THE PHENOMENON OF THE SQUARE KUFIC SCRIPT: THE CASES OF ĪLKHĀNID ISFAHĀN AND BAHRĪ MAMLŪK CAIRO Volume One: Text Tehnyat Majeed St. Hugh’s College Thesis submitted for the degree of Doctor of Philosophy Faculty of Oriental Studies University of Oxford Hilary Term 2006  CHAPTER TWO WHAT IS SQUARE KUFIC? TERMINOLOGY, ORIGINS, AND CONCEPTUAL FRAMEWORKS INTRODUCTION With an understanding of the formal principles, stylistic characteristics and the various types of Square Kufic, we can now examine the script from a more literary and historical perspective by referring to its terminology. This terminology is gleaned from a few monographs on Square Kufic and a number of obscure citations and passages that appear in disparate writings. For the most part, this meagre literature never deals directly with the issue of terminology, yet a number of different terms emerge over time and the preference of one over another is never justified. The question then arises, if scholarship has largely neglected terminology, is it truly significant for a better understanding of the script? In response to this important question, this chapter hopes to demonstrate how the terminology for Square Kufic reflects several concepts that are rooted in the function and meaning of the script, and also to direct us to possible sources for the origin and development of the script. Let us, first, look at the nature of our sources. It must be stated at the outset that no comprehensive textual tradition exists for Square Kufic, and therefore, there is hardly a living debate on the topic.1 The frequency of the secondary writings is too irregular, and the size, too paltry. Thus, even to contemplate the task of making meaningful use of the fragmentary literature available poses several dilemmas. For instance, the absence of a textual dialectic blurs the distinction between primary and secondary source. It also leads to much repetition and the use of stagnant approaches to the study of the Square Kufic script, even though the accounts range from different centuries. For simplicity’s sake, as most studies deal directly with the physical 1Most texts on calligraphy deal with the scribal and chancellery practices and are not concerned with the function and application of scripts in the monumental architectural context.  CHAPTER TWO evidence, the primary sources are the Square Kufic inscriptions themselves, and therefore, all textual sources can be considered secondary and tertiary evidence. Moreover, in order to understand the type of texts involved, we can divide these into two broad groups, namely, traditional and western. The ‘traditional’ group consists of sixteenth- to twentieth-century Persian accounts, of which a few draw upon references to Square Kufic contained in thirteenth/fourteenth- century Arabic texts, but in a rather uncritical and incidental manner. The second group under the ‘western’ label comprises eighteenth- to twentieth-century contributions from Europe and America. The ‘western’ literature, in general, does not refer to the traditional sources, and draws marginally on contemporary western ones. Overall, for both traditional and western sources, the visual physical evidence remains the primary criterion for the study of the script and for the formation of its nomenclature. The disjunctive and highly superficial nature of this literature imposes methodological problems for its analytical study. Given the nature of the sources, this chapter, in the first section, presents a short historiography of the chief secondary sources, and in the second section, mainly attempts to develop a theoretical framework of conceptual structures that emerge from placing major terms in their individual historical, cultural, and literary contexts. This hypothetical framework has practical implications for it principally aims to ground the reader in the several concepts relevant to the script, even though in the course of orientation, the framework tends to collapse chronological and even cultural barriers. This process hopes to enhance both the appreciation and understanding of the empirical evidence for Square Kufic contained in the following chapters. In this way, the study attempts to integrate the fragmentary and rather incoherent literature into a thematic matrix 66  CHAPTER TWO of terms that reveal some of the complex interactions of the Square Kufic script with its numerous abstract contexts. SECTION ONE BRIEF OVERVIEW OF LITERATURE Combined together, the traditional and western sources containing the wide-ranging terminology for Square Kufic are divided into three broad categories. Each category comprises a limited number of relevant and significant texts. The following overview introduces these important references with a brief appraisal of their main concerns, merits, and shortcomings. The three categories of writings on Square Kufic are: • Specifically Square Kufic includes short studies dedicated to the script • Square Kufic in Architectural Case Studies is a group of distinct monographs covering the epigraphic programmes of individual monuments. The Square Kufic inscriptions are not the mainstay of this series, but are, nonetheless, given relative importance as part of the overall decorative inscriptional scheme • Square Kufic in General Surveys consists of compendia where the focus may or may not be the study of historical epigraphic material and where Square Kufic is only mentioned in passing, especially in the effort to define the script SPECIFICALLY SQUARE KUFIC The earliest piece dedicated to Square Kufic was by the French Orientalist J. Marcel in the Journal Asiatique (1833).2 Almost half a century passed before the topic again received some attention in the two short Bulletin de l’Institut Égyptien publications by Rogers-Bey and W. Innès, respectively.3 These writings were by a team of experts who belonged to a series of French scholarly commissions sent to document the material culture of Egypt after Napoleon’s invasion in 1798. Therefore, the evidence for Square Kufic in this documentation was obviously only from Egyptian cities, such as Cairo and 2Marcel 1833, ‘Notice sur quelques inscriptions koufiques d’un genre singulier’, Journal Asiatique, tome XII, II série, Paris, pp.226-232. 3Rogers-Bey 1881, ‘Inscriptions en Caractères Coufique Carrés’, Bulletin de l’Institut Égyptien, pp.100-106; Innès 1890, ‘Inscriptions Arabes en Caractères Carrés’, Bulletin de l’Institut Égyptien, pp.61-67. 67  CHAPTER TWO Alexandria, and did not take into account similar inscriptions elsewhere in the Islamic world. As valuable as these contributions were for the material recorded, they contained several methodological problems that seriously limited the scope for further research. Their approach was predominantly palaeographic with some stylistic analyses. Documentation, for them, was a discrete process involving only the physical evidence that, on the one hand, divorced the inscriptions from their immediate architectural and cultural context, and on the other, contained incomplete information. Such a decontextualized and de-historicized method of representing the material, perhaps, not only reflected the lack of a consistent and well-formulated system for collecting inscriptions, but also revealed certain shortcomings and biases of the ‘orientalist’ scholar. While the enigmatic form and composition of these inscriptions intrigued scholars, the content often eluded them, even though for them, part of the charm in the encounter with Square Kufic lay in its decipherment. The Square Kufic ‘puzzles’, visually striking as highly geometricized images, represented mysterious codes waiting to be disclosed. Unfortunately, as intriguing as these ‘word puzzles’ may have appeared, these three French pieces failed to have an immediate impact. In 1920, Étienne Combe published an appraisal of some of the issues raised by Innès regarding the formal attributes of the script but ignored Innès’ objections concerning terminology.4 Combe’s essay, however, is a useful key for the earlier western documentation of Square Kufic inscriptions, as it systematically catalogues the evidence with information on content and location of inscriptions. More importantly, Combe brings in the evidence for Square Kufic on another important medium – coinage – thus extending 4 Combe 1920, ‘Notes d’Archéologie Musulmane’, Bulletin de l’Institut Français d’Archéologie Orientale, tome XVII, Imprimerie de l’Institut Français d’Archéologie Orientale: le Caire, pp.196–206. 68  CHAPTER TWO the scope of study for the script.5 Surprisingly, though, this piece never received any response. After an interval of more than forty years, three articles trickled in during the 1960s.6 The first in this series was highly disappointing. Lackany’s recourse to the redundant approach of anonymous lists of Square Kufic inscriptions was an intensely frustrating read, especially as he identified the inscriptions to be Qur’ānic but never provided either the verse citations or transcriptions. His preliminary remarks were even more exasperating as these perpetuated several old misconceptions about Arabic writing that had long been dispelled by Abbott’s in-depth analyses on the development of early Arabic writing.7 A negligible piece of writing, it is mentioned here to demonstrate that such isolated efforts undermined the interest in the topic and added no value to the discussion. With Paris-Teynac’s two articles featuring novel perspectives on possible meanings associated with the script, however, it appeared that the western scholarly tradition was somewhat redeemed. The author’s insightful comments shifted the focus from the formal stylistic discussion to a more symbolic philosophical one, although his comments remained schematic and needed to be developed to take on a more compelling position. In sum, however, these pieces too, concentrated predominantly on material in Egypt with a few inscriptions from North Africa. Still largely missing the evidence in the central Islamic lands, these writings gave the impression that Square Kufic was a predominantly Egyptian aesthetic. 5 Moreover, this link between the use of Square Kufic on monuments and coins is important because it brings in a direct ‘official association’ for the script. Even though, the majority of the monuments under consideration in this study are of dynastic patronage, in the case of coinage, the official dynastic connection is unequivocal. 6Lackany 1964, ‘Notes sur Quelques Épigraphes Coufiques d’Alexandrie’, Cahiers d’Alexandrie, pp.47-51; Paris-Teynac 1966, ‘The “Squared” Kufic Writing: Its Application in Artistic Decoration,’ The Islamic Review, January, pp.5-11; February-March pp.5-13. 7 Abbott 1939a, 1941. 69  CHAPTER TWO The geographical paradigm was modified only in 1975, when Makar conducted a comparative study of the Egyptian material with that from Iran and Anatolia. 8 Trained in western methods and techniques, she could not resist a formalist approach but creatively ventured within that to analyse compositional schemes in some detail. It was also the first attempt to show the relationship between compositional form and content of inscriptions. Similarly, Oman’s two studies on the large enigmatic Square Kufic panels within the Buqca-i Pīr-i Bakrān recalled the method of palaeographic and compositional analysis of earlier western writings on the script. 9 The balance tipped in favour of Iranian Square Kufic in the last quarter of the twentieth century when a number of Persian compilations on the script were generated from several decades of collaborative Iranian and western expeditions in the region, its building traditions, and architectural decoration. These Persian endeavours are photographic catalogues that serve primarily as pattern books. The writers construct detailed typologies for Iranian Square Kufic that are extremely tedious and complicated as only minor features distinguish the plethora of types. Their importance lies, perhaps, in design sketches and practices. Nonetheless, these publications are rich as illustrative and photographic surveys of the inscriptions, even though rather low on academic value.10 Only one of these catalogues discusses issues of origins, principles and formation of script by citing references contained in other contemporary Persian calligraphic surveys that, in turn draw upon earlier fourteenth- and fifteenth-century Persian texts.11 Even though the traditional pieces are not as widely accessible in terms of 8 Makar 1975, ‘Le Coufique Carré’, Nouvelle Revue du Caire, vol. 1, pp. 181-201. 9 Oman 1999, ‘Secondo Iscrizione in cufico “Quadrato” nella Moschea mausoleo di Pir-i-Bakran in Iran’, Quaderni di Studi Arabi, 17, pp.147-157; Oman 1998, ‘Il “Cufico Quadrato”: Tentativo di definizione delle tre varietà sinora riscontrate’ Quaderni di Studi Arabi, 16, pp.69-88. 10Ghouchānī 1985, Angular Kufic: On Old Mosques of Isfahān, Iran, text in Persian and English; Zamarashīdī 1991-92, Khatt-e Macqilī, Kāshīgārī-e Irān, Honar o Micmārī, Keyhān Š.1370, text in Persian; Mahmūd Māhir al-Naqsh 1991-92, Khatt-e Bannā’ī, Tehran Š.1370, text in Persian and English. 11 Zamarashīdī 1991-92. 70  CHAPTER TWO publication and lack academic rigour, they reflect age-old concerns about techniques and materials. It must also be remembered that despite the long textually barren interludes, the traditional aesthetic of the Square Kufic script has been preserved and continued as a practice in these regions. Finally in this category of writings dedicated to the Square Kufic script, I must mention an unpublished paper presented by Sheila Blair at a conference in 1996.12 Her study of the appearance of Square Kufic on significant monuments from the central Islamic lands compiles much of the earlier material in a systematic chronological order. Since then, however, a vacuum has existed in the literature on Square Kufic. SQUARE KUFIC IN ARCHITECTURAL CASE STUDIES For purely fortuitous reasons, the bulk of what constitutes the second category also focuses on the Iranian material. The study of architectural epigraphic schemes grabbed the attention of art historians who were trained in the languages and culture of this particular region. One of the earliest articles in this genre was by the Russian epigrapher Kratchkovskaya who documented a number of Square Kufic inscriptions in her article on the epigraphic programme of the fourteenth-century Īlkhānid Jāmic Masjid at Varāmīn (722H/1322).13 She also attempted to address the issues of origins for the script which she accurately surmised lay in brick construction on the basis of the earliest specimens existing in the eastern provinces of Iran.14 Miles followed Kratchkovskaya’s footsteps and catalogued the inscriptions in the Jāmic Masjid at Ashtarjān (715H/1315), the bulk of 12Blair (unpublished), ‘The Origin and Development of Square Kufic Script’, Actes du Colloque International; Université des Sciences Humaines, Départment d’Études Persanes, Strasbourg, ed. Hossein Beikbaghban, Tehran. 13Kratchkovskaya 1931, ‘Notices sur les inscriptions de la Mosquée Djoumca à Véramine,’ Revue des Études Islamiques, tome V, pp.25-58. 14 Thiswas of course, the origins of the ‘open’ type and not the ‘closed’ type. Kratchkovskaya does not make such a distinction of ‘open’ and ‘closed’ types of Square Kufic. Moreover, her article is mainly concerned with the material for Square Kufic in Iran and Central Asia. 71  CHAPTER TWO which, especially on the portal façade, are in Square Kufic.15 Pinder-Wilson brought to light the later much-quoted earliest example for Square Kufic, in his article on the famous minaret of Mascūd III, at Ghazna (492-508H/1099-1115).16 This Square Kufic inscription which has been carefully analyzed in the previous chapter, was as we recall, composed in a series of panels bearing parts of the commemorative titulature of the Ghaznavid Sultān Mascūd III.17 Similarly, the architectural case studies on Īlkhānid Sultān Öljeytü’s Mausoleum (704-13H/1304-13) at Sultāniyya and the Khānqāh and Shrine complex at Natanz (704-725H/1304-1325) conducted by Sheila Blair revealed the popular use of the script in fourteenth-century Īlkhānid Iran. 18 In sum, whereas these case studies do give an accurate impression of the proliferation of Square Kufic in Iran, they might mislead us into thinking that the rest of the Islamic world was empty of such material. This misrepresentation is, partly, created because of imbalances in the published literature and, mainly, because such case studies are region-specific, and therefore, do not take into account cross-regional parallels. Nonetheless, it is true that individual Iranian monuments are better covered for their overall epigraphic schemes (religious and secular inscriptions) than monuments in other regions and of course, the highest frequency of Square Kufic is found in Iran. Egypt, that had a rich inheritance of Square Kufic, appears to be marginalized in this category by the absence of such monographs. However, two architectural case studies are worth mentioning; one, by Meinecke on the Qubba al- Mansūriyya (683-84H/1285) and, the other by Ibrāhīm on the Zāwiya of Zayn al-dīn Yūsuf 15 Miles 1974, ‘The Inscriptions of the Masjid-i Jāmic at Ashtarjān,’ Iran, pp.89-98. 16 Pinder-Wilson 1985, ‘The Minaret of Mascūd III at Ghazni’ in Studies in Islamic Art, London, pp. 89-102. 17 See Chapter One, p.23, note 12. 18 Blair 1987, ‘The Epigraphic Program of the Tomb of Uljaytu at Sultaniyya: Meaning in Mongol Architecture,’ Islamic Art, II, pp.43-96; Blair 1986a, The Ilkhanid Shrine Complex at Natanz, Cambridge, Mass. Center for Middle Eastern Studies, Harvard University, pp.72-102. 72  CHAPTER TWO (697H/1298), both of which respectively recorded the Square Kufic panels within these edifices.19 In addition, of particular interest amongst the material for Egypt are a number of unpublished descriptive accounts of Mamlūk buildings, especially of Mamlūk portals, that contain useful photographic material for Square Kufic, as a couple of Bahrī and several Burjī entrance portals carry Square Kufic inscriptions.20 SQUARE KUFIC IN GENERAL SURVEYS The last category that contains, most often, only brief citations on Square Kufic consists of both traditional and western scholarly surveys with the earliest references traced back to late thirteenth- and early fourteenth-century Arabic manuscripts cited by Fazā’ilī in his calligraphic survey, along with some other post-medieval traditional texts.21 19 Meinecke 1971, ‘Das Mausoleum des Qalā’ūn in Kairo Untersuchungen zur Genese der Mamlukischen Architekturdekoration’, Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo (MDIK), Band 27, no.1, Tafel I-XII, pp.47-80; Ibrāhīm 1978, ‘The Zāwiya of Šaih Zain ad-Dīn Yūsuf in Cairo’, Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo, Band 34, Verlag Philipp von Zabern, Mainz/Rhein, Tafel 13-24, pp.79-110. 20 cAbdelRazik 1990, The Circassian Mamluk Monumental Entrances of Cairo, M.A. Thesis, AUC; Swelim 1986, The Complex of Sultān al-Mu’ayyad Shaykh at Bāb Zuwayla, M.A. Thesis, AUC; Roe 1979, The Bahrī Mamluk Monumental Entrances of Cairo: A Survey and Analysis of Intra Muros Portals 648-784/1250-1382, M.A. Thesis, AUC ; Makar 1972, Al-Sultānīyyā, M.A. Thesis, AUC. 21 Fazā’ilī 1971, pp.165-171. The sources cited by the author are in the following chronological order: 1. Nafāyis al-Funūn fī cArāyis al-cUyūn (compiled at the end of the seventh century hijra/ thirteenth century) (Islamic publications, Tehran Š.1377/1998-99). 2. Risāla-ye Khatt by cAbdallāh Sayrafī contemporary of Sultān Öljeytü (during the eighth century hijra/ fourteenth century) 3. Kitāb Sirāt al-Sutūr by Sultān cAlī Mashhadī (d.926H/1519) 4. Midād al-Khutūt by Mīr cAlī Haravī (d.951H/1544) 5. Midād al-Khutūt by Dervīsh Muhammad bin Dūst Muhammad Bukhārā’ī (d.995H/1586) 6. Tazkira-e Khushnavīsiyān (1239H/1823) by Ghulām Muhammad Haft Qalamī Dehlavī. 7. Nāma-yi Dānishverān 8. Khatt-o Khattatān 9. Musawwar al-Khatt al-cArabī (1388/1968) by Muhandis Nājī Zayn al-dīn 10. Ganjīna-yi Khutūt dar Afghanistān (1345H/1926). Unfortunately Fazā’ilī does not provide us with complete references, such as page numbers from these treatises and texts [Fazā’ilī 1971, pp.162-164]. 73  CHAPTER TWO The most popular excerpt is in Qādī Ahmad’s late sixteenth-century treatise on the history of Islamic calligraphers and painters. 22 Amongst the earliest western sources are Niebuhr’s Voyages en Arabie (1780) and Marcel’s documentation for the Description de l’Égypte (1800, 1828). 23 Niebuhr attached a handful of illustrations of Square Kufic inscriptions that he came across at Kūfa and Baghdād. Marcel who was aware of Niebuhr’s drawings,24 reproduced similar sketches for a number of Square Kufic inscriptions on the Islamic architecture of Egypt. It is obvious, however, that both scholars were merely intrigued by the odd formal and compositional aspects of the inscriptions, and not in historically situating the evidence. The nineteenth- and early twentieth-century European expansion in the Middle East produced a mass of such descriptive compendia for monuments and their decorative patterns, and therefore, especially in architectural documentation one is bound to encounter evidence for Square Kufic.25 With the advent of the major compilations on the epigraphy of the Islamic world, we would have hoped to have a complete record of Square Kufic inscriptions. 26 But sadly, this is not the case. Cultural biases influenced methodological approaches and resulted not only 22 Qādī Ahmad 1959. See Chapter One, pp.19-20 for this excerpt. 23Niebuhr 1780, Voyages en Arabie, Tab.XLIII, Tome II, de l’édition française d’Amsterdam; Marcel 1800, Mélanges de Littérature Orientale, le Kaire; Marcel 1828, Palæographie Arabe; ou, récueil de mémoires sur différens monumens lapidaires, numismatiques, glyptiques et manuscrits, présentant des inscriptions koufiques et karmatiques, dans tous les genres de caractères employés par les anciens Arabes. Première partie, l’Imprimérie Royale: Paris. 24 Niebuhr sketched the two small Square Kufic inscriptions that he found on the great mosque in Mesched cAlī, near Kūfa. He adds that he had seen several recent examples in this kufic style of Arabic writing in Baghdād [Niebuhr 1780, p.215]. 25d’Avennes 1877, L’art arabe d’après les monuments du Kaire, Paris, (trans) Erythraspis 1983, London; Creswell 1978, The Muslim Architecture of Egypt, Ayyūbids and Early Bahrite Mamlūks A.D. 1171-1326, volume II, Hacker Art Books, New York. 26van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, Première partie Egypte, Mémoires publiés par les membres de la Mission Archéologique Française au Caire sous la direction de M. Chassinat, Tome Dix-Neuvième, le Caire 1894, Paris 1903; Répertoire Chronologique d’Épigraphie Arabe, (eds.) Et. Combe, G. Wiet, J. Sauvaget, published by M. Cohen, Le Caire: Imprimerie de l’Institut Français D’Archéologie Orientale du Caire 1931-. 74  CHAPTER TWO in the utter neglect of Square Kufic, in particular, but also gave rise to a general disregard for the large bulk of religious inscriptions. While secular inscriptions received full attention with transcriptions, translations, locations, and commentaries, religious inscriptions, including Square Kufic, were either ignored or recorded inconsistently. To a large extent, this neglect of Square Kufic is symptomatic of certain pervasive attitudes, especially towards religious inscriptions, that had developed early in the field of Islamic epigraphy. As mentioned in the introduction to the dissertation, this was the legacy of the nineteenth- and twentieth-century project the Corpus Inscriptionum Arabicarum, formulated by Max van Berchem which considered religious inscriptions to lack historical information. Square Kufic inscriptions, wrote van Berchem “n’ont qu’une valeur artistique” and their content was too ‘banal’ for them to be included in any serious collection.27 Such was the dominant attitude that set the tone and model for van Berchem’s Corpus and of its successor the Répertoire Chronologique.28 The most frequently cited Square Kufic were the chār Muhammad panels in the Qubba al-Mansūriyya but, except for Meinecke’s article,29 not a single epigraphic compendium actually recorded the total number of panels within the edifice. Even though van Berchem records the presence of the panel carrying the name of Muhammad‫ ﺹ‬rendered in decorative geometric inlaid marble in a script which he called coufique carré, he offers no information on its precise location or on the total number of panels present in the monument.30 Like the case of our first category, the early concentration of references is within the nineteenth- and early twentieth-century surveys on Egypt, as Iran was still terra incognita. 27 van Berchem 1894, p.139. 28 Répertoire 1931-. 29 Meinecke 1971. 30 van Berchem 1894, p.139; Creswell 1978, p.193. 75  CHAPTER TWO In the 1930s, Godard’s Athār-e Irān and Pope’s Survey brought some of the Iranian evidence for Square Kufic to scholarly attention, but mainly through their photographic documentation.31 Apart from these, much later, short descriptions of the script appeared in the encyclopaedic volumes with entries on calligraphy and epigraphy.32 Amongst the traditional sources, Hunarfar covered the epigraphic schemes for the monuments of Isfahān in a comprehensive architectural survey that brought to light the majority of the Square Kufic inscriptions within this region.33 The most recent addition to western scholarship on Square Kufic was Blair’s Islamic Inscriptions that devoted a few pages tracing the development of the script and documenting the significant instances of Square Kufic inscriptions in the Islamic world, but avoided any discussion on the evolution of terminology.34 Last but not least, the documentation of historic inscriptions in Cairo, under the supervision of O’Kane, has included the evidence for Square Kufic and once published, will be a valuable source of photographs.35 A number of significant terms employed in this literature are tabulated as follows: 31Smith 1936, pp.313-360; Y. Godard 1936, pp. 367-369; and Pope 1939, 1079, 1747 for text dealing with Square Kufic. 32Sourdel-Thomine 1978, ‘Khatt’ in Encyclopaedia of Islam, Volume IV, 2nd edition, E. J. Brill, Leiden 1986, pp.1113-1122; Sourdel-Thomine 1986, ‘Kitābāt’ in Encyclopaedia of Islam, Volume V, 2nd edition, Leiden, pp. 210-216; Soucek 1979, ‘The Art of Calligraphy’ in The Arts of the Book in Central Asia: 14th-16th Centuries, ed. Basil Gray, UNESCO, 7-32; Yūsofī 1990, ‘Calligraphy (Khattātī Kŏšnevīsī), Encyclopaedia Iranica, vol. IV, fascicule 1, ed. Ehsan Yarshater, Routledge and Kegan Paul: London, New York, pp. 680-718. 33 Hunarfar 1971, Ganjīna-yi Āthār-e Tārīkhī-yi Isfahān, second edition, Isfahān: Kitābfurūsh-e saqafī. 34 Blair 1998, Islamic Inscriptions, Edinburgh University Press: Edinburgh. 35 O’Kane Documentation of the Inscriptions in the Historic Zone in Cairo [forthcoming 2006]. 76  CHAPTER TWO Major Terms for Square Kufic in Traditional and Western Sources Term for Square Kufic Region Scholar and Year Koufique Quadrangulaire/ Egypt (Cairo); Marcel 1828, d’Avennes 1877, 1983; G. & W. Coufique Quadrangulaire Algiers Marçais (1903); van Berchem (1894); Maslow (Tlemcen) (1937) Coufique Carré Egypt, Iran Rogers-Bey (1881); van Berchem (1894); Kratchkovskaya (1931); Y. Godard (1936) Caractères Carrés Egypt, Iraq Innès (1890) Chār cAlī Iraq Herzfeld, Sarre (1920) Hazārbāf Iran, Iraq Herzfeld, Sarre (1920); Herzfeld (1954, 1955); Ecriture Carrée Syria Herzfeld (1954) Naskhī Carré Syria, Anatolia Herzfeld (1955) Rectangular Naskhī Iran Smith (1936) Squared Kufic Egypt, Morocco Paris-Teynac (1966) Kufic Rectangles Egypt Creswell (1978) Rectangular Kufic Iran Flury (1939); Miles (1974) Square Kufic Iran Blair (1986), (1987), (1998); Sourdel-Thomine (1986); Soucek (1979) Macqilī / Mucaqqalī Iran Qādī Ahmad (16th century); Pinder-Wilson (1985); Ghouchānī (1984), Yūsofī (1990); Zamarashīdī (1991-92); Blair (1998) Kūfī Bannā’ī Iran Yūsofī (1990); Zamarashīdī (1991-92) Seal Script Iran Rogers (1997) al-Khatt al-Murabbac Egypt Marcel (1833) From this review of sources, we can see that we have no textual reference for Square Kufic from eleventh /twelfth centuries that would have been contemporary with the earliest physical evidence for the script. The references to Square Kufic contained in traditional sources that draw upon texts from the thirteenth to the sixteenth centuries, are extremely meagre and repetitive, for us to infer anything beyond what these state. A few give suggestions for the etymological roots for some of the terms such as macqilī. Most western studies are palaeographic analyses and never refer to traditional sources, and thus, no true discourse is ever generated. Moreover, there is never a collation of the archaeological evidence with written sources for Arabic epigraphy. Apart from the little that is offered by Innès, Herzfeld and Blair on terminology, there is almost nothing about the possible 77  CHAPTER TWO sources for the terms selected which to most scholars appeared to have no significant implication on the script or to have been self-explanatory.36 Even though Innès’ comments regarding terminology are very brief, these highlight the issue to be worthy of further consideration. Blair, by way of introduction, mentions a few of the important terms but does not go into any detail. The most unusual terms that are found in these texts are macqilī, square naskhī, chār cAlī, and ‘seal script’, while the most significant one for our purposes is, of course, the most current and widely used, Square Kufic. All other terms are modifications of one of these terms. SECTION TWO FRAMEWORKS FOR INTERPRETATION GENESIS OF THE TERM ‘SQUARE KUFIC’: THE KUFIC CONNECTION The eleventh and twelfth centuries were a time of extraordinary epigraphic creativity in the eastern Islamic lands, as designers vied to create new and ingenious methods of writing out their messages in fancy varieties of Kufic script. One script that survived the longest, probably because it is so effective in brick construction, is popularly known as square Kufic.37 Two unqualified statements that are also partially misleading are reinforced in this passage: first, that Square Kufic is a derivative of kufic; and second, that the script’s longevity and continuity over centuries is due to its affinity with architectural brickwork.38 In this section, we are only concerned with the first problem that deals with the term 36Innès 1890, p.62; Herzfeld 1956, p.357; Blair 1998, p.82. Their short comments on terminology are discussed further on in this chapter. 37 Blair 1998, p.82. 38 Thisis a problem with using a blanket term. The underlying assumption is that Square Kufic originated with decorative brickwork which is true in the case of the ‘open’ types which often covered large exterior revetments, however, we are still uncertain about the sources for the ‘closed’ panel types. 78  CHAPTER TWO Square Kufic and its relationship to the kufic script.39 According to Blair, the term Square Kufic developed out of convenience. Underlying this idea is the assumption that the script is related to kufic – another convenient generic designation for the different angular varieties of Arabic writing.40 Blair’s use of the term Square Kufic, however, is not without problems. For instance, her uncritical acceptance of the term, neither explains, nor justifies the choice of Square Kufic over other alternatives that she lists as square naskh, rectangular naskh, macqilī, and bannā’ī. Moreover, she neither cites earlier references from which she derives the term Square Kufic, nor elaborates on its evolution. 41 Unfortunately, with regard to terminology such neglect is quite typical, both in traditional and the western scholarship on Square Kufic. The problems that confront us, therefore, are: to trace the historical origins of the term Square Kufic; and more importantly, to investigate why it is considered an off-shoot of kufic and what are the implications of this connection between kufic and Square Kufic. Both the general entries on ‘khatt’ (calligraphy) and ‘kitābāt’ (epigraphy) in the Encyclopaedia of Islam employ the term ‘square kufic’ to refer to our particular script.42 As is the case with Blair, Sourdel-Thomine does not offer any information regarding the inception or evolution of the term. As we trace back the literature, we find that the term was coined by the nineteenth-century French scholar Rogers Bey, whose designation 39 The second problem of longevity and survival of Square Kufic cannot be attributed solely to its rendition in brick construction because much of the surviving material is actually in material other than brick. Moreover, whereas Square Kufic inscriptions in structural or decorative brick are common in Iran, they are a rare appearance elsewhere in Anatolia, Syria, Egypt and North Africa. Even in Iran, Square Kufic inscriptions are rendered in a variety of materials and techniques, such as stucco relief carving, tile faience mosaic, decorative plaster plugs, etc. 40 A popular misconception, as Abbott informs us that a number of angular and rounded scripts were known by different names during the first centuries of the hijra [Abbott 1939a, pp. 17-44]. Nonetheless it has to be admitted that it is easier and more convenient to use kufic to represent a number of angular varieties of scripts. 41 Blair 1998, p.82. 42 Sourdel-Thomine 1978, p.1122; and 1986, p.213. 79  CHAPTER TWO coufique carré was given epigraphic license by van Berchem in his monumental Corpus.43 The term ‘Square Kufic’ is then a literal translation of the French version.44 Coufique carré or Square Kufic became the most appropriate identification, during the twentieth century, for this unique epigraphic convention.45 A number of other closely related but rather cumbersome phrases such as koufique quadrangulaire, 46 squared kufic,47 kufic rectangles, 48 rectangular kufic49 had circulated but never gained much popularity.50 By the time Blair came to use it, Square Kufic was the customary nomenclature. At a glance, we can see that the word ‘kufic’ is the common denominator in all of these terms, but also that its adjunct status remains unexplained. Whether it appears before or after one of the geometric variants ‘square’, quadrangular’, or ‘rectangle, there is implicit consensus on its partnership with the square script. How did the kufic component become intrinsically linked with the script because none of the local terms in the late nineteenth and early twentieth century employed kufic with the various appellations for Square Kufic. 51 43 van Berchem 1894, p.139. He uses the term coufique carré when referring to the chār Muhammad panels within the Qubba al-Mansūriyya. 44 Rogers Bey 1881, Innès 1890, van Berchem 1894. 45 Thetwo respective publications in French by Kratchkovskaya (1931) and Makar (1975) retain the term coufique carré. 46 Marcel 1833, p.229. 47 Paris-Teynac 1966. 48 Creswell 1978. 49 Flury 1939, p.1747. 50Marcel 1800, p.14. This was the earliest known western attempt to define Square Kufic inscriptions and this led to a variety of English terms, such as, rectangular kufic. This particular denomination in French and in English was used by Prisse d’Avennes 1877; Maslow 1932; and Miles 1974. 51 Although we know recent Persian writings add kufic for example khatt kūfī bannā’ī, in Cairo the local term in the early nineteenth century was al-khatt al-murabbac [Marcel 1833, p.229]. And Pinder-Wilson mentions that the local term employed in Ghazna was mucaqqalī [Pinder-Wilson 1985, p.91]. 80  CHAPTER TWO What, then, is this ‘kufic’ connection with our script?52 At least two levels of association seem to exist between kufic and Square Kufic: one is at the formal level, thus more evident; the other is much more subtle, and operative on a more abstract conceptual plane. To investigate the first association, let us examine what the ‘kufic’ script represented for our nineteenth to twentieth-century sources. A historical overview of the development of Arabic writing undertaken by Abbott with reference to the Fihrist of the tenth-century copyist Ibn al-Nadīm reinstated ‘kufic’ as one of the early Arabic scripts classified under the khutūt al-masāhif, if not possibly the first formal Arabic script as Abbott speculates.53 Moreover, on the basis of the earliest influences of Syriac Estrangelo script on Arabic writing, she identifies this kufic to have been characterized by angularity and squareness where the vertical strokes tended to be short. 54 Other prominent features of this archaic kufic, as examined on early epitaphs and manuscripts, were the absence of diacritical marks and vowel accents. Its functional scope consisted of both its role as a lapidary script and, as a manuscript hand. Therefore, its functional domain extended to both monuments, as well as early Qur’ānic manuscripts. We can sum up the important physical attributes of this early type of kufic in the following manner: • Angularity and squareness • Absence of a systematic method for diacriticals and vocalization • Functioned both in the domain of monumental epigraphy and Qur’ānic calligraphy Angularity, squareness, lack of diacriticals and orthographical marks also characterize Square Kufic. Thus, there are striking similarities in the formal attributes of kufic and 52 Abbott points out that even though function and size of a script play an important role in its classification, other factors more truly calligraphic in nature, must have been influential too. “To discover and evaluate the role of these calligraphic factors a more intimate acquaintance with calligraphic terminology is needed” [Abbott 1941, p.95]. And the more one knows of these factors, the closer one comes to understand the use and meaning of a script. 53 Ibn al-Nadīm 1872, p.6, Ibn al-Nadīm 1970, p.11; Abbott 1939a, p.17. 54 Abbott 1939a, p.17. 81  CHAPTER TWO Square Kufic. One could surmise that Square Kufic being a later development in the evolution of Arabic scripts, displays obvious ‘kuficizing’ tendencies. However, two principles set Square Kufic apart from kufic: first, its strict organizing principle of negative and positive space, and second, for the majority of Square Kufic inscriptions the ‘unconventional’ arrangement of text. Although the first principle is unique to Square Kufic, the second can be applied to any script. The similarities, therefore, in formal characteristics, that is the usual system used for defining scripts, in general, and especially by western scholars of palaeography in particular, is the reason for the ‘kufic’ connection. In the Survey, Flury briefly defines rectangular kufic as ‘the last offspring of the unadorned Kufic script’ whose letter forms followed strict rectangularity which give the script the look of a ‘regular geometrical design’.55 Here this ‘unadorned Kufic script’ alludes to the archaic type of kufic rather than the contemporary eleventh- and twelfth- century kufic varieties which were highly ornate and can be regarded as evolved types incorporating a number of decorative flourishes and even curvilinear attributes. Whereas we see the reverse trend in Square Kufic which harks back to the more abstract, austere, simpler archaic kufic of the earlier centuries.56 In fact, Square Kufic inscriptions in the linear compositional setting closely resemble archaic kufic inscriptions except in certain areas where the Square Kufic tends to display stylized features where ligatures are elongated or twisted or turned in unconventional ways. The third functional aspect plays out yet another conceptual scheme. Similar to kufic, Square Kufic functions as a monumental script, yet unlike kufic, Square Kufic is hardly 55 Flury 1939, p.1747. 56 For instance the simple kufic writing seen in the inscription frieze within the seventh-century Umayyad Dome of the Rock and the similar script employed on Umayyad coinage. 82  CHAPTER TWO ever used as a Qur’ānic book-script.57 The extremely stylized formal attributes and the ‘unconventional’ compositional schemes of Square Kufic would not make it an appropriate script for copying the Qur’ān for such features rendered the text obscure rather than making it clear. However, it must be pointed out that, apart from a few exceptions, the bulk of Square Kufic epigraphy consists of pietistic religious expressions and Qur’ānic texts. Herein lies the subtle ‘kufic’ connection. Traditional sources classified kufic as one of the khutūt al-masāhif,58 that is one of the Qur’ānic scripts. The compelling evidence of early Qur’āns predominantly copied in angular scripts confirms this notion. As Abbott has perceptively surmised, the kufic script had come to be regarded as the sacred holy script of the Qur’ān.59 Even after Ibn Muqla’s reform when a variety of curvilinear scripts became popular for copying the Qur’ān, kufic was retained for Qur’ānic chapter headings. It is highly probable that western scholars did not pick up this particular notion or association in conjoining kufic with Square Kufic but it is certainly inherent in the traditional Persian sources. It cannot be purely coincidental that the Persian indigenous term for Square Kufic ̶ macqilī ̶ was, time and again, affiliated with kufic in all the traditional sources. For the Persian sources, especially of the Shīcī denomination, kufic was the esteemed 57 The appearance of Square Kufic on manuscripts and especially Qur’ānic manuscripts, is a much later development and even then its use is limited. For instance we have an example of a Tīmūrid illustrated manuscript documenting a Square Kufic inscription on a monument that represents the Kacba [Soucek 1979, p.20, II] [Figure 1]. Another example is of a sixteenth-century Ottoman Qur’ān whose frontispiece is decorated with a number of chapters and verses in Square Kufic while the main body of the Qur’ān is actually copied in a curvilinear script [Turks 2005, Item no.291, p.451] [Figure 2]. See also Karahisari’s Square Kufic Basmala [Safadi 1978, p.1] and the Square Kufic Painting in the Topkapi Saray Museum Hazine 2152, fol-9v [Grabar 1992, Plate 3]. 58 The Qur’ānic scripts were the khutūt al-masāhif, a category within the book hands khutūt al-warrāqīn which were of a rectilinear ductus while chancery hands were the khutūt al-kuttāb of a curvilinear ductus [Gacek 1989, p.144]. Ibn al-Nadīm lists the khutūt al-masāhif as al-Makkī, al-Madanī, al-Kūfī and al-Basrī scripts [Ibn al-Nadīm 1872, p.6]. It has also been noted that the larger scripts tended to be more angular whereas the smaller ones, more rounded [Gacek 1989, p.144]; , therefore monumental scripts were rectilinear while manuscript hands were more curvilinear – the material of the surface obviously influenced both the choice of script and its size [Abbott 1939a, p.16]. 59 Abbott 1939a, p.22; Ibn Taymiyya 1981, xiii, pp.420-421. 83  CHAPTER TWO script brought to perfection by cAlī ibn Abī Tālib. 60 Kufic was the script associated with the imām caliph cAlī’s political and spiritual centre at Kūfa. But kufic was also considered the source for all styles of writing in Arabic as Qalqashandī asserts:61 ‫ﻭ ﺍﳋﻂ ﺍﻟﻌﺮﺑﻰّ ﻫﻮ ﺍﳌﻌﺮﻭﻑ ﺍﻵﻥ ﺑﺎﻟﻜﻮﻓﻰّ ﻭ ﻣﻨﻪ ﺍْﺳﺘُﻨﺒِﻄﺖ ﺍﻷﻗﻼﻡُ ﺍﻟﺘﻰ ﻫﻰ‬ ‫ﺍﻵﻥ‬ The Arabic script (khatt) is the one which is now known as the Kūfic. From it evolved all present pens.62 While the western sources attach the ‘kufic’ component to Square Kufic because of the evident archaising features of the script, the traditional texts create a sacred association, especially a Shīcī one by placing the indigenous term in conjunction with kufic, as we shall see further on.63 MAcQILĪ, BANNĀ’Ī, HAZĀRBĀF: ARCHITECTURAL DECORATIVE VOCABULARY AND ITS DIFFUSION ACROSS MEDIA The three terms macqilī, hazārbāf, and bannā’ī are the major traditional Persian terms for Square Kufic. Most modern Persian authors use these terms interchangeably and consider them to be types of kufic.64 The emergence of the term macqilī is uncertain, but hazārbāf and bannā’ī are both known to be used within the vernacular of architectural decorative techniques. 60 Dūst Muhammad 1989, p.338; “None wrote better than that Holiness ̶ God’s blessing on him! ̶ and the most excellent kūfī is that which he has traced ̶ God’s peace on him!” [Qādī Ahmad 1959, p.54]. 61 al-Qalqashandī 1913-22, p.15. 62 Abbott’s translation of al-Qalqashandī’s text [Abbott 1939a, p.31]. 63“For, beginning with cAlī and Abū al-Aswad, the cIrāḳīs were soon going their own way in the writing of the Ḳur’ān, even to the extent of introducing new practices in vocalization, punctuation, and perhaps also ornamentation.... their script was soon to rival the Makkan and take the leadership in Ḳur’ānic scripts, so much so that in later centuries the phrase ‘Kufic script’ became almost synonymous with ‘Ḳur’ānic script’” [Abbott 1939a, p.22]. 64 Fazā’ilī 1971; Yūsofī 1990, p.686; Zamarashīdī 1991-92, p.6. 84  CHAPTER TWO Macqilī In the midād al-khutūt Mīr cAlī Haravī identifies the same main attributes of macqilī as described by Qādī Ahmad.65 Haravī adds at the end that the best type of macqilī is in a state of confinement ‘mahall-i tacaqqul’.66 Both macqilī and tacaqqul are formed from the Arabic root c-q-l. cAqala literally means to withhold, to restrain, to confine, to tether, to bind. Thus tacaqqul signifies a state of confinement, and restraint. Nouns formed from this verb such as cuqla and ciqāl mean bond, knot, and shackle. According to Lane, macqil is ‘a place to which one betakes himself for refuge, protection, preservation and therefore, signifies a fort, or a confined structure. 67 Fazā’ilī contends that these meanings correspond appropriately with the formal and compositional state of Square Kufic inscriptions where caqlun, the act of confinement, is evident in the frame that encloses a Square Kufic inscription. cUqla, the bond or shackle, resembles the straight lines that twist and turn on right angles to bond and restrain words. Thus Haravī’s macqilī is a state of confinement that tethers its letter-forms in interlocked or knotted bondage implying full restraint over the positive-negative traits to maintain balance and symmetry. In Afghanistan, the term is pronounced mucaqqalī and refers to ‘interlocking’.68 ‫ﺍﻣﺎ ﺩﺭ ﺩﯾﮕﺮ ﻧﺴﺨﻪ ﻫﺎ ﻫﻔﺖ ﺩﯾﮕﺮ ﻧﻮﺷﺘﻪ ﺍﻧﺪ ﺍﻭﻝ ﲤﻮﺩﯼ ﺩﻭﻡ ﺣﺠﺮﯼ ﺳﻮﻡ‬ ‫ﺭﻭﻣﯽ ﻣﻘﻠﻮﺏ ﭼﻬﺎﺭﻡ ﻛﻮﻓﯽ ﭘﻨﺠﻢ ﻣﻌﻘﻠﯽ ﺷﺸﻢ ﺟﻌﻔﺮﯼ ﻫﻔﺘﻢ ﮔﺮﺟﯽ ﻭ ﺍﯾﻦ‬ ‫ﺧﻄﻮﻁ ﻣﯿﺎﻥ ﻣﺮﺩﻡ ﻣﺘﺪﺍﻭﻝ ﺑﻮﺩﻩ ﻭ ﺑﯿﻜﺪﯾﮕﺮ ﻣﯿﻨﻮﺷﺘﻪ ﺍﻧﺪ ﻭ ﻗﺒﻞ ﺍﺯ ﺁﻧﻜﻪ ﺧﻂ‬ ‫ﻓﺎﺭﺳﯽ ﺑﺮ ﺭﻭﯼ ﻛﺎﺭ ﺁﯾﺪ ﻭ ﻛﺎﺭ ﮔﺎﻩ ﻋﺎﻟﻢ ﭼﻮﻥ ﺍﯾﻦ ﺯﻣﺎﻥ ﺑﺠﻮﺍﻫﺮ ﺧﻄﻮﻁ ﻭ‬ ‫ﻧﻘﻮﺵ ﺑﺪﯾﻌﻪ ﺑﯿﺎﺭ ﺍﯾﺪ ﺧﻂ ﻣﻌﻘﻠﯽ ﺑﻮﺩ‬ 65 Qādī Ahmad 1973, p.12; Qādī Ahmad 1959, p.53. 66 Fazā’ilī 1971, p.163. Haravī’s definition of macqilī is similar to Qādī Ahmad’s version: ّ‫ ﺑﻬﺘﺮﯾﻦ ﺧﻂ‬،‫ﻭ ﺩﺭ ﻗﺪﯾﻢ ﺧﻂّ ﻣﻌﻘﻠﯽ ﻣﻌﻤﻮﻝ ﺑﻮﺩﻩ ﻭ ﻣﺠﻤﻮﻉ ﺁﻥ ﺳﻄﺢ ﺍﺳﺖ ﻭ ﺍﺻﻼً ﺩﻭﺭ ﻧﺪﺍﺭﺩ‬ ‫ﻣﻌﻘﻠﯽ ﺁﻧﺴﺖ ﻛﻪ ﺳﻮﺍﺩ ﻭ ﺑﯿﺎﺽ ﺁﻧﺮﺍ ﺗﻮﺍﻥ ﺧﻮﺍﻧﺪ ﻭ ﻣﻌﻘﻠﯽ ﺑﺮﺍﯼ ﺁﻥ ﮔﻮﯾﻨﺪ ﻛﻪ ﻣﺤﻞّ ﺗﻌﻘّﻞ ﺍﺳﺖ‬ 67 Lane 1863, pp. 2113-2116. 68 Pinder-Wilson 1985, p.91. 85  CHAPTER TWO But there were seven other scripts in use. First tamūdī, second hijrī, third rūmī, fourth kūfī, fifth macqilī, sixth jacfrī, seventh girjī. And these styles were common amongst men and were in use one after the other. Before the fārsī script came into use, and the universe was embellished during this time, with wonderful scripts and tracings, there was the macqilī script 69 The obscure origin of macqilī is one of the integral issues concerning this term. As mentioned above, most sources highlight the derivative nature of kufic from macqilī. Qādī Ahmad reverses this order in his historical listing of important scripts in the world and places macqilī immediately after kūfī.70 He does not elaborate any further on the relationship between macqilī and kūfī. However, in a footnote, Minorsky, the translator and editor of Qādī Ahmad’s treatise, on the authority of Dr. Bayani adds that macqilī was named after the canal Macqil in Basra and that it was a script that resembled kufic.71 This is also confirmed by Thackston with reference to macqilī in Dust Muhammad’s preface to Bahram Mirza’s album written in 1544.72 Even though, we know that several early varieties of Arabic scripts were named after places allegedly where these originated, kufic being one of them, we have no information why a script would be called after a particular canal. In short, it is liable to be a Kūfa-Basra topos where both cities were known to have set standards in writing and because of their major role in the evolution of Arabic writing.73 In fact, this association with Basra has several other aspects to it. For instance, the phrases tamrun macqilīyun or rutabun macqilīyun denoted the special type of dates grown in Basra. Another source for macqilī is in relation to a Basran figure Mackil ibn 69 Qādī Ahmad 1973, p.12 for Persian text. 70 Qādī Ahmad 1959, p.53. 71 Qādī Ahmad 1959, p.53 note 125. 72 Thackston as the editor and translator of this piece, adds in the footnote that the Nahr al-macqil at Basra in southern Iraq is the source for the macqilī script. [Dūst Muhammad 1989, p.338]. 73 Abbott 1939a, p.17. 86  CHAPTER TWO Yesar. All these links between macqilī and Basra and macqilī and kufic recall the early development of the Arabic script. Abbott called this writing the north Arabic script as it originated in the North Arabian Nabataean region and travelled down to the Ḥijāz via Basra and Kūfa. 74 This may also be the reason why macqilī is not only mentioned in tandem with kufic, but according to some of the sources, is used almost interchangeably with kufic. For instance, when cAbdallāh Sayrafī75 draws parallels between the angular traits of muhaqqaq with those of kūfī o macqilī ( ‫ﭘﺲ ﻣﺸﺎﺑﻬﺖ ﺑﺎ ﺧﻂ ﻛﻮﻓﯽ ﻭ ﻣﻌﻘﻠﯽ ﺑﯿﺸﺘﺮ ﺩﺍﺭﺩ ﺍﺯ‬ ‫)ﺟﻬﺖ ﺳﻄﺤﯿّﺖ‬, we get a sense of this intimate affiliation. From these accounts, it almost appears as if macqilī was an ancient script associated with Basra and very similar to kufic. To venture even further, might it be suggested that macqilī and kufic on their close similarities were derived from a single source that came to be known by different names in the two cities of Basra and Kūfa? It is then understandable why Qādī Ahmad ignores the chronological sequence and places macqilī after kufic whereas Dūst Muhammad whose work preceded that of Qādī Ahmad, in the historical sequence of scripts lists macqilī as the predecessor of kufic: Thereafter, Yacrub ibn Qahtan produced the Kufic style from Macqili, and he was the inventor of the Kufic script; nonetheless, it reached perfection at the glorious hand of the Prince of the Faithful and Imam of the Pious, the Conquering Lion of God, cAli ibn Abi Talib.76 The ancient historicization of macqilī is interesting because the earliest reference to it dates back to a single late thirteenth-century source which serves as a basis for later 74 Abbott1939a, pp.14, 17. This was one of the routes traced for the origins of Arabic writing into the Ḥijāz. The other was a direct influence from southern Ḥaurān into the Ḥijāz. 75 The Risāla-ye Khatt of cAbdallāh Sayrafī was written in the fourteenth-century during the reign of Īlkhān Sultān Öljeytü [Fazā’ilī 1971, p.163]. 76 Dūst Muhammad 1989, p.338. 87  CHAPTER TWO interpretations.77 However, this late thirteenth-century text does not contain the word macqilī and instead contains a reference to an ancient script called camlīqī.78 On the basis of similarities in the description of the two styles of writing, Fazā’ilī has inferred that macqilī and camlīqī were one and the same script. While we do not come across the term macqilī in ibn al-Nadīm’s Fihrist, neither do we have camlīqī listed in his compilation. Anyhow, even if we accept that there was an ancient script called macqilī and it shared several traits in common with kufic, this archaic macqilī, obviously would have appeared quite different from the medieval stylized macqilī or what we have referred to as Square Kufic.79 Though an interesting aspect of macqilī as highlighted by Sultān cAlī Mashhadī (d. 92H/1519) was its prominence as an architectural script. 80 Incidentally, Yūsofī who defines macqilī as the ‘unadorned form of Kufic’ draws on this link with kufic as the ‘mother script’ ̶ the source from which macqilī evolved. This is especially apparent in the few paragraphs he devotes to macqilī under the sub-heading of ‘Kufic’.81 The question, however, remains as to how, when and why our stylized Square Kufic script acquired the appellation macqilī? From the passages in the scant traditional sources, we sense an underlying conscious or semi-conscious attempt to provide a historical and sacred lineage 77 The anonymous treatise Nafāyis al-Funūn fī cArāyis al-cUyūn actually contains the term camlīqī’ instead of macqilī within a similar listing of scripts where it is mentioned that kūfī was derived from this ancient camlīqī script. Fazā’ilī contends that this camlīqī became later corrupted to macqilī and proposes that camlīqī could have been the better-known script himyarī ̶ the script of the south Arabian tribes [Fazā’ilī 1971, p.162]. 78Nothing more is mentioned about this script. Perhaps it was associated with the Amalekites, a nomadic people in southern Palestine. 79Fazā’ilī 1971, p.164. The following sources cited by Fazā’ilī contain the term macqilī and associate it with kufic. Nafāyis al-Funūn fī cArāyis al-cUyūn (Š.1377/1998-99); Risāla-ye Khatt by cAbdallāh Sayrafī; Kitāb Sirāt al-Sutūr by Sultān cAlī Mashhadī; Midād al-Khutūt by Mīr cAlī Haravī; Midād al-Khutūt by Dervīsh Muhammad; Tazkira-e Khushnavīsiyān (1239H/1823) by Ghulām Muhammad Haft Qalamī Dehlavī copies that muhaqqaq resembles cibrī and kūfī and macqilī; Musawwar al-Khatt al-cArabī (1388/1968) by Muhandis Nājī Zayn al-dīn on the authority of ibn Athīr in the Usd al-ghāba and with reference to the Subh al-Acshā, cites macqilī as a type of kufic. 80 Zamarashīdī 1991-92, p.6. 81 Yūsofī 1990, p.686-687. 88  CHAPTER TWO for the script. This preoccupation with ‘prototypes’, ‘inventors’, and ‘religious figures’ suggests tendencies first, to establish a hierarchy, secondly, to reaffirm the sacredness and prestige of kufic and macqilī, and lastly, in the Iranian context, to realign the sacred lineage of the calligraphic tradition with cAlī ibn Abī Tālib and, thus, tie it securely to a Shīcī milieu. Bannā’ī and Hazārbāf Bannā’ī is from the word bannā that means ‘builder’ or ‘mason’ or ‘bricklayer’ in Persian.82 Bannā’ī is used to refer to the ‘construction of buildings’.83 In the wider architectural context, it came to be applied to the builder’s technique of patterned brick masonry that produces ornamentally enriched designs over the structural fabric. The term was used for Square Kufic inscriptions, in particular, the ‘open’ types where the epigraphic pattern was woven into the structural fabric through special brick-laying techniques. This, according to Kratchkovskaya, was the pure form of Square Kufic, or as we would define it as the bannā’ī type in which the epigraphic decoration was linked to the structural and architectonic elements of the building.84 The Mascūd III Ghazna example (492-508H/ 1099-1115) would then be a purer form of the bannā’ī Square Kufic type.85 From this we extrapolate that later bannā’ī types, especially adapted in tile faience mosaic where it formed a purely ornamental revetment that had lost its intrinsic connection with the structural fabric, would be regarded as the impure form of Square Kufic. However, the term bannā’ī was still retained even after it had abandoned the architectonic element of the brick bonding technique. Modern Persian authors generally refer to the script as khatt-i 82 Wulff 1966, p.108. 83 Bromberger 1989, p.709. 84Kratchkovskaya does not actually use the term bannā’ī but I include it to distinguish between other types of Square Kufic. 85 See Chapter One Figure 2b. 89  CHAPTER TWO bannā’ī or kūfī bannā’ī, perhaps to draw the subtle distinction between epigraphic and non-epigraphic bannā’ī. It is no wonder, then, that the Encyclopaedia Iranica’s entry on bannā’ī only defines it as ‘construction’ without ever mentioning that it is also a term used to denote the ‘builders’ script’.86 Hazārbāf is the other local Persian term used for ornamental brick decoration. It literally means a ‘thousand interweavings’ because the technique gives the impression of a woven structural fabric with a herringbone pattern produced by the step bonding technique. It is possible that the term came from textiles because it literally implies a technique used for making textiles. Like bannā’ī, hazārbāf is not exclusively used for Square Kufic inscriptions, but also denotes non-epigraphic geometric designs in brickwork. In fact, the earliest instance of hazārbāf was non-epigraphic that appeared on the Baghdād gate at Raqqa.87 The transmission of terminology across media, like motifs is not unusual and its dissemination from textiles to architecture is not inconceivable. Two conditions allowed for such exchanges. First, although the organization and operations of the guild system have been little studied, the evidence of diffusion of similar motifs, structural designs, techniques across media indicates that such transference and exchanges were certainly taking place.88 In such a case of transmission, the diffusion of terminology would not be surprising either. Second, in the medieval context, especially with regard to Mongol and Turkic dynasties, who despite sedentarization were still itinerant and nomadic in spirit, the importance of textiles for not only ceremonials, rituals, and diplomatics but palatial tents 86Bromberger 1989, ‘Bannā’ī’ in Encyclopaedia Iranica, vol.III, ed. Ehsan Yarshater, Routledge & Kegan Paul: London and New York, pp.709-712. 87 Wulff 1966, p.118; Creswell 1978, p.185. 88 It is documented that in sixteenth-century Ottoman Turkey, painters and illuminators designing manuscripts would also be employed to design the decorations of imperial buildings [Rogers 1992, p.231]. 90  CHAPTER TWO has been demonstrated cogently by Thomas Allsen.89 Indeed, the evidence of Square Kufic on textiles, even though fairly limited and a much later phenomenon, provides a further link between architecture and textiles.90 NASKHĪ CARRÉ AND THE PROPORTIONED SCRIPT Naskhī Carré was first designated by Herzfeld in the following passage with reference to two engraved panels found on the minbar within the Jāmic Taghribirdī (797H/1394) in Aleppo:91 L’examen de ces exemples montre que cette écriture est dérivée du naskhi, et que la designation “coufique carré” est incorrecte.92 Whilst Herzfeld accepts both hazārbāf and chār cAlī as other substitutes for this type of writing, his total departure from conventional terminology in his rejection of the term coufique carré is at the same time puzzling and provocative. However, the earliest credit for such a dismissal of coufique carré in favour of naskhī actually goes to Walter Innès in 1890.93 But Herzfeld does not refer to Innès at all in his work. In his earlier writings such as the Archaeologische Reise (1920) and the first volume of the Matériaux pour un Corpus (1955), Herzfeld used the term ‘tchār ‘Alī” en écriture 89 Allsen 1997. 90Square Kufic inscriptions are seen on a number of Safavid carpets and on Ottoman talismanic shirts. Interestingly, for the medieval period, we have instances of chār cAlī type of motifs appearing on western ecclesiastical vestments that were produced as far north-west as Austria under the supervision of the Queen Kunigunde in the thirteenth century [Figure 3]. 91 Unfortunately, I have not come across any photograph of this minbar in published sources. 92 Herzfeld, 1956, p.357. 93 ... les caractères carrés pouvant, en outre, représenter tout aussi bien des caractères neshi, il n’y a pas de raisons pour classer toute inscription en caractères carrés, qui porte de ces signes, parmi celles du genre coufique. Innès disagreed with classifying all of these square inscriptions under the genre ‘kufic’ as some of the traits displayed by them were more akin to those of naskhī [Innès 1890, p.63.] However, this was because Innès interpreted the mask fillers within the composition as diacriticals, attributes employed in naskhī. 91  CHAPTER TWO carré,’.94 Even when the text of the panel did not contain the name of cAlī, as the author informs us, it was the designation given to this script in the East.95 In the second volume of the Matériaux (1956), Herzfeld presents his revised and more accurate term naskhī carré.96 Unfortunately, we know nothing more about his reasons for considering the script to be a derivative of naskhī instead of kufic – contrary to western popular opinion. However, it appears that amongst certain scholars a tacit understanding of Herzfeld’s contention for a naskhī derivation appears to have existed, for Smith used a similar but modified version called rectangular naskhī.97 In terms of formal qualities, both kufic and Square Kufic have a mutual rectilinear resonance. Naskhī carré is not only an extremely divergent term, but it has the effect of creating a dynamic tension with the pairing of two words that have opposite connotations: the angular square (carré) versus the rounded naskhī. In this case, the retention of square (carré) serves as a counterpoint. An extremely strange designation from every single formal and physical aspect of the script that in no way displays the slightest inclination towards roundedness. Let us split the designation square naskhī, and cast a glance at how the naskhī script is interpreted in early twentieth-century scholarship. As a rival to kufic, naskhī had come to represent the antithesis of all angular scripts – the curvilinear ductus.98 This is clearly attested by al-Qalqashandī who notes that naskhī represented the rounded type of script and was a rounder style of writing than the ghubār 94 Herzfeld 1955, p.94; Herzfeld 1920, II, p.308. 95 Herzfeld 1920, p.308. 96 Herzfeld, 1956, p.357. 97 Smith 1936, p.323. 98 Abbottobserves that this was a mistaken view prevalent amongst Arabic sources and which was later on “adopted by the western Arabists, who further confused the issue by applying the term naskhī to cover all rounded scripts” [Abbott 1939a, p.34]. 92  CHAPTER TWO and riqca scripts.99 In the case of kufic versus naskhī, Abbot states that the two scripts originally associated with specific individual scripts had come to be applied in the case of kufic to all scripts predominantly angular, and naskhī to those that were rounded.100 Moreover, Abbott informs us that naskhī had always coexisted with the earliest kufic type in manuscripts. From the extant evidence, naskhī was a administrative script that did not make an appearance in sacred Qur’ānic manuscripts before the eleventh century. Its efflorescence on monumental architecture followed almost a century later in the twelfth century under the Zengids and their successors the Ayyūbids.101 One thing is certain, though, that the eleventh-century Qur’ānic naskhī and the twelfth-century monumental one, was very different from the archaic naskhī of the early centuries. The turning point for naskhī was often regarded to be the time of Ibn Muqla’s reform on calligraphic and scribal practices in the tenth century. It is perhaps this reform of the so-called ‘proportioned script’ that may possibly hold the key to the missing link in understanding Herzfeld’s choice of square naskhī over square kufic. Unlike the previous cases where the tendency was to recall the archaic type of kufic, Herzfeld’s square naskhī referred to the post Ibn Muqla ‘reformed’ naskhī rather than its archaic form. The khatt al-mansūb, often mistranslated as the proportioned script, was actually a system or method of setting Arabic letter shapes to relative geometric mathematical proportions.102 It is uncertain which particular Arabic script was reformed to this method but scholars such as Schroeder put forward the hypothesis that the reform was applied to 99 Abbott 1939a, p.37. Abbott also mentions that the earliest reference to the term naskhī as employed for a specific pen or style of writing is in the Fihrist where ibn al-Nadīm refers to a script called qalam al-nassākh [Abbott 1939a, p.37; ibn al-Nadīm 1872, p.9]. 100 Abbott 1941, p.66. 101 Tabbaa 1994. 102 Abbott 1939b, p.70. 93  CHAPTER TWO the cursive scripts,103 even though the literary evidence cited by Nabia Abbott, on the authority of Ibn Khallikān, clearly suggested that it was the kufic script.104 If Herzfeld subscribed to Schroeder’s hypothesis that naskhī was the ‘proportioned’ script, or the script set out to this method, then we may begin to understand how the highly geometricized and seemingly proportioned attributes of Square Kufic would have struck some resonance with the ‘proportioned script’. Another remark is appropriate here. Herzfeld’s view about the relationship of Square Kufic to naskhī cannot be definitively substantiated, but it is strengthened by the evidence of architectural naskhī. Is it a coincidence or an accident of survival that the appearance of this ‘reformed’ naskhī and of Square Kufic on medieval architecture is nearly simultaneous? Herzfeld must have observed that the advent of naskhī on monuments coincided with the earliest attempts of Square Kufic epigraphy on architecture. On the basis of its highly geometric image, Herzfeld may have possibly surmised that the roots of, and the organizing principles for, the Square Kufic script are more similar to those of the reformed naskhī than to medieval kufic, that was mostly of a highly ornate variety. This hypothetical relationship between naskhī and Square Kufic can be taken one step further. The proportioned script was based on the proportions of the alif which served as the basic unit for generating the remaining letter shapes. The alif of the khatt al-mansūb was based on the measuring dot system; and depending on the type of ductus, the alif was 103 Schroeder 1937, p.234. 104Ibn Khallikān (608-81H/1211-82) is unequivocal that the reform of Ibn Muqla was actually applied to kufic rather than to naskhī. Abbott mentions that he records that ‘it was Abū cAlī ibn Muḳlah who first took the present system (of written characters) from the (style of) writing of the people of Kūfa, and brought it under its actual form’. So on the authority of Ibn Khallikān, Abbott concludes that the khatt al-mansūb was not a new script but a system of mathematical proportions to which each letter of the Arabic language was set in calligraphic writing thus bringing them all into proportional relationship with the basic alif, and she adds that it was in part at least devised from kufic forms [Abbott 1939b, p.83]. 94  CHAPTER TWO made up of five, seven or nine dots. 105 If we look at Square Kufic inscriptions and attempt to deconstruct formally the script, we see that the alif is generally also based on these number of dots, that is unit squares [Figure 4]. However, this is where the semblance ends for the remaining letter shapes either were of varying proportions that do not appear to be set against the height of the alif or were of fixed shapes such as the wāw. However, if this was Herzfeld’s reasoning for the similarities in the ‘measured method’ between naskhī and Square Kufic, then he was incorrect for even the earliest type of monumental kufic, that is on the Dome of the Rock (72H/691-92) was based upon the measured dot system. CHĀR cALĪ AND THE MANDALA: UNIVERSAL SACRED ARCHETYPES AND MAGICAL SYMBOLS The term chār cAlī is Persian in origin. Chār means four and cAlī refers to the fourth orthodox caliph and the first Shīcī imām cAlī ibn Abī Ṭālib. Although chār cAlī signifies a specific type of Square Kufic inscription, in the early twentieth century, according to Herzfeld, it was generally applied as a term for Square Kufic epigraphy in Persia and Baghdād.106 In Herzfeld’s use of the generic chār cAlī, an appendage ‘en écriture carrée’ always follows. Incidentally, ‘écriture carrée’ translates into al-khatt al-murabbac which was the local term for Square Kufic in nineteenth-century Cairo.107 None of the traditional sources, however, contain such a general use of chār cAlī. It is possible that even Herzfeld’s generic chār cAlī referred to perhaps only the ‘closed’ types and not all types of 105In his description of the formation of the cursive script, Huart begins with naskhī whose letter shapes are developed on the system of the rhomboid dot ˗ the fundamental unit of measurement. This method is modeled after the system that he found in the Persian manuscript called the Miftāh al-Khutūt ‘The Key to the Scripts’ by Rizā cAlī Shāh Qadrī written in India in 1249H/1833 [Huart 1972, p.5]. This dot, whether rhomboid or square, is also the basic unit of measurement for designing Square Kufic inscriptions. ‫ﻛُﻞﱡ ﺣَﺮﻑٍ ﻣُﺮَﻛّﺐ ﻣِﻦ ﺍﻟﻨُﻘﻄَﺔ‬ ‘Each letter is formed from the dot.’ According to a tradition ascribed to the Prophet‫[ ﺹ‬Atlagh 1992, p.163]. 106 Herzfeld 1920, vol. II, pp.157, 158, 179, 231, 308. 107 Marcel 1833, p.229. 95  CHAPTER TWO Square Kufic inscriptions. For it is difficult to understand how the bannā’ī or hazārbāf types would go by the appellation chār cAlī, in particular if these contained historical secular texts.108 In any case, Herzfeld is careful to make the distinction between chār cAlī and the ‘vrai tchār Ali’, the latter being based on ‘la croix gammée’ – the swastika.109 This distinction emphasizes the very curious design feature of the inscription that sets apart the chār cAlī type from all other Square Kufic inscriptions. Unlike our previous terms that were based on formal and technical qualities, the term chār cAlī is directly embedded in its content which carries the name of cAlī, and indirectly in composition, that is represented by the number four or the four arms of the cross. In short, the chār cAlī has several peculiarities: (i) it is repeated four times, (ii) it is contained in a square (iii) the text forms a pin-wheel that, most often, moves clockwise. The arrangement of four cAlīs tied to a central square disc creates a circular movement within a square frame and, therefore, gives the impression of a design with a rotating circle inscribed within a square. There are a number of variations to the chār cAlī type but this was the most widespread standard pattern. From its visual design and formal compositional setting, it is obvious that although it is a text, it was not meant to be read as an inscription but rather to be seen as a symbolic image. While we are able to follow the logical emergence of the diagonal composition, as developing from the stepped brick-bonding technique, the origins of the circular setting within a square does not naturally arise from ornamental brickwork. Moreover, while our earliest monumental Square Kufic is, unequivocally, the hazārbāf or bannā’ī ‘open’ specimen from the minaret of Mascūd III (492-508H/1099-1115) at Ghazna, the earliest 108 Theearliest instances of Square Kufic are the Ghaznavid minaret at Ghazna and the Ghūrid mosque at Herāt ̶ specimens that bear commemorative texts with royal titulature [Pinder-Wilson 1985, 2001; Hillenbrand 2002]. 109 Herzfeld 1955, p.101. 96  CHAPTER TWO incidence of the chār cAlī motif predates this.110 Ghouchānī records the earliest surviving example for the chār cAlī carved into a wooden window frame within a funerary structure commissioned by the Būyid cAdud al-Dawla Fanā Khosrow in 363H/973 [Figure 5].111 Unfortunately, Ghouchānī does not provide any more information on this specimen and the photographic reproduction is not clear enough to examine the exact pattern but, according to the author, the carved inscription contains the name of cAlī which is repeated four times within a lozenge. A pair of such identical lozenges decorates the upper part of the window frame.112 This tenth-century appearance of the chār cAlī in wood instead of brick, raises, once again, the question of its origins. The closest parallels to the chār cAlī design are to be seen outside the Islamic world. There are two distinguishing features of the chār cAlī: (i) the circle inscribed within a square and (ii) the cross-shaped setting of the four cAlīs. The circle within a square order is a prominent geometric symbol for the Indian or Buddhist mandala [Figure 6];113 and the formation of the four cAlīs resembles the swastika motif which has its ancient roots in Indian Vedic cosmology where it forms a mandala. The similarities between the chār cAlī and the swastika pattern were observed by both Herzfeld and Paris-Teynac. In Herzfeld’s opinion, it was the chār Muhammad type that resembled this archetypal motif. Fascinated 110 This is not actually a monumental inscription but it certainly appears in the architectural context making it all the more relevant. 111Ghouchānī 1985, p.4, plate E. The windows are at the moment in the Persepolis Museum where these have been attributed to the Safavid period. Ghouchānī, however on stylistic bases, attributes the windows to the Būyid period. 112 There appears to be a longer inscription in a more ornamental variety of kufic within the lower part of the window frame. Incidentally Ghouchānī does not call this inscription chār cAlī but simply uses the Persian term macqilī which he translates in English as structural kufic [Ghouchānī 1985, p.4]. 113 The typical Tibetan mandala is composed of the square and the circle. A mandala is a geometric image used for meditation. The idea of using basic geometrical shapes combined in certain schemes has two principles governing it: first, a concentration on geometric shapes sharpens perception and creates a distinctive mindshift because such basic linear shapes are abstract and do not exist in nature and, second, geometric shapes such as the square, circle, and triangle and their combined permutations over centuries had acquired a cosmological symbolism and one of integration that was inherited by traditional cultures and religions across the world. 97  CHAPTER TWO by the composition and format of the chār Muhammad inscription on the Turba al- Rukniyya, Sālihiyya (621H/1224), in Damascus with ‘the name of Muhammad written four times in the turning movement of a swastika’, Herzfeld remarked that ‘in Persia it is usually the name cAli that is written in that way, whence the name chār Ali’ for that type of script’.114 Paris-Teynac also noticed the swastika as the dominant motif in the chār cAlī type and he tried to ascertain the possible intrinsic meanings of the design. He concluded that the motif carried Sūfī symbolism.115 In the combined motif of the circle and the square, the circle usually represented the cosmos and the archetypal plane while the square, was a symbol of the earth, the material plane. The circle is also timeless in quality, having no beginning or end. The concept of the circle inscribed within a square mandala then symbolised the ‘cohesion and integrality of the cosmos’ which upon contemplation would lead one towards integrating the material and the spiritual. Thus, the mandala was a multivalent symbol representing divine harmony, protection, and centering. Its physical manifestation is seen in numerous ancient religious contexts. The form of the third-century BC Buddhist stupa at Sanchi is a three- dimensional representation of such a mandala [Figure 7]. Similarly, it is this square-circle mandala that is the organizing principle of space in the late tenth-century temples at Khajuraho, and the same typical mandala appears in Hindu paintings for religious contemplation [Figure 8].116 It was also used as a mnemonic device in the transference of sacred architectural knowledge. The Chinese Luo Shu pattern which is generally a numerical arrangement into a magic square when represented non-numerically and linked 114 Herzfeld 1946, pp. 24-25. 115 Paris-Teynac 1966. 116 Crouch 2001, p.40, Figures 2.9, 5.5. 98  CHAPTER TWO transforms into a swastika pattern [Figures 9a-c]. 117 As ancient symbols of the earth and heaven, the square and the circle combined to represent the ‘universal sign of the transformation of matter into spirit, or death into life, of time into eternity’.118 A closer but more abstract parallel to the circle within square is the reversed order of square within circle that is created from an aerial view of the tawāf at Mecca – the circumambulation of the Kacba [Figure 10].119 However, a major difference must be pointed out: the tawāf around the Kacba is an anti-clockwise movement, while most chār cAlī and chār Muhammad compositions are set in a clockwise direction. 120 Burkhardt integrates the physical manifestation of the Kacba with its cosmological significance in observing that the four corners of the Kacba approximately correspond to the cardinal points, “doubtless because the cardinal points mean, in the Arab concept, the four ‘corner pillars’ (ARKAN) of the universe.” The four rotating arms of the swastika creating a circular movement within a square are perhaps also better understood in this interpretation by Burkhardt: The centre of the terrestrial world is the point intersected by the ‘axis’ of heaven: the rite of circumambulation (tawāf), around the Kacba which is to be found in one form or other in the majority of ancient sanctuaries, is then seen to reproduce the rotation of heaven around its polar axis.121 The square and the circle are also images of perfection since they are perfect shapes, in complete harmony and balance. The square nimbus like the circular halo was used to represent saints, and noble men who symbolized perfect conduct. We find the use of the 117 Berglund 1990, p.23. 118 Snodgrass 1985. 119 The Kacba literally means the ‘cube’ and is a physical approximation of a cubic form in reality. It functions as the centre ˗ the axis mundi to which Muslims in all parts of the world orientate themselves for religious prayer. The anti-clockwise direction of the tawāf (circumambulation) of the Kacba by pilgrims is found at times mirrored in chār cAlī compositions. 120 The clockwise setting of text is also a natural function of the fact that Arabic is read from right to left. 121 Burkhardt 1976, p.4. 99  CHAPTER TWO square halo in early Christian and Jewish portrayals of holy men, saints and prophets [Figure 11].122 The obvious parallels between the chār cAlī and the mandala are inherent in the geometrical form, the repetitive pattern of squares and circles, and in the centre as a point of reference. Moreover, just as the foundation of Square Kufic is based on the square dot, the mandala originates from the dot at the centre. In summarizing the meanings inherent in Buddhist mandalas used during the late eighth and early ninth centuries in Tibet, Bjerken identifies their five fundamental functions: (1) as meditative device, (2) as scholastic schema and mnemonic device, which he explains as the cosmos represented in miniature where the mandala connects the micro with the macro (3) as object of devotion once a mandala is represented spatially, for example, at sacred religious sites (4) magical function where the image or object is used to ward off negative energies and (5) ideological function representing an idealized hierarchy of status and power. 123 If the chār cAlī did approximate the role of a mandala in certain contexts, can we then attribute some or all of these functions to this particular type of Square Kufic? The number four has its own symbolism. It can represent as a square the material earth with its four cardinal directions. 124 And the superimposition of a name could be inferred as representing the spiritual and material authority of that figure on earth. Thus the chār cAlī could have magical talismanic qualities associated with it, since objects and images carrying apotropaic symbols were commonly employed for protection against evil, harm, and calamities. Simultaneously the chār cAlī could also have carried meanings of 122 Ladner 1983. 123Bjerken 2005, p.9. The last function, Bjerken argues was effectively employed by the Buddhist mission in Tibet to create a feudal polity during the late eighth and early ninth centuries. 124 Berglund 1990, p.253. 100  CHAPTER TWO power and influence on the material and mundane plane. The numerous examples of Square Kufic inscriptions found on later artefacts and seals testify to the magical qualities of the script at least in these particular cases [Figure 12]. 125 The parallels are there but the literary evidence lacking. Whether the chār cAlī derived from the mandala and the swastika, we cannot be sure but its widespread appearance in Iran under the Shīcī dynasties and the high incidence of its counterpart chār Muhammad does suggest that it was certainly more than just a text being repeated four times. Its arrangement, order, and frequency all suggest potent symbolic qualities. Under the Shīcī Būyids, the chār cAlī makes its debut; for later Fātimids, the name of cAlī was seen rendered in geometric patterns repeated six times and often juxtaposed with the name of Muhammad [Figure 13].126 THE SEAL SCRIPT AND CHINESE CONNECTIONS The seal script is not a widely applied term for Square Kufic. Blair mentions this term but she does not give any reference to who exactly employed or even invented the term.127 Close similarities between Square Kufic and the script on certain Chinese seals, however, have been observed [Figure 14]. The evidence, in fact, shows that a number of Īlkhānid official chancellery seals contained Square Kufic inscriptions, moreover, the script was also employed on talismanic and protective seals from at least the later Īlkhānid period. We 125 It is not surprising then that the Square Kufic script had come to acquire some sort of magical talismanic quality as it was employed in later centuries on magical artefacts and textiles for its apotropaic features. “Square Kufic was reserved for Koranic passages and pious phrases, as on a wooden Koran box or a talismanic shirt” [Blair 1998, p.100] – this was under the Ottomans during the sixteenth century. We find chār cAlī on a steel hand mirror produced during the fourteenth century in south-east Turkey [Turks 2005, pp. 395, Cat. no.73]. 126 As on the façade of the mosque of al-Aqmar in Cairo where the centred cAlī is linked to five Muhammads – designs that for Bierman were more Ismacīlī than simply Shīcī such as concentric circles with the bull’s eye etc [Bierman 1998, p.68.]. So the name of cAlī and Muhammad had taken on several renditions in geometric interlace and star patterns under the Fātimids – pentagrams, six-pointed stars, eight-pointed interlaced stars, stellar strapwork patterns etc. Blair and Bloom, however, regard Bierman’s theory on the ideological underpinnings of these symbols to be based on selective use of evidence [Blair 2003, p.174]. 127 Blair 1998, p.88. 101  CHAPTER TWO have the example of an official seal impression attributed to Īlkhān Anūshīrvān which carries within the central square a Square Kufic inscription with lā ilāha illā allāh muhammad rasūl allāh abū bakr, cumar, cuthmān, calī ( ‫ﻻ ﺍﻟﻪ ﺍﻻّ ﺍﻟﻠّﻪ ﻣﺤﻤﺪ ﺭﺳﻮﻝ ﺍﻟﻠّﻪ ﺍﺑﻮ ﺑﻜﺮ ﻋﻤﺮ‬ ‫[ )ﻋﺜﻤﺎﻥ ﻋﻠﻲ‬Figure 15].128 Similarly, a seal with apotropaic powers was used in the name of Shaykh Abū Ishāq (351-424H/963-1033) the founder of the Sūfī Kāzerūniyya order in the Fārs region on documents that guaranteed the Shakyh’s baraka and protection to sea-farers and travelers (dated 757H/1356) [Figure 16].129 The earliest comparison made between Square Kufic and Chinese seal script was by Marcel who observed that these strange inscriptions revealed traits “qui semblent offrir une ressemblance presque identique avec les cachets d’antiques caractères Chang fang ta tchouan, dont un grand nombre de manuscrits chinois et même quelques-uns de leurs livres imprimés nous offrent l’empreinte”.130 Thus, he makes a connection with the script of Yüan seal impressions found on a number of Chinese documents. Flury corroborated this display of common features between Square Kufic and certain Chinese scripts and attributed it to the influx of Chinese artistic and cultural influences into the Islamic world. He contended that even if the origins of Square Kufic script did not lie in Chinese scripts, they certainly must have had a strong influence on the development of Square Kufic.131 Herzfeld had a similar stance where he also reiterated the similarities of Square Kufic with the script used on Chinese seals and official stamps.132 Given the evidence, he claims that 128 This is the Diez A Folio 74 in Berlin. 129 Ibn Battū ta 1983, p.95; Soudavar 1992, pp.79-80. 130 Marcel 1833, pp. 226-227. 131Flury 1939, p.1747. Chinese aesthetic influences had been prevalent for centuries but the magnitude and frequency of these influences increased manifold with the advent and the establishment of the Mongol Īlkhānid dynasty in the mid-thirteenth to mid-fourteenth century. The Īlkhānids had strong diplomatic, political, cultural and ethnic connections with the Mongol Yüan dynasty in China. See Chapter One, pp. 56-58. 132 Smith 1936, p.323. 102  CHAPTER TWO naskhī carré’s origins were more likely to be found in the elaborate Iranian brickwork which had been in vogue for some centuries prior to the infiltration of the Chinese script into the Islamic world.133 Nevertheless, he does not completely dismiss the possibility of influence and sources because relations between the Islamic world and China had been well established for some centuries. 134 Amongst the western scholars on calligraphy, the Chinese connection is mentioned in Soucek’s article which briefly compares Square Kufic with the severe rectilinear type of script found engraved on the seals of fourteenth-century rulers in Iran. She cites the example of a seal impression on a document issued at Tabrīz in 773H/1372 bearing texts with Qur’ānic phrases and religious invocations in this script [Figure 17].135 Incidentally, as her essay is strictly on calligraphy, she fails to draw any connections between the script of these seal impressions with that on monuments. Sourdel-Thomine and Blair also join the bandwagon of possible Chinese influences on Square Kufic.136 Aesthetic quality and concealed religious sense were all that counted, and they explain the birth of varieties such as ‘square kufic’ which had been influenced by certain Chinese characters and which succeeded in engendering, from Central Asia to Turkey, complex and rigid types of lace-work where the distinctive fundamental features of ancient Arabic writing are barely recognisable. 137 Blair contributes a little to the area as she links the stucco panel type – that is the ‘closed’ types to the Yüan seals in Pagspa script 138 used in the Īlkhānid chancelleries. On scrutiny, we realize that, since Marcel’s profound insights and correlations, the discussion 133 Herzfeld 1955, p.14. 134 Herzfeld 1955, p.101. 135 Soucek 1979, p.10. 136 Sourdel-Thomine 1978, p.1122; 1986, p.213-214.; Blair 1998, p.85. 137 Sourdel-Thomine 1978, p.1122. 138 This has been discussed in Chapter One pp.56-58. 103  CHAPTER TWO has actually not moved much further. However, Blair’s remark on the Chinese Pagspa influencing a particular type of Square Kufic is noteworthy because although the ‘closed’ type was already in existence before the invention of the Pagspa, it nonetheless could have been the reason for the wider application, diffusion and popularity of stucco ‘closed’ types under the Īlkhānids. In essence, we return to our earlier distinction between the ‘open’ and ‘closed’ types that despite their possibly similar origins may have been influenced differently. This, in turn, may have also separately affected their application, function and meaning. The ‘closed’ types are single occurrences, sometimes carrying complex texts. Their strategic placement emphasizes singular attention. On the contrary, the ‘open’ types usually carry simpler texts that are then repetitively employed to cover large surfaces. Although it is not within the scope of this dissertation, it would be interesting to examine in later research the compositional schemes of the Yüan seals in Pagspa and the Īlkhānid seals to see if comparable parallels exist with Square Kufic types. A highly schematic comparison of the Tabrīz seal impression on the 773H/1372 [Figure 17] and the large lozenge Square Kufic inscription bearing Qur’ānic verses attributed to the fourteenth-century Muzaffarid restoration within the interior of the south domed chamber in the Great Mosque of Isfahān, reveals some similarities in compositional layout [Figure 18]. In the first case, the larger square is divided into a smaller central square with surrounding rectangular divisions, while in the second case, within the larger lozenge, the text is arranged similarly in blocks of smaller squares and rectangles. 139 This rectilinear resonance is even present in the Chinese seals in the Uighur script employed by the Īlkhāns Öljeytü and Abū Sacīd, and the Square Kufic inscriptions of the period. Just as the rhythm of negative and positive space is so integral to Square Kufic script, one of the 139Such quadrangular divisions are quite common in later Square Kufic inscriptions of the Tīmūrids and the Ṣafavids. 104  CHAPTER TWO distinguishing qualities of a fine Chinese seal is the elegant balance of red and white spaces within a restricted area.140 This reminds us of Qādī Ahmad’s remarks about the finest type of macqilī writing where the black is distinguishable from the white.141 However, this area of the influence of Chinese seals on Square Kufic requires further investigation, therefore, until then, all similarities and comparisons drawn by western scholars remain tenuous.142 CONCLUDING REMARKS This discussion on terminology has covered most of the significant terms used for Square Kufic and has in the process highlighted the problems and issues confronting the existing literature on the topic. A major problem encountered is the generic use of terms in traditional sources and in their subsequent interpretation or lack of, by western scholarship. For instance, a term such as macqilī is highly problematic, especially when we begin to analyze its place in the historical development of Arabic script and find that the question of formal discrepancies arises in Qādī Ahmad’s and Dūst Muhammad’s description of the historicized macqilī and the contemporary version of the macqilī script. Similarly, a term 140Lai 1976, p.89. The significance of emptiness and negative space is a major theme of the ancient Chinese text of Lao-Tzu which contains a passage illustrating the essential and intrinsic relationship between negative and positive space. Thirty spokes meet in the hub. Where the wheel isn’t is where it’s useful. Hollowed out, clay makes a pot. Where the pot’s not is where it’s useful. [Le Guin 1998, p.14] 141 3 Previous to the time when Persian writing came to be used and the universe was embellished, as in our days, with wonderful letters and tracings, there was the macqilī writing which consisted of straight (sathī?) lines with no rotundity (daurī) in it; the best macqilī writing is that in which one can distinguish blackness from whiteness. [Qādī Ahmad 1959, p.53]. 142 Another medium for receiving such direct influence on Square Kufic would be that of Īlkhānid coinage – after all seals and coinage are both official aspects of dynasties and would reflect greater role of the script and perhaps even a certain cross appropriation of function and meaning. 105  CHAPTER TWO like ‘kufic’ is quite obscure itself. The sources hardly ever specify precisely the kufic type under consideration. In addition, the problem of using a blanket term is compounded by the absence of illustrations of scripts and inscriptions in traditional sources. Therefore, it is difficult to collate the physical evidence with the fragmentary descriptions. However, this situation informs us about the different cultural methods of transmitting artistic and calligraphic knowledge in traditional societies where its dissemination is oral and the emphasis is on the practical and the practice rather than on the need to document the theory and philosophy behind its function and meaning. Another noticeable feature is the obvious predominance of terms in Persian. This is not surprising since certainly the origins for the ‘open’ type of Square Kufic lie in architectural brick decoration for the earliest examples documented are at Ghazna (492-508H/1099-1115), Bukhārā (12th century) and Herāt (c.597H/1201) ̶ in the eastern and Persian speaking lands. Even though, the appearance of the ‘closed’ types in Mosul (568H/1172) and Mardin (572H/1176/77) was more or less simultaneous, the majority of specimens in both types from the twelfth century onwards are to be found in Iran ̶ the land of the Persians. Thus, it would be fair to say that the Persians took to this aesthetic more than others in the Islamic world and therefore, their contribution to the frequency of the script is not only unprecedented, but also lies in its development demonstrated by the highly variegated and innovative material that embellished their architecture. No single reason explains this preference which was motivated by multiple combinations of technical, artistic, cultural and historical factors. We also see that any symbolic value associated with the Square Kufic script is more pertinent in the case of the ‘closed’ types, in particular, the chār cAlī and the chār Muhammad specimens which in their stylized geometric centralized mandala-like rotating 106  CHAPTER TWO composition have been specially suggestive of cosmological and metaphysical meanings, and simultaneously, of magical and ideological significance. Apart from these, the ‘closed’ spiral types especially under the thematic category of sacred names such as al-Asmā’ al- Ḥusnā, the Rāshidūn, and the Muhammad wa Chārdah Macsūmīn also reflect potent symbolic and iconographic qualities. The circular and the spiral schemes are closely interrelated; one could say that the spiral is the three-dimensional rendition of the two- dimensional circular type. Moreover, the swastika structure within the circular type also pertains to a movement that is, in reality, spiral in motion. The relationship of the text to the circular and spiral compositional schemes can be simultaneously one of universality and particularity in Islam. For instance where the chār cAlī, the chār Muhammad, and al- Asmā’ al-Ḥusnā could be regarded as universal Islamic symbols, the Rāshidūn and the Muhammad wa Chārdah Macsūmīn have a more sectarian denominational character carrying at times even particular political subtexts. Thus, from the perspective of meaning and symbolism that emerges from this study of terminology, there is an evident hierarchy present in the material for Square Kufic inscriptions. This area is explored in greater depth in the following chapters once we attempt to contextualize historically and culturally the evidence. 107