is com- monly seen to be employed with four meanings. For instance, in the sense of [desig-  Reusing, Adapting, Distorting? 291 Once it is established that atha refers to something preceding the BS, Bhāskara wonders what this is and he refers in a prima facie view to the study of the ritual part of the Veda (not of Pūrva Mīmāṃsā, then, but of its object, the Brāhmaṇa prescriptions): tatra kim api pūrvavṛttaṃ vaktavyam. tad ucyate pūrvaṃ karmavicāro vṛtto ’nantaraṃ brahmajijñāsā prārabhyate. atrāhuḥ. nāyaṃ vṛttavarti- ṣyamāṇasambandha upapadyate dharmajijñāsāyāḥ prāg apy adhītave- dāntasya24 brahmajijñāsopapatteḥ. na cātra kramasya pratipādukaṃ pramāṇam asti […]. na cātrāṅgāṅgibhāvaḥ prayājādivat. nāpy adhi- kṛtādhikāro godohanādivat. phalajijñāsyabhedāc ca. abhyudayaphalaṃ dharmajñānaṃ niḥśreyasaphalaṃ brahmajñānaṃ.25 sādhyaś ca dhar- mas tatra jijñāsyaḥ siddharūpaṃ brahmātra. codanāpravṛttibhedāc ca. (Dvivedin 1903: 1–2; Kato 2011: 2)26 nating] an immediate succession in: ‘Having eaten, now she goes away.’ As another example, [the word atha is used] in the sense of requiring something which has oc- curred previously when, having said something before and desiring to choose another option, [an author] says ‘now, this is [my] opinion.’ Or in the sense of [designating] auspiciousness. […] Or in the sense of introducing a topic, e.g., ‘now, the teaching about words starts.’ Among these [four uses, the word atha] [here] does not mean the introduction of a topic, since the desire to know the brahman does not need to be intro- duced. Nor does [the word atha] mean auspiciousness, since auspiciousness cannot be connected to the meaning of the sentence because it is understood just from hearing [it], like the sound of a drum. [Only] the meaning of immediate succession is taken up here, because there is necessarily an immediate succession since there is the require- ment of something that has previously occurred.” 24 Dvivedin 1903 reads: prāgavyadhītavedāntasya. The reading prāg apy adhītavedān- tasya is found in Kato 2011 and in Kato’s unpublished edition. The parallel with Śaṅkara’s text quoted above, section 3.1.3.1, further favors the reading chosen by Kato. 25 Dvivedin 1903 reads: mokṣajñānam. The reading brahmajñānam is accepted in Kato 2011 and in Kato’s unpublished edition (it is supported by all available manuscripts). 26 “In this regard, one needs to say which thing that happened before [is presupposed by the desire to know the brahman]. Therefore [we] say: First the investigation of the ri- tual action is done, immediately after that the desire to know the brahman starts. In this regard, [others] (like Śaṅkara) say: This connection of something future with some- thing which has already occurred is not possible, [for three reasons, namely, first] be- cause the desire to know the brahman is possible for one who has studied the Upaniṣads also before the desire to know the dharma. And it is also not the case that there is here an instrument of knowledge determining that there is a sequence [in the study of the Brāhmaṇas and then the Upaniṣads] […] Nor is there a relation of prin- cipal and subordinate, as in the case of the pre-sacrifices [which are subordinate to the main sacrifice]. Nor is it the case that [only] the one who is responsible [for the inves- tigation of one part, i.e., the Brāhmaṇas] can be responsible [for the investigation of the brahman], as in the case of milking cows [when milk is required for a sacrifice and the 292 Elisa Freschi A long reflection on this topic follows, whereby the Upaniṣadic saying tat tvam asi is mentioned. Then: etat pūrvavṛttam etadānantaryam athaśabdenocyate. saty etasmin sā- dhanacatuṣṭaye brahmajijñāsopapadyate nānyatheti. ataḥśabdo hetvar- thaḥ yataḥ śrutir eva karmaṇāṃ kṣayitvaṃ darśayati. (Dvivedin 1903: 2; Kato 2011: 2–3)27 In this connection, Bhāskara also mentions the need to collect both (samuc- caya) ritual actions and knowledge, instead of choosing just one or the other: ato hetor brahmajijñāsā kartavyeti. atra brūmaḥ yat tāvad uktaṃ dhar- majijñāsāyāḥ prāg api brahmajijñāsopapatter iti, tad ayuktam. atra hi jñānakarmasamuccayān mokṣaprāptiḥ28 sūtrakārasyābhipretā. (Dvive- din 1903: 2; Kato 2011: 3)29 This leads to the idea that the Brāhmaṇa prescriptions need to be studied; a long discussion about this follows (Dvivedin 1903: 3–6), in which an ob- jector also mentions the Pūrva Mīmāṃsā: one who is responsible for the sacrifice is thus automatically held responsible also for the milking]. The second reason [for denying that atha means ‘after the study of the Brāhmaṇas’] is that there is a difference concerning the result [of the investigation] and the object one desires to know (in Pūrva and Uttara Mīmāṃsā). [As for the result,] the knowledge of dharma results in heaven. The knowledge of the brahman results in the summum bonum (i.e., liberation). And [as for the object one desires to know], the dharma, which has yet to be established, is what one desires to know there (in the Pūrva Mīmāṃsā), whereas the brahman, which has an already established form, [is what one desires to know] here (in the Uttara Mīmāṃsā). The third reason [for denying that atha means ‘after the study of the Brāhmaṇas’] is that there is a difference in the activity of the injunction [promoting the two different investigations].” 27 “Through the word atha it is said that something has happened before, i.e., that [the desire to know should take place] immediately after it. Once these four kinds of proofs have been established, the desire to know the brahman is possible, not otherwise. The word ataḥ means a reason, because the very Veda shows that the ritual actions are pe- rishable.” 28 Kato 2011 and Kato’s unpublished edition read: kṣemaprāptiḥ (against Dvivedin 1903 and all manuscripts but with the support of a variant recorded in van Buitenen’s un- published edition; about this, see Kato 2009–2010; for Kato’s reasons for this choice, see Kato 2011: 252). 29 “Because of this reason the investigation of the brahman must be done. In this regard we say that what has just been said, namely that the investigation of the brahman is possible also before the investigation of the dharma, is wrong. In fact, here the author of the sūtra meant that final emancipation is attained through the accumulation of knowledge (of the brahman) and ritual action.” (On the accumulation of knowledge and ritual action in Bhāskara, see Kato 2012).  Reusing, Adapting, Distorting? 293 nanu ca mīmāṃsāyām eva dvādaśalakṣaṇyāṃ śabdavyāpāraḥ kṛtsno nirūpitaḥ. (Dvivedin 1903: 5; Kato 2011: 7)30 Within this discussion, mention is made of the compound brahmajijñāsā: brahmaṇo jijñāseti karmaṇi ṣaṣṭhī kartṛkarmaṇoḥ kṛtīti viśeṣavidhānāt. (Dvivedin 1903: 6; Kato 2011: 10)31 And, with a phrasing similar to that of Śaṅkara: jñātum icchā jijñāseti prakṛtyarthapradhāneyam icchā. dhātoḥ karmaṇa iti dhātuvācyasyārthasya īpsitatvasmaraṇāt. tat punar brahma prasiddham aprasiddhaṃ vā. (Dvivedin 1903: 7; Kato 2011: 11)32 3.2 Conclusions on the commentaries ad Brahmasūtra / Pūrvamīmāṃsā- sūtra 1.1.1 To sum up, some elements (atha in the sense of designating ānantarya ‘immediate succession,’ ataḥ in the sense of designating hetu ‘cause,’ dhar- ma- or brahmajijñāsā being a ṣaṣṭhī tatpuruṣa ‘dependent determinative compound in which the former member is linked to the latter by a relation which would be expressed outside the compound through the genitive case’) are shared by all the commentaries but there is a huge quantity of textual material which, in terms of its structure and content, is uniquely shared by Rāmānuja and Veṅkaṭanātha. The latter also happens to neglect or implicitly refute similar topics (most notably, Śaṅkara’s and Bhāskara’s claim that there is no need to know dharma before starting the brahmajijñāsā ‘investigation into brahman’). 30 “[Obj.:] But the entire function of language has been explained in the twelve chapters of the [Pūrva] Mīmāṃsā.” 31 “In ‘the desire to know the brahman’ the genitive ending denotes the object (genitivus objectivus), because of the specific [grammatical] rule ‘[The sixth case ending is employed after a stem] meaning agent or object, when [used] along with a kṛt affix’” (Aṣṭādhyāyī 2.3.65). 32 “The jijñāsā is the desire to know. This desire regards principally the meaning of the root (i.e., jñā- ‘to know’), because it is recollected [in the tradition of Grammar] that the meaning expressed by the root is what is mostly desired (i.e., it conveys the syntactical object), due to [the rule ]‘[The desiderative affix is optionally introduced] after a verbal root meaning [an action which is] the syntactical object [of a verbal stem expressing desire when both the actions have the same agent].’ Besides, is the brahman well known or not?” The rule being evoked is A 3.1.7 (for an English translation, see Katre 1989). 294 Elisa Freschi 4 The Śrībhāṣya and the Seśvaramīmāṃsā: A shared agenda concerning aikaśāstrya As hinted at above (section 1), it is my opinion that Rāmānuja and Veṅka- ṭanātha shared a similar agenda, insofar as both aimed at broadening the hori- zons of their theological school by taking into account a different position. To state this concretely, they both aimed at showing that their school and another one (Vedānta or Pūrva Mīmāṃsā) formed an ekaśāstra, a ‘single śāstra.’ Indeed, the doctrine of aikaśāstrya refers to Pūrva and Uttara Mīmāṃsā (i.e., Vedānta) also in the ŚrīBh, but there it seems to be grounded on a reference to the unity of their contents, i.e., the Brāhmaṇa and the Upaniṣad parts of the Veda (see Marlewicz 2007). Moreover, the ŚrīBh addresses the problem of the Pūrva Mīmāṃsā as a separate philosophical school only peripherally.33 Rāmānuja’s position might seem difficult at first sight, insofar as it is possible that Śrī Vaiṣṇavism was not (yet) recognized as a Vedāntic school in his time. However, I could not detect any traces of a struggle to make Śrī Vaiṣṇavism acceptable as a Vedāntic school in the ŚrīBh itself, which seems to have been written from the perspective of an author with an already well- established audience. It does not appear to have been in need of addressing the worries of Vedāntins who did not want too much religious devotion in their philosophy. Nor does the ŚrīBh address the worries of Tamil Vaiṣṇavas who did not want Vedānta to become an essential part of their religion (if such Vaiṣṇavas ever existed). Instead, the ŚrīBh quotes a lot from the Upa- niṣads, possibly in order to show that Viśiṣṭādvaita Vedānta is their best inter- preter. Further, it discusses at length issues such as the oneness of brahman. This could mean that the ŚrīBh’s target audience consisted of Viśiṣṭādvaitins ante litteram, Advaitins or generic Vedāntins. Perhaps Rāmānuja even came from their ranks, since this would explain the smoothness of his ŚrīBh. In this sense, Rāmānuja’s task of making Śrī Vaiṣṇavism part of Vedānta becomes less difficult, insofar as Rāmānuja introduced Viśiṣṭādvaita as a sort of Ve- dānta within a landscape (that of Vedānta) which was still open enough to ac- commodate new attempts at being reconfigured. As for Pūrva Mīmāṃsā, some of Rāmānuja’s arguments were further developed by Veṅkaṭanātha,34 33 Śaṅkara and Bhāskara also referred to Pūrva Mīmāṃsā, but they did not elaborate on the unity of the two Mīmāṃsās nor did they use the term aikaśāstrya; see Kato 2012, section 3. 34 The short passage on Jaimini’s atheism in the Vedārthasaṅgraha, which Veṅkaṭanātha expands upon in the prologue of SM (p. 5 of Veṅkaṭanātha 1971), is worthy of a short note: “In order to avoid the lack of faith in ritual action of people who have not heard the Upaniṣads (aśrutavedānta), some excessive statements (ativāda) have been  Reusing, Adapting, Distorting? 295 but in general his way of understanding aikaśāstrya seems to consider Pūrva Mīmāṃsā more from the perspective of its object (the Brāhmaṇas) or as a reservoir of exegetical rules (which he thoroughly applies in his works), so that even on this side no conflict had to arise. The same attitude towards Pūrva Mīmāṃsā is also evident for post-Rāmā- nuja but pre-Veṅkaṭanātha authors such as Vātsya Varadaguru. In his Tattva- nirṇaya, he – like Rāmānuja in his ŚrīBh – quoted frequently from the Upani- ṣads and the BS, whereas the Pūrva Mīmāṃsā is called upon whenever ex- egetical points are discussed but not for its specific contents (see the Index of references and quotations from the PMS in Stark 1990). By contrast, in the case of Veṅkaṭanātha’s attempt to incorporate Pūrva Mīmāṃsā, it is easier to detect that he was facing internal opponents, since he wanted to integrate Pūrva Mīmāṃsā as a philosophical system and not just as a technique. Accordingly, in the SM the apologetic intent of showing that the Pūrva Mīmāṃsā was originally not atheist is quite explicit already in the title. In a related work, the Mīmāṃsāpādukā, the presence of divergent opinions within Viśiṣṭādvaita Vedānta is even more evident, since Veṅkaṭanātha needs to defend his choice (see MP 7–10) and recur to his predecessor’s (mostly implicit) support (MP 11). To sum up: Yāmuna was the first to attempt to justify the validity of the (Vaiṣṇava) Pāñcarātra sacred texts by discussing their epistemological status and to discuss Vedāntic topics. Rāmānuja proposed a Viśiṣṭādvaita form of Vedānta. Veṅkaṭanātha re-formed this Viśiṣṭādvaita and made it into the phi- losophical counterpart of a Vaiṣṇavism which incorporated Pāñcarātra and the Āḻvārs’ hymns, while at the same time being open to other Indian darśa- nas. Thus, Rāmānuja founded a Vedānta school, whereas Veṅkaṭanātha foun- ded a wider philosophical version of Vaiṣṇava Viśiṣṭādvaitism. Perhaps given the universalism of Vaiṣṇavism, rooted in the idea of the singleness of God‒Viṣṇu, Veṅkaṭanātha was able or needed to show that there was no real contradiction between Śrī Vaiṣṇavism and other philosophical systems. 4.1 Similarities between the treatment of aikaśāstrya in the Seśvara- mīmāṃsā and the Śrībhāṣya To begin with, some basic similarities are quite evident in the discussions about aikaśāstrya in the ŚrīBh and the SM, insofar as both texts used in the Devatādhikaraṇa so that one puts faith in ritual action as such. Thus, the de- finitive conclusion of those who know the Veda is that this is all a single treatise (śāstra)” (see below, section 5). 296 Elisa Freschi – justify the aikaśāstrya through the internal connections holding together the chapters of PMS-(SK)35-BS, – justify the differences between the PMS, (SK) and BS and the fact that they do not lead to different śāstras because the differences are compared with those distinguishing the various books of the PMS. Just as the twelve books of the PMS deal with different topics and are yet part of the same śāstra, the PMS, (SK) and BS deal with different topics but are still part of the same śāstra. Here is the initial description of the aikaśāstrya in the ŚrīBh and the SM: ataḥ pratipipādayiṣitārthabhedena ṣaṭkabhedavat adhyāyabhedavac ca pūrvottaramīmāṃsayor bhedaḥ. mīmāṃsāśāstram – “athāto dharmajijñāsā” ity ārabhya, “anāvṛttiḥ śabdād anāvṛttiḥ śabdāt” ity evamantaṃ saṅgativiśeṣeṇa viśiṣṭakramam. (ŚrīBh ad BS 1.1.1)36 “athāto dharmajijñāsā” ity ārabhya, “anāvṛttiḥ śabdād anāvṛttiś śabdāt” ity evamantaṃ viṃśatilakṣaṇaṃ mīmāṃsākhyam ekaṃ śāstram. sā hi mīmāṃsā […] pratividyāsthānam avāntaropa- kārādibhedasya vyavasthitatvāt na tathāpi bhedo darīdṛśyate. kāṇ- ḍabhedas tu syāt ṣaṭkādibhedanyāyāt. smṛtipurāṇavat maharṣibhiḥ pṛthagaparisaṃkhyānāc ca. aikaprabandhyaṃ tu saṃgativiśeṣair viśiṣṭakramatvād iti bhagavatā bhāṣykāreṇaivābhāṣi. (SM ad PMS 1.1.1)37 35 The SK is a possible intermediate section within the single Mīmāṃsā treatise, con- sisting of the PMS, the SK and the BS. See immediately below and sections 4.2 and 4.2.1. 36 “Therefore, the distinction between the Pūrva and the Uttara Mīmāṃsā [is explained] due to the distinction between the purposes they want to convey, like the distinction between the two [groups] of six [books] (in the PMS) and like the difference between the books. The Mīmāṃsā system, from ‘Now, because of that, the desire to know dharma [starts]’ (PMS 1.1.1) until ‘There is no return, according to the [sacred] text; there is no return, according to the [sacred] text’ (UMS 4.4.22), has a specific sequence due to the specific connection between the topics.” 37 “There is a single treatise which starts with ‘Now, because of that, the desire to know dharma [starts]’ (PMS 1.1.1) and ends with ‘There is no return, according to the [sacred] text; there is no return, according to the [sacred] text’ (UMS 4.4.22). It is composed of 20 chapters and is called Mīmāṃsā. And this is the [single] Mīmāṃsā. […] because for each field of knowledge distinctions due to intermediate assistances, etc., have been fixed (and in the case of this single field there is a single intermediate assistance, i.e., the same Vedic hermeneutic rules). [Although there are some further specifications*], nonetheless it is not the case that [in the case of PMS, SK, UMS] a split is seen, even if intensively looked for. However, there might well be a distinction in the [different] parts [only], according to the way (nyāya) of the distinction among  Reusing, Adapting, Distorting? 297 Careful readers may have already noticed that, unlike in the textual passages examined above, here Veṅkaṭanātha explicitly recurs to the authority of Rāmānuja (called Bhagavat Bhāṣyakāra ‘revered commentator [of the BS]’). An even more careful look shows that this is probably not a coincidence. In fact, Veṅkaṭanātha’s text differs more fundamentally from Rāmānuja’s than in the previous examples. Veṅkaṭanātha explicitly states that the Mīmāṃsā consists of twenty chapters, which means the PMS (8 chapters), the BS (4 chapters) and the intermediate Saṅkarṣakāṇḍa (henceforth SK, 4 chapters38). The attribute viṃśatilakṣaṇa is introduced directly after a literal reuse of Rāmānuja; while it seems only ornamental, it has the crucial function of in- troducing a broader concept of Mīmāṃsā. In contrast, Rāmānuja had spoken of Pūrva and Uttara Mīmāṃsā in the dual form and mentioned the number sixteen referring to the PMS (which, thus, needs to include also the SK) slightly later (see below, section 4.2). Thus, for Rāmānuja the opposition between Pūrva and Uttara Mīmāṃsā is quite clear (and perhaps not meant to be fully overcome), although the Veda is unitary and the study of the Brāhmaṇas (and perhaps also of their exegesis in the Pūrva Mīmāṃsā) is needed as a prerequisite for studying the Upaniṣads and the BS. In contradistinction to Rāmānuja’s approach, Veṅkaṭanātha wanted to achieve a stronger concept of aikaśāstrya and a broader Mīmāṃsā śāstra that also included the SK as a separate component. The background of this enter- prise is probably Veṅkaṭanātha’s systematizing project with this newly con- ceived Mīmāṃsā śāstra as its center.39 4.2 The Saṅkarṣakāṇḍa The explicit inclusion of the SK within the single Mīmāṃsā śāstra as con- ceived by Veṅkaṭanātha meant the possibility of opening Rāmānuja’s philos- the two [groups] of six [books] (in the PMS) and the other [divisions in adhyāyas, adhikaraṇas, etc,] (i.e., according to the division into six plus six chapters of the PMS, respectively dedicated to archetype and ectype), and because they are not listed separately, like [instead] the [various] smṛtis and purāṇas, by the great ṛṣis (and, thus, there is no evidence for a deeper split). By contrast, the fact that it is a single treatise is due to the fact that the sequence is specified by the distinct connections between the topics (saṃgati, a member of each adhikaraṇa) [reaching from the beginning of the PMS to the end of the UMS] ‒ so said the highly revered author of the Commentary (i.e., Rāmānuja).” (SM ad PMS 1.1.1) *This concessive meaning is achieved if we pre- suppose, e.g., evaṃ saty api. 38 See section 4.2 for the role of this work within the two Mīmāṃsās, and see section 4.2.1 for the extant SK. 39 I am currently working on a larger research project on this topic as part of the FWF project V 400-G 15. 298 Elisa Freschi ophy (I am not talking about his personal religious life, which was most like- ly deeply Vaiṣṇava) to non-Vedāntic theistic works, such as the Pāñcarātra texts and the hymns of the Āḻvārs. This move could only work if it were backed by a prestigious tradition, and the SK already had a long history of references by theistic authors indeed (see section 4.2.2). Of particular rele- vance in this context is its being mentioned by the Vṛttikāra, an unknown but revered forerunner of Rāmānuja who was quoted both by him and by Veṅka- ṭanātha in this context (see also above, section 1). The Vṛttikāra quote runs as follows: saṃhitam etac chārīrakaṃ jaiminīyena ṣoḍaśalakṣaṇeneti śāstraika- tvasiddhiḥ. This BS has been accorded with the Mīmāṃsā of Jaimini, which en- tails sixteen chapters. Thus, the unity of the teaching is established. Here, again, the differences between the texts of Rāmānuja and Veṅkaṭanātha that surround the quote are crucial. Rāmānuja writes: vakṣyati ca karmabrahmamīmāṃsāyor aikaśāstryam, “saṃhitam etac chārīrakaṃ jaiminīyena ṣoḍaśalakṣaṇeneti śāstraikatvasiddhiḥ” iti. And [the Vṛttikāra]40 will say that the Karma- and Brahma-Mīmāṃsā are a single teaching: “This BS has been accorded with the Mīmāṃsā of Jaimini, which entails sixteen chapters. Thus, the unity of the teaching is established.” Here, the stress is clearly on two parts of the Mīmāṃsā teaching that need to be accorded. By contrast, the parts are clearly three for Veṅkaṭanātha: “saṃhitam etac chārīrakaṃ jaiminīyena ṣoḍaśalakṣaṇeneti śāstrai- katvasiddhiḥ” iti vṛttigranthaś ca pratibandhyādiyuktigarbhaḥ. And the passage of the Vṛttikāra: “This BS has been accorded with the Mīmāṃsā of Jaimini, which entails sixteen chapters. Thus, the unity of the teaching is established” contains [reasonings] such as an argument per absurdum. The pratibandhi argument Veṅkaṭanātha tries to detect in the Vṛttikāra’s words would run as follows: No one denies the unity of the PMS and the SK. Denying the unity of these two and the BS would be like trying to split the SK and the PMS. To show that these two cannot be divided, it is said that they were both composed by the same author. 40 Mentioned with this title a few words before this passage.  Reusing, Adapting, Distorting? 299 It is very unlikely that the Vṛttikāra really had such a sophisticated argu- ment in mind, given that the attribution of the SK to Jaimini was widespread and that Rāmānuja did not comment on the passage according to Veṅka- ṭanātha’s interpretation. More probably, Veṅkaṭanātha forced his agenda onto that of his important forerunner because he could not ignore his quote (possibly because his entire audience knew it and would have confronted him with it). In this case, the act of reusing was (most probably) intended to distort the given text. 4.2.1 The extant Saṅkarṣakāṇḍa The extant SK as it has been edited and published (in 1894, 1963, 1965 and twice in 2009) is a set of approximately 465 sūtras divided into 4 adhyāyas ‘books’ containing 16 pādas ‘chapters’ and 386 adhikaraṇas ‘sections’ (Ver- poorten 1987: 6). The text tends to follow the PMS, since it elaborates on ritual matters, as do the last books of the PMS. It also shares a lot of termi- nology with the PMS. The SK was commented on once by Devasvāmin (see Subrahmanya Sastri 1965; Devasvāmin might be the same author who wrote a commentary on the Āśvalāyanaśrautasūtra and on the Āśvalāyana- gṛhyasūtra41) at a relatively early time (11th century?42) and once in the 18th century by the polygraph Bhāskararāya (see Śāstrī 1894).43 A separate manu- script of the first pāda of the first adhyāya of the SK was edited indepen- dently from the commentaries in Sarma 1963. Notwithstanding the same sequence of topics, in the various editions the numbering and wordings attributed to the single sūtras of the SK differ widely, with sūtras not accepted by one or the other.44 41 This hypothesis could be reinforced by the fact that, as acknowledged also by those who support the authenticity of the SK, this “is more in the nature of the Kalpasūtras,” since, unlike the PMS, it “has not got any separate principle to enunciate and, there- fore, is a miscellaneous supplement” (Ramaswami Sastri 1933: 297). 42 Tentative date by Kane (Kane 1962, vol. 1.2: 591–593). 43 For further commentaries, see below, n. 51. 44 Subrahmanya Sastri 1965 has a double numbering, possibly in order to reflect differ- ent ways of setting sūtras apart from the commentary. The first numbering encom- passes far fewer sūtras than the second. However, even in this numbering the corres- pondences between the editions are far from complete. For instance, the sūtras labelled 4, 6–7, 11, 13, 15 in the first numbering of Subrahmanya Sastri 1965 are absent in Kupalapati 2009, whereas the sūtra labelled SK 1.1.5 in Kupalapati 2009 is absent in Subrahmanya Sastri 1965. The sūtra labelled SK 1.1.12 in Kupalapati 2009 is placed in Subrahmanya Sastri 1965 after the sūtra following it in Kupalapati 2009. Sarma 1963 and Subrahmanya Sastri 1965 are much more similar but still not iden- tical, as different readings are frequent, sūtras are segmented in different ways (as for example the beginning of SK 1.2) and some sūtras are absent in one or the other (for 300 Elisa Freschi Śabara most probably refers to the extant SK in his commentary ad PMS 10.4.32: vidhinigamabhedaḥ prakṛtau kṛtaḥ. sviṣṭakṛdvikāraś ca vanaspatir iti saṅkarṣe vakṣyate “Prescription and mantras have been distinguished in the [passage on the] archetype sacrifice. And it will be said in the Saṅkarṣa: ‘The Vanaspati is an ectype of the Sviṣṭakṛt sacrifice.’” The reference might be to SK 2.4.39, reconstructed in Subrahmanya Sastri 1965 as sviṣṭakṛdvikāre yājyāyāṃ devatānigamāḥ syuḥ prakṛtyupabandhāt “Given that the [Vanaspati sacrifice] is an ectype of the Sviṣṭakṛt, in this sacrificial text there are mantras about deities, because [they] have been employed in the case of the archetype sacrifice (i.e., in the case of the Sviṣṭakṛt).”45 The commentary by Devasvāmin on this verse starts with sviṣṭakṛdvikāre vanaspatau yājyāyāṃ devatānigamāḥ syuḥ neti vicāryate “It is investigated whether, given that the Vanaspati sacrifice is an ectype of the Sviṣṭakṛt, there are mantras about deities in this sacrificial text” (Subrah- manya Sastri 1965). A similar passage is found in ŚBh ad PMS 12.2.11, where Śabara says nanu naiva paśor haviṣkṛd asti, auṣadhārthā, avahananārthā vā, ya- thā patnī tulyavac chrūyata iti saṅkarṣe vakṣyate “[Obj.:] But, in fact there is no person preparing the oblation at the animal sacrifice. [R.:] The wife [of the person commissioning the sacrifice] is in charge of the purpose of [preparing] the vegetables. It will be said in the Saṅkarṣa that ‘like, rather, the wife, who is in charge of threshing [the grains], is directly mentioned in the sacred texts [as] equal [to the sa- crifice commissioner].’” This could correspond to avahananārthā vā yathā patnī tulyaḥ śrūyate (SK 1.1.36) “like, rather, the wife who is in charge of threshing [the grains] is directly men- tioned in the sacred texts [as] equal [to the sacrifice commissioner].”46 example, sūtra 1.1.42 in Sarma 1963 is absent in Subrahmanya Sastri 1965 and sūtra 1.2.33 in Subrahmanya Sastri 1965 is absent in Sarma 1963). Further examples of differences are mentioned in fns. 45 and 46. 45 The text is identical in Kupalapati 2009, which however counts the sūtra as 2.4.20. Tātācārya 2009, as usual in this edition, reproduces only the beginning of the sūtra, namely sviṣṭakṛt. 46 Subrahmanya Sastri 1965 has avahananārthaṃ vā yathā patnī tulyā śrūyate (SK 1.1.36). The tulyaḥ found in Sarma 1963 seems in fact unjustifiable, since the term  Reusing, Adapting, Distorting? 301 The fact that Śabara used the future tense may indicate that he considered the SK to be a later part of the PMS, although it is by no means clear that saṅkarṣe refers to the title of a work. I did not find any other case of iti fol- lowed by the title of a work and then by vakṣyate in the ŚBh, although vakṣ- yate ‘it will be said’ is frequently used to refer to later passages of the PMS and/or of the corresponding ŚBh. Kumārila does not comment on these references, nor could I find any other reference to the SK by authors prior to the 9th century apart from an interesting passage by Śaṅkara in his commentary on BS 3.3.43 where he says: “Therefore it has been said in the Saṅkarṣa: ‘Verily the deities are many, because they are distinctly known’” (tad uktaṃ saṃkarṣe nānā vā de- vatā pṛthagjñānāt iti). Here the sūtra referred to is clear, namely SK 2.2.15, nānā vā devatā pṛthaktvāt “Rather, the deities are many, since they are sepa- rated.”47 This is a key element for the present study, and I shall come back to it. Personally, I would locate the SK in the Śrauta Sūtra-Pūrva Mīmāṃsā milieu, in the sense that – unlike many Pūrva Mīmāṃsā texts – it deals pri- marily with technical details rather than with more general problems. Moreo- ver, its topics and terminology tend to conform to a stage of the development of Pūrva Mīmāṃsā which might be slightly later than or slightly different from Jaimini’s. For instance, tantra ‘simultaneous application’ is discussed within the same constellation as in the PMS: prakṛti ‘archetype ritual,’ eka- kāla ‘same time’ (SK 1.1.6) and sādhāraṇa ‘common’ (SK 1.4.42), and it is opposed to abhyāsa ‘repetition’ in SK 1.3.16 and āvṛtti ‘repeated perfor- mance’ in SK 1.1.9, as in later and technical Pūrva Mīmāṃsā treatises. By contrast, it is not opposed to prasaṅga ‘automatic involvement’ as happens to be the case in the ŚBh, but also not to āvāpa ‘separate application of a ritual element,’ which is the typical counterpart of tantra in the PMS.48 Thus, the hypothesis that the text is ancient but does not belong to the more speculative part of the Pūrva Mīmāṃsā – represented by Jaimini and Śabara – appears all the more credible. Therefore, it makes sense that Śabara occasionally refers to it or that Someśvara does so once (NS ad TV ad PMS 1.3.3, Sāstrī 1909: 145, l. 2749) but that no more Pūrva Mīmāṃsā energies are dedicated to it. needs to refer to the feminine “wife.” I could trace the sūtra in neither Kupalapati 2009 nor Tātācārya 2009. 47 The sūtra is identified as 2.2.15 in Subrahmanya Sastri 1965, Kupalapati 2009 and Tātācārya 2009. 48 On all these terms and their interaction, see Freschi and Pontillo 2013. 49 nanu saṅkarṣe vrīhibhir iṣṭvā vrīhibhir eva yajeta yavebhyo yavair iṣṭvā yavair eva yajetāvrīhibhya iti vākyam udāhṛtya kim anenāgrayaṇābhyāso vidhīyate […]. It is not 302 Elisa Freschi One might recall the complements to Pāṇini’s Aṣṭādhyāyī, such as the Dhātupātha, in the sense that Śabara seems to refer to the SK as an appendix of the PMS which does not need specific exegetical attention (and, in fact, he did not comment on it). As for its commentators, Devasvāmin may have been the same author who commented on the Āśvalāyanaśrauta- and -gṛhyasūtra and thus could have been interested in ritual texts like the Kalpasūtras, so that his attention to the SK could be part of the same interest. Bhāskararāya (or Bhāskara), the late commentator of the SK, was trying to complete Khaṇḍadeva’s treatment of the PMS, since he felt (perhaps because of the pressure of Vedānta?) that a complete Mīmāṃsā work also had to address the SK. Nonetheless, his Bhāṭ- ṭacandrikā explicitly states that the text lacks the connection (saṅgati) among adhikaraṇas ‘sections’ which, as seen above (section 4.1), is among the key reasons in favor of aikaśāstrya mentioned by Veṅkaṭanātha and Rāmānuja.50 Nakamura (1983: 393) lists four extant commentaries on the SK. Apart from the two mentioned here, the other two are one in manuscript form by Rāja- cūḍāmaṇi Dīkṣita51 and one by Khaṇḍadeva himself;52 both seem to be inac- cessible. To sum up, generally Pūrva Mīmāṃsā authors are hardly or not at all in- terested in the SK, whereas several authors among the Uttara Mīmāṃsakas deal at length with the status of the Mīmāṃsā śāstra and of the SK. By the 10th century it had become almost a hot topic in parts of Vedānta. An inscrip- tion of “Anur (Chingleput district, Tamil Nadu) of 999 A.D.” (Verpoorten 1987: 6) describes the single śāstra made of Pūrva and Uttara Mīmāṃsā as consisting of 20 books, i.e., presumably, the 12 of the PMS, the 4 of BS and the 4 of SK. As evidence of the persisting importance of the SK for Śrī Vaiṣ- ṇavas, it was recently reprinted twice (Kupalapati 2009 and Tātācārya 2009), both times within an explicit Śrī Vaiṣṇava framework, namely within clear whether Someśvara is quoting the SK here or rather referring to it. That the latter might be the case is further suggested by the fact that I could not identify the sūtra in the extant SK (Sarma 1963 and Subrahmanya Sastri 1965). 50 ata eva saṅkarṣe na pratyadhikaraṇaṃ saṃgatyapekṣā (beginning of the Bhāṭṭacan- drikā, l. 3, quoted in Ramaswami Sastri 1933). I have been unable to obtain a copy of the text of the Bhāṭṭacandrikā until now. 51 Nakamura has this information third-hand, since he writes that “according to Ramas- wami, this work is inserted in Hultzsch, Reports of South Indian MSS, Vol. II, No. 1489” (Nakamura 1983: 396, n. 12), and reports the title of the work commented upon by Rājacūḍāmaṇi Dīkṣita as Saṃkarṣanyāyamuktāvali (Nakamura 1983: 394). 52 This information is hardly believable, given that it contrasts with Bhāskararāya’s need to complete Khaṇḍadeva’s work.  Reusing, Adapting, Distorting? 303 series dedicated to Śrī Vaiṣṇavism, by Śrī Vaiṣṇava editors and with Śrī Vaiṣ- ṇava maṅgalas and illustrations. 4.2.2 The Saṅkarṣakāṇḍa-devatākāṇḍa Coming from the same period of the Anur inscription mentioned above, Veṅkaṭanātha and other authors and testimonies (as far as I know, mostly from South India) explicitly situate the SK after the PMS and before the BS and hold different views regarding its authorship. Why did a quite unsyste- matic text, which might have been an appendix of the PMS composed to account for further minor issues, become so central for the Vṛttikāra, Rāmānuja and Veṅkaṭanātha? Why did they decide to explicitly focus on it? A sort of Kalpasūtra discussing trivial ritual matters could not have been of particular interest for any of them (Veṅkaṭanātha’s SM never discusses ritual details). In contrast, they presented the SK as a means of introducing the topic of the deity (devatā). This is also the context in which Śaṅkara in- serted his reference to the SK, so that the connection of the SK with the topic of deities appears to predate Rāmānuja and Veṅkaṭanātha. In this manner, the SK became a way to make theism present in the unitary Mīmāṃsā śāstra from the time of the composition of its foundational texts. And here we find a small conundrum. In fact, the SK is often called madhyamakāṇḍa ‘middle part,’ as opposed to the karmakāṇḍa ‘part on ritual action’ and brahmakāṇḍa ‘part on the Brahman,’ or to the pūrvakāṇḍa ‘former part’ and uttarakāṇḍa ‘latter part,’ i.e., the PMS and the UMS. In addition to these names and its other name of Saṅkarṣa(ṇa) Kāṇḍa, however, the SK is also called devatā- kāṇḍa ‘part on deities’ or upāsanākāṇḍa ‘part on veneration.’ These designa- tions do not appear together, so that it seems clear that the basic assumption for authors interested in these topics was that there were (at most) three basic texts of the Mīmāṃsā śāstra. A further significant detail is that the content of the extant SK corres- ponds neither to the appellation devatākāṇḍa nor to the function ascribed to it by Veṅkaṭanātha and other Vedāntins, i.e., the discussions of deities, later to be subsumed within the brahman in the UMS. By contrast, the SK-devatā- kāṇḍa referred to by such Vedāntins as Vedānta Deśika in his SM fits nicely in a progressive scheme: In this interpretation the PMS deals with karman, the SK with deities, and the UMS with brahman. In fact, later authors, from Appayya Dīkṣita (see section 4.2.4) to the contemporary editor of the SK (Kupalapati 2009), had to deal with the ap- parent inconsistency of the name devatākāṇḍa applied to a text that speaks of different topics. Sannidhāna Sudarśana Śarmā Kupalapati, after having quoted Appayya, proposes the following solution: 304 Elisa Freschi Through these words of the author of the Parimala [i.e., Appayya], one understands that this SK is called devatākāṇḍa only because of the many eulogies of the deities, and not because it reflects upon the own form, nature, and [attributes] of the deities. This is clearly understood. (Kupalapati 2009: v)53 4.2.3 Quotations from the Saṅkarṣakāṇḍa Which sūtras are attributed to the one (the extant SK) or the other (let us call it “SK-devatākāṇḍa”)? Śabara (two sūtras), Someśvara (one sūtra) and Śaṅkara (one sūtra) mention sūtras also (more or less, see section 4.2.1) found in the extant SK, whereas later Vedāntins either do not quote anything at all or quote a) the sūtra quoted by Śaṅkara with a small variant (as does Rāmānuja, ŚrīBh ad 3.3.43). b) the same three (or four) theistic sūtras referring to Viṣṇu that are not found in the extant SK (but quoted by Veṅkaṭanātha in his Śatadūṣaṇī54 53 etaiś ca parimalakāravacanaiḥ idam avagamyate yat devatāprastāvaprācuryād eva asya saṅkarṣakāṇḍasya devatākāṇḍatvavyavahāraḥ na tu devatāsvarūpasvabhāvādi- nirūpakatayā iti spaṣṭam eva avagamyate. 54 devatākāṇḍaṃ ca karmakāṇḍaśeṣatayā bhāṣyakāraiḥ parigṛhītam, “tad uktaṃ saṅ- karṣa” iti tatratyasūtrāṇi codāharanti. tasya ca kāṇḍasyopasaṃhāre “ante harau tad- darśanāt” iti devatākāṣṭhām pradarśya “sa viṣṇur āha hi” iti sarvadevatārādhanānān tatparyavasānāya tasya sarvāntarātmatvena vyāptim pratipādya “tam brahmety āca- kṣate taṃ brahmety ācakṣate” iti tasyaiva vedāntavedyaparabrahmatvopakṣepeṇopa- saṃhārāt, sāmānyato ’pi viśeṣataś ceśvaraḥ prastuta iti (Śatadūṣaṇī 3 [aikaśāstrya], Anandacharlu Vidyāvinod 1903: 18). “Also the Devatākāṇḍa is accepted by the Commentator* as a supplement to the Karmakāṇḍa, [as] the sūtras in it** illustrate, e.g., ‘This has been said in the Saṅkarṣa.’ And God is praised in a general form (as brahman) or in a particular one (as a personal God), because in the summary of this (SK) part [of the unitary Mīmāṃsā system], [the author] showed the rank of God [within Mīmāṃsā] with ‘At the end (of the Mīmāṃsā system), in the sense of Hari (God), because this has been shown’ and he conveyed with ‘For he, Viṣṇu, said’ the pervasion of him (Hari), who [pervades] everything by means of being its inner self, in order to bring to completion the [acts of] pleasing addressed to all deities. Then, he summarized [the whole] by hinting at the status of supreme brahman, which is known through the Upaniṣads, of Him (God), with: ‘That (God) is called brahman, that is called brahman.’” *Who is this Bhāṣyakṛt? Veṅkaṭanātha usually used this title to refer to Rāmānuja, who however does not appear to have referred to the SK. The three alleged quotes from the SK are also not found in the extant work of the other author who might be referred to as bhāṣyakṛt in this context, namely in Śabara’s Bhāṣya. The editor’s commentary in Tātāyāryadāsa 2004 reproduces the opinion of some who identify him with Upavarṣa, who is believed to have written a commentary (albeit  Reusing, Adapting, Distorting? 305 and in the Tattvaṭīkā,55 by Madhva in the Anuvyākhyāna and by his com- mentator Jayatīrtha). I counted three to four sūtras because Madhva’s attribution is less clear. In fact, his Anuvyākhyāna mentions a sūtra, namely sa viṣṇur āha hi “In fact, Viṣṇu Himself said,” but does not attributes it directly to the SK- devatākāṇḍa, but rather to a Devaśāstra ‘Treatise about God.’56 Jaya- tīrtha’s Nyāyasudhā commentary on Madhva explains that the quote comes from a Devatāmīmāṃsāśāstra ‘Treatise on the Mīmāṃsā about God,’ which starts with athāto daivī (scil. jijñāsā?) “Now, because of that the divine [desire to know?] starts” and ends with tam brahmety ācakṣate “This (Viṣṇu) is called brahman (in the UMS).”57 Madhva’s mention of a called Vṛtti) on the whole Mīmāṃsā system. **The reference point of tatratya is difficult to determine, given the doubtful identity of the bhāṣyakṛt. I could not detect any such sūtra in the PMS, and Śabara used verbs in the future tense when referring to the SK; see section 4.2.1. 55 evaṃ tarkite karmaṇi saṅkarṣaṇakāṇḍe* caturlakṣaṇyā tattatkarmārādhyadevataiva svarūpabhedaguṇaprakarṣaiḥ nirakṛṣyata**. tatsamāptau ca “ante harau taddarśa- nāt” “sa viṣṇur āha hi” “taṃ brahmety ācakṣate taṃ brahmety ācakṣate” iti vicāra- yiṣyamāṇam upacikṣipe iti tattvavṛddhāḥ***. (Tattvaṭīkā ad ŚrīBh ad BS 1.1.1, ad v. 240, Aṇṇāṅgarācārya 1941: 31, Mahādeśika 1938: 78). *Variant reading: saṅ- karṣakāṇḍe. **Variant reading: nirākṛṣyata [sic] (as in Aklujkar 2012: 208, but since all editions attest the ending -ta, I see no need to emend nirakṛṣyata, an imperfect passive of niḥkṛṣ- in the sense of ‘was extracted’. ***Aklujkar (2012: 208) argues that this should be emended to tattvavidaḥ, which seems in fact much smoother. Transla- tion: “In this way, after having reflected on the ritual action, the deity which is pleased through those various ritual actions was extracted in the Saṅkarṣaṇa part, in four chap- ters, by [examining] (the deity’s) nature, distinction (from the world and the individual souls) and excellent qualities. And at the end of it [the author] hinted at what he was about to examine [in the UMS] with ‘At the end (of the Mīmāṃsā system), in the sense of Hari (God), because this has been shown’, ‘For he, Viṣṇu, said’, [and] ‘That (God) is called brahman, that is called brahman’. So say the real seniors (the respected members of our school).” 56 Cf. Madhva’s Anuv 1.1.81a: svayaṃ bhagavatā viṣṇur brahmety etat puroditam || Anuv 1.1.80 || sa viṣṇur āha hītyante devaśāstrasya tena hi | (Siauve 1957: 23, Pandu- rangi 2002: 777), “In fact, it has been said before by the Venerable [author of the sūtras] himself that the brahman is Viṣṇu. In fact, therefore [it is said] at the end of the Devaśāstra ‘In fact, Viṣṇu Himself has said.’” Jayatīrtha explains that purā ‘before’ means “before our sūtra,” that is, before the UMS. 57 kin tu daivīmīmāṃsāśāstrasyānte sa viṣṇur āha hi. taṃ brahmety ācakṣate iti sūtradva- yena viṣṇuḥ brahmety uditam. tadanantaram eva cāthato brahmajijñāseti […] athāto daivītyādir udāhṛtaḥ “But, through the two sūtras at the end of the treatise on the Mīmāṃsā about deities, namely ‘In fact, Viṣṇu Himself said’ and ‘This (Viṣṇu) is called brahman (in the UMS)’, it is said that Viṣṇu is the brahman. And immediately after this [sūtra] comes ‘Now, after that, the desire to know the brahman starts’. […] 306 Elisa Freschi devaśāstra could refer to the SK-devatākāṇḍa before it was identified or confused with the SK, whereas Jayatīrtha could represent the next stage of this process (see below, section 4.2.6). Veṅkaṭanātha does not mention the alleged first sūtra of this text, but has the other two preceded by ante harau taddarśanāt “At the end (of the Mīmāṃsā system), in the sense of Hari (God), because this has been shown.” As Veṅkaṭanātha in the Śatadūṣaṇī puts it (see n. 54), the expression tam brahmety ācakṣate indeed creates a smooth transition to the BS. c) Four more ślokas from some saṅkarṣaṇasūtreṣu that do not occur in the extant SK are found in Utpala Vaiṣṇava’s commentary (called Spanda- pradīpikā) on the Spandakārikā.58 These final sūtras are never quoted in the otherwise compact and homo- geneous tradition referring to the SK-devatākāṇḍa, and they may in fact ra- ther derive from a Saṅkarṣaṇa-Pāñcarātra text, as suggested in Sanderson 2009a, in the section “The Kārkoṭas and Pañcarātra Vaiṣṇavism.” 4.2.4 The Saṅkarṣakāṇḍa and Advaita Vedānta I have already mentioned the lack of interest of Pūrva Mīmāṃsā authors in the SK. This is mirrored by a similar lack of interest among Advaita Vedānta authors. Apart from the sūtra mentioned by Śaṅkara in his commentary on The [beginning] is announced as ‘Now, after that, [the desire to know] the deity starts.’ ” (Pandurangi 2002: 777‒778) 58 Ad SpK 26: anyac ca saṅkarṣaṇasūtreṣu – svātmaikaniṣṭhaṃ cidrūpaṃ bhāvābhāva- pariṣkṛtam | svasaṃvedanasaṃvedyaṃ prakṛtyātītagocaram || iyaṃ yoniḥ smṛtā vipra mantrāṇāṃ pratyayātmikā | te mantrā varṇarūpeṇa sabāhyābhyantaroditāḥ || naiṣkāli- kapadāvasthāḥ karaṇānīva dehinām | prayuktāḥ sarvakāleṣu siddhyante vīryayoga- taḥ || “And moreover, [it is said] in the Saṅkarṣaṇasūtras: ‘The form of consciousness, which is installed in itself alone, and is prepared through presence and absence, | is perceivable through self-awareness, and its sphere of knowledge lies beyond nature || This source of the mantras is recollected, o sage, to consist of cognition | These mantras, which appear externally and internally in the form of phonemes || rest on the undivided level. Like the [sense] organs of the embodied beings, | when they are employed, [the mantras] are successful at all times because of the connection with vigour ||’” Ad SpK 29: saṅkarṣaṇasūtreṣv api ‒ yenedaṃ dṛśyate viśvaṃ draṣṭā sarvasya yaḥ sadā | dṛśyaś carācaratve yaḥ sa viṣṇur iti gīyate || iti “And also in the Saṅkarṣaṇasūtras: ‘The one by whom the whole is seen, who is always the observer of everything | and who can be seen in [all things] movable and immovable, he is called Viṣṇu’” (Dyczkowski 2000: 35, 41). I am very grateful to Lubomír Ondračka, who sent me a scan of these verses. These passages have been translated in Dyczkowski 1994: 161, 164.  Reusing, Adapting, Distorting? 307 BS 3.3.43, there is almost no evidence of any acquaintance with the text of the SK. The Advaita Vedāntins who do mention the SK do not cite the theis- tic sūtras quoted above, nor do they stress its theistic character, but they might have been exposed to a tradition of such interpretations. This is espe- cially possible in the case of Madhusūdana Sarasvatī, who had been a Vaiṣṇa- va and remained a bhākta before (chronologically) and besides being an Advaitin. Although he seems to refer to the SK only second hand (he pro- vides no quotes), he reports on both its position within the PMS and its theis- tic character. In this connection, it is possible that Madhusūdana and other authors (see section 4.2.7) attributed the SK to Jaimini because they needed to insert it into a ready-made scheme, where only the PMS and UMS had a place (we will see that at a later stage, the SK was also linked to the UMS). In fact, the passage on the SK looks like an excursus inserted within a framework that only had room for the PMS and UMS: In this way, the Mīmāṃsā is also of two types: Karmamīmāṃsā and Śārīrakamīmāṃsā. Of these, the Karmamīmāṃsā is made of twelve chapters, from the first sūtra to anvāhārye ca darśanāt (i.e., the sūtra later numbered PMS 12.4.40), [and] it was composed by the revered Jaimini. Within it, the purpose of the twelve chapters are, in sequence, [the discussion on] the means to know dharma [and the other eleven topics up until] the automatic involvement. In the same way, the Saṅkarṣakāṇḍa consists of four chapters and was composed by Jai- mini. And this (SK), which is well known by the name of Devatā- kāṇḍa, because it teaches a ritual called upāsana ‘worship,’ is just a part of the Karmamīmāṃsā. In the same way, the Śārīrakamīmāṃsā has four chapters, starting with athāto brahmajijñāsā (i.e., the sūtra later numbered BS 1.1.1) and ending with anāvṛttiḥ śabdāt (i.e., the sūtra later numbered BS 4.4.22).59 The lack of mention of any sūtra of the SK seems to hint at the fact that Madhusūdana had no access to the work. 59 evaṃ mīmāṃsāpi dvividhā karmamīmāṃsā śārīrakamīmāṃsā ceti. tatra dvādaśādhyāyī karmamīmāṃsā “athāto dharmajijñāsā“ ityādiḥ “anvāhārye ca darśanāt” ity antā bha- gavatā jaimininā praṇītā. tatra dharmapramāṇam […] prasaṅgaś ceti krameṇa dvāda- śānām adhyāyānām arthaḥ. tathā saṃkarṣaṇakāṇḍam apy adhyāyacatuṣṭayātmakaṃ jai- minipraṇītam. tac ca devatākāṇḍasaṃjñayā prasiddham apy upāsanākhyakarmapratipā- dakatvāt karmamīmāṃsāntargatam eva. tathā caturādhyāyī śārīrakamīmāṃsā “athāto brahmajijñāsā“ ityādiḥ “anāvṛttiḥ śabdāt” ity antā […] (Prasthānabheda [by Madhusūdana Sarasvatī] 1912: 11). 308 Elisa Freschi Another partial exception to the rule that the SK was only popular within Viśiṣṭādvaita Vedānta occurs in the work of Appayya Dīkṣita (possibly 1520–1593), a polymath writing – apart from his critical work on belles-let- tres (alaṅkāraśāstra) – from the point of view of Pūrva Mīmāṃsā (Vidhira- sayana) and Advaita Vedānta (e.g., Joshi 1981, Śrīkaṇṭha and Appayya Dīkṣita 1908, Appayya Dīkṣita 1890). The exception is only partial because Appayya, though a Śaiva, was also traditionally considered close to Viśiṣṭā- dvaita Vedānta due to some of his theological tenets and his openness towards devotion to Viṣṇu.60 Appayya chiefly quotes the extant SK.61 How- ever, he also mentions its title “Devatākāṇḍa,” which he justifies as being the result of the presence of many discussions about deities in the SK. In an Advaita Vedānta work, his sub-commentary on Śaṅkara’s UMS- Bhāṣya, Appayya Dīkṣita discusses the topic of the desire to know dharma (dharmajijñāsā). In this context, an objector argues that it cannot be the case that PMS 1.1.1 announces the whole śāstra, since it cannot be said that, after having announced the investigation of the whole Veda, one willfully abandons the investigation of one part of it. Nor can it be the case that Jaimini, the best of the great ṛṣis, after having composed the twelve chapters [PMS] for the sake of investi- gating dharma, having noticed that there he had not put some rules in the sūtras, for the sake of collecting these [rules], composed the SK, which is a supplement to the twelve chapters [PMS].62 Appayya’s answer displays once again that his approach was not purely Advaita Vedānta, since he emphasized the aikaśāstrya of the PMS and UMS, as did Rāmānuja. However, he does not return to the SK, although he reused 60 On the Śaiva affiliation of Appayya, see Bronner 2007; on his devotion to Viṣṇu, see Okita’s contribution in this volume. 61 Aklujkar writes that “Appayyadīkṣita cites several sūtras of the SK and at least one passage from D-S [=Devasvāmin’s commentary, EF] (SSS 1965: iii, 87, 259)” (Ak- lujkar 2012: 214). 62 na hi kṛtsnavedārthavicāraṃ pratijñāya tadekadeśavicāras tyakto buddhipūrvam iti vaktuṃ śakyam, na vā dharmavicārārthaṃ dvādaśalakṣaṇīṃ kṛtvā, tatrāsūtritān kāṃścin nyāyān ālakṣya tatsaṃgrahārthaṃ dvādaśalakṣaṇīśeṣaṃ* saṅkarṣakāṇḍaṃ api kṛtavato maharṣivarasya bhagavato jaimineḥ (Parimālā ad Kalpataru ad Śaṅkara’s Bhāṣya ad BS 1.1.1, Joshi 1981: 50). *Joshi 1981 reads dvādaśalakṣaṇīṃ śeṣam, which is semantically and grammatically unacceptable. Aklujkar quotes the second part of the same passage but without the initial na vā and without any hint at the fact that it occurs within a pūrvapakṣa ‘passage stating a prima facie view’ (Aklujkar 2012: 214). However, he quotes from Appayya’s Brahmasūtrakalpataru- parimālā, Nirnaya Sagara Press (Bombay 1917): 50, which I could not yet access.  Reusing, Adapting, Distorting? 309 arguments already found in Veṅkaṭanātha (SM ad PMS 1.1.1 and Śatadūṣaṇī 3), i.e., the unity of the śāstra is not invalidated by the fact that there are two authors (Jaimini and Bādarāyaṇa). 4.2.5 The Saṅkarṣakāṇḍa and the Pāñcarātra A further significant element to be taken into account when reconstructing the history of the SK is its connection with the Pāñcarātra. Śaṅkara mentions the SK in his UMS-Bhāṣya within a doctrinal discussion (should Vāyu and Prāṇa, which are identified in the same way, be equated?), which fits the con- ceptual framework of the SK. However, it is easy to imagine that theistic authors might have later used the same passage as the basis for theistic argu- ments, given that the knowledge of god(s) is mentioned. Further, Kanazawa 1989 drew my attention to a very interesting passage by Mukunda Jhā Bakhśī, the editor of Rāghavabhaṭṭa’s Padārthādarśa (15th century), who in his Bhūmikā discusses the sources of dharma and mentions the following section among various others: Likewise, in this regard, due to the distinction between ritual acts, ve- neration and knowledge, three parts have been composed, of which the part on ritual actions (karmakāṇḍa) has been fully described in the Pūrva Mīmāṃsā [Sūtra] and other [similar texts] by Jaimini and other [Pūrva Mīmāṃsā authors], the part on veneration (upāsanākāṇḍa) has been fully described in the Pañcarātra and other similar [texts] by Nārada, etc., and the part on knowledge (jñānakāṇḍa) has been fully described in the Uttara Mīmāṃsā by Vyāsa and other [Vedāntic au- thors].63 At the beginning of the text edited by Bakhśī, Rāghavabhaṭṭa attributes the upāsanākāṇḍa (i.e., the SK) to Nārada and the UMS to Vyāsa64 (who is con- 63 yathātra karmopāstijñānabhedāt kāṇḍatrayam upanibaddham, yatra karmakāṇḍaṃ pūr- vamīmāṃsādau, jaiminyādibhiḥ. upāsanākāṇḍaṃ ca pañcarātrādau nāradādibhiḥ jñāna- kāṇḍaṃ cottaramīmāṃsāyāṃ vyāsādibhir upavarṇitam (Bakhśī 1981, Bhūmikā: 1). 64 tatra sarvāsu śrutiṣu kāṇḍatrayaṃ karmopāsanābrahmabhedena. tatra jaiminiprabhṛtibhiḥ samyaktayā vivṛtam. idam upāsanākāṇḍaṃ nāradādibhir brahmakāṇḍaṃ bhagavadvyāsādibhir iti smṛ(śru)timūlakatā* asya pratyakṣopalabdhā (Bakhśī 1981: 1–2) “As for all the Vedas, there are three parts, according to the distinction between ritual action, veneration and brahman [as the topics of the three Vedic parts]. Among them, the [part on ritual action] has been clearly explained by Jaimini, etc. This part on veneration has been [explained] by Nārada, etc., the part on the brahman has been explained by the venerable Vyāsa, etc. Thus, it is perceivable that these (three parts) are based on the Veda.” *The parentheses are in the edited text. 310 Elisa Freschi sistently identified by Viśiṣṭādvaitavedāntins as Bādarāyaṇa; see, e.g., Veṅ- kaṭanātha’s Śatadūṣaṇī 3). Who is this Nārada? Nārada figures in the key role of narrator in many Pāñcarātra texts, from the Pārameśvara Saṃhitā to the Sātvata Saṃhitā and the Ahirbudhnya Saṃhitā, and he is mentioned as an ekāntin (a follower of the Ekāyana Veda, see Rastelli 2003) in the MBh, Nārāyaṇīya Parvan, which is the most ancient text referring to Pāñcarātra (MBh 12.334.1).65 Further- more, the name Nārada is connected with the Vaiṣṇava milieu and figures together with Saṅkarṣana in the guruparamparā leading to Vyāsa in the (Vaiṣṇava) Hayagrīvopaniṣad (Kanazawa 1989: 41). The connection of the SK with the Vaiṣṇava (and perhaps Kaśmirian) milieu and, thus, with the Pāñcarātra could be reinforced by Utpala Vaiṣṇava’s quotes (on which see the end of section 4.2.3), the last of which appears very close to a quote from the “Pāñcarātropaniṣad” and can be found among further quotes attributed to the “Pāñcarātra.” 4.2.6 Conclusions on the Saṅkarṣakāṇḍa To sum up, the extant SK does not seem to properly fulfil the role assigned to it by theistic Vedāntin authors. How can this be accounted for? Three possi- ble answers can be suggested. 1. Possibly theistic Vedāntin authors used the name of a text that was as- sumed to be part of the unitary Mīmāṃsā Śāstra but was either lost or scarcely known (remember the lack of quotations from the extant SK in Veṅkaṭanātha). The same authors would then have mistakenly confused the text, namely the SK, with a different text that fulfilled a role they needed to see fulfilled, i.e., that of introducing God to the Mīmāṃsā sys- tem. Perhaps Kanazawa is right in pointing out that the very name SK might have helped due to the importance of Saṅkarṣaṇa in the Pāñcarātra vyūha doctrine (Kanazawa 1989: 40). The fact that at a certain point in time, the non-Ekāyanaveda group within the Pāñcarātras was known with the name of Saṅkarṣaṇapāñcarātras may also have contributed, so that one might have been prone to connect Saṅkarṣaṇa with the faction of Pāñcarātrins more favorable to the Mīmāṃsā interpretation of the Ve- das.66 In fact, the edition of Veṅkaṭanātha’s Tattvaṭīkā, if at all reliable, 65 A Nāradapāñcarātra has been preserved, but it is a rather late text (see Leach 2012: 21). 66 For the two groups of Pāñcarātrins, see Leach 2012: 48–49 and Sanderson 2009a, section “The Kārkoṭas and Pañcarātra Vaiṣṇavism.” I am grateful to Robert Leach for having drawn my attention to this point. On the possible conflation of the Mīmāṃsā-  Reusing, Adapting, Distorting? 311 uses the appellation Saṅkarṣaṇa for the SK (but not so the editions of his SM and Śatadūṣaṇī, where Saṅkarṣa is again found). It is still difficult to tell how and when exactly this superimposition of the one text on the other took place, but, as already hinted at, it seems to have taken place in Vedānta-Pāñcarātra milieus and Śaṅkara may have played a major role in it, since he quotes from the extant SK, though in the context of a theologi- cal discussion.67 It might, thus, have been Śaṅkara who involuntarily made the SK no longer associated with sheer technical discussions, but rather with theo- logical ones. In other words, before Śaṅkara there may have been a tech- nical SK and a different theistic text (perhaps only a few sentences). If one accepts Jayatīrtha’s authority, the latter work already had a Vedāntic flavor, and one might speculate that it had been used by Vaiṣṇavas (per- haps: Pāñcarātrins) to vindicate the Vedānta status of their system. Śaṅkara’s quote from the former SK in a context where one could have expected the latter may have caused confusion between the two, a confu- sion which was very much welcomed by non-Advaita Vedāntins and which harmonizes nicely with other tripartitions (e.g., the one between karman, jñāna and bhakti). 2. A second explanation would require the assumption that no SK-devatā- kāṇḍa ever existed and that some theistic Vedāntins artfully manipulated the evidence regarding the SK. But due to the favor rei principle and since attestations regarding it range well beyond the borders of an intercon- nected group of people (who could have conspired together), this expla- nation is less likely than the other two at the present stage of research. 3. Last, it is possible that there existed a tradition of interpreting the extant SK in a theistic way and that it was in this connection that other theistic sūtras were attributed to it. This hypothesis clashes with the fact that no sūtras of the extant SK have been transmitted together with those of the SK-devatākāṇḍa, and with the fact that the sūtras attributed to the SK- devatākāṇḍa are of a completely different nature and thus cannot be lost parts of the extant SK. Nonetheless, the hypothesis is probably right in pointing out that the confusion between the ritualistic SK and the SK- devatākāṇḍa happened quite early. Ānandagiri’s explanation of the name saṅkarṣa, for instance, refers to the technical contents of the extant SK and the Pāñcarātra-Saṅkarṣaṇa, see also section 4.2.3. 67 After I had completed this study, V. N. Pandurangi sent me a contribution of his (with the title Devataakaanda of Kashakrtsna/Paila) on the SK in the Mādhva tradition, where he concludes that the SK-devatākāṇḍā existed as a separate text. Unfortunately, I could not find out whether the paper has been or will be published. 312 Elisa Freschi but then calls it devatākāṇḍa (saṅkarṣyate karmakāṇḍastham evāvaśiṣṭaṃ karma saṃkṣipyocyate iti saṅkarṣo devatākāṇḍam, “The Devatākāṇḍa is called Saṅkarṣa because it summarizes what is left [of] the ritual action, which is dealt with in the portion on the ritual action (e.g., the PMS),” commentary on Śaṅkara’s Bhāṣya on UMS 3.3.43). Thus, following Śaṅkara, a mostly technical text came to be used in devotional contexts because of its mention of devatās. The mention of devatās within the SK only served ritual purposes (as in the PMS), so that there was no intrinsic reason to choose the SK as a theistic text. But it was certainly easier to adopt a lesser known work like the SK for one’s purposes than the well- known PMS, given that Pūrva Mīmāṃsā was still active and well known as an atheist school. This might also be what is hinted at by Madhusūdana Sarasvatī in his Prasthānabheda (see section 4.2.4), where he makes sense of the SK by saying that it teaches upāsanākhyakarman, i.e., “ritual [as it makes sense within the Pūrva Mīmāṃsā] called veneration (upāsa- nā).” The Pūrva Mīmāṃsā milieus seemingly remained unaffected by this move (remember Someśvara’s reference to the extant SK as late as the 12th century and the general lack of interest in the SK, especially the SK-devatākāṇḍa). Later Advaita Vedānta authors acknowledged the existence of the SK but did not emphasize its theistic content. One might imagine that they had to acknowledge it because of the growing popularity of the name “SK” among theistic Vedāntins, but that they had no interest in admitting the SK- devatākāṇḍa into their systems. 4.2.7 The authorship of the Saṅkarṣakāṇḍa As for the authorship of the SK, some witnesses (see below and cf. Kana- zawa 1989: 40) connect it to a person named Saṅkarṣa(ṇa). Several witnesses speak of the SK as having been authored by Jaimini (as for example, the perhaps first commentator of the SK, Devasvāmin,68 the Vṛttikāra’s quote mentioned by Rāmānuja and Veṅkaṭanātha, the Prapañcahṛdaya,69 Khaṇḍadeva in the Bhāṭṭadīpikā and Bhāskara Rāya in what he considered the prosecution of the latter work, and Madhusūdana Sarasvatī [see section 4.2.4]). By contrast, the Sarvasiddhāntasaṅgraha attributes both the UMS and the SK to Vyāsa (see Kanazawa 1989: 41), and this might be also the position of Jayatīrtha: 68 According to Lariviere 1981, but Aklujkar (2012: 215–216) convincingly argues that the colophon Lariviere refers to is a modern coinage by the editors of his commentary. 69 “After 900 AD, later limit undecided” (Aklujkar 2012: 212).  Reusing, Adapting, Distorting? 313 After that [the desire to know the brahman starts]. Because the Re- vered [author of the UMS] had composed the last two sūtras of the Devatāmīmāṃsā (i.e., the SK).70 The Sarvamatasaṅgraha might depend on the Sarvasiddhāntasaṅgraha when it also attributes both the SK and the UMS to Vyāsa. However, it adds a new nuance, that is, it states that both works deal with brahman, saguṇa and nirguṇa, respectively. Vedānta Deśika refers to the author of the SK as Kāśakṛtsna, the name of an ancient Mīmāṃsaka of whom no other work has been preserved but whose views are referred to in the PMS and the UMS. In a learned and insightful study (Aklujkar 2012), Ashok Aklujkar has sup- ported this attribution with the references to Kāśakṛtsna in the PMS and the UMS and in the Mahābhāṣya, but also on the basis of the fact that: 1. The attestations (e.g., in the Vṛttikāra’s quote preserved in the ŚrīBh and in the SM) regarding the SK as jaiminīya only mean ‘connected to Jai- mini’s PMS’ and not “authored by Jaimini” (Aklujkar 2012: 205–207 and p. 210). 2. The name “Saṅkarṣa” does not need to refer to the author. At any rate, all testimonies mentioning a person named Saṅkarṣa regard him as the com- mentator of the SK, not its author (Aklujkar 2012: 212 and p. 223). 3. Some of the authors referring to Jaimini as the author of the SK, namely Khaṇḍadeva, Śambhubhaṭṭa and Bhāskararāya, constitute a mutually con- nected group. Therefore, these authors are not independent witnesses.71 4. Kāśakṛtsna is a suitable candidate for the authorship of the SK because he is an ancient Mīmāṃsaka, and there is no other way to explain the attri- bution of the SK to him (Aklujkar 2012: 210), given that he was neither well known nor influential. In favor of Aklujkar’s hypothesis, it can be added that even a quote he and K. V. Sarma, the editor of the editio princeps of the SK sūtras, consider evi- dence in favor of Jaimini’s authorship, namely Śambhubhaṭṭa’s Prabhāvalī 70 It is not clear whether Jayatīrtha wants to attribute only the last two sūtras to Bādarā- yaṇa (as claimed in Siauve 1957: 23). In contrast, I am inclined to think that the ex- pression antimasūtradvayam ‘the last two sūtras’ does not have an exclusive purpose (“just the last two sūtras”), but rather stresses the connection of these two sūtras with the following UMS. 71 Aklujkar (2012: 215) further points out that the three authors probably based their attribution of the SK to Jaimini on Devasvāmin’s commentary on the SK. This claim, however, contradicts Aklujkar’s accurate demonstration that Devasvāmin did not re- gard Jaimini as the author of the SK; see above n. 68. 314 Elisa Freschi on Khaṇḍadeva’s Bhaṭṭadīpikā, should in my opinion be interpreted in a different way. The text runs as follows: The twelve chapters [PMS] composed by Jaimini and beginning with “Now, because of that the desire to know the dharma starts,” and the four chapters composing the Saṅkarṣaṇakāṇḍa constitute the Karma- mīmāṃsā. The four chapters [UMS] composed by Vyāsa and begin- ning with “Now, because of that, the desire to know the brahman starts” is the Śārīrakamīmāṃsā.72 In other words, Śambhubhaṭṭa attributes the PMS to Jaimini, whereas he leaves the SK without an author. The conceptual unity of PMS and SK does not mean that they share the same author, as proved by the fact that the ex- pression ‘composed by Jaimini’ is clearly put within the clause about the PMS. Furthermore, there are no (or few) explicit arguments against Kāśakṛtsna’s authorship of the SK. K. V. Sarma in his Introduction to Sarma 1963 uses the following arguments against the authorship of Kāśakṛtsna (and in favor of that of Jaimini): – He endorses K. C. Chatterjee’s view that “though the SK is fairly early, […] it is still difficult to hold that Kāśakṛtsna is its author.” – Vedānta Deśika himself did not uphold Kāśakṛtsna’s authorship, as shown by the fact that he quotes the Vṛttikāra’s statement approvingly. However, in fact, Veṅkaṭanātha reuses the Vṛttikāra’s passage for his own agenda and as further evidence in favor of the unity of the Mīmāṃsā śāstra exactly insofar as it states – at first sight erroneously – that Jaimini composed the PMS and the SK (see above, section 4.2). And as for the argument that the SK cannot have been authored by Kāśakṛtsna because the work is “fairly early,” this is also not a strong one. In fact, Sarma himself maintains that the SK was authored by Jaimini, who is also not dated later than Kāśakṛtsna, since both are mentioned as authorities in the PMS and the UMS. Nonetheless, a few further points may be added to Aklujkar’s learned re- construction (which, in Aklujkar’s own explanation, limits itself to the extant SK and does not take the SK-devatākāṇḍa into account73). 72 athāto dharmajijñāsā ityādinā jaiminipraṇītā dvādaśādhyāyī saṅkarṣaṇakāṇḍātmikā ca- turādhyāyī ca karmamīmāṃsā, athāto brahmajijñāsā ityādinā vyāsapraṇītā caturādhyāyī śārīrakamīmāṃsā ca (Krishna Sastri 1987: 43). I am grateful to the University Li- brary of Marburg for providing me a copy of this text. 73 “In almost all of my sentences ‘Saṅkarṣakāṇḍa (= SK)’ stands for the text found in Subrahmanya Sastri’s edition [Subrahmanya Sastri 1965] and in an indirect, second-  Reusing, Adapting, Distorting? 315 1. Veṅkaṭanātha clearly needed the SK-devatākāṇḍa for his attempt to dem- onstrate the unity of the Mīmāṃsā śāstra and its theistic core. 2. In this connection, the attribution of the SK to Vyāsa, with the underlying assumption that the same brahman was dealt with from two different perspectives, might have suited him best. 3. However, his illustrious predecessor, the Vṛttikāra, had attributed it to Jaimini. 4. This attribution was not a viable solution for Veṅkaṭanātha, who had a difficult task in front of him, namely to convince the Viśiṣṭādvaitins of the legitimacy of his attempt to broaden Viśiṣṭādvaita Vedānta through the claim that it formed the same śāstra as the Pūrva Mīmāṃsā. The SK was meant to work as a bridge. Attributing it to Jaimini would have just elimi- nated the possibility of it being a mediation, pushing it back into the Pūrva Mīmāṃsā field. 5. It is, moreover, possible that the attribution to Vyāsa was not yet current at the time of Veṅkaṭanātha (the dates of the Sarvasiddhāntasaṅgraha and of the Sarvamatasaṅgraha have not been fixed, but the latter was proba- bly acquainted with the former and refers to Rāmānuja, see Kanazawa 1989: 33). 6. By contrast, another attribution was surely known to Veṅkaṭanātha (since he mentioned it in his Adhikaraṇa Sārāvali, v. 15), namely Parāśara Bhaṭṭa’s attribution of the SK to Kāśakṛtsna in his Tattvaratnākara.74 Veṅkaṭanātha’s unnatural interpretation of the Vṛttikāra’s quote derives from these premises which, possibly, led Veṅkaṭanātha to resort to an intermediate position among the various ones he may have been exposed to at his time. Attributing the SK to Jaimini would not have helped Veṅkaṭanātha’s attempt to justify the aikaśāstrya of Pūrva Mīmāṃsā and Vedānta. The attribution of the SK to Vyāsa would have raised suspicions among Vedāntins. The attri- bution to Kāśakṛtsna was in this sense a useful intermediate solution. ary way in the edition of Bhāskararāya’s complementation of Khaṇḍadeva’s Bhāṭṭa- dīpikā [Śāstrī 1894]. There are passages in Viśiṣṭādvaita literature in which the SK is understood as a text primarily or entirely devoted to the discussion of deities. It is also spoken of as a completely or almost completely lost text. Signs of ambiguity and ab- sence of first-hand knowledge also exist. This description is inapplicable to the SK I have in mind in almost all of the present essay” (Aklujkar 2012: 191, n. 1). 74 In the fragment No. 56 in Oberhammer 1979: karmadevatābrahmagocarā sā tridhod- babhau sūtrakārataḥ | jaiminer muneḥ kāśakṛtsnato bādarāyaṇād ity ataḥ kramāt || “This [Mīmāṃsā], which has within its scope ritual action, deities and brahman ap- peared in three ways, according to the author of the sūtras, | that is, respectively, the sage Jaimini, Kāśakṛtsna and Bādarāyaṇa.” See also Aklujkar 2012: 194–200. 316 Elisa Freschi 5 Yāmuna and the Seśvaramīmāṃsā: Shared textual material As seen above (section 2 and section 3), Veṅkaṭanātha frequently refers to Rāmānuja. On much more rare occasions, Veṅkaṭanātha needs the support of both Yāmuna and Rāmānuja. This is the case when he has to explain away the alleged atheism of Jaimini. Here he happily quotes Yāmuna and Rāmā- nuja on a related topic, namely, the emphasis on rituals in the PMS: uktaṃ hy āgamaprāmāṇye bhagavato jaimineḥ karmaṇaḥ phalopanyā- saḥ karmaśraddhāsaṃvardhanāya75 iti. vedārthasaṅgrahe ’pi aśrutave- dāntānāṃ karmaṇy aśraddhā mā bhūd iti devatādhikaraṇe ’tivādāḥ kṛ- tāḥ karmamātre yathā śraddhāṃ kuryāt iti sarvam ekaṃ śāstram iti ve- davitsiddhāntaḥ.76 (Veṅkaṭanātha 1971: 5)77 Rāmānuja’s passage (embedded in a long rebuttal of the Prābhākara view) goes more in Veṅkaṭanātha’s direction, since it explicitly emphasizes the unity of Pūrva and Uttara Mīmāṃsā. In contrast, the passage in Yāmuna does not aim directly at Pūrva Mīmāṃsā, since it rather defends the compatibility of Vedas and Pāñcarātra through the (Mīmāṃsā) device of option (vikalpa) and discusses in this connection the problem of possible contradictions be- tween various commentators (bhāṣyakāra78). His answer is that if one care- fully investigates something there are no contradictions. As the case of Jai- mini shows, one can reconcile seemingly contradictory statements if one considers that a certain author might have emphasized the role of the Vedas (presumably over that of the Pāñcarātra) in order to keep people from disres- pecting them. 75 Cf. the almost identical passage in the Āgamaprāmāṇya: yathaiva hi bhagavato jaimineḥ karmaphalopanyāsaḥ karmaśraddhāsaṃvardhanāya (Śāstri 1937: 67). 76 Cf. the almost identical passage in the Vedārthasaṅgraha: aśrutavedāntānāṃ karmaṇy aśraddhā mā bhūd iti devatādhikaraṇe ’tivādāḥ kṛtāḥ karmamātre yathā śraddhā syād iti sarvam ekaśāstram iti vedavitsiddhāntaḥ (Raghavachar 1978: 157). 77 “It has been said in [Yāmunācārya’s] Āgamaprāmāṇya: ‘The statement that the fruit [comes] out of the ritual action (karman) (and not out of God) by the Revered Jaimini has the purpose of promoting faith in ritual action.’ And also in [Rāmānuja’s] Vedārthasaṅgraha: ‘In order to avoid the lack of faith in ritual action of people who have not heard the Upaniṣads (aśrutavedānta), some excessive statements (ativāda) have been used in the Devatādhikaraṇa, so that one should put faith in the ritual action as such. Thus, the definitive conclusion of those who know the Veda is that this all is a single treatise (śāstra).” 78 Commentators on what? The context makes one think that Yāmuna is referring to commentaries on the (U)MS.  Reusing, Adapting, Distorting? 317 However, as already hinted at, there is much more shared textual material between the SM and Yāmuna’s Ātmasiddhi. This material is however never identified as such. Moreover, it is also often embedded within prima facie views (!). Particularly telling is the case of Yāmuna’s endorsement of yogipratyakṣa, a kind of intellectual intuition through which one can directly access non-sensory matters. Yāmuna’s yogipratyakṣa plays an important role especially in his extolment of his grandfather Nāthamuni, praised as a great yogin and a great bhākta who could directly access God’s reality through yogipratyakṣa (see the beginning of Yāmuna’s Stotraratna). In contrast, the Pūrva Mīmāṃsā strongly opposes (in PMS 1.1.4 and in its commentaries) the epistemological role of yogipratyakṣa, since it maintains that the Veda is the only key to accessing super-sensory matters, such as dharma and heaven (see McCrea 2009). Veṅkaṭanātha comments on PMS 1.1.4 completely supporting the Pūrva Mīmāṃsā point of view, but adds at the end some unexpected verses in which he states that what has been said only applies to the yogins’ pratyakṣa, i.e., to the extraordinary intellectual in- tuition of yogins, but not to God’s intellectual intuition (called aiśādhyakṣa). In this context, Veṅkaṭanātha reuses Yāmuna’s words in passages repro- ducing the views of an opponent (pūrvapakṣin) in SM ad PMS 1.1.4. sarvaṃ hi sātiśayaṃ niratiṣayadaśām anubhavad dṛṣṭaṃ viyatīva pari- māṇam, sātiśayaṃ ca kākolūkagṛdhrādiṣu pratyakṣam īkṣitam iti tad api tathā bhavitum arhati. iyaṃ ca jñānasya parā kāṣṭhā yā sarvago- caratā, adhikaviṣayatayaiva hi jagati jñānāni parasparam atiśerate. […] tan nu pratyakṣavijñānaprakarṣaḥ kalpito ’pi vaḥ | svagocaram ati- kramya nānyad āskandituṃ kṣamah || 6 || tathā hi – rūparūpitade- kārthasamavāyiṣu cākṣuṣaḥ | prakarṣo bhavituṃ yukto dṛśyamānapra- karṣavat || (Āgamaprāmāṇya, beginning)79 79 “[Obj.:] In fact, everything liable to improvement is seen to experience a level which has no further abode [beyond itself], like in the case of the sky the measure [of which is bigness without anything bigger]. And sense-perception has been seen to be liable to improvement in the case of crows, owls, vultures, etc. [each of which sees better than the preceding one]. Therefore, also this [intellectual intuition] can be in the same way. And this is the upmost level of knowledge: the fact of regarding everything. For, in the world cognitions surpass each other insofar as they have a larger content. […] [R:] Therefore indeed, the intensity in the perceptual cognitions, although it has been post- ulated by you | cannot go beyond its own field and invade another one || 6 || To elabo- rate: the intensity of the visual sense faculty can occur in regard to things in which there is inherence of the visible content in a visible substrate, and things inherent in this single object (visible content), just like the intensity in what is seen [which can increase, but will not go beyond its precinct] ||”. 318 Elisa Freschi kākolūkagṛdhrādiṣu ca indriyaprakarṣatāratamyaṃ dṛśyate. tāra- tamyavatāṃ ca kāṣṭhāprāptir upalabdhā. tataś ca prakarṣatāratamyaṃ kvacid viśrāntaṃ tāratamyarūpatvāt parimāṇatāratamyavad ity evaṃ pratīndriyaṃ vā pratyakṣaprakarṣatāratamyam iti sāmānyena vā pra- yoktavyam. evaṃ sthite tadviśramasya sarvaviṣayatām antareṇānupa- patter atīndriyaṃ sarvaṃ kasyacid aindriyakatvam iti siddhyati. (SM ad 1.1.4)80 Does Veṅkaṭanātha simply re-use, because Yāmuna’s formulation was easily available? Does the fact that Veṅkaṭanātha does not reuse Yāmuna literally imply that he does not need his readers to be aware of the reuse, possibly because he does not want them to see his own thoughts as being close to those of Yāmuna?81 80 “[Obj.:] Furthermore, one commonly experiences a gradation in the intensity of the sense faculties, as in the case of crows, owls, vultures, etc. [each of which sees better than the preceding one]. And one perceives that [among] those who take part in this gradation, one reaches the [highest] level. And hence, [the inference] ‘Somewhere the gradation in intensity comes to an end, because of its nature of gradation, like the gra- dation of measures’* can either be applied in this way to every sense faculty (starting from sight, as in the example mentioned above), or it can be applied in general in the form of ‘the gradation in intensity of direct perception.’ If this is the case (i.e., if these two are the only alternatives), every super-sensuous [object] is established as sensory in relation to someone [e.g., the perception of small ants is sensory for one who has well-trained eyes], because its (of the gradation of intensity of direct perception) rest would not be logically possible without (before) the fact that everything [has become] its content.”* This might refer to the fact that in ancient India the belief prevailed that a highest number exists. In the Padārthadharmasaṅgraha on VS, parimāṇas are said to be of four types: aṇu, mahat, dīrgha and hrasva. I suppose that the gist of the argument is that mahat (or dīrgha) should exhaust the possible magnitude (or length) of some- thing. For a possible source of the argument, see Pātañjalayogaśāstra 1.25, where the highest degree of knowledge of God is inferred on the basis of the fact that there must be a limit, just as in the case of measures (parimāṇavat). I am grateful to Philipp Maas for having pointed this passage out. Also the Pātañjalayogaśāstravivaraṇa on Pātañja- layogaśāstra 1.25 equates the peak of knowledge with omniscience (see Harimoto 2014: 10–11). See also Maas’ chapter in the present volume, section 4.2.1. 81 Interestingly, Rāmānuja also pursued a similar procedure with regard to Yāmuna, since he reused his works more often than he acknowledged such reuses; see Mesquita 1971: 4.  Reusing, Adapting, Distorting? 319 6 Conclusions Veṅkaṭanātha reuses Rāmānuja for two reasons: 1. because he appreciates his work and shares its main points, so that he does not feel the need to formulate again what has already been said per- fectly by Rāmānuja: this leads to unacknowledged reuses; 2. because he needs Rāmānuja’s prestige and indirect support for his innova- tions, i.e., for broadening the scope of the Mīmāṃsā system: this leads to explicit quotations or acknowledged reuses. In the case of the Vṛttikāra, it is impossible to detect whether Veṅkaṭanātha shared his views, given the scarcity of his fragments. Instead, the prestige of the Vṛttikāra seems to be the main reason for the forced interpretation of his SK quote (see above, section 4.2). Veṅkaṭanātha’s relationship to Yāmuna (and possibly other forerunners) is different, insofar as Veṅkaṭanātha probably appreciated Yāmuna’s formu- lations enough to want to re-use them, but he probably did not want to be understood as being too close to Yāmuna. Accordingly, reuses are often un- acknowledged. Possible reasons for this have been suggested in section 2 and are based on Veṅkaṭanātha’s disagreement with Yāmuna’s ultimate agenda, especially in its anti-Pūrva Mīmāṃsā part. The status of the SK is an interesting test for Veṅkaṭanātha’s adherence to his own tradition. As one can see, the result is that Veṅkaṭanātha was enlarg- ing (perhaps even pushing) Rāmānuja’s system to include elements Rāmānuja had not explicitly meant to embrace, such as the Pāñcarātra and the Pūrva Mīmāṃsā. 320 Elisa Freschi References Abbreviations Anuv Anuvyākhyāna, by Madhva, in Siauve 1957 BS Brahmasūtra (or Uttaramīmāṃsāsūtra), by Bādarāyaṇa, in Kato 2011 and Joshi 1981 MP Mīmāṃsāpādukā, by Veṅkaṭanātha, in Veṅkaṭanātha 1971 NS Nyāyasudhā on the Tantravārttika, by Someśvara, in Sāstrī 1909 PMS Pūrvamīmāṃsāsūtra, by Jaimini, in Frauwallner 1968 (PMS 1) and Subbāśāstrī 1929‒1934 ŚBh Śābarabhāṣya on the PMS, by Śabara, in Frauwallner 1968 (PMS 1) and Subbāśāstrī 1929‒1934 SK Saṅkarṣakāṇḍa, in Śāstrī 1894, Sarma 1963, and Subrah- manya Sastri 1965 SM Seśvaramīmāṃsā, by Veṅkaṭanātha, in Veṅkaṭanātha 1971 SpK Spandakārikā, by Vasugupta, in Dyczkowski 2000 ŚrīBh Śrībhāṣya on the BS, by Rāmānuja, in Mahādeśika 1938 UMS Uttaramīmāṃsāsūtra (or Brahmasūtra), by Bādarāyaṇa Primary and Secondary Sources Aklujkar 2012 Aklujkar, Ashok. “Authorship of the Saṅkarṣakāṇḍa.” Devadattīyam. Johannes Bronkhorst Felicitation Volume. Ed. François Voegeli et al. Bern: Lang, 2012. 191–227. Anandacharlu Vidyāvinod 1903 = Viśārada Anandacharlu Vidyāvinod, ed. Śata- dūṣaṇī [by Vedānta Deśika] with the Commentary En- titled Caṇḍamāruta by Śrīman-Mahācārya. Calcutta: Asiatic Society, Baptist Mission Press, 1903. Aṇṇāṅgarācārya 1941 Aṇṇāṅgarācārya, Prativādibhayaṅkara, ed. Śrībhāṣyā by Rāmānuja with Tattvaṭīkā by Veṅkaṭanātha: Śrī- bhāṣyavyākhyā Tattvaṭīkā. Kañcīvaram: Granthamālā- kāryāla, 1941. Śrīmadvedāntadeśikagranthamālā. Appayya Dīkṣita 1890 Appayya Dīkṣita. Siddhāntaleśasaṅgraha with Extracts of the Śrīkṛṣṇālaṃkāra of Acyutakṛṣṇānandatīrtha. Ed. Gaṅgādhara Śāstrī Mānavallī. Benares: E. J. Lazarus and Co., 1890. Vizianagram Sanskrit Ser. Bakhśī 1981 Bakhśī, Mukunda Jhā, ed. Śrīlakṣmaṇadeśikendravira- citaṃ Śāradātilakam śrīmadrāghavabhaṭṭakṛta ‘padār- thādarśa’ vyākhyāsahitam. Vārāṇasī: Caukhambhā Saṃskṛta Saṃsthāna, 1981. Kāśī Saṃskṛta Grantha- mālā 107. Bronkhorst 2007 Bronkhorst, Johannes. “Vedānta as Mīmāṃsā.” Mīmāṃsā and Vedānta: Interaction and Continuity. Ed. Johannes Bronkhorst. Delhi: Motilal Banarsidass, 2007. 1–91.  Reusing, Adapting, Distorting? 321 Bronner 2007 Bronner, Yigal. “Singing to God, Educating the People: Appayya Dīkṣita and the Function of Strotras.” Journal of the American Oriental Society 127.2 (2007): 1–18. Duraiswami Aiyangar and Venugopalacharya 1942 = Duraiswami Aiyangar, M. and T. Venugopalacharya, eds. Śrī Pāñcarātrarakṣā of Śrī Vedānta Deśika: Crit. Ed. with Notes and Variant Readings and with an Introd. in English by G. Srini- vasa Murti. Madras: The Adyar Library, 1942. The Adyar Library Ser. 36. Dvivedin 1903 Dvivedin, Vindhyeshvarī Prasāda, ed. Brahmasūtra with a Commentary by Bhāskarāchārya. Benares: Vidya Vi- las Press, 1903. Chowkhambā Sanskrit Ser. Dyczkowski 1994 Dyczkowski, Mark. The Stanzas on Vibration: The Span- dakārikā with Four Commentaries. Transl. with Introd. and Exposition. Varanasi: Dilip Kumar Pub- lishers, 1994. Dyczkowski 2000 —, ed. Spandapradīpikā. A Commentary on the Span- dakārikā by Bhagavadutpalācārya. Varanasi: Indica Books, 2000. Frauwallner 1968 Frauwallner, Erich, ed. Materialien zur ältesten Er- kenntnislehre der Karmamīmāṃsā. Wien: Verlag der Österreichischen Akad. d. Wiss., 1968. Kommission für Sprachen und Kulturen Süd- und Ostasiens; Philo- sophisch-Historische Klasse; Sitzungsberichte. Freschi 2012 Freschi, Elisa. “Proposals for the Study of Quotations in Indian Philosophical Texts.” Religions in South Asia 6.2 (2012): 161–189. Freschi 2015 —, ed. The Re-use of Texts in Indian Philosophy. Spec. issue of Journal of Indian Philosophy 43.2–3 and 4–5 (2015). Freschi and Kataoka 2012 = Freschi, Elisa and Kei Kataoka. “Jayanta on the Vali- dity of Sacred Texts. Annotated English Translation and Study.” South Asian Classical Studies 161 (2012): 1– 55. Freschi and Pontillo 2013 = Freschi, Elisa and Tiziana Pontillo. Rule-extension- strategies in Ancient India: Śrautasūtra, Mīmāṃsā and Grammar on tantra- and prasaṅga-principles. Wien: Lang, 2013. Harimoto 2014 Harimoto, Kengo. God, Reason, and Yoga: A Critical Edi- tion and Translation of the Commentary Ascribed to Śaṅ- kara on Pātañjalayogaśāstra 1.23–28. Hamburg: Depart- ment of Indian and Tibetan Studies, Universität Hamburg, 2014. Indian and Tibetan Studies 1. 322 Elisa Freschi Haspelmath 2014 Haspelmath, Michael. The Generic Style Rules for Linguistics. Web. 11 February 2016. <https://www. academia.edu/7370927/The_Generic_Style_Rules_for_ Linguistics>. Joshi 1981 Joshi, K. L., ed. The Brahmasūtra Śaṅkara Bhāṣya with the commentaries Bhāmatī [by Vācaspati Miśra], Kalpa- taru [by Amalānanda Sarasvatī] and Parimala [by Ap- payya Dīkṣita]. Ahmedabad: Parimal, 1981. Parimala Sanskrit Ser. 1. Kanazawa 1989 Kanazawa, Atsushi. “Notes on the Saṅkarṣakāṇḍa: Under Stimulus from the Article by Lariviere.” Acta Asiatica 57 (1989): 31–44. Kane 1962 Kane, Pandurang Vaman. History of Dharmaśāstra. An- cient and Mediaeval Religious and Civil Law in India. Vol. 5.2. P o o n a : Bhandarkar Oriental Research Inst., 1962. Kato 2009–2010 Kato, Takahiro. “Bhāskara’s Brahmasūtrabhāṣya – An Unpublished Edition by J. A. B. van Buitenen.” Wiener Zeitschrift für die Kunde Südasiens 52–53 (2009–2010): 295–305. Kato 2011 — “The First Two Chapters of Bhāskara’s Śarīrakamī- māṃsābhāṣya: Critically ed. with an Introduction, Notes and an Appendix.” Diss. Halle: Universität Hal- le, Philosophische Fakultät, 2011. Kato 2012 — “Bhāskara’s Concept of Jñānakarmasamuccaya.” In- dologica Taurinensia 38 (2012): 1–17. Katre 1989 Katre, Sumitra M., ed. Aṣṭādhyāyī of Pāṇini. 2nd e d . (1st ed. Austin 1987). Delhi: Motilal Banarsidass, 1989. Krishna Sastri 1987 Krishna Sastri, Ananta, ed. Bhāṭṭadīpikā [by Khaṇḍade- va], with the Commentary Prabhāvalī of Śambhu Bhaṭṭa. 2nd ed. (1st ed. Bombay 1921–1922). Delhi: Sri Satguru, 1987. Sri Garib Dass Oriental Ser. 50. Krishnamacharya 1938 = Ananthacharya, V. and V. Krishnamacharya, eds. Śrī- Bhāṣya or Brahmasūtrabhāṣya by Bhagavad-Rāmā- nuja with Ten Commentaries. Krishnapuram: The Madras Sanskrit Book Depot, 1938. Kupalapati 2009 Kupalapati, Sannidhāna Sudarśana Śarmā, ed. Śrīmaj- jaiminimunipraṇītaṃ saṅkarṣakāṇḍam. Śrībhāskara- praṇīta bhāṭṭacandrikayā saṃvalitam. Tirupati: Śrī- veṅkateśvaravedaviśvavidyālayaḥ, 2009. S. V. Vedic University Ser. 27.  Reusing, Adapting, Distorting? 323 Lariviere 1981 Lariviere, Richard W. “Madhyamamīmāṃsā – The Saṅ- karṣakāṇḍa.” Wiener Zeitschrift für die Kunde Südasiens 25 (1981): 179–194. Leach 2012 Leach, Robert. “Textual Traditions and Religious Identities in the Pāñcarātra.” Diss. Edinburgh: Univer- sity of Edinburgh, 2012. Mahādeśika 1938 Mahādeśika, Śrīraṅgaśaṭhakopayatīndra, ed. Tattvaṭīkā [by Veṅkaṭanātha]. Śrībhagavadrāmānujaviracitaśrībhā- ṣyasya vyākhyā kavitārkikasiṃhasarvatantrasvatantra- śrīnigamāntamahādeśikaiḥ anugṛhītā. Madras 1938. Śrīvaiṣṇavasiddhānta Grantharatnamālā 2. Marlewicz 2007 Marlewicz, Halina. “The Question of the Unity of Kar- makāṇḍa and Jñānakāṇḍa According to Rāmānuja.” Studies in Hinduism 4. On the Mutual Influences and Relationship of Viśiṣṭādvaita Vedānta and Pāñcarātra. Ed. Gerhard Oberhammer and Marion Rastelli. Wien: Verlag der Österreichischen Akad. d. Wiss., 2007. 75– 88. McCrea 2009 McCrea, Lawrence. “‘Just Like Us, Just Like Now’: The Tactical Implications of the Mīmāṃsā Rejection of Yogic Perception.” Yogic Perception, Meditation and Altered States of Consciousness. Ed. Eli Franco. Wien: Verlag der Österreichischen Akad. d. Wiss., 2009. Beiträge zur Kultur- und Geistesgeschichte Asiens 64. Mesquita 1971 Mesquita, Roque. “Das Problem der Gotteserkenntnis bei Yāmunamuṇi.” Diss. Wien: Universität Wien, 1971. Mesquita 1980 — “Yāmuna’s Vedānta and Pāñcarātra: A Review.” Wiener Zeitschrift für die Kunde Südasiens 24 (1980): 199–224. Nakamura 1983 Nakamura, Hajime. A History of Early Vedānta Philo- sophy. Pt. 1. Transl. into English by Trevor Leggett. Delhi: Motilal Banarsidass, 1983. Religions of Asia Ser. 1. Narasimhachary 1998 Narasimhachary, Mudumby. Contribution of Yāmunā- cārya to Viśiṣṭādvaita. 2nd ed. (1st ed. 1971). Hydera- bad: Sri Jayalakshmi Publications, 1998. Neevel 1977 Neevel, Walter G. Jr. Yāmuna’s Vedānta and Pāñca- rātra: Integrating the Classical and the Popular. Mis- soula, Montana: Scholars Press, 1977. Harvard Disser- tations in Religion 10. Oberhammer 1971 Oberhammer, Gerhard. Yāmunamunis Interpretation von Brahmasūtram 2, 2, 42–45. Eine Untersuchung zur 324 Elisa Freschi Pāñcarātra-Tradition der Rāmānuja-Schule. Wien: Hermann Böhlaus Nachf., 1971. Oberhammer 1979 — Materialien zur Geschichte der Rāmānuja-Schule I. Parāśarabhaṭṭas Tattvaratnākaraḥ. Wien: Verlag der Österreichischen Akad. d. Wiss., 1979. Olivelle 1998 Olivelle, Patrick. The Early Upaniṣads: Annotated Text and Translation. Oxford: Oxford University Press, 1998. Pandurangi 2002 Pandurangi, Krishnacharya Tamanachary, ed. Śrī Jaya- tīrthaviracitā Nyāyasudhā ṣaṭṭippaṇīsahitā. Vol. 2. Ban- galore: Dvaita Vedanta Studies and Research Foun- dation, 2002. Prasthānabheda [by Madhusūdana Sarasvatī] 1912 = Madhusūdana Sarasvatī. Pra- sthānabheda. Srirangam: Sri Vani Vilas Press, 1912. Raghavachar 1978 Raghavacar, S. S., ed. Vedārtha-Saṅgraha of Śrī Rāmānu- jācārya. [Text and] English Translation. Mysore: Sri Ramakrishna Ashrama, 1978. Ramaswami Sastri 1933 Ramaswami Sastri, V. A. “The Saṃkarṣa Kāṇḍa – A Ge- nuine Supplemement to the Pūrvamīmāṃsā Śāstra.” In- dian Historical Quarterly 9 (1933): 290–99. Rastelli 2003 Rastelli, Marion. “The Ekāyanaveda in the Pāñcarātra Tradition.” Web. 10 April 2015. <http://www.ikga.oeaw. ac.at/Online_Texte#MarionRastelli>. Rastelli forthcoming — “Yoga in the Daily Routine of the Pāñcarātrins.” Yoga in Transformation. Historical and Contemporary Perspectives on a Global Phenomenon. Ed. Karl Baier, Philipp A. Maas and Karin Preisendanz. Sanderson 2009a Sanderson, Alexis. “Kashmir.” Brill’s Encyclopedia of Hinduism, Vol. 1: Regions, Pilgrimage, Deities. Ed. Knut A. Jacobsen. Boston: Brill, 2009. 99–126. Handbuch der Orientalistik. Zweite Abteilung, Indien, vol. 22. Sanderson 2009b — “The Śaiva Age ‒– The Rise and Dominance of Śaivism during the Early Medieval Period.” Genesis and Development of Tantrism. Ed. Shingo Einoo. Tokyo: Institute of Oriental Culture, University of Tokyo, 2009. 41–349. Sarma 1963 Sarma, K. V., ed. Saṅkarṣa Kāṇḍa Sūtras of Jaimini. Critically Edited with Introduction and Appendix. Ho- shiarpur: Vishveshvaranand Vedic Research Institute, 1963. Vishveshvaranand Indological Ser. 18. Sāstrī 1909 Sāstrī, Mukunda, ed. Nyāyasudhā ṭīkā bhaṭṭasomeśva- raviracitā. A Commentary on Tantravārtika by Some- śwara Bhaṭṭa. Benares: Vidya Vilas Press, 1909. Chow-  Reusing, Adapting, Distorting? 325 khamba Sanskrit Ser. 45–50, 52, 53, 55–57, 60, 71, 73, 107, 121, 133. Śāstrī 1894 Śāstrī, Rāma Miśra, ed. Saṅkarṣa Kāṇḍa [with the Bhāṭṭacandrikā commentary by Bhāskararāya]. Bana- ras: Pandit Reprint, 1894. Śāstri 1937 Śāstri, Rāma Miśra, ed. Āgamaprāmāṇyam by Śrī Yamunachārya Swamin. Vārāṇasī: Tārāyantrālaya, 1937. Siauve 1957 Siauve, Suzanne, ed. La voie vers la connaissance de dieu (Brahma-Jijñāsā) selon l’Anuvyākhyāna de Madh- va. Pondichéry: Institut Français d’Indologie, 1957. Publications de l’Institut Français d’Indologie 6. Śrīkaṇṭha and Appayya Dīkṣita 1908 = Śrīkaṇṭha and Appayya Dīkṣita. Brahma- sūtrabhāṣya of Śrīkaṇṭha: With the commentary Śivār- kamaṇidīpikā of Appaya Dīkṣita, ed. with Sūtrārthacan- drikā by Halasyanatha Sastri. Ed. Halasyanatha Sastri. Bombay: Nirṇayasāgara Press, 1908. Bharatee Mandi- ram Sanskrit Ser. 1. Stark 1990 Stark, Sylvia. Vātsya Varadagurus Tattvanirṇaya. Teil 1: Kritische Textedition. Wien: Verlag der Österreichi- schen Akad. d. Wiss., 1990. Steinkellner 1988 Steinkellner, Ernst. “Methodological Remarks on the Constitution of Sanskrit Texts from the Buddhist Pra- māṇa-Tradition.” Wiener Zeitschrift für die Kunde Süd- asiens 32 (1988): 103–129. Subbāśāstrī 1929‒1934 Subbāśāstrī, ed. Śrīmajjaiminipraṇitaṃ Mīmāṃsādar- śanam. Poona: Ānandāśramamudrāṇālaya, 1929‒1934. Subrahmanya Sastri 1965 = Subrahmanya Sastri, S., ed. Saṅkarṣa Kāṇḍa of Sage Jaimini with the Bhāṣya of Devasvāmin. Madras: Univer- sity of Madras, 1965. Tātācārya 2009 Tātācārya, Śaṭhakopa Rāmānuja, ed. Saṃkarṣakāṇ- ḍam. Jaiminīyaṣoḍaśādhyāyyāḥ śeṣacaturadhyāyīsvarū- pam. Śrībhāskarabhaṭṭacandrikopetam. Chennai: Śrīraṅ- gaṃ Śrīmadāṇḍavanāśramaḥ, 2009. Tātāyāryadāsa 2004 Tātāyāryadāsa, Śaṭhakoparāmānuja, ed. Śatadūṣaṇī [by Veṅkaṭanātha]. Chennai: Srirangam Srimad Andavan Ashrama, 2004. Thibaut 1890 Thibaut, George, ed and trans. Vedânta-sûtras with the Commentary by Saṅkarâkârya. Oxford: Clarendon Press, 1890. Torella 2008 Torella, Raffaele. “From an Adversary to the Main Ally. The Place of Bhartṛhari in the Kashmirian Śaiva Advaita.” Linguistic Traditions of Kashmir: Essays in Memory of Paṇḍit Dinanath Yaksha. Ed. Mrinal Kaul 326 Elisa Freschi and Ashok Aklujkar. Jammu: D.K. Printworld, 2008. Chap. 19, 508–524. Veṅkaṭanātha 1971 Veṅkaṭanātha. Seśvaramīmāṃsā-Mīmāṃsāpaduke, Ses- wara Mimamsa and Mimamsa paduka. Ed. Uttamur T. Viraraghavacharya and Nainaracarya. Madras: Ubhaya Vedanta Granthamala, 1971. Verpoorten 1987 Verpoorten, Jean-Marie. Mīmāṃsā Literature. Wiesba- den: Harrassowitz, 1987. A History of Indian Litera- ture 6.3.5.  If You Don’t Know the Source, Call it a yāmala: Quotations and Ghost Titles in the Ṛgvedakalpadruma Cezary Galewicz The hitherto probably unpublished, if ever seen in manuscript, Rudrayāmala – a powerful phantom of an authoritative text of amorphous body – appears to have haunted the imagination of many an author in the past and continues to do so in the present.1 The peculiar way in which the name Rudrayāmala happens to surface throughout a good part of historical religious literature classified as Tantra had been noticed in passing by scholars and earned the Rudrayāmala the label of a ghost title. One of the notorious peculiarities of the Rudrayāmala seems to be its ability to attract alien textual matter, or to appear in other texts through extensive quotations from, or chunks of, its elusive textual body, never to be seen in full. Accordingly, the concept of a locus of ascription was introduced (Goudriaan and Gupta 1981: 24) in order to explain the way in which the title Rudrayāmala has been used in other texts, their authors or editors ready to ascribe to it almost any sort of authori- tative formulation they needed to present their texts in robes of prestige lent by an old textual source. Instances of either alleged quotations from or point- ing to the Rudrayāmala as a remote source of a traditional formulation or a valid definition can be seen not only in texts claiming a place within this or that Tantric tradition. A custom of referring to or invoking the name of Ru- drayāmala in circumstances in which a source for an authoritative quotation is required and one’s memory fails to recall one quickly appears still to per- sist among the members of a brāhmaṇa community from the Konkan littoral.2 The concept of an absent source of scriptural authority that must have been at work in such cases deserves closer attention. The present chapter attempts to 1 As far as my knowledge reaches, all printed editions as well as all manuscripts featur- ing the name Rudrayāmala refer to a completely different text, a work that some sources call the Uttararudrayāmala. 2 Claimed to be current among the members of the community of Karhade brāhmaṇas by a descendant of W. L. Paṇśikar, the editor of the SSRS (personal communication 2012). 328 Cezary Galewicz map the network of intertextual relationships among a group of early modern Sanskrit works referring to the Rudrayāmala and to each other from within both manuscript and early print cultures. It focuses on mutual borrowings among the little known works of the Ṛgvedakalpadruma, the Ṛgvedadaśa- grantha and the Yāmalāṣṭakatantra. Moreover, the chapter attempts to arti- culate the concept of an absent source and to reconstruct its functional modalities within the regional and trans-regional spaces of interaction be- tween different ideas of Vedic scripture held by competing learned com- munities of brāhmaṇas in Western India at the turn of the last century. “People are always getting credit for someone else’s words,” concludes Gary S. Morson in the prologue to his recent work, in which he depicts a common practice in contemporary America in which a quotation is identified by “recalling a maxim that if you don’t know the source of a moral exhortation, assume it comes from Shakespeare, Franklin, or the Bible ….”3 It may appear that we are presented with a cultural parallel in the case of the Ṛgveda- kalpadruma (ṚKD), whose author identifies one of the numerous sources of his quotations as Rudrayāmala (RY).4 There would not be anything special in that – after all, isn’t it the case that most formulations aspiring to authority in Sanskrit religious literature tend to use some sort of back-looking quotation or reference to corroborate their standpoints – if not for one thing: the ṚKD quotes from the RY along with the colophons (see below, p. 339) that identi- fy the quoted passages as belonging to an altogether different source. The situation is also different, because the RY, unlike Shakespeare, Franklin or the Bible, seems to represent a radically different concept of (quasi) textu- ality: we are not sure whether it ever existed in any form at all as a text with established integrity. On philological grounds it may seem enough to con- clude by voicing a judgment that would classify the case as “an instance of false attribution.” But the author does not hide another attribution visible in the colophon concluding the quoted passage.5 How should we look at the in- tentional logic of such an act of reference (or attribution, or ascription)? In 3 Morson 2011: 1–2. 4 On the identity of the Rudrayāmala, see Goudriaan and Gupta 1981, and below, p. 345. 5 This long quotation starts from folio 18A, line 3, and runs through folio 24B. The ṚKD ascribes it to the Rudrayāmala without indicating any specific coordinates. However, the quoted passage can be identified as YAT 9–15, including proper colophons after each of the seven quoted paṭalas stating the subject matter and running number of paṭalas within the YAT (no mention of the Rudrayāmala whatsoever). The quotation ends with: …iti śrīmad yāmalāṣṭakatantre upaśāstrāṅgaśāstrasvarūpavarṇanaṃ nāma pañcadaśa paṭalaḥ | iti triskandhargvedakalpadrume rudrayāmaloktargvedādisva- rūpanaṃ [ṚKD, folio 24B, lines 4–5].  If You Don’t Know the Source, Call it a yāmala 329 other words: is it a yāmala, or a generic name of a class, that the author of the ṚKD actually refers to, or one particular such text, namely the Rudrayāmala? 1 The Ṛgvedakalpadruma The ṚKD appears to be a rather late if not recent text that survived at least in two collections, of which only one is known to me.6 It has never been pub- lished. Preserved in its manuscript form at the Chhatrapati Shivaji University in Kolhapur, the ṚKD has three distinct parts (skandha), each fitted with a colophon of its own. The ṚKD represents the little-known world of pre-early modern and early modern secondary treatises that most often appear to have strived to summarize or attempt to look afresh at mostly technical aspects of the use of the Vedic mantras. As Keśava Māṭe, the author of the work, him- self states in the maṅgalācaraṇa, he wrote the ṚKD as a treatise (racana) on the procedure of the study (adhyayana) of the Ṛgveda “so that the Vedic students do not reach anything like the [state of] a barren and motionless pillar bringing the fall into the pit of sin and death.”7 The three parts (skan- dha) of the work are, respectively, a general introduction (upodghāta) into several aspects of the study of the Ṛgvedic mantras, the rules for ceremonial reading of the Ṛgveda (pārāyaṇavidhi), and the procedures for the ritual of reading the mantras with offerings to the fire (saṃhitāhomapaddhati). Several times Keśava claims the novelty of his work, stating, for instance, that what he writes is a new manual (paddhati) compiled after a thorough stu- dy (vicāra) of all pertinent authoritative texts that should be taken into con- sideration.8 And true to his word, the ṚKD is teeming with references and quotations. The works and authors referred to, or quoted, include the Ṛkprātī- śākhyā, the Mādhyandinaprātiśākhya, the Sarvānukramāṇikā, the Chāndo- gyopaniṣad, the Bṛhaddevatā, Yāska’s Nirukta, the Śāṅkhāyanasūtra, the Ṛgvidhāna, the Mahābhāṣya, the Yājñavalkya, the Mānavadharmaśāstra, the Smṛtisārasamuccaya, the Uvaṭabhāṣya and Sāyaṇabhāṣya, the Kūrmapu- rāṇa, the Nṛsiṃhapurāṇa, the Paraśurāmakalpa, Rāmavājapeyin, the Prayo- gaparijāta and many others. Yet the longest quotation in the ṚKD, found 6 For the other, see Aithal 1993: 191–93; cf. also Galewicz forthcoming b. 7 ṛgvedādhyāyanaprakāraracanāṃ kurve satatāṃ tuṣṭidāṃ vedādhyetraghamṛtyugarta- patanasthān vāditān āptaye [ṚKD, folio 1, line 3]. Keśava seems to refer here to the motif of the unstudied Veda being compared to a barren wooden post (see Sāyaṇa’s VBhBhS, p. 44,30–35, and Yāska’s Nirukta 1. 18 on Ṛgvedasaṃhitā 10.71). 8 See, for instance, the colophon to Skandha 3: śāstre śākalaśākhayānigaditaṃ homam vicāryādhunā kurve ’haṃ navapaddhatiṃ pravitatāṃ śrīkeśavākhyodvijaḥ. 330 Cezary Galewicz from folio 18A through folio 24B, comes from the source identified by Keśava as RY. The name Rudrayāmala surfaces in the ṚKD four times, starting already with a short citation on folio 1B. 2 The concept of the daśagrantha It appears that the intention behind the quotation on folio 1B was to establish the concept of the daśagrantha, the decalogy9 or the set of the so-called Ten Books (relative to the Ṛgveda [śākala-śākhā]). These were understood as a canonical and authoritative set of texts necessary for the true experience of the Ṛgvedic tradition, a sort of governing framework seemingly meant to sta- bilize and legitimate the actual practice of traditionally studying and using the Ṛgvdic mantras. The concept of the daśagrantha resembles a ready-made and all-inclusive package of the source texts together with the manuals necessary for their proper reading and decoding. From earlier occurrences of the con- cept of the daśagrantha, we can see that to a certain extent the idea must have remained fluid. Thus, according to one source (Bhāratīya Saṃskṛtikośa) the daśagrantha is said to include: 1. saṃhitā, 2. brāhmaṇa, 3. padakrama, 4. āraṇyaka, 5. śikṣā, 6. chandas, 7. jyotiṣa, 8. nighaṇṭu, 9. nirukta and 10. aṣṭā- dhyāyī. The same source suggests that the daśagrantha is a concept of some antiquity: it ascribes it to Vyāḍi, according to whom the daśagrantha should include, however, a slightly different set: saṃhitā, brāhmaṇa, āraṇyaka, śik- ṣā, kalpa, aṣṭādhyāyī, nighaṇṭu, nirukta, chandas and jyotiṣa.10 Especially noteworthy is the absence of sūtra from both lists. The idea of the daśagrantha seems even today to retain some currency among Maharashtra brāhmaṇas as the publication and circulation of the Ṛg- vedadaśagrantha (ṚDG) suggests.11 The notion of a set of “ten books” (daśa- grantha) has been recognized by the Mahārāṣṭraśabdakoś (and can also be seen on contemporary popular websites). This source, however, does not 9 The term appears in several works taken into consideration in this essay, either in the singular (as the compound daśagrantha) or in the plural (daśagranthāḥ). 10 Bhāratīya Saṃskṛtikośa IV, 1994:37. It is not certain, however, whether the Bhāratīya Saṃskṛtikośa refers to Vyāḍi the grammarian, who predated Pātañjali, or to another, much later, author of the same name, who wrote on the Ṛgvedavikṛtis as late as in the 16th century CE. For differently arranged sets of the ten books, cf. also Müller 1867, or the contemporary website of the Hindujagruti: http://www.hindujagruti.org/hinduism/ knowledge/article/correct-way-of-chanting-vedic-mantras.html#2. See also Galewicz forthcoming a and forthcoming b. 11 For more on the idea of the daśagrantha and on the daśagranthis, see Galewicz 2014.  If You Don’t Know the Source, Call it a yāmala 331 mention a sūtra among the ten members of the set, while the ṚDG takes sūtra to refer to the [Āśvalāyana]gṛhyasūtra.12 The concept of the daśagrantha was also noticed by F. Max Müller, who mentions it in his Lectures on the Origin and History of Religions (quoting a letter from one of his informants from Pune): “a student of a Rigveda-śakha, …. if sharp and assiduous, takes out eight years to learn the Dasagranthas, the ten books, which consist of (1) The Samhitā of the hymns, (2) the Brāhmaṇa, . . . (3) the Āraṇyaka, the forest book, (4) the Grihya-sūtras, (5–10) the six Angas….”13 As is apparent, the informant on whom Max Müller relied must have had access to a source expressing a similar point of view as that of Vinayak H. Ghaisāsa, the editor of the ṚDG. Thus, regardless of the exact identity of the ten constituent members of the canonical set of Ten Books, the concept of the daśagrantha appears to have established itself regionally in Maharashtra as a powerful idea and a point of reference for those seeking legitimacy for their teachings on the use of the Ṛgveda mantras and associated religious and scholarly practices.14 One such practice, considered important by insiders, must have been the so-called vedapārāyaṇa, or the ceremonial reading-cum-study of the Ṛgvedasaṃhitā in its entirety. 2.1 Keśava Māṭe’s interpretation of the daśagrantha The concept of daśagrantha appears very important for the author of the ṚKD, Keśava Māṭe, who takes it as the foundation upon which he constructs his own hierarchy of priorities concerning the practice of reading and using the Ṛgveda mantras. He points to a twofold source for the concept, indicating that it is of considerable antiquity (having been proclaimed by the earlier teachers – pūrvācāryair ukta) but that it can also be seen in a work of Rāma- vājapeyin – apparently a more recent author. The title of that author’s work, however, remains unspecified; the reference is given only generically as a grantha in which the words of the earlier spiritual teachers concerning the 12 Ṛgvedadaśagrantha, folio 2A. 13 Müller 1878: 161. Cf. Galewicz forthcoming b: 27. 14 For contemporary associations of the concept with the competition for control over Vedic education centers; see Galewicz 2014. The idea of the ten “holy” books of the Ṛgveda appeared at times to be new; see Bhagawat 1899: 234: “The modern Brahmin going a step further, or rather descending a step lower, holds even the Shrauta-sūtras with the remaining five Angas … to be coeternal, calling these the Ten Books (Dasha Granthas), and taking special care to commit them to memory (even without under- standing a single syllable) …” Cf. also contemporary Maharashtra-based websites con- cerned with the preservation of Vedic tradition, such as: http://www.hindujagruti. org/hinduism/knowledge/article/correct-way-of-chanting-vedic-mantras.html#2 332 Cezary Galewicz concept of the ten books were to be seen.15 Thus the claim for authority concerning the idea of the daśagrantha is laid out by Keśava in a double manner: referring to a historical work by Rāmavājapeyin while at the same time reaching back for the authority of unspecified ancients (who are seen as voicing their ideas in the book by Rāma). Containing a gloss by Keśava on his own words in an elaborate maṅgala introduced with a homage to Cidam- baraguru and Śaunaka as “truly knowing the essence of many a book” (anekagranthārthatattvavid), the passage in question reads: And these books had been spoken of by ancient teachers in the work of Rāmavājapeyin [as] [the corpus of] mantras and brāhmaṇas, the aṅgas, the ṛṣyādyanukrama and declared [there to form] the ten books that shed light on the meaning of the Veda. (te ca granthāḥ rāmavā- japeyīgranthe pūrvācāryair uktāḥ mantrabrāhmaṇam aṅgāni sūtram ṛṣyādyanukramaḥ daśagranthā ime proktā vedārthapratibodhakā iti. ṚKD 1, folio 1B, line 9.) Keśava then continues with his gloss by identifying the generic terms of the enumeration, according to which we may now reconstruct the list as con- taining the following set of ten “books”:16 1 saṃhitā + 1 brāhmaṇa + 6 aṅgas + 1 sūtra (= prātiśākhya) + 1 ṛṣyādyanukrama (= sarvānukramaṇikā) = 10 granthas 2.2 The sūtra within Keśava’s daśagrantha The identifications decided upon by Keśava (represented by equation marks in parentheses above) may not have been obvious enough at the time, since Keśava apparently needs an authoritative corroboration for them. It is here that one of the quotations from the RY fits in. For reasons he does not dis- close at first, Keśava seems to need to establish the Ṛkprātiśākhya as one of the “Ten Books” or the decalogy of the RV. And establish this he must, since other sources for the concept of daśagrantha seem not to have been as will- 15 Perhaps Rāmavājapeyin, the author of the commentaries on the Śulbasūtras, is meant here. I touch upon the problem of his identification in my forthcoming Galewicz forth- coming b. A hint for interpreting this reference might perhaps be drawn from the evi- dence of the Kavīndrācāryasūcipatram, where the manuscripts of a work called Rāma- vājapeyī are listed as item No. 709 among other texts labeled as works on the know- ledge of kuṇḍas, or fire-pits and rituals pertaining to them. See Kavīndrācāryasūcī- patram, p. 13. 16 asyārthaḥ mantrasaṃhitābrāhmaṇam nāma mantrādiviniyojakaṃ aṅgāni śikṣākalpo vyākaraṇam nirūktam cchando jyotiṣam iti ṣaṭ sūtraṃ prātiśākhyam tatrāpi sūtravya- vahārāt… ṛṣyādy anukramasarvānukramaṇikā [ṚKD I, folio 1B, lines 10–12].  If You Don’t Know the Source, Call it a yāmala 333 ing as he was to include the Ṛkprātiśākhya within the fold of the fixed set of the “Ten Books.” This can be seen in the recently published Ṛgvedadaśa- grantha (Pune 1986), which explicitly uses the concept for its title and apparently acknowledges sūtra to be a member of the set of the ten granthas. However, it takes the word sūtra to refer rather to the [Āśvalāyanīya] Gṛh- yasūtra, which accordingly finds its place among the ten books actually printed in the edition. This understanding of the sūtra-component of the da- śagrantha goes back at least to the 1910 edition of a text entitled Sasvāhāka- raprayoganirṇayā samantrakośā ca ṛksaṃhitā (SSRS), which was prepared for the Nirṇaya Sāgara Press by the indefatigable Vāsudev Lakṣman Paṇśī- kar.17 This compendium does not offer any indication of the source of the idea. But it does include a reference to the ṚKD. What is more, the second edition of the SSRS incorporates within its comprehensive body one chapter of the Yāmalāṣṭakatantra (YAT),18 which appears to be the same work that Keśava quoted and identified as RY. Nevertheless, the editor understands the sūtra element of the set in a noticeably different way than Keśava. Also other earlier evidence for such an understanding makes the choice of Keśava somewhat unorthodox, if not revolutionary – as we have seen in section 2.1, for instance, some sources did not feature sūtra in their lists at all. Why did Keśava need to understand sūtra in the way he did? A clue to the intended primacy of prātiśākhyas in general and the Ṛkprātiśākhya in partic- ular can be seen in Keśava’s way of introducing the idea of the daśagrantha to his readers. He does so while glossing his own words from the opening, in 17 The SSRS contains a collection of technical manuals concerning the ritualized proce- dures of study, preservation, re-memorization and practical use of the Ṛksaṃhitā along with the text of the latter accompanied by selections from Sāyaṇa’s bhāṣya. The actual number of these manuals happens to differ considerably between the two editions of the SSRS. 18 Burnell recognized the YAT as an independent work (Burnell 1880: 205), as does the New Catalogus Catalogorum (vol. 22, p. 37). Direct references to the YAT in Tantric works are for the most part absent; except for the relatively recent ṚDG, which con- tains YAT 13, and the ṚKD incorporating YAT 9–15 (but ascribing it to RY), one ref- erence by the NCC (Vamakeśvaratantra, Catalogue of Sanskrit MSS in the India Of- fice, p. 883) contains the name yāmalāṣṭaka, but it clearly refers to the concept of a group of eight yāmalas, not an independent work by that name. I was unable to locate the other NCC reference, namely that to the Devīpurāṇa (which is a part of the Kāli- kāpurāṇa). However, a copy of the YAT is listed as belonging to Kavīndrācārya’s famous collection. See Kavīndrācāryasūcīpatram (entry 1767 on the list) and Gode 1945. Indirect references or quotations from the YAT can, however, be found in Bhās- kararāya’s commentary on the NSA and in the Ṛgvedakalpadruma (both texts, how- ever, refer to the Rudrayāmala instead of YAT; see Galewicz 2011: 128). For more on the YAT, see Galewicz 2011. 334 Cezary Galewicz which he pays homage to two specific personalities: Cidambaraguru (in plural) and Śaunaka. As he explains to his readers, he respects the two be- cause they are true wise men, versed in Prātiśākhya(s) and other granthas needed for understanding the artha of the Veda: satāṃ prātiśākh[y]ādyane- kagranthārthatattvavidāṃ (ṚKD I, folio 1B).19 Thus, from the very begin- ning, Keśava takes the concept of the daśagrantha to include a prātiśākhya and, in what follows, with the help of internal evidence or external authority he does everything he can to prove that the concept of the Ten Books of the Ṛgveda [Śākalaśākhā] indeed includes the prātiśākhya. The passage he de- cides to use for this purpose does not specifically name any of the known prātiśākhyas. It does, however, vaguely indicate that the tenth element is a sūtra. Most of the text of ṚKD I is devoted to proving that it is not only useless, but it is even detrimental to use the Ṛgvedic mantras in an inexperienced, unprofessional way, that is, without thorough technical knowledge. This knowledge is preserved in technical manuals that safeguard the correct pro- nunciation as well as internal correspondence between the three indispensable elements of āṛṣa, chandas and devatā. The same implies that reciting the Ṛgvedic mantras without a thorough knowledge of their three defining ele- ments may prove futile, if not even harmful, to their users. These three ele- ments, the name of the inspired author (ṛṣi), the meter (chandas) and the presiding deity (devatā), are understood as concerning each and every mantra of the saṃhitā. Several folios of the ṚKD meticulously explain why and how the three (and especially the first two) are indispensable for securing the meaning-and-purpose (artha) and the efficacy of the mantra. This whole discussion ends on folio 17A with the Ṛgvedaprātiśākhya of Śaunaka being established as the exclusive primary source for the knowledge of these three elements. It now becomes clear why Keśava was insistent on the prātiśākhya being considered one of the “Ten Books” necessary to under- stand and use the Ṛgvedic mantras. What is more, it stands to reason that Keśava must have thought of, or intended to, present his position and his work, the ṚKD, as being somehow derived from the Ṛkprātiśākhya. This can be concluded from the wording of the colophon to the first of the three books (skandha) that comprise the ṚKD. For fear of overextending the book, the respected Ṛgvedakalpadruma ceases here, produced as it is at Keśava’s tree watering pool with three 19 kiṃ kṛtvā vedagaṇaṃ vedasamūhaṃ cidambarākhyagurūn śrīśaunakaṃ ca natvā kīdṛ- śīm ityākāṅkṣāyām āha satāṃ prātiśākhyādyanekagranthārthatattvavidāṃ (RKD I, fo- lio 1B, lines 8–9).  If You Don’t Know the Source, Call it a yāmala 335 trunks containing the introduction, [the method of] invocation, and the knowledge of reciting [the Ṛgvedic mantras]. … Oh Gods on Earth (= brāhmaṇas?), you should resort to this (tree = ṚKD), which is easily accessible, for the sake of obtaining the fruit born from pronouncing the [correct] svaras. It has leaves consisting in its own wise formula- tions and roots watered by the streams of sayings of the sages included in the Prātiśākhya and its like.20 (ṚKD I, folio 40A, lines 3–5: granthavistṛtibhayād uparamyate śrimān ṛgvedakalpadruma iha, jani- taḥ keśavasyālavāle triskandhaiḥ tair upodghātahavanapaṭhanajñā- nagarbhair upetaḥ. … bhūdevāḥ prātiśākhyādyṛṣivacanajalaiḥ sikta- mūlasvakḷptayuktinyāyachadas taṃ svarajananaphalāptyai sulābhaṃ śrayadhvam).[21] What then, we may ask, makes Keśava refer so frequently to the elusive RY? Why does he quote from this particular source so extensively? He quotes it first in connection with the very concept of the daśagrantha. Does the RY indeed corroborate such an understanding? If so, does this mean that the RY or the passage quoted as allegedly stemming from it could be seen as somehow connected with the idea of the Ten Books, the decalogy of the Ṛgveda? On closer look, just a few lines earlier in the same passage, which can be identified as Paṭala 10 of the YAT (here attributed by Keśava to RY Paṭala 10), it reads: In this regard, it is said that there are ten divisions (bheda) in the Ṛg- veda, and the number of ṛk-stanzas [to be recited] is proclaimed as amounting to ten thousand and to ten hundred and to eighty and to a 20 I thank Philipp Maas for suggestions concerning the understanding of this passage. 21 [A note by the editors: The editors suggested in an email to the author (September 2015) to emend the text -jalaiḥ siktamūlasvakḷptayuktinyāyachadas taṃ to -jalaiḥ sik- tamūlaṃ svakḷptayuktinyāyachadanaṃ. These emendations solve the following two syntactical problems that make the text as it is edited now ungrammatical: (1.) In the emended text, the word -jalaiḥ is the agent of the adjective sikta- in the bahuvṛīhi- compound siktamūlaṃ “having roots.” In this way, the syntactical rule that “any word standing outside a compound may form a grammatical relationship only with a com- pound as a whole” (Coulson 1976: 91) is not violated. (2.) -chadanaṃ, as the final part of a further bahuvṛīhi, is the object of the imperative pl. śrayadhvam, of which the vocative bhūdevāḥ is the grammatical subject. The grammatically correctly emended text translates as “Oh Gods on Earth, you should resort to the Wish-Fulfilling-Tree-of- the-Ṛgveda (i.e., the ṚKD), the roots of which are watered by the streams of sayings of the sages included in the Prātiśākhya and so on, the leaves of which consist of the wise formulations that I conceived myself, and which is easily accessible for the sake of ob- taining the fruit resulting from pronouncing svaras (correctly).”] 336 Cezary Galewicz quarter. (tatra bhedās tu ṛgvede daśa caiva prakīrtitāḥ | ṛcāṃ daśa- sahasrāṇi ṛcāṃ daśaśatāni ca | ṛcām aśītiḥ pādaś ca ṛksaṃkhyā pari- kīrtitā |) Thus, it stands to reason that even if the ten bhedas of RY 10 (=YAT 10) refer to the same idea as the ten granthas do, the actual components of the set differ from one source to the other. What is more, the extant YAT (with its many corrupted fragments in the available transcripts) gives a rather vague, if at all recognizable, idea of the ten elements:22 the ṛgveda (sapramāṇa)23 + āraṇ[y]as + ṛgvedabrāhmaṇa + 6 aṅgas + 6 upāṅgas + prātiśākhya + saṃhitāsūtra. We can see, however, that the elusive RY, exactly as Keśava wants it to be seen or glimpsed through his own text, does recognize the prātiśākhya, too, as somehow belonging to the concept of the tenfold Ṛgveda. 3 The Rudrayāmala as quoted in the Ṛgvedakalpadruma The ṚKD is replete with quotations and references. While the Ṛkprātiśākhya remains by far the source most often resorted to, the longest single quotation offered by Keśava is one ascribed to the RY. It follows an exposition on the details and reasons for the divisions into the branches, or vedaśākhās, in general and the śākhās of the RS in particular (iti ṛgvedaśākhābhedāḥ). The passage happens to be directly introduced with a formulation containing the 22 YAT 10, 42–43 reads: tatra bhedās tu ṛgvede daśa caiva prakīrtitāḥ | ṛcāṃ daśasahas- rāṇi ṛcāṃ daśaśatāni ca | ṛcām aśītir vā daśa ca ṛksaṃkhyā parikīrtitā | ṛgvedaḥ sa- pramāṇo ’yaṃ sahasrāṇi caturdaśa || āraṇāni dvisāhasragranthamānāni pārvati | ca- tuḥsahasrasaṃkhyānaṃ ṛgbrahmāṇam udāhṛtam | śīkṣā kalpovyākaraṇaṃ niruktaṃ jyotiṣaṃ tathā || chandaś ceti ṣaḍaṅgāni trisahasrāṇi kṛtsnaśaḥ | padaṃ pratipadaṃ chandaḥ jñānaṃ svarasamudbhavaḥ || nyāyatarkas tu mīmāṃsety upāṃgāny uditāni ṣaṭ | eṣāṃ sārdhaikasāhasraṃ granthamānaṃ pracakṣate || tasyaivaprātiśākhyam* ca saṃhitāsūtram eva ca | ubhayor ardhasāhasraṃ granthamānaṃ prakīrtitam || evaṃ vyāsena ṛgvedaḥ saṃvibhakto mahātmanā | 23 The Ṛgveda, qualified here with the adjective sapramāṇa, is said to count as many as fourteen thousand granthas. If the word grantha stands for a cluster of thirty-two akṣaras, then here the total number of akṣaras would amount to 448,000. This being a number considerably larger than the number commonly given, perhaps it refers to a collective body of texts, including the saṃhitā. And perhaps other elements in this body of texts make up for the missing three elements of the daśagrantha in this enume- ration.  If You Don’t Know the Source, Call it a yāmala 337 suggestion that the citation to follow is an abridgment (saṃgṛhyate) of the original wording of the RY: And this is incidentally how the essence of the Ṛgveda and other Ve- das proclaimed in extenso in the RY is now hereby abridged. (atraiva prasaṅgād ṛgvedādinām svarūpaṃ vistareṇa rudrayāmaloktaṃ saṃ- gṛhyate |)24 Is Keśava here again deploying a strategy for a double authority? We have already seen this when he embedded the sayings of the ancient teachers within the historical work by Rāmavājapeyin. Does Keśava wish to imply that a recent work summarizes the ancient knowledge of the ghostly RY?25 This time Keśava does not expressly bring the title of YAT to the mind of his readers. Nonetheless, rather surprisingly in these circumstances, the cited text proves to incorporate its own [original?] coordinates in the form of colophons after the succeeding chapters (paṭalas), with the last one reading: Thus ends the fifteenth chapter in the honorable Yāmalāṣṭakatantra entitled “A Description of the Nature of Minor and Additional Śāstras.” Thus ends [the chapter devoted to] the Description of the Nature of the Ṛgveda and the like voiced by the Rudrayāmala in the Wondrous Tree of the Ṛgveda (ṚKD) consisting of three chapters (iti śrimadyāmalāṣṭakatantre upaśāstrāṅgaśāstrasvarūpavarṇanaṃ nāma pañcadaśaḥ paṭalaḥ | iti triskandhaṛgvedakalpadrume rudrayāma- loktaṛgvedādisvarūpaṃ |).26 Thus, what we get in final analysis here amounts almost to a formula of triple framing: the frame of the YAT is enclosed within that of the RY and the RY, within that of the ṚKD. But Keśava does not tell his readers whether he considers the YAT part of the RY. This relatively innocent “proclamation of incorporation” is followed by a passage devoted to the khilas, praiṣas and nivid formulas, with reference to the discussion on the division of Vedic “schools” (śākhābheda). With this, the establishment of the Ṛkprātiśākhya as 24 ṚKD I, folio 18A, line 3. 25 We cannot but continue to ask questions here: Was Keśava actually familiar with the world of later Tantra literature with its specific concepts of textuality and titles indi- cating the ingestion of works within earlier works in a succession leading to a distant source of unquestioned authority? If he knew the YAT, as he must have because he quoted from it, how did he conceive of its textual status? Did he actually take it to be an abridgment of the distant RY? I leave these questions open due to current lack of sufficient evidence. 26 ṚKD I, folio 24B. 338 Cezary Galewicz one of the Ten Books necessary for the true and comprehensive experience of the Ṛgveda seem to have been secured and Keśava goes on to indulge in technical pyrotechnics: he analyzes a great number of doubtful readings and resolves intricacies of mantra identification after vague hints that this needs the very Ṛkprātiśākhya to be authoritatively decided. His expertise, no doubt genuine, is apparently recognized in later works, as for example the already mentioned ṚDG, which quotes several times from Keśava’s ṚKD. In his dexterous use and handling of arguments and counterarguments, quotations from and references to the RY seem to play an important, if not the key role as prestigious sources for his own claims to authoritative judgments. And these sources focus on the tradition running from the early Vedic exegetical works like the Bṛhaddevatā and the Nirukta and the linguistic analysis of the prātiśākhya type. It might seem superfluous to reach for the authority of the enigmatic RY since those of the prātiśākhya were clearly at hand for Keśava. Yet perhaps the idea of the daśagrantha must have been so influential that he felt it necessary to establish the Ṛkprātiśākhya as one of the Ten Books rec- ognized as being part of the true tradition of Ṛgvedic study. The RY seems to have provided an authority to ground the inclusion of the Ṛkprātiśākhya in the daśagrantha. How can we understand the logic of such a strategy? What made the idea of the RY as a remote source of authority so appealing? 4 The Rudrayāmala and the yāmalas The vast religious literature of India generally referred to as Tantra features a number of interesting concepts of textuality and ideas of textual integrity that should perhaps be given greater attention on theoretical grounds. A good number of works included within the broad category of Tantra tend to display in their titles a generic name indicating an affiliation with one or the other particular religious tradition. Thus, contemporary scholarship often takes it for granted that titles featuring the name saṃhitā should be considered as belonging to Vaiṣṇava (Pañcārātra) Tantra, while those exhibiting the names āgama or tantra should be taken as part of the Śaiva or Śākta divisions, re- spectively. While it is rather difficult, if not impossible, to categorically deli- neate any sort of objectivity, the classification itself as well as the suggested implications in choosing to use one of the three was certainly not without purpose. And the appearance of detailed and conceptually complex classifi- cations of scripturally acknowledged texts in relatively early works of these religious traditions must have been recognized as points of reference for later works to come. There must have been indigenous ideas of textual division  If You Don’t Know the Source, Call it a yāmala 339 behind such classifications; these ideas escape us today, as do the concepts of textual boundaries that made these ideas possible. But these concepts must have played an important role in how religious traditions built their identities. If we reverse the perspective and take these categories not as ready-made and unproblematic building blocks, but rather as the result of choices made by the users of these texts, then a theoretically interesting point of departure emerges. By using a particular generic name, and thus alluding to the con- vention it presupposes, a title may be taken as a claim to rightfully deserving such a name or as an act of aspiration on the part of authors or users of par- ticular texts. Taking inspiration from the theoretical propositions of Gerard Genette (Genette 1987), we might say that titles as well as opening and clos- ing formulas may be taken as important forms of mediation between authors or editors (in our case: copyists and redactors) and text users, in the way that gates, porticos and thresholds have long been understood in architectural forms. From this point of view, a title containing the word saṃhitā, āgama or tantra makes from the very beginning a certain claim on how the text it is introducing is presented to its readers or users. This must perhaps be presup- posed also with regard to other generic names within the three broad currents constituting the starting point of these classifications. Of direct interest for me here is the generic name yāmala, and that of the Rudrayāmala in particu- lar. Simplifying an otherwise complex matter, we might say that yāmalas, in the plural, appear in relatively early classifications as a group of scriptural texts holding a place within the bigger agglomeration of the so-called Bhaira- vas, which are usually held to be sixty-four in number.27 The classifications that can be traced within the historic Śaiva literature acknowledge an impor- tant place for this group of works and suggest that they be taken as texts of prestige. They are understood there as a set of texts usually numbering eight and forming together one of the eight sub-groups of the superstructure of sixty-four Bhairavas.28 Though the actual names on the list differ to some extent from one source to another, the name Rudrayāmala usually appears. This suggests the existence of a stable tradition of acknowledging the impor- tance of this particular yāmala. 27 See, among others, Bhattacharya 1982, and Goudriaan and Gupta 1981. 28 Some of the sources, however, hold yāmalas to be six in number. See, for instance, the opinion ascribed to Hemacandra in the Ṡabdakalpadruma Vol. IV, p. 41. An additional clue regarding the identity of the RY may perhaps be the Kavīndrācāryasūcīpatram (p. 20), which features a manuscript titled Rudrayāmalatantram listed together with four other manuscripts containing the name yāmala in their titles (Viṣṇuyāmalatantra, Brahmayāmalatantra, Śivayāmala and Devīyāmala). 340 Cezary Galewicz For this short chapter, however, the actual classification is not of primary interest, but rather the fact that the prestigious title Rudrayāmala seems to have spilled over the limits of Tantra into other domains of traditional know- ledge. It can be found among the late early-modern works of a regional va- riety of the Maharashtra Brāhmanic tradition. A possible reason could be that a particular person or community may have engaged in both Vaidika (by birth) and Tāntrika (by choice) traditions of this or that form, thereby hybri- dizing the two traditions. 5 Textual identity reconsidered In premodern India, with its changing and hybrid socio-political and religious order, hybridization not only tolerated but also inspired new forms of know- ledge production, knowledge that favored mixed realms of authority and developed new hierarchies and classifications. The entities suggested by the classification schemes like that of the Vaidika Tantras feature not only in the extant manuscripts of the YAT, where they might be suspected of being more a theoretical concept than reality. But they are also confirmed by evidence such as a list of the actual manuscripts in the collection of the 17th-century polymath Kavīndrācārya Sarasvatī.29 This introduces new aspects of intertex- tuality, with sometimes blurred or undecided boundaries and various types of ingestion, incorporation and inclusion. Here, claims for textual authority find new forms and logic that move beyond the simple categories of originality, plagiarism, borrowing or interpolation. Help in understanding this complexity might be found in studies con- cerned with similar historical moments in other geographical and cultural milieus, such as the otherwise controversial hypothesis of A. Johns (Johns 1998), according to which “neither the fixity, nor originality nor piracy could be taken as intrinsic to the body of a text but rather as attributed qualities at- tached to texts by their users.” Comparable to this might be the colophons in the case of the ṚKD and its quotations, which may be taken not so much as factographic statements but rather as acts of aspiration or claims. This phe- nomenon happens to have been given a most interesting, though short, treat- ment in Goudriaan and Gupta 1981. In discussing colophons, these authors speak of three types of loci of ascription, namely the famous old tantras, le- 29 See Kavīndrācāryasūcīpatram, pp. 25–26. For an attempt to understand the rationale behind this list, see Galewicz 2011: 125–27.  If You Don’t Know the Source, Call it a yāmala 341 gendary texts, and cover titles of texts or traditions that may have once ex- isted but have now disappeared.30 The case of the ṚKD and its intertextual links to the RY and the YAT seems, however, to be different from the case mentioned above (also dis- cussed in Goudriaan and Gupta 1981) and involving various loci of ascrip- tion, although it is not without a parallel: it would be difficult to imagine that it was intended to inscribe the ṚKD, a rather recent work, within a particular current of Tantric traditions. At least the ṚKD does not make any statement of that kind in its colophons. Its title deserves attention, no doubt, but in a different way than those of earlier religious works claiming a place under the umbrella of a prestigious text of the past. Nonetheless, here is probably something similar at work: making use of the associations expected to be raised by the very sound of the name Rudrayāmala. What kind of associa- tions could these be? The type of textuality represented by the RY does not seem to have received enough theoretical analysis. First of all, within the culturally determined framework of an altogether different concept of textual identity, integrity and authorship, we have no choice but to accept that texts act by themselves, just as living beings do.31 6 What does the name Rudrayāmala stand for? In their History of Tantric Literature, Goudriaan and Gupta try to articulate the specific ontological status of a text suggested by the name Rudrayāmala: The Rudrayāmala is perhaps the most mysterious of all Yāmalas. It is encountered everywhere, yet always vanishes after closer inspection. It is even uncertain if an original Rudrayāmala ever existed, despite the fact that the title figures in all old lists of Yāmalas.32 The name Rudrayāmala appears to have haunted texts, books and imagi- nations for several centuries now (those of modern scholars included), and it is perhaps one of the best examples of a class qualified by the above-men- tioned authors as “ghost titles.” This formulation, otherwise very apt and well-chosen in my opinion, suggests however that ghost titles remained empty forms, shallow containers ready to accept anything that was poured in. 30 See Goudriaan and Gupta 1981: 24. 31 This presupposition, problematic as it is, is applicable for series of works for which almost no author is named, as is the case for Tantra in the premodern era. 32 Goudriaan and Gupta 1981: 47. 342 Cezary Galewicz But can only this be said about them? Do their titles or appellations remain empty labels? Or are we rather dealing here with the idea of an amorphous entity that claims to represent a textual tradition whose boundaries, but also form and content, remain fluid? Fluid to the extent that we can understand them as representing a concept quite different from that of a definite text or book? Whether this concept resembles that of a receptacle ready to contain any and all contents is uncertain. As already mentioned above, Goudriaan and Gupta consider titles like Rudrayāmala to have functioned as what they call loci of ascription, used by later or minor textual traditions seeking their place among the acknowledged families of Tantric texts with more stabilized pedi- grees. This seems a convincing proposal for conceptualizing a genre of “ghost titles” within the broad context of Tantric literature. But it appears that the strategy of using such name titles, as well as the concept thereof, tran- scends the boundaries of the textual traditions that can be qualified as Tantra. As the evidence of the ṚKD, ṚDG and related texts shows, we find the title Rudrayāmala also in late commentarial and technical texts composed at the onset of the modern era, that is, in works linked to the regionally understood tradition of the Veda and its practical use. 7 Tantricized Veda or Vedicized Tantra? The fact that the ṚKD – a work from the modern and quite localized Vedic tradition of Maharashtra – uses a vehicle like that of the RY to authoritatively support certain ideas it fails to support through clearly Vedic sources points not only to how brahmanized a local Tantra tradition could be or how tantri- cized the local Vedic tradition appears. It also shows how profoundly and mutually dependent textual imagination, derived both from the Veda and the Tantra, became in the premodern era, especially in the hands of those who disclosed or claimed affiliation to both. For a contemporary reader this means that the two streams usually thought of as being separate can be seen as mu- tually accentuating or complementing each other. Focusing on textuality is perhaps not enough to describe such cases. Probably a community of users, a textual community, should be taken as a point of reference when investigat- ing the mutual relations between the two allegedly separate traditions. Thus, one might reconstruct a Vedic‒Tantric continuum of practice and belief to supply a background against which puzzling phenomena belonging to one or the other stream or situated at the crossroads of the two might be judged. The custom of citing the name Rudrayāmala in circumstances when one’s memory fails to recall the source of an authoritative statement has continued  If You Don’t Know the Source, Call it a yāmala 343 to persist among members of a brāhmaṇa community in the Konkan area. The concept of an absent source of scriptural authority that must be at work in such cases needs further clarification. As a humble contribution to this project, I avail myself of this opportunity to indicate a few sources that are apparently connected to a localized Maharashtra Ṛgvedic tradition, especially those for which the concept of the daśagrantha remains a basis. This concept finds its material reference in the ṚKD, whose author seeks its origins in the work (grantha) of Rāmavājapeyin. What is more important for this study, however, is the logic behind the practice of not only referring to the Rudrayāmala, but quoting it. While there is practically no way to prove that the passages quoted in the ṚKD do not come from a manuscript bearing the name RY, they can actually be traced to another ghost-like title, namely that of the YAT, which though absent from early traditional classifications does survive in several manuscripts and transcripts, attested at least from the time of the library of Kavīndrācārya (late 17th century) as having existed in a more or less fixed form. 8 Quotations and loci of ascription If we inquire into the conceptualization of the re-use of texts and ideas in the context of the ṚKD and related texts, perhaps the term “recycling” will help clarify the actual practice reflected here. The recycled textual matter attri- buted by the ṚKD to the elusive RY and actually to be found in the YAT proves to represent a remote and external view of the entire development of traditionally recognized knowledge systems. These had been envisioned in a neatly organized architectural pattern with an internal hierarchy that reflected. In our case, the organizing pattern that can be seen is the purāṇic model of emanation from a primeval being to the specific knowledge systems of avai- dika and vaidika tantras, among which the YAT also locates itself. As a well- structured network of relations, this model must have offered a convincing point of reference for orienting oneself at the time of the ṚKD’s composition. The purāṇic model of emanation was indeed one of the many sources either directly quoted or referred to by the ṚKD with the purpose of supporting its own authority on matters regarding the ritual performance of the Ṛgvedic mantras. As such, the ṚKD proved successful in establishing itself as an authoritative text in the field. Among later works, for example, the ṚDG recognizes the ṚKD as one of its authoritative sources. Is it at all possible to conceptualize this practice of relegating or re-labe- ling a recycled portion of a text to a particular locus of ascription? What func- 344 Cezary Galewicz tions could such a practice fulfill? Perhaps what we have here is just another manifestation of the old practice of looking for the lost meaning of the Veda, the vedārtha, once the object of heated argument between commentators, from the predecessors of Yāska and evidenced in Yāska’s Nirukta and con- tinuing at least to the times of Sāyaṇa in 14th-century Vijayanagara. The new manifestation was perhaps influenced by the tradition of including Ṛgvedic mantras in newly contextualized magical practices, as exemplified here by the Ṛgvidhāna (a source strongly present and much quoted in the ṚKD) on one hand, and the Tantra on the other. It fed on the presupposition that man- tras must yield their supposed powers if properly handled. It involves a quest for the key to tapping the inner efficacy of a foundational text considered to be powerful beyond doubt if only we have access to it, if only we know how to open its gates so that it yields the precious insights it hides. An important indication in support of such an understanding is the traditional practice of constructing “know-how” secondary texts, to which the ṚKD itself belongs. This type of practical manuals sought authority by quoting either well-known or mysterious reference sources or loci of ascription. Ghost titles seem to represent, so to speak, a reversed order of textuality: they are not a label put on a mass of words, but a ready-made receptacle of predefined authority awaiting texts that seek its authority to pour into.33 The old problem of where, how and through whom to find the meaning of the Veda gave rise historically to several concepts of how to rescue its suppo- sedly lost meaning and aim. The idea of a complete and closed set of author- itative texts sufficient for thoroughly studying and understanding the Veda can be traced back at least to Patañjali’s Mahābhāṣya: “A Brahmin should learn and understand the Veda with its six limbs as his disinterested duty” (brāhmaṇena niṣkāraṇo dharmaḥ ṣaḍaṅgo vedo ’dhyeyo jñeyaḥ iti MBh I.1.18 f. [in Kielhorn 1962]). The notion of a set of knowledge disciplines needed for the thorough comprehension of the Veda appears to culminate in the idea of the so-called fourteen “fortresses of knowledge” (vidyāsthānas), expanded later to eighteen elements. Its formulation can be seen in Yājñ. Smṛti I.3, Viṣṇupurāṇa III.6.27 and a number of other later works, including Sāyaṇa’s introduction to his commentary on the Ṛgvedasaṃhitā (VBhBhS, p. 44,15–17). While the point of reference for the last mentioned work remains the Mīmāṃsā concern with locating the meaning of a real or hypothetical śrauta-ritual, that of the ṚDG appears to be informed rather by the techni- calities of domestic ritual applications in which knowledge of the Veda 33 The concept of location in electronic readers as opposed to fixed page numbers and contents of a codex may perhaps serve as a parallel.  If You Don’t Know the Source, Call it a yāmala 345 serves as practical know-how for skillful individual ritualists within the domestic or temple environment.34 When trying to account for the logic behind using ghost titles like Rudra- yāmala, we must perhaps assume that this name rang overtones of consider- able prestige in the imaginations of those who came into contact with the ṚKD. Otherwise the link would not have worked at all and the strategy would have been futile. What sort of overtones might these have been? Could it be that the very sound of the name of the RY was believed to stimulate as- sociations with a comprehensive knowledge system represented by the con- cept of the complete set of eight yāmalas? Neither the ṚKD nor the ṚDG confirm this directly. The outsourcing of authority and prestige might be used as a working concept to describe the practice of quoting ghost works. In this context it is perhaps of interest to look inside the YAT for its own under- standing of its links to the RY. A connection is indeed vaguely suggested in YAT 22, which places itself within the group of yāmalas by insinuating that it has the status of an essence-bringing restatement or an essential abridge- ment. 9 Spatial topography of ideas The many questions arising from reading the ṚKD must await their answers for a time when we know more about the topographic distribution, location and trajectories of intellectual elites, ideas and texts in the premodern space represented by the ecumene of the Maratha confederacy. It no doubt consti- tutes a meaningful historical background for reconstructing the meaning of works such as the ṚKD. A methodological examination of the spatial defini- tions and geographies of the ṚKD would allow at least some of the names appearing in its colophons (Śāhunṛpa, Cidambaraguru, Saptarṣidurga) to be traced through the spatial topography of places included in the body of its text: these would include places known today as Satara, Ghulhasur, Sawanur and, broadly, Maharashtra, northwest Karnataka and Deccan. Also the space formed by the symbolic social transactions in the unique pre-colonial Mara- tha world is still to be studied. The pre-modern and early modern scholarly communities stemming from several distinct groups of Maharashtra brāhma- ṇas deserve more attention regarding their specific local, regional and trans- regional activities. In these competing learned communities, the concept of the daśagrantha and the RY seem to have been used as elements in their 34 I focus more on these developments in Galewicz 2014. 346 Cezary Galewicz strategies to successfully lay claim to the authoritative use of the Ṛgvedic mantras in the changing and modernizing social configurations of Maharash- tra and India.35 Instead of presenting a conclusion to this chapter, let me offer one more humble thought concerning a general problem of reading early modern works like Keśava’s Ṛgvedakaladruma. What I aimed at in this short essay was by no means a psychology of reuse. Nor was it a reconstruction of any sort of personal logic of a particular author’s choices or decisions. My aim was rather an archeological exploration of work conventions and an investigation of how authors played with these conventions. Uncovering such conventions might help to read and make sense of the historical dimension of texts from particular periods or schools of thought. These conventions and textual strate- gies were either taken for granted by authors and their readers, or they were deployed by the former to communicate something to the latter. Without the tools for decoding these conventions, any attempt to make sense of these texts will remain a misplaced endeavor. References Abbreviations BhSK Bhāratīya Saṃskṛtikośa see Jośi 1994 MBh Mahābhāṣya see Kielhorn 1962 NCC New Catalogus Catalogorum NSA Nityaṣoḍaśikārṇava ṚDG Ṛgvedadaśagrantha ṚKD Ṛgvedakalpadruma RY Rudrayāmala SSRS Sasvāhākaraprayoganirṇayā samantrakośā ca ṛksaṃhitā VBhBhS Vedabhāṣyabhūmikāsaṃgraha YAT Yāmalāṣṭakatantra Primary Sources Bhāratīya Saṃskṛtikośa see Jośi 1941 Burnell 1880 Burnell, Arthur C. A Classified Index to the Sanskrit Ma- nuscripts in the Palace at Tanjore: Prepared for the Madras Government. London: Trübner, 1880. 35 For the rival Brahmin community of pre-modern South India, see for instance Wash- brook 2010 and O’Hanlon 2010.  If You Don’t Know the Source, Call it a yāmala 347 Jośi 1994 Jośi, Mahādevaśāstri, ed. Bhāratīya Saṃskṛtikośa. Part 4. Tantraśāstra te nāṇī. Pune: Bhāratīya Saṃskṛtikośamaṇ- ḍala, 1994. Kavīndrācāryasūcīpatram = see Sastry 1921. Kielhorn 1962 Kielhorn, Franz and Kashinath Vasudev Abhyankar, eds. The Vyākaraṇa Mahābhāṣya of Patañjali. Poona: Bhan- darkar Oriental Research Insitute, 1962. Mahārāṣṭraśabdakoś Mahārāṣṭraśabdakoś, vibhagacavathā (4). Ed. Y. R. Date, C. G. Karve, Aba Cāndorkar and C. S. Dātār. New Delhi: Asian Educational Service, 1998 (originally published 1932‒–1938). New Catalogus Catalogorum = New Catalogus Catalogorum. Alphabetical Register of Sanskrit and Allied Works and Authors. Vol. 22. Ed. S. Dash. Chennai: University of Madras, 2011. Nirukta The Nighaṇṭu and the Nirukta. Ed. and trans. L. Sarup. Delhi: Motilal Banarsidass, 1984 (originally published 1920–1927). Nityaṣoḍaśikārṇava Vāmakeśvaratantrāntargatanityaṣoḍaśikārṇavaḥ śrībhās- kararāyonnītasetubandhākhyāvyākhyānasahitaḥ. Ed. H. N. Āpte. Puṇyapattana: Ānandāśrama 1908. Ṛgvedadaśagrantha Ṛgvedadaśagrantha. Pune: Ghaisāsa Vedapāṭhaśālā, Va- santa Press, 1986. Ṛgvedakalpadruma Ṛgvedakalpadruma, MS N191, Cat. No 45, in the Cata- logue of MSS in the Manuscript Library of Chatrapati Shahuji Maharaj University, Kolhapur. Śabdakalpadruma Śabdakalpadrumaḥ. Caturtho bhāgaḥ. Ed. Rājā Rādhā- kānthadeva. Vārāṇasī: Vidyā Vilās Press, 1967. Sasvāhākaraprayoganirṇayā samantrakośā ca ṛksaṃhitā. Ed. Wāsudev Laxman Śās- trī Paṇśīkar. Mumbai: Nirṇaya Sāgara Press, 1930. Śrīkaṇṭhī (Śrīkaṇṭhīyasaṃhitā) = “The Śrīkaṇṭhī.” In: Hanneder 1998: Appendix 1. 237–268. Tantraratna Pārthasārathimiśraviracitam […] Tantraratnam. Ed. G. Jhā. Vārāṇasī: Vidyā Vilās Press, 1930. The Princess of Wales Saraswatī Bhavana texts 31. Tantrasāra Tantrasāra of Abhināvagupta. Ed. M. R. Shāstrī. Bombay: Research Dep., Jammu & Kashmir State, 1918. Vedabhāṣyabhūmikāsaṃgraha = Vedabhāṣyabhūmikāsaṃgraha. Ed. Pt. Bāladeva Upādhyāya. Varanasi: Chowkhamba, 1958 (originally published 1934). Viṣṇupurāṇa The Critical Edition of the Viṣṇupurāṇam. Ed. M. M. Pathak. Vadodara: Oriental Institute 1997–1999. Yājñavalkyasmṛti Yājñavalkyasmṛti of Yogīśvara Yājñavalkya: With the Commentary Mitākṣarā of Vijñāneśvara, Notes, Variant 348 Cezary Galewicz Readings etc. Ed. Narayan Ram Acharya, Bombay: Nirṇa- ya Sāgara Press, 1949. Yāmalāṣṭakatantra Yāmalāṣṭakatantra, MS No B6594 (Devanāgarī transcript from an unidentified source) in the Adyar Library, Madras. Other: http://www.hindujagruti.org/hinduism/knowledge/article/correct-way-of-chanting- vedic-mantras.html#2 Secondary Sources Aithal 1993 Aithal, Parameswara, K. Veda-Lakṣaṇa. Vedic Ancillary Literature. Delhi: Matilal Banarsidass, 1993 (originally published in 1991). Bhagavat 1899 Bhagavat, Ramakrishna Rajaram. “Three Interesting Vedic Hymns.” Journal of the Bombay Branch of the Royal Asiatic Society, 20 (1899): 234–56. Bhattacharya 1982 Bhattacharya, Narendra Nath. History of the Tantric Reli- gion. Delhi: Manohar, 2005 (originally published in 1982). Bretelle-Establet 2010 Bretelle-Establet, Florence, ed. Looking at it from Asia. The Processes that Shaped the Sources of History of Science. Dordrecht: Springer, 2010. Chitnis 2000 Chitnis, Krishnaji Nageśvarao. The Nawabs of Savanur. New Delhi: Atlantic Press, 2000. Coulson 1976 Coulson, Michael. Teach Yourself Sanskrit. London: Hod- der, 2003 (originally published in 1976). Deshpande 1997 Deshpande, Madhav. “Building Blocks or Useful Fictions: Changing View of Morphology in Ancient Indian thought.” India and Beyond: Aspects of Literature, Mean- ing, Ritual and Thought. Essays in Honour of Frits Staal. Ed. Dick van der Meij. Leiden: International Institute for Asian Studies, 1997. 71–127. Dyczkowski 1988 Dyczkowski, Mark. The Canon of the Śaivāgama and the Kubjikā Tantras of the Western Kaula Tradition. Albany: SUNY, 1988. Eaton 2005 Eaton, Richard. A Social History of the Deccan, 1300– 1761. Eight Indian Lives. Cambridge: Cambridge Univer- sity Press, 2005. Galewicz 2006 Galewicz, Cezary. “The Fourteen Strongholds of Know- ledge: On Scholarly Commentaries, Authority and Power in 14 c. South India.” Texts of Power. The Power of the Text. Readings in Textual Authority across History and Cultures. Ed. Cezary Galewicz. Krakow: Homini, 2006. 141–164.  If You Don’t Know the Source, Call it a yāmala 349 Galewicz 2009 — A Commentator in Service of the Empire. Vienna: Sammlung de Nobili, 2009. Galewicz 2011 — “All the Books You Need to Understand the World around You.” Text Divisions and Classifications of Know- ledge in Literary and Epistemic Cultures of Pre-Modern India. Ed. Cezary Galewicz. Krakow: Institute of Oriental Studies, 2011. 125–162. CIS 13. Galewicz 2012 — “Texts and Communities: The Manuscripts of the Lost Yāmalāṣṭakatantra.” Aspects of Manuscript Culture in South India. Ed. Saraju Rath. Leiden: Brill, 2012. 119– 138. Galewicz 2014 — “A Socio-textual Ecology of the Ṛgvedadaśagrantha.” Rethinking Western India: The Changing Contexts of Cul- ture, Society and Religion. Eds. Dušan Deák and Daniel Jasper. New Delhi: Orient Blackswan, 2014. 48–71. Galewicz forthcoming a — “The Power of the Printed Veda.” Vedic Śākhās: Past, Present, Future. Eds. Julieta Rotaru and Jan Houben. Har- vard: Harvard University Press, forthcoming. 6–42. Har- vard Oriental Ser., forthcoming. Galewicz forthcoming b — “The Ṛgvedakalpadruma of Keśava Māṭe.” Proceed- ings of 6th Int. Vedic Workshop Calicut, 2014. Eds. M. Witzel, B. Narayanan and P. V. Bhattiripad. Forthcoming. Ganeri 1999 Ganeri, Jonardon. Semantic Powers. Meaning and the Means of Knowing in Indian Philosophy. Oxford: Clarendon Press, 1999. Genette 1987 Genette, Gérard. Paratexts: Thresholds of Interpretation. Trans. from French by J. E. Lewin. Cambridge: Cam- bridge University Press, 1997 (originally published in 1987). Gode 1945 Gode, Parshuram Krishna. “Some Evidence about the Lo- cation of the Manuscript Library of Kavīndrācārya Saras- vatī at Benares in A.D. 1665.” Jagadvijayachandas of Kavindracharya. Ed. C. Kunhan Raja. Bikaner: Anup Sanskrit Library, 1945. xlvii–lvii. Goudriaan and Gupta 1981 = Goudriaan, Teun and Sanjukta Gupta. Hindu Tantric and Śākta Literature. Wiesbaden: Harassowitz, 1981. A History of Indian Literature 2.2. Hanneder 1998 Hanneder, Jürgen. Abhinavagupta’s Philosophy of Revela- tion. Groningen: Forsten, 1998. Johns 1998 Johns, Adrian. The Nature of the Book. Chicago: The Uni- versity of Chicago Press, 1998. Klostermaier 1999 Klostermaier, Klaus K., ed. Sarvadarśanasaṅgraha As- cribed to Mādhavācārya. Chapter 16: Śaṃkaradarśanam. Chennai: Adyar Library and Research Center, 1999. 350 Cezary Galewicz Knorr Cetina 1999 Knorr Cetina, Karin. Epistemic Cultures. How the Sciences Make Knowledge. Cambridge, Mass.: Harvard University Press, 1999. Malamoud 1977 Malamoud, Charles. Le svādhyāya. Récitation personelle du Veda. Taittirīya-Āraṇyaka livre II. Paris: Institut de civilisation indienne, 1977. Minkowski 2010 Minkowski, Christopher. “Sanskrit Scientific Libraries and their Uses: Examples and Problems of the Early Modern Period.” In: Bretelle-Establet 2010: 81–114. Morson 2011 Morson, Gary Saul. The Words of Others. From Quota- tions to Culture. New Haven: Yale University Press, 2011. Müller 1898 Müller, Friedrich Max. Lectures on the Origin and Growth of Religion as Illustrated by the Religions of India. London: Longmans, Green, and Co., 1898. O’Hanlon 2010 O’Hanlon, Rosalind. “Lineages of Non-Brahman Thought in Western India c. 1600–1850.” (Mahatma Phule Me- morial Lecture, University of Calcutta 2010. 27 October 2016. <https://www.academia.edu/5773846/Lineages_of_ Non-Brahman_Thought>). Parameshwaranand 2004 Parameshwaranand. Encyclopaedia of the Śaivism. Delhi: Sarup and Sons, 2004. Patil 2002 Patil, Shankaragouda Hanamantagouda. Community Do- minance and Political Modernization: The Lingayats. New Delhi: Mittal Publications, 2002. Pollock 2001 Pollock, Sheldon. “New Intellectuals in Seventeenth-cen- tury India.” Indian Economic and Social History Review 38.1 (2001): 3–31. Pollock 2008 — “Is There an Indian Intellectual History? Introduction to ‘Theory and Method in Indian Intellectual History.’” Journal of Indian Pilosophy 36 (2008): 533–542. Roberts 1971 Roberts, John. “The Movement of Elites in Western India under Early British Rule.” The Historical Journal 14.2 (1971): 241–262. Sanderson 2001 Sanderson, Alexis. “History Through Textual Criticism in the Study of Śaivism, the Pañcarātra and the Buddhist Yoginītantras.” Les Sources et le temps. Sources and Time: A Colloquium, Pondicherry, 11–13 January 1997. Ed. F. Grimal. Pondicherry: Institut Français de Pondi- chéry/École Française d’Extrême-Orient, 2001. 1–47. Publications du département d’Indologie 91. Sastry 1921 Sastry, R. Anantakrishna, ed. Kavīndrācāryasūcipatram. Kavindracharya List. Baroda: Central Library, 1921.  If You Don’t Know the Source, Call it a yāmala 351 Sengupta and Ali 2011 Sengupta, Indra and Daud Ali. Knowledge Production, Pedagogy and Institutions in Colonial India. New York: Palgrave Macmillan, 2011. Srinivasavaradan 1976 Srinivasavaradan, K. “Śrīyāmalāṣṭakatantram (Manuscript Notice).” Sri Venkateswara University Oriental Journal 19 (1976): 49–58. Urban 2003 Urban, Hugh. Tantra: Sex, Secrecy, Politics and Power in the Study of Religion. Berkeley: University of California Press, 2003. Vishvanath 2007 Vishvanath, K. T. “Daśagranthi Tradition of Vedic Chant- ing.” Veda. Society. Modernity. Eds. C. M. Neelakandhan and K. A. Ravindran. Kunnamkulam: Panchangam Press, 2007. 74–82. Washbrook 2010 Washbrook, David. “The Maratha Brahmin Model in South India: An Afterword.” Indian Economic and Social History Review 47.4 (2010): 597–615.  Section 4: Reuse from the Perspective of the Digital Humanities  Methodological and Practical Remarks on the Question of Reuse in Epic Texts* Sven Sellmer Introduction Looking at the Sanskrit epics1 from the angle of “reuse,” the following chap- ter has three aims. Because reuse is quite a novel conception, its application in the context of epic studies will first be discussed on a general, methodo- logical level. Next, some examples of different types of reuse will be pre- sented and briefly discussed. Finally, the practical question will be raised of how to find possibly reused passages in the first place, given the vast amount of text that constitutes the two great epics. As to the type of reuse being investigated, I will look at the phenomenon in a traditional philological way; I will not be talking about the reuse of ideas or other abstract entities, but about concrete textual material (from single lines to dozens of stanzas). Before the question of reuse can be addressed, it is necessary to say some words about the history of the epic tradition. I am well aware of the fact that the historical development of the Sanskrit epics is in many details a hotly debated topic, but it is hoped that the general outlines I am going to sketch will be acceptable to most scholars.2 The earliest texts in the lines of tradition that have led to the large epics we have today are probably the result of oral poetry, comparable to other ancient and contemporary oral traditions that scholars have found in many regions and languages. At this point it is impor- ∗ My sincere thanks go to both editors for their many valuable questions, remarks and corrections. 1 Specifically, this chapter deals with the Mahābhārata (henceforth Mbh). Its title is nevertheless justified, because the methodological assumptions and analytic methods presented here can be applied to both epics. 2 Generally, I accept the picture depicted by John Brockington (1998, pp. 18–28). It should be added that there are also scholars, most notably Alf Hiltebeitel, who claim that the Mbh originated as a written work and, accordingly, downplay the importance of its oral background. Nevertheless, as explained in n. 14 below, the approach to reuse taken in this chapter also makes sense in the framework of this hypothesis. 356 Sven Sellmer tant to note that “oral poetry” in the technical sense is more than just non- written poetry – rather, its main characteristic is the fact that it is created (not recited!) during a live performance. This way of presentation is only possible if the poets (or bards, as they are often called) are in command of a traditional poetic language that provides a large repertoire of patterns (to use a very general term) fitted to the metrical and other exigencies of the poetic form in use. Many stylistic characteristics of oral poetry can be directly linked to this special production process, most obviously the abundance of formulaic expressions and the frequent use of “fillers.” But the epic tradition probably ceased to be pure oral poetry in this sense at a rather early stage, and there was a gradual shift from improvised oral composition to other forms that involved writing. Many things will certainly remain unclear about this process,3 but in any case we have some of the final products: the different versions and subversions of the great epics that have been copied and handed down as manuscripts for many centuries after becoming fixed.4 For my present purpose it is crucial that from a structural point of view the difference between the output of purely oral and post-oral poetry is not particularly big, due to the fact that the authors of the later genre continued to use the traditional style more or less faithfully. Therefore, it is perhaps ap- propriate to dub their method of composition “oral style poetry.” As a matter of fact, we cannot be sure that any part of the epic texts we have today is a written version of a real oral performance, or even comes close to one – per- haps the epics consist entirely of oral style poetry. For the time being, I propose to conceive the historical developments leading to the composition of the epics tentatively in the following simplified way: During the oral phase (i.e., at least for several generations) a means of poetic expression was formed on the basis of Sanskrit that should be regarded as a distinct sub-variety: epic Sanskrit.5 Linguistically, epic Sanskrit differs from standard Sanskrit, not only by containing certain non-Pāṇinian verb forms, sandhis, etc.,6 but most importantly by the existence of a large number 3 An outline of this development is proposed in Brockington 2000. As to Hiltebeitel’s different theory, see n. 14. 4 As a result of a purely technically motivated decision, the text used for the analyses presented here is the electronic text of the critical edition. It would have been best to use as many different versions of the Mbh as possible, if more texts had been available in digital form. 5 This way of looking at the status of the language used in the epics is epitomized in Oberlies’ Grammar of Epic Sanskrit (2003). 6 These features are comprehensively recorded in Oberlies 2003. Unfortunately, the metrical structure of epic Sanskrit and its importance for grammatical questions has, in my estimation, received too little attention in this otherwise admirably complete work  Methodological and Practical Remarks on the Question of Reuse 357 of metrically prompted quasi-syntactic patterns.7 This sub-variety of Sanskrit continued to be used for centuries for creating texts in the epic tradition, even at a time when the original oral tradition had long died out. In a non-oral environment, epic Sanskrit certainly underwent some changes, e.g., many of the patterns that once were essential for oral composition but had acquired the role of mere conventions for the post-oral poets faded away; nevertheless, many other patterns survived and can be put to use for the question of reuse, as I shall show below. As these oral features play a greater role in the narra- tive than in the so-called didactic passages of the Mbh, I will focus on non- didactic parts of the text. 1 Epic reuse The question of reuse takes on different forms in different contexts; therefore it seems appropriate to start with some reflections on its epic incarnation. Generally, one might perhaps say that “reuse” is a post-postmodern concept,8 in the sense that for its sake, the author – after having been buried by Roland Barthes (1967) and considered dead by this thinker’s postmodern followers for a long time – must be resurrected, because there simply is no reuse with- out a reuser. This move has especially important consequences for literary works like the epics, as opposed to, say, philosophical treatises. When inter- preting a work of literature where a character cites another text, we normally try to understand this reference against the backdrop of our understanding of this literary personage. In a way, one could say that the character in question does indeed reuse the quoted text;9 but this is not the kind of reuse at stake now. Rather, in order to apply the reuse paradigm to epic texts, a historical and, from the point of view of literary theory, somewhat naïve approach must be undertaken, one that focusses on the authors or redactors of single pas- sages to understand their methods or motives.10 This is a difficult task, be- (cf. ibid., pp. XXX‒XLI). 7 Compared with the situation in Homeric studies, not much work has been done on the formulaic language in the Sanskrit epics, though there do exist several very valuable studies; for example those by Brockington, Vasil’kov, Grincer, von Simson and Smith, to name just the most important scholars. (For detailed references, see the bibliogra- phies of Brockington 1998 and Brockington 2000 as well as my recent study Sellmer 2015). 8 See also the remarks in the introduction to this volume (pp. 11–13). 9 Cf. Hiltebeitel’s remarks about “Bhīṣma’s citation apparatus” (2011, p. 23). 10 Such an approach has a certain pedigree in epic studies, as the following remarks by Brockington on the distribution of Vedic material in the Mbh show: “The extent to 358 Sven Sellmer cause, as a rule, we know very little about the authors, or even if we are talk- ing about one person or a group, etc. It is therefore methodologically advisa- ble to avoid speculations about authorship and to stay at a general level where it is safe to say that the epic poets practiced the reuse of at least two types of textual elements: formulaic patterns of the epic language and passages of texts belonging to non-epic traditions. The first type that uses material from inside the epic tradition I will call “internal reuse” to distinguish it from the second, “external” type. 1.1 Internal reuse What in terms of the reuse paradigm can be dubbed internal adaptive reuse is more frequently called “creative use” of linguistic material. In this sense, it is a borderline case of reuse. Therefore I will confine myself to giving a few examples. First we have to identify the pattern being reused. For the present purpose it is best to leave aside finer distinctions and to work with the fol- lowing rough categories:11 1) Repetitions: identical or nearly identical passages (often one śloka or more) that are used twice or a few times. 2) Fixed formulas: stereotyped phrases (typically of a length between one and four pādas) that regularly appear with only minimal variation in cog- nate contexts. 3) Formulaic expressions with variable elements; in this type only the main part of a phrase is fixed, whereas in certain “slots” different words may be inserted. 4) Flexible sentence patterns that feature more than one dimension of varia- tion (examples will be given below). 1.1.1 Repetitions In contrast to the second type, i.e., fixed formulas, repetitions do not belong to the stock repertoire of the epic language, but are rather characteristic for which such exact or more distant citation of Vedic literature clusters in the philosophi- cally oriented parts of the Mahābhārata is very striking and no doubt attests the efforts made by the authors of these passages to buttress the authority of their compositions by this means” (1998, p. 14). 11 I deal much more thoroughly with this typology in the book mentioned in n. 7. The ex- pression “the epic language” entails a simplification, because inside the epic tradition it is possible to distinguish not only sub-traditions of the Mahābhārata and the Rāmā- yaṇa, but also different schools of these sub-traditions. This explains the very uneven distribution of certain formulas (cf. J. D. Smith 1987, pp. 609–611).  Methodological and Practical Remarks on the Question of Reuse 359 one author or for one small sub-tradition. If repetitions have their origin in non-epic texts, they must be treated (also) under the heading of “external reuse.” In practice, however, the distinction between the two types of recur- ring elements under enquiry is not always very sharp. As a rule one can say that repetitions are rather long, not frequent, and tend to appear compara- tively close to each other,12 but it must be admitted that all of these criteria are vague. The repetitions appearing close to each other are, for the most part, either the result of a kind of echo effect (i.e., the tendency to repeat words and phrases used a short time ago) or rhetorically motivated refrains. Those repetitions appearing in a greater distance, in turn, often seem to result rather coincidentally from textual growth and not from conscious reuse. In some cases, however, they may also fulfil a special function of internal adaptive reuse: the role of linking distant passages. Thus the following well-known verse is found twice. dharme cārthe ca kāme ca mokṣe ca bharatarṣabha | yad ihāsti tad anyatra yan nehāsti na tat kva cit || Bull among Bhāratas, whatever is here, on Law, on Profit, on Plea- sure, and on Salvation, that is found elsewhere. But what is not here is nowhere else (tr. van Buitenen 1978). It occurs once at the beginning of the epic (1,56.33) and once almost at the end of the work (18,5.38). In this way the stanza has – whether on purpose or by coincidence – two functions. In addition to its primary function of ex- tolling the comprehensiveness of the Mbh, it also serves as a pair of pa- rentheses enclosing a gigantic ring composition. 1.1.2 Fixed formulas Fixed formulas are regularly reused, though in the vast majority of cases not adaptively, but as stock elements in narrative situations of similar types; e.g., the phrase tad adbhutam ivābhavat appears as the final pāda of a śloka 77 times to mark an event as astonishing and extraordinary; the hemistich atrāpy udāharantīmam itihāsaṃ purātanam is used 110 times to introduce an alle- gedly ancient story, etc. 12 E.g., about one third of the 1333 pairs of identical hemistichs appear at a distance of less than 500 lines from each other. (The critical edition, without appendices and star passages, features 141,900 hemistichs in anuṣṭubhs.) 360 Sven Sellmer 1.1.3 Formulaic expressions The third type of reused elements has a “built-in” ability to adapt to changing contexts, as it were. As a common example one may cite a group of phrases that mark the beginning of a fight. These always occupy the first pāda of a line and end with the words samabhavad yuddham, so that the first two syl- lables remain to be variously filled. How the verse is completed depends on the number of fighters. If it deals with a single combat, the genitive dual tayoḥ is used (18 times); in the case of a larger group of warriors the line starts with the neutral tataḥ (11 times), and the involved heroes are then nor- mally enumerated in the following line. One may assume that historically one version of a given formulaic ex- pression was the prototype for the whole group of related phrase patterns, so that this family can be considered a petrified product (so to speak) of adaptive reuse. 1.1.4 Flexible patterns The fourth type of reused elements is somewhat similar to the third, but much more flexible due to the fact that, in addition to the possibility of exchanging an element for a metrically equivalent one, two more features ensure a high adaptive ability: a) The pattern as such has a general form (e.g., hero A hit hero B with ar- rows) and is concretized and realized during a stepwise construction process in which first the central, essential elements are fixed; only then are different peripheral elements chosen among a set of alternatives and added in accordance with the metrical situation resulting from the choice of the essential building blocks. b) In addition to choosing the metrically most suitable peripheral element, the poet may also change the word order in the framework of possibilities offered by the realized pattern. These patterns are useful for the poet in order to solve, with maximum ease, problems resulting from the fact that the central elements, most importantly the names of the heroes involved in an event, may have a very different me- trical shape, as in the following examples: senāpatiḥ suśarmāṇaṃ śīghraṃ marmasv atāḍayat (07,013.035ab) The general struck Suśarman quickly in the vital spots. śalyas tu navabhir bāṇair bhīmasenam atāḍayat (06,109.004ab) Śalya struck Bhīmasena with nine arrows.  Methodological and Practical Remarks on the Question of Reuse 361 These two lines illustrate two general rules that apply to one-liners of the type “A hit B.” 1) Tetrasyllabic objects are put into the first pāda if the subject is also tetra- syllabic. 2) The object is located in the second pāda if it has three or more syllables and the subject is disyllabic. It seems that rules of this type belong to the core features of the epic lan- guage, but the mechanics of Sanskrit epic versification are still little known.13 Generally, flexible sentence patterns can be regarded as belonging to the main tools of adaptation at the disposal of the epic poet. 1.2 External reuse and its detection In view of the many unclear points concerning the textual history of the Mbh, it is quite difficult to define precisely what it would mean for a textual pas- sage to be incorporated into the text from “outside” at different points of its development. So I propose simplifying the problem by supposing the sketchy model presented above: after an initially oral phase of composition, one or more versions of a work that at one point of time acquired the name Mahā- bhārata must have been known in a more or less fixed form. Accordingly, whoever further augmented this work (be it a poet or a redactor) must have done so either by composing new lines himself or by incorporating ready- made passages from other texts, possibly with small changes.14 In the latter case, the imported lines had to consist of meters fitting for the Mbh, i.e., first of all anuṣṭubhs, sometimes also triṣṭubhs, plus some rare meters. Now, in many texts “imported” passages are often marked as such, sometimes attri- buted to a source text or even to an author. In the epics, however, this is only exceptionally the case; mostly we are dealing with covert reuse. This raises 13 Note that in the given examples, as in most other cases, the rules are conventional, not the direct outcome of the metrical properties of the involved elements. Theoretically, it is also possible to construct an equivalent line with suśarmāṇam in the second pāda (e.g., *senāpatir bhruvor madhye suśarmāṇam atāḍayat), etc. 14 This general formulation is even compatible with the theory propounded by Hiltebeitel – who rejects the idea that an oral Mbh ever existed (though, of course, he does not deny the existence of oral poetry) and thinks that the Mbh was composed as a written text in a “short period of one or two generations […] between 150 B.C. and the year Zero […] as a work of composite authorship” (2011, pp. 11–12) – because the mem- bers of this hypothetical writing committee must have relied heavily on the reuse of different materials. 362 Sven Sellmer the fundamental question of how to find such passages in the first place.15 In this regard, the possibility to work on electronic texts with the help of com- puters has opened new perspectives that are not yet widely known; I shall present some of these in the following part of this chapter. There are two methods that are commonly employed for the purpose of finding reused passages: the first is based on direct evidence, whereas the second draws upon indirect evidence. When applying the latter method, some scholars focus on content, others work with formal criteria. Methods using direct evidence depend on the explicit ascription of a passage to an author or text, be it in a concrete or general way (e.g., by using anonymous references like ity eke, ity āhuḥ, ucyate, etc.). Indirect content-related criteria are typi- cally used in such cases where a passage is considered to be inserted on the grounds that it does not fit into the context, is unconnected to the plot, etc. – such arguments contain a certain amount of subjectivity and so are typically inconclusive if not supported by further evidence. Lastly, methods using indirect formal criteria are based on observations of marked differences between a certain passage and the surrounding text with regard to some ob- jective, formal features, such as for example metrical structure, word order, usage of particles, etc.16 As a general word of caution it should be underscored that neither of the methods mentioned is able to prove that a passage is a reused one – it can only suggest that this is the case. (This is even true for the method of direct evidence, because the ascription to a different author may well be fictional.)17 For any final proof one must be able to present a parallel passage in another text that can be demonstrated to be older; then the younger passage must be a reused version either of the older passage, or of a third, still older, common source. If this is not possible (as in most real-world cases), the second best solution is a strategy of accumulative evidence that uses arguments derived with as many of the above methods as feasible in any given case. In the following part of the present chapter, I would like to briefly present three methods of the indirect formal type that use quantifiable textual features 15 The biggest collection of identified quotations (including some covert ones) can be found in Hopkins 1901 (pp. 1–57). The remaining challenge now is to identify those quotations that the great American scholar could not possibly have found with the tra- ditional methods available to him. 16 This approach is an offshoot of stylometry, a branch of text analysis that goes back to the 19th century, but today can be much more successfully applied thanks to develop- ments in IT. 17 It is enough to mention the well-known case of Madhva’s Mbh “quotations” (Mesquita 2007). On this topic, see also Okita’s contribution to the present volume.  Methodological and Practical Remarks on the Question of Reuse 363 in order to identify passages that “stick out” in the context of their closer environment or even against the backdrop of the whole Mbh. These methods work with the following features, which shall be explained one after the other:18 – vocabulary – usage of heterotopes – metrical patterns 1.2.1 Unusual vocabulary As far as vocabulary is concerned, perhaps the most straightforward way to detect candidates for reused passages would be to look for a comparatively high density of words (in the sense of lemmata) that are rarely used in the Mbh. Unfortunately, this method cannot be realized at present, because flec- tion, sandhi and homonymy make it a difficult task for a computer to identify lemmata in a Sanskrit text. One would need a lemmatized text with resolved sandhis of the whole Mbh, but such a version is not yet available;19 in the meantime one has to look for rare strings (i.e., sequences of characters), which also yields reasonably good results.20 In order to achieve quantifiable results, the following calculation method was used. Every string that occurs with a frequency freq ≤ 521 receives a “rarity value” rv of 1/freq; all other strings are given a rv of 0. In the next step the rarity values of lines are calculated by simply adding up the string values. Now we can compare the rarity value of whole passages by letting the computer calculate the average of “windows” (i.e., passages) of a predefined length that move through the whole text. The plot below shows the results for windows with a length of 10 lines; for the sake of readability only the pas- sages with the highest (rv > 3.0, 142 observations) and lowest (rv < 0.1, 159 observations) mean rarity values are displayed, in such a way that vertical lines represent the difference from the mean rarity value of all anuṣṭubh lines 18 For the analyses that I present in this chapter, I only used the 141,940 anuṣṭubh lines of the main text of the critical edition and disregarded the so-called star passages and ap- pendices. 19 Fortunately, Oliver Hellwig’s DCS project already includes a lemmatized and POS tagged text with resolved sandhi for more than half of the Mbh, and it is hoped that the rest will be available in the near future (see Hellwig 2012). 20 In some cases, string search results may be improved by using regular expressions and fuzzy matching, so that mere sandhi variants of one and the same form (e.g., devo and devas) are not treated as two distinct entities. 21 The number five is a compromise between clarity and quantity of results: with a lower threshold only a few lines would be classified as having a high rarity value, whereas a higher threshold would multiply their number and thus yield a less clear picture. 364 Sven Sellmer (1.00). Accordingly, those pointing upwards have a higher than average rarity value (concretely, rv – 1.00), whereas lines pointing downwards have a lower one (1.00 – rv); dotted vertical lines mark parvan boundaries. Figure 1: Rarity values. A thorough analysis of this one graph alone would be too lengthy a task for this chapter, so I shall restrict myself to a brief discussion of the three highest peaks, which represent the following passages: – 3,155.36–48. In these lines that start the description of the Gandhamādana Forest, which the Pāṇḍavas are just entering, we find lists of the trees and birds of the forest. – 6,10.13–34. This passage also consists in a list. It contains the rivers that Saṃjaya mentions to Dhṛṭarāṣṭra in the course of his description of the earth.  Methodological and Practical Remarks on the Question of Reuse 365 – 9,45.1–28. In the last passage mentioned, the “mothers” associated with Skanda are enumerated. These three passages show two things: firstly, that looking for clusters of unusual words is a suitable method for finding unique passages; on the other hand, the “harvest” suggests that a high density of unusual words is often a rather weak indication that we have actually found an imported passage, because the phenomenon might be simply due to the fact that the topic as such is unusual, or that we have to do with a list that, thanks to its very structure, is more favorable to a higher concentration of rare words than narr- ative passages. Nevertheless, lists are items that may well be part of certain traditions of wisdom and devotion and so might be incorporated into the epic as useful treasures of information. Therefore, it would certainly be worth- while to have a closer look at passages identified with the help of the method just described and to search for parallels and possible sources, e.g., in Purāṇic literature. 1.2.2 Exceptional heterotopes The second way to detect exceptional passages is to look at the density of rare heterotopes of common word forms. First, a terminological remark may be in order: “heterotope” is a term I use for a linguistic entity that has gone largely unnoticed so far, though it is of crucial importance for the analysis of ślokas. Because of the fact that all words in the epics (and many other me- trical texts) tend to appear at certain metrical positions more often than at others, it is useful to regard the same strings that appear at different places of a hemistich as different entities.22 Using the Greek words for “different places” (ἕτεροι τόποι) and also thinking of a similar conception in chemistry, where we have isotopes (variants of the same chemical element), I deemed “heterotope” a fitting term. A convenient way to refer to heterotopes, also borrowed from the chemists, is to use upper indices; the syllables counted are the 16 syllables of an ardhaśloka. So, 9rājñā would be the heterotope of the string rājñā starting on the 9th syllable of a hemistich, and 10rājñā the hete- 22 The first person to address and give a name to the phenomenon of the uneven distribu- tion of words in ślokas was, to my knowledge, Ingalls (1991). Trying to systematize and quantify his observations, I introduced the term “heterotope” and proposed a quantitative measure of polarization – see Sellmer 2013a, 2013b. It is important to de- fine heterotopes on the basis of strings, not of words, because strings that are sandhi variants of the same word form (e.g., devaḥ) can have different metrical properties (e.g., deva and devo). Combining data about polarization with lexical and syntactic in- formation would be an important next step, once the tagged Mbh mentioned in n. 19 is fully available. 366 Sven Sellmer rotope starting on syllable no. 10, etc. But while the first one is found 49 times in the Mbh, the second appears only once, which illustrates why these items should be kept apart. Now, the point of the second method is to look precisely at the distribution of heterotopes like 10rājñā, which are rare not only in absolute but also in relative terms, i.e., compared to the most frequent heterotope of the whole group. This can be seen in the following table, which shows the frequency of the heterotopes of rājñā that occur in the Mbh; the reference item in our case is 9rājñā because it is the most frequent. Table 1: Heterotopes of rājñā starting syllable frequency except. value 1 20 59.18 2 3 93.88 3 16 67.35 4 8 83.67 6 14 71.43 7 46 6.12 9 49 0.00 10 1 97.96 11 25 48.98 Based on the relative frequencies one can calculate an exceptionality value for heterotopes (see column 3 of Table 1). In order to obtain these values I use the following formula: = 1− 100 Here fh stands for the frequency of a given heterotope (e.g., 25 in the case of 11 rājñā), fmax for the frequency of the most commonly occurring heterotope of the same string (which in the example above is 9rājñā with 49 occurrences). The resulting values range from 0 (no exceptionality, i.e., the most frequent heterotope) to close to 100 (very high exceptionality).23 In practice, one has 23 It should be stressed that this method is solely descriptive and does not deal with the reasons for the polarization patterns, which in some cases are rather obvious (though in most cases they are not). So it is certainly to be expected, for instance, that ca does not occur in the first syllable, because this particle is enclitic, that comparatively few hete- rotopes start at the second syllable, etc. The statistics applied in this section, however, are simply based on the frequencies of heterotopes as such and aim at discovering pas- sages that are exceptional qua featuring rare or exceptional heterotopes. The relative  Methodological and Practical Remarks on the Question of Reuse 367 to refine this method slightly because it makes little sense to call a heterotope “exceptional” when – though relatively rare in comparison to the standard heterotope – it is nevertheless frequent in absolute terms (i.e., 2ca makes up only 3.3% of all ca heterotopes, but still occurs 825 times in the Mbh). It seems advisable therefore to restrict the notion of exceptionality to hete- rotopes that are both of low frequency and belong to frequently occurring strings. These conditions are not clearly defined, so one has to find suitable thresholds for each individual text on a trial-and-error basis. For the Mbh quite interesting results can be achieved by including only exceptionality values of heterotopes that occur less than 10 times, whereas the respective base string features a frequency higher than 100. By adding up the exceptio- nality values of all heterotopes verse by verse, one can obtain a list of the most exceptional ślokas. Among these, there are a considerable number of verses that contain self-contained and general claims, so that it is quite possi- ble that they originally belonged to a free-floating body of gnomic utterances. Consider the following examples: sā bhāryā yā gṛhe dakṣā sā bhāryā yā prajāvatī | sā bhāryā yā patiprāṇā sā bhāryā yā pativratā || (1,68.39) She is a wife who is handy in the house, she is a wife who bears child- ren, she is a wife whose life is her husband, she is a wife who is true to her lord (tr. van Buitenen 1973). kruddho hi kāryaṃ suśroṇi na yathāvat prapaśyati | na kāryaṃ na ca maryādāṃ naraḥ kruddho ’nupaśyati || (3,30.18) An angry man does not perceive his task correctly, full-hipped wo- man; the raging man sees neither task nor limit (tr. van Buitenen 1975). For these two verses I was unable to find close parallels in the older lite- rature,24 but both appear in contexts where several unconnected verses of a gnomic type with a similar topic are used one after the other for rhetoric pur- poses, which makes it probable that the author(s) of these passages drew on conventional wisdom. frequency of heterotopes typically depends on many factors, but this is a different question. Here we are only concerned with finding exceptional passages. In a second step, one may then, in each case, try to explain the unusually high number of rare hete- rotopes. But explanations have no bearing on the purely statistical feature of “exceptio- nality” as defined here. 24 But cf. Garuḍapurāṇa 1,108.16cd; 1,108.18; Hitopadeśa I 197. 368 Sven Sellmer 1.2.3 Specific metrical patterns A particularly valuable method for the statistical analysis of epic texts uses the distribution of metrical patterns. Such an analysis is possible, because the sequences of light and heavy syllables in certain positions of an anuṣṭubh verse allow for some variation.25 Basically, one can apply simple methods, e.g., those that are based on the relative frequency of single patterns, but also more complex ones like the analysis of variance or cluster algorithms that take into account the frequencies of multiple patterns at the same time.26 One of the great advantages of a metrical approach is that it works to a large ex- tent independently of the dimension of content.27 In the context of the present article, it may suffice to discuss one simple example. The metrical sequence – ⏑ ⏑ ⏒ in syllables 5–8 of odd pādas (the so-called second or bha-vipulā) appears 4,479 times in the critical edition of the Mbh (i.e., in about 3.2% of lines, because each line consists of one odd and one even pāda) and generally is distributed quite evenly throughout the epic. There are, however, some two dozen rather short passages where it occurs with comparatively high frequency. One of these is the story of how Indra cunningly decapitated the Asura Namuci with foam (because he had promised not to kill him with anything wet or dry) in 9,42.28–9,42.37 – here we have five occurrences of the second vipulā in twenty lines, which amounts to a line frequency of 25%.28 This exceptional density raises the suspicion 25 For a convenient overview of the possible variations in anuṣṭubhs, see Steiner 1996. Though this chapter mainly deals with the classical anuṣṭubh, for the most part the rules presented are also valid for the epics. More details may be found in the literature given there, to which (as far as Buddhist texts are concerned) Balk 2011 should be added. 26 After the first attempts in the 19th and early 20th century (Hopkins 1901, Oldenberg 1909 etc.), pioneers of large-scale statistical metrical analyses in more recent times have been R. M. Smith (1960), Yardi (1986, 1994) and Tokunaga (1995). Unfortunate- ly, the first two scholars did not have an electronic text at their disposal and so had to collect their data manually. Smith, looking at the ratio of the different vipulā types, worked with quite fine-grained data, but did not use very convincing statistical methods. Yardi, on the other hand, collected rougher data but nevertheless managed to achieve highly significant results by applying a variance analysis. Tokunaga, in turn, did process an electronic text with the help of a computer, but based his statistics on whole parvans, which led to rather uninteresting results because these units are simply too large. 27 There are exceptions, though. E.g., the very frequent occurrence of a certain important word with an unusual metrical structure may influence the statistics of a whole pas- sage. 28 The search that yielded this passage as one result was conducted with the method of moving averages, where the frequency of a feature is measured in a “window” of n  Methodological and Practical Remarks on the Question of Reuse 369 that we are dealing with a reused passage here. Of course, as always, other explanations, like the personal style of a particular author, etc., are also possible. Still, the possibility of a borrowing is also suggested by the fact that the Indra Namuci myth, which has no direct relation to the main plot, is told by the narrator Vaiśaṃpāyana as one of the many stories connected to the holy sites on the bank of the river Sarasvatī visited by Balarāma. It is quite probable that numerous tales of a similar type existed outside the Mbh tradi- tion as local religious lore and were reused in the epic for various purposes.29 To be sure, even striking metrical peculiarities cannot prove on their own that a passage was imported from outside, but they certainly may serve as addi- tional indicators and so should always be taken into account when discussing such questions. Conclusion “Reuse” in the context of epic studies may mean different things. For metho- dological reasons, it seems important to make a distinction between internal and external reuse. Internal reuse works with different types of elements that form part of the epic tradition, and so really is a paradigm situated between the conceptions of “use” and “reuse.” It can only be properly understood against the backdrop of the general mechanisms of epic versification, a ques- tion that is still under-researched. Questions of external reuse, on the other hand, have been important in epic studies, often classified under the heading of terms like “interpolation.” Unfortunately, the developing discussions in many cases have proved to be inconclusive, because the arguments used have sometimes been based mainly on subjective assessments and insufficiently supported by objective observa- tions. Perhaps the statistical approach presented here may contribute to clari- fying some discussions and opening up new perspectives. Moreover, this could also hold true for texts outside the genre of epic literature, because lines that moves through the text (line 1 – line n, line 2 – line n + 1, and so on). In or- der to detect passages of different length, the size of the “windows” should be varied. 29 In the present case, there is a further detail that might suggest a non-epic origin of this passage: the text features the same expression meaning “by foam” (apāṃphenena) as found in two Vedic passages that refer to Indra’s feat (Ṛgveda 8.14.13, Atharvaveda, Śaunaka-recension, XX 29.3). This parallel makes it quite probable that the author consciously reused this combination of words – which, due to the genitive apām, has a certain archaic ring – directly from the Vedas or from some third source harking back to the Vedas. 370 Sven Sellmer methods like the ones proposed here are also applicable to other texts and corpora, especially if these are composed in meters with a flexible structure, like anuṣṭubhs, and are large enough to allow for meaningful statistical ana- lyses. To be sure, the methodological tools must be further improved, but even now it can be said that the detection of possibly reused, or, in any case, unusual passages is one of the fields where the help of the computer for the philologist may prove to be particularly fruitful.30 References Balk 2011 Balk, Michael. Untersuchungen zum Udānavarga unter Berücksichtigung mittelindischer Parallelen und eines tibetischen Kommentars. Marburg: Indica et Tibetica, 2011. Indica et Tibetica 53. Barthes 1967 Barthes, Roland. “La mort de l’auteur.” Le bruissement de la langue.” Ed. Roland Barthes. Paris: Seuil, 1984. 61–67. Essais critiques / Roland Barthes 4. Brockington 1998 Brockington, John L. The Sanskrit Epics. Leiden: Brill, 1998. Handbuch der Orientalistik 2.12 Brockington 2000 — “The Textualization of the Sanskrit Epics.” Textuali- zation of Oral Epics. Ed. Lauri Honko. Berlin: de Gruyter, 2000. 195–215. Trends in linguistics 128. van Buitenen 1978 van Buitenen, J. A. B., ed. The Mahābhārata. Vol. 3. Book 4. The Book of Virāṭa. Book 5. The Book of the Effort. Chi- cago: Chicago University Press, 1978. Garuḍapurāṇa Garuḍapurāṇa of Kṛṣṇadvaipāyana Vyāsa. Ed. by Shri Ramtej Pandey. Bombay: Veṅkaṭeśvarasṭīma Mudraṇā- laya, 1906. Hellwig 2012 Hellwig, Oliver. “DCS. The Digital Corpus of Sanskrit.” Heidelberg (2010-2012). URL http://kjc-fs-cluster.kjc.uni- heidelberg.de/dcs. Hiltebeitel 2011 Hiltebeitel, Alf. “Weighting Orality and Writing in the Sanskrit Epics.” Reading the Fifth Veda: Studies on the Mahābhārata: Essays. Vol. 1. Eds. Vishwa Adluri and Joydeep Bagchee. Leiden: Brill, 2011. 3–29. Hitopadeśa Kale, M. R., ed. Hitopadeśa by Nārāyaṇa. With a Sanskrit Commentary “Marma-prakāśikā” and Notes in English. Delhi: Motilal Banarsidass, 61967. 30 I would be glad to distribute the tools I applied for the analyses used in this chapter to all interested persons. If you would like to obtain them or discuss related topics, please do not hesitate to contact me by e-mail: sven@amu.edu.pl.  Methodological and Practical Remarks on the Question of Reuse 371 Hopkins 1901 Hopkins, E. Washburn. The Great Epic of India. Charac- ter and Origin of the Mahabharata. Rpt. Delhi: Motilal Banarsidass, 1993. Ingalls 1991 Ingalls, Daniel. “The Mahābhārata: Stylistic Study, Com- puter Analysis, and Concordance.” Essays on the Mahā- bhārata. Ed. Arvind Sharma. Leiden: Brill, 1991. 19–56. Brill’s Indological Library 1. Janert 1967 Janert, Klaus L., ed. Hermann Oldenberg. Kleine Schrif- ten. Teil 1 und 2. Wiesbaden: Steiner, 1967. Glasenapp- Stiftung 1.1. Mesquita 2007 Mesquita, Roque. Madhvas Zitate aus den Purāṇas und dem Mahābhārata: Eine analytische Zusammenstellung nicht identifizierbarer Quellenzitate in Madhvas Werken nebst Übersetzung und Anmerkungen. Wien: Sammlung de Nobili, 2007. Publications of the De Nobili Research Library 34. Oberlies 2003 Oberlies, Thomas. A Grammar of Epic Sanskrit. Berlin: de Gruyter, 2003. Indian Philology and South Asian Studies 5. Oldenberg 1909 Oldenberg, Hermann. “Zur Geschichte des Śloka.” Nachr. v. d. Königl. Ges. d. Wiss. zu Göttingen, Philologisch-hist. Kl. (1909): 219–246. Rept. in Janert 1967: 1188–1215. Sellmer 2013a Sellmer, Sven. “Words in Metrical Milieux – A Computa- tional Approach to a Neglected Field.” Recent Researches in Sanskrit Computational Linguistics. Eds. Malhar Kul- karni and Chaitali Dangarikar. New Delhi: DK Printworld, 2013. 301–320. Sellmer 2013b — “Computer-Aided Detection of Unusual Passages in the Mahābhārata.” CEENIS Current Research Ser. Vol. 1. Eds. Danuta Stasik and Anna Trynkowska. Warszawa: Dom Wydawniczy ELIPSA, 2013. 44–54. Sellmer 2015 — Formulaic Diction and Versification in the Mahā- bhārata. Poznań: Wydawnictwo Naukowe UAM, 2015. R. M. Smith 1961 Smith, Ronald Morton. “Ślokas and Vipulās.” Indo-Ira- nian Journal 5 (1961): 19–35. J. D. Smith 1987 Smith, John D. “Formulaic Language in the Epics of India.” The Heroic Process: Form, Function, and Fantasy in Folk Epic: The Proceedings of the International Folk Epic Conference, University College, Dublin, 2–6 Septem- ber 1985. Eds. Bo Almqvist et al. Dublin: Glendale Press, 1987. 591–611. Steiner 1996 Steiner, Roland. “Die Lehre der Anuṣṭubh bei den indi- schen Metrikern.” Suhṛllekhāḥ. Festgabe für Helmut Ei- 372 Sven Sellmer mer. Ed. Michael Hahn. Swisttal-Odendorf: Indica et Ti- betica, 1996. 227–248. Indica et Tibetica 28. Tokunaga 1995 Tokunaga, Muneo. “Statistic Survey of the Śloka in the Mahābhārata.” Memoirs of the Faculty of Letters, Kyoto University 34 (1995): 1–37. Yardi 1986 Yardi, M. R. The Mahābhārata, its Genesis and Growth: A Statistical Study. Poona: Bhandarkar Oriental Research Institute, 1986. Yardi 1994 — The Rāmāyaṇa, its Origin and Growth: A Statistical Study. Poona: Bhandarkar Oriental Research Institute, 1994.